proceedings of the twelfth international symposium … · international symposium on structural...

PROCEEDINGS OF THE TWELFTH INTERNATIONAL SYMPOSIUM ON

STRUCTURAL ENGINEERING

VolumeⅠ

Edited By Hongping Zhu, Bin Huang and Jiping Ru

Science Press

科

学出版社

职教技术出版中心

www.abook.cn

Responsible Editor: Jialin Ren Anqi Tong

PROCEEDINGS OF THE TWELFTH INTERNATIONAL SYMPOSIUM ON STRUCTURAL ENGINEERING

Copyright © 2012 by Science Press, Beijing

Published by Science Press

16 Donghuangchenggen North Street

Beijing 100717, P.R.China

Printed in Beijing

All right reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form on by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permi- ssion of the copyright owner. ISBN 978-7-03-035787-8

International Scientific Committee

Chairman: Hongping Zhu, Huazhong University of Science and Technology, P.R.China

Members: (In Chinese alphabetical order) S.C.S. Cai, Louisiana State University, USA Nawawi Chouw, The University of Auckland, New Zealand Jie Cui, Guangzhou University, P.R.China Yang Ding, Tianjin University, P.R.China Xiuli Du, Beijing University of Technology, P.R.China Qin Fang, PLA University of Science and Technology, P.R.China Hanbin Ge, Nagoya University, Japan Linhai Han, Tsinghua University, P.R.China Hong Hao, University of Western Australia, Australia Bin Huang, Wuhan University of Technology, P.R.China Zongming Huang, Chongqing University, P.R.China Weiliang Jin, Zhejiang University, P.R.China Hui Li, Harbin Institute of Technology, P.R.China Jie Li, Tongji University, P.R.China Aiqun Li, Southeast University, P.R.China Dapeng Li, National Natural Science Foundation of China, P.R.China Dawang Li, Shenzhen University, P.R.China Hongnan Li, Dalian University of Technology, P.R.China Qiusheng Li, City University of Hong Kong, P.R.China Xiaojun Li, Institute of Engineering Mechanics, CEA, P.R.China Zhongxian Li, Tianjin University, P.R.China Weiqing Liu, Nanjing University of Technology, P.R.China Wenguang Liu, Shanghai University, P.R.China Yong Lu, University of Edinburgh, UK Guowei Ma, Nanyang Technological University，Singapore Z. John Ma, Tennessee State University, USA Jianguo NIE, Tsinghua University, P.R.China Ditao Niu, Xi’an University of Architecture and Technology, P.R.China Jinping Ou, Dalian Institute of Technology, P.R.China Weixin Ren, Hefei University of Technology, P.R.China Jiping Ru, National Natural Science Foundation of China, P.R.China Gangbing Song, University of Houston, USA Zhong Tao, Fuzhou University, P.R.China Jun Teng, Harbin Institute of Technology, P.R.China Jinguang Teng, Hongkong Polytechnic University, P.R.China Xintang Wang, Ningbo University, P.R.China

科学出版社


www.abook.cn

·iv·

Bo Wu, South China University of Technology, P.R.China Chengqing Wu, The University of Adelaide, Australia Zhishen Wu, Ibaraki University, Japan Yan Xiao, University of Southern California, USA Feng Xing, Shenzhen University, P.R.China Youlin Xu, Hong Kong Polytechnic University, P.R.China Bin Xu, Hunan University, P.R.China Qingshan Yang, Beijing Jiaotong University, P.R.China Zhiwu Yu, Central South University, P.R.China Yunfeng Zhang, University of Maryland, USA Yangang Zhao, Kanagawa University, Japan Xiaolin Zhao, Monash University, Australia Yun Zhou, Guangzhou University, P.R.China Hongping Zhu, Huazhong University of Science and Technology, P.R.China

Local Organizing Committee

Chairman

Bin Huang, (Wuhan University of Technology, P.R.China) Members

Huiming Fang, Fei Gao, Rongxiong Gao, Hanbing Luo, Hui Luo, Li Li, Lin Li, Xiaohong Long, Yu Miao, Dansheng Wang, Shun Weng, Shishu Xiong, Kun Ye, Yong Yuan, Aizhu Zhu, Junjie Zheng, Wei Zhang, Yaoting Zhang, Zhongxian Zhang (Huazhong University of Science and Technology, P.R.China)

Muyu Liu, Shujin Li, Yuanyou Xia (Wuhan University of Technology, P.R.China)

Secretariats Yong Yuan, Shun Weng (Huazhong University of Science and Technology, P.R.China)

Preface The Twelfth International Symposiums on Structural Engineering (ISSE-12) will be held at Huazhong University of Science and Technology, Wuhan, P.R.China, on November 17-19, 2012. ISSE-12 is organized by Huazhong University of Science and Technology and Wuhan University of Technology, sponsored by National Natural Science Foundation of China (NSFC), and supported by many related international and national organizations.

In today’s world, structural engineering has become an increasingly important issue due to safety, economic and environmental incentives. To provide an international forum for young experts from the research and practicing engineering communities, the previous eleven International Symposiums on Structural Engineering were successfully held in Leshan, Harbin, Shanghai, Beijing, Shenyang, Kunming, Tianjin, Xi’an, Fuzhou, Changsha and Guangzhou between 1990 and 2010. Due to developments over the two decades, this series of structural engineering symposiums possess distinct characteristics and features, which continue to attract many young and middle aged elite academics among Chinese people.

As an on-going series, the objective of the twelfth symposium is aimed at presenting and publishing the most recent researches and developments in structural analysis, design, construction, maintenance and disaster mitigation, discussing the implementation and development of new tools and technologies for professional application to safe and sustainable infrastructure, and further promoting international collaboration, cooperation, and mutual understanding.

More than 240 full papers from more than 20 countries were accepted for publication in the proceedings of ISSE-12. The contributions of all the authors to the proceedings are really appreciated.

We would like to thank all the participants and authors for their contributions. The special contributions from all the keynote speakers and the invited guests are gratefully appreciated. We also would like to express our sincerely thanks to all the members of the International Scientific Committee as well as the Local Organizing Committee, whose dedication is the key for the success of this symposium. Particular appreciation is extended to National Natural Science Foundation of China for sponsoring the conference, Huazhong University of Science and Technology and Wuhan University of Technology for hosting the conference. We really hope that all the participants will have a good time and enjoy their stay in Wuhan.

Hongping Zhu, Huazhong University of Science and Technology, P.R.China

Bin Huang, Wuhan University of Technology, P.R.China Jiping Ru, National Natural Science Foundation of China, P.R.China

September 5, 2012 科

学出版社


www.abook.cn

TABLE OF CONTENTS

Preface

Volume Ⅰ

Keynote Paper

The Next Challenge in Earthquake Engineering: Robust and Disaster Resilient Structures ..........................................................................................................................3 Stephen Mahin

Automated Model Construction Forseismic Disaster Assessment of Pipeline Network of Lifeline..........................................................................................................9 M. Hori, W. Lalith, S. Tanaka，T. Ichimura

Seismic Performance of Masonry-Infilled RC Frames..........................................................26 P. Benson Shing, Ioannis Koutromanos, Andreas Stavridis

Framework of Wind-Vehicle-Bridge Interaction Analysis and its Applications.............36 C.S. Cai, Wei Zhang, Xianzhi Liu, Wei Peng, S.R. Chen, Y. Han, J.X. Hu

A New Test Method for FRP-To-Concrete Bond Behaviour under Impact Loading....................................................................................................................53 J.F. Chen, X.Q. Li, Y. Lu, I.M. May, T.J. Stratford, R. O’sullivan, A. Sheil

Recent Research Developments in Ductile Fracture of Steel Bridge Structures .............61 Hanbin Ge, Lan Kang, Kei Hayami

Seismic Induced Pounding of Bridge Structures: an Overview...........................................78 Hong Hao, Kaiming Bi

Seismic Analysis and Design of Structure with Metallic Dampers ....................................88 Hongnan Li, Gang Li

Stochastic Viscous-Damage Constitutive Model for Concrete............................................95 Jie Li, Qiaoping Huang

Dynamic Interaction between Water and Bridge Pier under Earthquake Excitation..................................................................................................................106 Zhongxian Li, Yulong Li, Xin Huang, Ning Li

Dynamic Effects in Structural Response to Impact and Blast Loads ...............................121 Yong Lu

New Structural System for Earthquake Resilient Design....................................................131 Xilin Lv, Yuanjun Mao, Yun Chen, Jingjing Liu, Ying Zhou

Research on Seismic Perpormance and Failure Mode of High-Rise Diagrid Tube Structures ..............................................................................................................145 Jun Teng, Zuohua Li, Yanbo Han

·iv·

A New Approach to Derive Normalised Pressure Impulse Curves for Elastic Members Against External Blasts ........................................................................157 Jonathon Dragos, Chengqing Wu

Mutliple Purposes of Long-Gauge Fiber Optic Sensors Towards Structural Health Monitoring.....................................................................................................163 Zhishen Wu, Jian Zhang, Yongsheng Tang

The Study of a Self-Healing Cementitious Composite Using Urea Formaldehyde/Epoxy Resin Microcapsule...................................................................176 Biqin Dong, Ningxu Han, Ming Zhang, Xianfeng Wang, Hongzhi Cui, Feng Xing

Earthquake Hazard Mitigation Strategies and Countermeasures of Highway Infrastructure................................................................................................................183 W. Phillip Yen

Isolation Performance Study of the Hongkong-Zhuhai-Macao Bridge Engineering ....................................................................................................................................191 Hongping Zhu, Chulong Chen, Aizhu Zhu, Yong Yuan

Smart Materials and Structures

Numerical Analysis of Fire Response of Spatial Pre-Stressed Steel Truss in Fire ...................................................................................................................................205 Xintang Wang, Hongliang Sun, Pingxin Sun, Jinyi Zhang

Prediction of Long-Term Strength of Concrete based on Artificial Neural Network.............................................................................................................................211 Xiaoming Yang, Dan Shi

Effect of PVA Fiber on the Freezing and Thawing Behavior of Concrete .....................216 Fei Xu, Ju Chen, Weiliang Jin, Jihua Zhu

Study on the Effect of Natural Climate on the Corrosion of Rebar in Concrete............222 Jianhua Jiang

The Influence of Steel Reinforcement on Creep and Shrinkage of Concrete Columns ...................................................................................................................227 Tiejun Liu, Leilei Guo, Dujian Zou

Static Behaviour of Floating Piers under Unilateral Loads ................................................231 Yujun Qi, Fubin Zhang, Weiqing Liu

A Thermodynamical Consistent Microplane Plastic-Damage Model for Concrete....................................................................................................................................237 Jianying Wu, Shilang Xu

Smart Aggregates Used for Seismic Stress Monitoring in Concrete Structures ............243 Shuang Hou, Haibin Zhang, Jinping Ou

科

学出版社


www.abook.cn

·v·

Reinforcement of Damaged Frame Joint with Carbon Fiber..............................................249 Zhengfei Chang, Zhiyong Yang, Jun Hu,Gang Wang

Influences of Obstruction Action of Reinforcing Bar on Chloride Diffusion and Corrosion Initiation ..............................................................................................................254 Chengming Lan, Jie Yuan, Hui Li

Experimental and Numerical Study on Seismic Behavior of Circular Steel-Concrete-CFRP-Concret Solid Hybrid Columns........................................................261 Yongjun Liu, Dong Wang, Yu Wang

Preliminary Study on Finite Element Simulation of Steel- Asphalt Composite Isolation Layer..........................................................................................................267 Fei Yao, Jing Shao, Shouping Shang

Study on Distribution of Steel Corrosion Products in Concrete Structures ....................272 Yingyao Wu, Yuxi Zhao, Weiliang Jin

Shear Behavior of RC Bems Containing Corroded Steel Bars ..........................................278 Xin Xue, Seki Hiroshi, Yu Song

Comparitive Analysis on Z-Direction Qualities of Structure Steels Among Different Codes ..............................................................................................................284 Yuanqing Wang, Yuanyuan Zhang, Yongjiu Shi

Research on Impact Toughness of Q460C High-Strength Steel at Low Temperature..........................................................................................................................290 Yuanqing Wang, Yun Lin, Yannian Zhang, Gang Shi, Yongjiu Shi

The Impact of Cracks on Different Age of Concrete Small-Sized Hollow Block Masonry ..............................................................................................................................295 Yin Wang, Zhijian Shi, Yadi Hu

Innovatively Designed Externally FRP Reinforced Expansive Concrete Beams..........298 Qi Cao, Zhongguo John Ma

Behavior of SERC Short Columns with H-Rolled Steel under Axial Compression ..................................................................................................................................305 Chenxia Wang, Ziyan Guo

Hydration of Ultrafine and Ordinary Portland Cement at Early Ages .............................310 Jikai Zhou, Yuanyuan Yan, Xudong Chen, Minquan Zhou, Xiuming Jiao

Effect of Aggregate on Creep Behavior of Self-Compacting Concrete...........................316 Surong Luo, Pengfei Chao

Applied Study on the Inorganic Thermal Insulation Material in Concrete .....................321 Zhao Lin , Zhu Li , Wenjing Wang , Xiaoqing Bai , Yu Zhang , Gang Ma

·vi·

Structural Control and Disaster Prevention

Applying Discrete Control Theory to Develop an Explicit Integration Algorithm with Unconditional Stablility and Controllable Numerical Damping for Real-Time Testing...........................................329 Cheng Chen

Plate Size-Dependent Structured Behaviors of Magneto-Rheological Fluid Suspensions .........................................................................................................................336 Yongbo Peng, Jie Li, Roger Ghanem

Seismic Damage Control of Hybrid Structure Using MR Dampers.................................342 Longhe Xu, Yang Lv, Zhongxian Li, Yang Ding

Study on Isolating Horizontal Earthquake of One Structure System ...............................348 Biao Wei, Gonglian Dai, Qingyuan Zeng

The Three-Dimensional Magnetic Field Finite Element Analysis on the New Viscoelastic Damper with Adjustable Stiffness ..............................................354 Long Huang, Jianwei Tu

LQG Predicted Control Based on RBF Neural Network ....................................................359 Yanhui Liu, Ping Tan, Fulin Zhou, Yongfeng Du , Weiming Yan

Camber Analysis and Control Methods Research for Prestressed Concrete Beam ..............................................................................................................................366 Yugui Cao, Panfeng Wang

Experimental Research on Seismic Behavior of Spatial Joints in a Composite Frame Consisting of CFST Crisscross Section Columns and Steel Beams ..........................................................................................................370 Chengxiang Xu , Xiaoqiang Liu, Jicheng Zhang

Study on the Energy Distribution and Collapse-Resistant Capacity of RC Frames with Nonlinear Viscous Dampers ..................................................................376 Zhiwei Miao, Zhaoyun Qiu, Aiqun Li

Earthquake Ground Motion Prediction and Its Infleucen on Building Structures in Bhutan....................................................................................................382 Hong Hao, Choki Tashi

Simulation Analysis of Rockfall Impact on Reinforced Concrete Pier Column based on Holmquist-Johnson-Cook Model ..................................390 Hu Xu, Lei Lv, Zhibin Zhong, Zhixiang Yu

Protective Measure of Friction Base-Isolated Building With Limit Device Subjected to Near-Fault Earthquake ..........................................................................396 Jian Fan, Xiaohong Long

科学出版社


www.abook.cn

·vii·

Optimal Design of Viscoelastic Dampers Coupling Adjacent Unequal-Height Structures under Earthquakes....................................................402 Hongping Zhu, Xiao Huang

Structural Analysis and Integrity

Evolution of Probability Distributions in Different Constrction Stages —the Simulation of a Foundation Pit...................................................411 Haiying Wang, Hongwei Zheng, Jing Cao, Zhigang Song

Study on Occupant Comfort Mode in Wind Excitation Using Fuzzy Probability Method ..............................................................................................417 Dayang Wang, Yun Zhou, Yong Zhu, Xuesong Deng

Canonical Correlation Theory Applied in Reliability of Structural System...................423

Weidong Chen, Jiancao Li, Yanchun Yu,Wenmiao Yang,Fengchao Zhang

A Two-Step Kalman Estimation Approach for the Identification of Nonlinear Structural Parameters ................................................................428

Y. Lei, Y.Q. Jiang, Y. Liu

Nonlinear Static Analysis of Space Trusses Using Vector Mechanics of Structure................................................................................................................433 Qiubin Ni, Yuanfeng Duan, Edward C.Ting, Boqing Gao

Statistical Analysis of Dynamic Characteristics of Large Span Cable-Stayed Bridge Based on the Recursive Stochastic Finite Element Method................................................................................................................440 Bin Huang, Liping Zhu, Roohollah Fallahi Seresh

Static Reliailityb Analysis of Structures with Random Parameters based on Non-Orthogonal Polynomial Expansion ..........................................449 Bin Huang, Weidong Zhang, Roohollah Fallahi Seresh

Simulation of Collapse Process of Masonry Walls based on Expanded Discrete Element Method..................................................................................455 Wangxi Zhang, Jiechao Zhao, Meng Liu,Weijian Yi, Yan Xiao

An Application Study on a Forced Settlement for Rectifying Leaned Structure by Loess Ground Stiffness Softening Method ......................................461 Jun Zhou, Yu Song

A Research on the Full State Function Constitutive Relation Model...............................466 Zhenhai Wei, Mengshu Wang, Dingli Zhang

Connecting Parameters Optimization on Unsymmetrical Twin-Tower Structure Linked by Sky-Bridge.......................................................................470 Huangsheng Sun, Hongping Zhu, Xiao Huang

·viii·

Computational Modeling of Realistic Experiment Boundary Conditions and Restraints ...........................................................................................................477 Liping Kang, Roberto T. Leon, Xilin Lu

An Efficient Algorithm for the Probability Density Evolution

Analysis of Strong Nonlinear Structures.................................................................................484

Jianbing Chen, Shenghan Zhang, Jie Li

Inelastic Time-Dependent Structural Analysis Method for High-Rise Buildings ..............................................................................................................490 Xin Zhao, Bingquan Yu

Finite Element Analysis of Steel Tube Strengthened Damaged Bridge Piers under Lateral Loads...........................................................................498 Xueqiong Li, Jun Deng

Redundancy-Based Safety Performance Optimization of Large-Scale Glass Facade Structures .......................................................................................503 Jinming Ma, Don Chen, Xiaoguang Zhang, Zhenguo Sui

Interface Connection Method for Multi-Scale Simulation of Linear Structures...........................................................................................................................508 F.Y. Wang, Y.L. Xu

Influence Factors to the Workability of Thermal Insulation Glazed Hollow Bead Concrete and Mix Design Using The Orthogonal Design Method .....................................................................514 Wenjing Wang, Zhu Li, Lin Zhao, Xiaoqing Bai, Yu Zhang, Gang Ma

Performance Analysis of Sliding Mode Control for Real-Time Hybrid Test .........................................................................................................520 Bin Wu, Huimeng Zhou

Structural Design, Construction and Management

Approach Analysis on the Technological Construction Process of Corrugated Arch Metal Roof ................................................................................................529 Fuju Gao, Xiliang Liu

High Pier and Long-Span Continuous Rigid Frame Bridge Construction Linear Control Analysis .....................................................................................534 Changxi Liu, Aihong Shi

Reconstruction of the Closure Segment of Main Span in Existing Prestressed Concrete Cable-Stayed Bridges ..........................................................539 Hongjiang Li, Jinquan Zhang, Wanheng Li, Hanliang Wu

Research and Application on Performance-Based Optimization of Concrete Structure Construction ..........................................................................................546

科学出版社


www.abook.cn

·ix·

Ye Tian, Nanguo Jin, Xianyu Jin, Bingliu Wu

Classification System of Structures for Wind-Resistant Design........................................556 Jiansheng Zhang, Yue Wu, Di Wu, Guangen Zhou

Structural Design of the Mencius Grand Theatre..................................................................560 Wei Liu, Dayi Ding

The Deformation Mechanism of a Multistoried Building Raft Foundation Near a Deep Excavation and its Safety Control ..............................................564 Aijun Yao, Huanfang Chen, Xuejia Yang

Comparative Study on Design Methods for Steel Structural Members Strengthened by Increasing the Section Area While under Load in Different Codes......................................................................................570 Yuanqing Wang, Ruixiang Zhu, Guoxin Dai, Gang Shi

Finite Element Analysis and Design Method Study on the Interactive Buckling of Welded Steel Members under Axial Compression ...................577 Gang Shi, Wenjing Zhou, Yuanqing Wang, Yongjiu Shi

Progress in Application and Research of High Strength Steel Structures .......................583 Gang Shi, Huiyong Ban, Yongjiu Shi, Yuanqing Wang

Optimal Shape Design of Several Kinds of Membrane Structures...................................587 Bingbing San , Yue Wu

Construction and Operation of the Construction Site Management Model based on the Lean Construction Theory.....................................................................592 Gang Li, Guoqiang Hu, Jiayao Zhao

Reconstruction of Human Induced Walking Load based on Multi-Segmental Rigid Model...................................................................................................598 Mengshi Zhang, Jun Chen, Yixin Peng

Design and Nonlinear Analysis of a New Skip-Floor Shear Wall Structure ....................................................................................................................604 Jianxin Liu, Pei Li, Hongnan Wang, Meichun Zhu,

Zhihong Zhang, Ying Wang, Hongwei Wang

Information Entropy-Based Optimal Sensor Configuration for Continuous-Coordinate Structures .....................................................................................609 Tao Yin, Dianqing Li, Hongping Zhu

Structural Condition Assessment Subject to Environment and External Excitation

Behaviour of Concrete-Filled Steel Tubular Frames with Rocking-Walls under Cyclic Loading .....................................................................................619 Bo Wang, Jingfeng Wang, Xudong Gong, Xiang Li, Hongwei Cheng

·x·

Effects of Railway Wind Fence on the Aerodynamic Forces of Train and Fence ..........................................................................................................625 Bo Li, Zhi Xu, Qingshan Yang, Shaohua Feng

Impact Factors of the Bridge under Moving Vehicular Loads...........................................631 Xinfeng Yin, Xiao Hu, C.S.Cai, Yang Liu

Relationship Between Time-History Characteristics of Input Energy and Structural Displacement Response in SDOF Systems...................................637 Zhefeng Liu, Kui Chen , Qiong Zhou

Finite Element Analysis of Corroded Reinforced Concrete Bridge Piers under Cyclic Loading ..........................................................................................644 Haijun Zhou, Xi Xu, Kaizhi Lv

Numerical Simulation on Sandwich Panels with Rotational Friction Hinge Device Against Blast Loading.......................................................................650 Wensu Chen, Hong Hao

Wind Tunnel Measurements of Aerodynamic Forces on Vehicles and Bridges under Crosswinds .................................................................................658 Yan Han, Steve C. S. Cai, Zhengqing Chen, Jiexuan Hu, Chunguang Li

Loads Generated By Human Walking: Experiments and Numerical Modeling ....................................................................................................................673 Jun Chen, Yixin Peng, Ting Ye

Loads Generated by Human Jumping: Experiments and Numerical Modeling.............679 Jun Chen, Ling Wang, Bo Chen, Shixin Yan

Finite Element Analysis of Foam Protected RC Members Against Blast Loads .....................................................................................................................686 Chengqing Wu, Hamid Sheikh

Dynamic Analysis Method for Fire-Induced Progressive Collapse of a Planar RC Frame ..................................................................................................................692 Yi Li, Xinzheng Lu, Weiming Yan, Lieping Ye, Shicai Chen

Study on Wind Loads of a Typical Super-Tall Building.....................................................698 Lunhai Zhi, Q.S. Li

Thermal Analysis and Design of Over-Long High-Rise Concrete Buildings ................703 Hongdong Zhang, Hongqiang Fang

Analysis on Tower Crane under Wind Load by ANSYS....................................................709 Yongqiang Gu, Wenfeng Wang , Shaodong Guo

Failure Mode and Critical Temperature of Tubular Y-Joints under Fire Condition ...................................................................................................713 Fei Gao, Shaobo Chu, Miao Zhang, Enxu Lu, Xingquan Guan, Zhigao Wu

科

学出版社


www.abook.cn

·xi·

Numerical Simulation of Wind Loads and Wind Environment in Super-Star Hotel at Sanya Phoenix Island ..............................................................................720 Zhupan Li, Xi Wang, Yiping Zhang

Wind-Induced Vibration and Damage Detection for High-Rise Building ......................723 Xin Zhao, Zhenshan Lin

Analysis on Vertical Tank Fluid-Solid Coupling Vibration and Base Isolation Dynamic Response .................................................................731 Yicheng Jiang, Chunming Liu, Zhenguo Li, Yinghou He

Substructural Identification of Nonlinear Restoring Force and Dynamic Loadings with Partially Unknown Excitations ............................................739 Jia He, Bin Xu

Dynamic Performance Study on the Super High Damping Rubber Isolation Bearings Depend on the Real-Time Substructure Hybrid Loading System......................................................................................745 Liu Yi,Yong Yuan , Demi Ai

Dynamic Load Assessment with Out-Put Information of a Substructure Only .............752 Y. Ding, S. S. Law, B. Wu, B. Y. Zhao

Volume II

System Applications and Field Tests

Selection of Durable Closure Pour Materials for Accelerated Bridge Construction .....................................................................................................................761 Peng Zhu, Zhongguo John Ma

Experiment, Analysis and Design of Semi-Rigid Joints to Concrete Filled Steel Tubular Columns ..................................................................................770 Jingfeng Wang, Wenbin Xing, Xudong Gong

Approximate Calculation on Deflection of Reinforced Concrete Box Girders Considering the Effect of Shear Lag and Shear Deformation .......................................................................................................780 Lixia Lin, Nahong Ding, Yuanhai Zhang, Yaping Wu

Experimental Research on Fundamental Mechanical Performance of Straw Concrete Block.............................................................................................................784 Kun Qian, Tian Xia, Liguang Xiao

Research and Applicant of Solar Energy-Prefabricated Bamboo Pole House ...............789 Bo Shan, Li Gao, Zhi Li, Yan Xiao, Zheng Wang

Priliminary Research on Behavior of Composite Joints with CFST Column to Steel Beam under Long-Term Loading ..................................................795 Wenda Wang, Pengpeng Zhang, Aihua Zou

·xii·

Preliminary Research on Fire Resistance of Concrete-Filled Steel Tubular Columns to Composte Beam Frame.........................................................................801 Jingxuan Wang, Wenda Wang, Zhongmao He

Experimental Study on the Connector and Bearing Performance of Lightweight Aggregate CFT Composite Beam........................................807 Zhongqiu Fu, Bohai Ji, Danwen Shen, Duanduan Zhao, Yadong Dong

The Experimental Research on Flexural Performance of CLT Deck...............................813 Yongjian Liu, Meizhen Fu, Yujuan Liu, Xiaoyu Gao

Analysis of the Torsional Stiffness of the Dumbbell-Shaped Concrete-Filled Steel Tubular Members .................................................................................819 Jiangang Wei, Bin Fu, Baochun Chen

Calculation of Interface Slips in Duplex Steel Reinforced Concrete Column by Principle of Minimum Potential Energy..........................................824 Bin Liang, Xiaomin Liu, Rui Mao

Research on the Shear Strength of Reinforced Concrete Interior Joint Based on Bayesian Theory ................................................................................828 Tao Wu , Ningning Wang, Dayuan Li , Xi Liu, Guohua Xing

Shear Strength of Reinforced Concrete Column Based on the Bayesian Theory ...........................................................................................................................835 Tao Wu, Jie Tang, Xi Liu, Dayuan Li

Finite Element Analysis on the Mechanical Properties of CFST Frame with Dampers....................................................................................................................843 Fengming Ren, Yun Zhou, Jianwei Liang, Xiuli Du

Analysis of Flexural Capacity of the High-Strength Cold-Formed Ultra-Thin-Wall Steel Truss Beam.................................................................850 Hao Luo, Qian Gu, Guangyue Ma

Experimental Investigation and Design of Concrete-Filled Cold-Formed Thin-Walled Steel Tubes Subject to Bending..............................................855 Aoyu Jiang, Ju Chen, Weiliang Jin

Stress & Strain Analysis of Specially Shaped RC Column in Elastic-Plastic State .................................................................................................................862 Bin Li, Yin Fen, Junjie Wang, Xingsheng Yu

Investigation on the Fatigue Life Prediction of Reinforced Beams under Complex Environment of Chloride Ion and Stray Current........................867 Mengcheng Chen, Kai Wang, Zhen Qin, Li Xie

科

学出版社


www.abook.cn

·xiii·

Analysis on Fatigue Behavior of Orthotropic Steel Deck with Pavement in Suspension Bridge............................................................................872 Ce Chen, Bohai Ji, Rong Liu, Muye Yang, Hanjiang Xu, Hirofumi Maeno

Perception Effect Analysis of Smart Pot Rubber Bearing under Vertical Load......................................................................................................881 Bin Huang, Yonghui Feng, Roohollah Fallahi Seresh, Xingqi Song

The Distortion Contrast Analysis of Inverse Structure and General Structure Asphaltum Pavement .................................................................................886 Jianhong Gao

The Experimental Study on Mechanical Properties of Steel Crane-Girder Before and After Reinforcement .....................................................................890 Yang Wen, Cong Li

Experiment Study on the Influence Factors of Thermal Stress of Building Steel Elements.............................................................................................897 Hongrui Yang

Research on Shear Capacity for Concrete-Filled Rectangular Steel Tubular Frame Joint with Inner Strengthening Rings................................................902 Chunyan Gao, Kuahai Yi, Bin Li

Simplified Calculation for the Stiffness of RC Simple-Supported Cracked Beams under Uniform Loads..................................................907 Wangxi Zhang, Zhiming Yang, Xiaoan Xie,Weijian Yi, Yan Xiao

Test Study on Reinforced Concrete Beams Strengthened with Prestressed Near Surface Mounted CFRP Strips .........................................................913 Hui Peng, Jianren Zhang, Yang Liu, Xianfeng He, C.S. Cai

Mechanical Property Analysis of Four Types of Concrete Piles.......................................922 Jianlei Mu, Jikai Zhou, Fengchen Zhang

The Nonlinear Dynamic Stability of Arch Structures under Harmonic Loading Action ..........................................................................................................928 Dongping Wu, Yunli Chen

Study on Stability Capacity Correction Coefficient of Coupler Steel Tube Falsework ..................................................................................................934 Changming Hu, Xiaozhou Fan

Geometric Nonlinear Analysis of Long-Span Composite Girder Cable-Stayed Bridge with Three Pylons During Construction Stage .........................................................................................................939 Muyu Liu, Xiaoguang Deng

·xiv·

Deterioration Mechanism of Reinforced Concrete Columns Suffered Corrosion and Freeze-Thaw Cycles.......................................................947 Xiaoning Zheng, Bo Diao, Yang Sun, Yinghua Ye

Deformation Development Mechanism Analysis of Tubular T-Joints under Impact Loading.............................................................................955 Hui Qu, Chao Xu

Deflection Analysis of Z-Core Sandwich Panels under Bending in Weak Direction........................................................................................................960 Xuegao Wang, Yongbo Shao, Mingjuan Cui

Compressive Behavior of Square and Rectangular High-Strength Concrete-Filled FRP Tubes ............................................................................965 Togay Ozbakkaloglu

Mesoscale Modelling of Concrete under Split Tension.......................................................971 Xiaoqing Zhou, Yong Xia

Distributed Load Identification for Vertical Wind Turbine Straight Blade ....................977 Jinghua Lin, Youlin Xu, Songye Zhu, Yong Xia

Theoretical and Experimental Study on the Long-Term Behavior of Crushed Limestone Sand Concrete Beams......................................................983 Yi Zheng, Yaoting Zhang, Yiyan Xu, Yuancheng Peng

Experimental Study of the Dynamic Behaviour of High Damping Rubber Bearing Isolator............................................................................................992 Zhenxue Zhang, Yong Yuan, Hongping Zhu, Wei Wei

Experimental Study on Fire Resistance of Composite Slabs Filled with Demolished Concrete Blocks ...............................................................................999 Jing Duan, Bo Wu, Shuyi Zhang

Numerical Study on the Behavior of RC Columns Considering Confinement and Strain Rate Effects.....................................................................................1006 Xiang Zeng, Bin Xu

Time-Dependent Behaviors of Prestressed Concrete Track Beams under Sustained Loads .................................................................................................1012 Weichen Xue, Ting Liu

Analytical Model for Bar-End Cover Separation Failure in RC Beams Strengthened with Near-Surface Mounted FRP2............................................1018 S.S. Zhang, J.G. Teng, J.F. Chen

Behavior of Reinforced Concrete Columns under Different Loading Histories .....................................................................................................1024 Weijian Yi, Yi Liu, Yun Zhou

科学出版社


www.abook.cn

·xv·

Experimental Investigation of Ductile Failure of Steel Bridge Piers .............................1031 Lan Kang, Hanbin Ge, Shinki Hada

Effect of End Bending Moment on the Global Stability of Buckling Restrained Braces .....................................................................................................1039 Junxian Zhao, Bin Wu, Chia-Ming Uang, Jinping Ou

Seismic Analysis and Vibration Mitigation

Aeoline Vibration Test for Twin Bundle Conductors ........................................................1049 Li Li, Huajin Cao, Yicheng Jiang, Yuankun Chen

Dynamic Analysis and Seismic Performance Evaluation of Fluid-Filled Tank Considering Liquid Sloshing..................................................................1055 Shanpo Jia, Youqing Zhao, Chengxiang Xu, Jike Tan

Earthquake Responses of Plane Frame Structures to Wave Passage Excitations ....................................................................................................................1060 Tielin Liu, Yangyang Sun, Wenbo Chen, Yingchun Jiang

Relations between Ground Motion Intensity Measures and Story Drift Ratio of Super High-Rise Buildings..........................................................1066 Xiao Lu, Xinzheng Lu, Lieping Ye, Wankai Zhang

Seismic Reliability Analysis and Optimization Design of Urban Water Distribution Network.....................................................................................................1072 Wei Liu, Jie Li

Shaking Table Tests of a Free-Standing Cultural Relic Strengthened by Magnet Blocks .............................................................................................1077 Qian Zhou, Weiming Yan, Jinbao Ji

The Pseudo Static/Dynamic Response Components of Large Structures under Non-Uniform Seismic Excitations ..........................................................1084 Guoliang Zhou

Combinatorial Innovation in Bridge Seismic Research.....................................................1093 Jian Zhong, Wancheng Yuan, Kai Wei

Structural System and Seismic Performance of Three-Tower Suspension Bridges ....................................................................................................................1099 Xinjun Zhang, Hailing Sun

Test on Seismic Behavior of I-Shaped Section Superimposed Semi-Precast Reinforced Concrete Walls .............................................................................1105 Xun Chong, Xianguo Ye, Ning Li, Lin Xu

Experimental Studies on Seismic Performance of Subsidiary Piers of Long-Span Cable-Stayed Bridge with Energy Dissipation ...............................1110 Limin Sun, Jun Wei, Wen Xie

·xvi·

The Experimental Study on Self-Adaption Minimum Controlling Synthesis (MCS) Algorithm of AMD System...............................................1116 Xiaofeng Lin, Jianwei Tu

An Analysis on the Seismic and Windinduced Response in the Roof Maching Room of the Ship-Lifter ..........................................................................................1122 Guang Que, Jianwei Tu

Seismic Analysis of Rectangular Hollow Section RC Piers Including Flexure-Shear Coupling Behavior ..........................................................................................1129 Ning Li, Zhongxian Li, Lili Xie

Seismic Analysis of a 600kW Wind Turbine in Frequency Domain and Time Domain .......................................................................................................................1137 Liang Ji, Lei Zhu, Xiaoqin Yao

Time-History Analysis of Seismic Response of Chengdu East Railway Station Subjected to Non-Uniform Ground Motions.........................................1142 Yongjiu Shi, Bo Zhao, Yuanqing Wang, Yangjiang, Zhihua Chen

Seismic Response Analysis of Large-Size Liquid Storage Tanks Considering Liquid-Solid Coupling .......................................................................................1148 Yuan Zhang, Youhai Guan

Seismic Damage Analysis of Chinese Ancient Wooden Buildings under Wenchuan Earthquake ...................................................................................................1151 Qian Zhou, Weiming Yan, Jinbao Ji

Investigation of Collapse Pattern of Seismic Isolated Structure Subject to Multi-Dimensional Earthquake............................................................................1158 Yongfeng Du , Xiaohu Wang

Experimental and Numerical Modelling of Composite Crisscross Concrete-Filled Steel Tubular Columns Subjected to Earthquake Loading .................1164 T.Y. Yang, Dorian P. Tung, Chengxiang Xu

The Hysteretic Characteristic with P-Δ Effect and Influence on the Collapse Resistance Capacity of Structure under Earthquakes.................................1170 Renjie Liang, Jing Wu, Hanbin Ge, Xuechen Lei

Seismic Performance of Moment Connections between Cold-Formed Steel Sections.....................................................................................................1180 Fubo Cao, Gentain Zhao

Seismic Behavior of Long-Span Connected Structures under Multi-Supported and Multi-Dimensional Earthquake Excitations..................................1185 Wei Lin, Shanghong Chen, Ai Qi

科

学出版社


www.abook.cn

·xvii·

Seismic Fragility Analysis Based on Reliability of Reinforced Concrete Frame Structures.............................................................................1190 Qiaoyun Wu, Hongping Zhu

Spectrum Displacement based Seismic Fragility Curves for China High Rises ........................................................................................................................1196 Fan Wu, Xin Y. Yang

Seismic Behavior of FRP-Concrete-Steel Double Skin Tubular Columns...................1201 Yunita Idris, Togay Ozbakkaloglu

Seismic Response Control of Reticulated Shell by Using Semi-Active Friction Damper .....................................................................................1208 Bo Chen, Kaikai Lu, Pengyun Li, Chunfang Song

The Benefit of Upliftable Structures in Earthquakes..........................................................1216 Nawawi Chouw, Xiaoyang Qin

A Practical Computation Formula for Seismic Isolation Coefficient in the Horizontal Direction of LRB Base-Isolated Building Structures ..........................................................................................1224 Rong Fu, Kun Ye, and Li Li

Experiments Study on the Seismic Behavior of Slitted Shear Wall Structure with Permanent Thermal-Insulation Wall form under the Low Cyclic Load......................................................................................................1229 Yuanzhen Liu, Zhu Li, Shangsong Qin, Li Xu

Passive Optimal Control of Adjacent Structures under Earthquake Excitation................................................................................................................1235 Dongdong Ge, Hongping Zhu, Qisong Miao, Peng Jin

Underground Structures, and Soil-structure Interaction

New Methods of Safety Evaluation for the Rock/Soil Mass Surrounding Tunnel under Earthquake..................................................................................1247 Xuansheng Cheng , Yingren Zheng, Ruirui Tian, Xiuli Du

Status of Seismic Analysis Methods for Traffic Tunnel and Their Applicability Suggestions in China .....................................................................1257 Chuan He, Ping Geng, Qixiang Yan, Kun Feng

Seismic Analysis and Performance Evaluation of Shield Tunnels in Transverse Direction .............................................................................................1272 Zhiyi Chen, Hao Shen, Yong Yuan

Performance Study on Prestressed and Precast Concrete Segment Joint for Shield Tunnel.............................................................................................1281 Fengjun Liu, Jiying Qin, Hehua Zhu

·xviii·

A New Concept of Concrete Constitutive Model: the Fractal-Microplane Model........................................................................................................1286 Hanliang Wu, Yuanfeng Wang

Analysis on Three Dimensional Response of Utility Tunnel under Inconsistent Seismic Excitations...................................................................1292 Qingxia Yue, Xin Zhang

Free Span Detection of Submarine Pipeline Based on Wavelet Method.......................1297 Chao Wang, Hongping Zhu

Corrosion-Weathered Red Layer Characteristics and Treatment Study of Subway Station Excavation.....................................................................................1303 Liming Peng, Dan Zhao, Juan Huang

Research and Analysis on Stability of Bolt Reinforced Tunnel in Anisotropic Layered Rock Mass with Low Inclination Angle Stratification ...............................................................................................1308 Dan Li, Shiwei Bai , Hao Chen, Binwei Xia

Structural Mechanics, and Structural Strengthening

Analysis of Live Load Nonlinear Effect of Composite Girder Cable-Stayed Bridge with Three Towers..............................................................................1321 Feng Wang

Analysis of the Mechanical Properties and Stength Evaluation of Tension Legs of TLP ............................................................................................................1327 Shisheng Wang, Bin Xie, Dongyun Huang , Xichong Yu

Analysis on Structure System of Bucking-Restrained Braced on Steel Frame............1332 Zhan Wang, Zhongcai He, Jianian He

Experimental Research on Seismic Behavior of Concrete-Filled Rectangular Steel Tubular Planar Frames...............................................1337 Bin Li, Xia Li, Chunyan Gao

Finite Element Study on Load-Carrying Performance of Double C Steel Axially-Loaded Short Column with Gusset Plates...........................1342 Ming Chen, Yang Sun

Complex Movement and Restoration of a Heritage Masonry-Timber Building........................................................................................................1347 Erjun Wu, Zhu Chen, Wang Jianyong

Nonliner Static Analysis on SRC-RC Vertical Hybrid Frame Structure.......................1353 Kai Wu, Ming Yang, Pingzhou Cao, Hongpeng Ge, Li Zhang

科

学出版社


www.abook.cn

·xix·

Test on Mechanical Behavior of Reinforced Recycled Coarse Aggregte Concrete Beams ..........................................................................................1358 Zongping Chen, Jie Fan , Chunheng Zhou, Jiangmei Wang

Dynamic Performance of Semi-Rigid Steel Frame on Macroscopic Elements ........ 1363 Tao Wang, Zhan Wang, Junjie Feng

Overload-Induced Progressive Collapse Simulation for a Reinforced Concrete Arch Bridge .......................................................................................1368 Shengnan Huang, Xinzheng Lu

Performance of Integral Abutment Wall Supported with Concrete Piles......................1374 Zhao Liu, Xiaoqin Jin, Z. John Ma

Load Tests and Bearing Capacity Evalution of Spatial Torsional Tied-Arch Bridge.....................................................................................................1380 Wei Huang, Lei Yang

Bearing Capacity of the Straw Concrete Lightweight Steel Composite Wall ..............1385 Dianzhong Liu, Ying Yu

Mechanism of Progressive Collapse Resistance of Reinforced Concrete Spatial Frame Structure ...........................................................................................1390 Jingang Xiong, Yan Li, Zhaoqiang Wu, Gengfeng Hou, Yinong He

Experimental Study on Mechanical Behavior of Reinforced Recycled Aggregate Concrete Columns under Compression Loading ..........................1395 Zongping Chen, Wei Zheng, Peihuan Ye

Hysteresis Performance Analysis of Reinforced Concrete Thin-Walled Piers with Rectangular Hollow Cross-Section under Bi-Directional Low Cyclic Loading Based on Fiber Beam-Column Element .............................................................................................................1401 Zhanghua Xia, Xueyang Huang, Zhouhong Zong

Experimental Study on Fatigue Behavior of Welding in -5°C for Steel Structure of Railway Bridge....................................................................................1407 Xinxin Zhao, Li Wang, Yuling Zhang, Xiaoguang Liu, Yongjie Pan

Experimental Study on the Shear Mechanical Behavior of Steel Reinforced Recycled Coarse Aggregate Concrete Beams.................................1413 Zongping Chen, Yuliang Chen,Wudang Ying,Ming Zhong

Analysis of Column-Supported Steel Silo Structure Stability based on the Finite Element Method......................................................................................1418 Zhiyong Yang, Shunhui Liu, Song Zhao, Jun Hu

·xx·

Behavior of Partially Encased Composite Columns Subjected to Eccentric Compression......................................................................................1422 Gentian Zhao, Fubo Cao, Shan Wang, Xin Wan

Analysis on Mirror Lateral Displacement and Steel Frame Connection Strength of Collector ...........................................................................................1427 Junbin Gu, Jiaogen Zhu

Theoretical and Experimental Study on Unconsolidated Isolation Structure and Unconsolidated Bearing.................................................................1431 Wenguang Liu, Wenfu He, Qiaorong Yang, Lusun Wei,Yan Guo

Contrast Test Investigation of Different Multi-Layer Composite Structures under the Simulated Fragments Impact of 150mm Red Bomb...................................................................................................1440 Shuang Li, Tianyun Wang

Experimental Research on Bearing Capacity of Steel Reinforced Recycled Aggregate Concrete Columns ..........................................................1445 Zongping Chen, Ming Zhong , Ling Li, Yuliang Chen

Pile-End Pressure Grouting Technology Effect on the Carrying Capacity of Bored Pile...............................................................................................................1451 Yuqin Feng, Lu Zhang

Reliability-Based Estimates of Dynamic Load Allowance for Capacity Rating of Prestressed Concrete Girder Bridges under Different Road Surface Conditions .............................................................1455 Lu Deng, C.S. Cai, Michele Barbato

The Influence Analysis of Creep Factors on Long-Term Deformation of PC Continuous Rigid-Frame Bridge.........................................................1464 Zhifang Lu, Muyu Liu

The Simplification Calculation Method of the Beam-Column Joint Rotation of Frame Structures under Static Load .......................................................1468 Long Yang, Lin Li, Hongping Zhu

Finite Element Analysis of the Bearing Properties of the Floor Structure of Precast Inclined Support Cold-Formed Steel Building.............1474 Xiyue Liu, Yuanqing Wang, Gang Shi, Ming Liu

Free Vibration of a Simple Cable-Network-Damper System...........................................1480 Haijun Zhou, Xia Yang

科

学出版社


www.abook.cn

·xxi·

Structural Health Monitoring, and Sensing Technique

Dynamic Nonlinear Analysis of the Shanghai Pudong Shangrila Hotel Extension Engineering ................................................................................1489 Hu Qi, Yungui Li, Xilin Lu

Multi-Scale Damage Model and Its Application to RC Structures.................................1496 Xiaodan Ren, Jie Li

Study on Structural Health Monitoring System of Xihoumen Bridge ...........................1500 Zhiqiang Liu

Application of the Laser Doppler Vibrometer for Structural Testing: Case Studies.................................................................................................................1505 Kaoshan Dai, Bin Zhao, Xiaosong Ren, Qingjun Chen

Exploratory Research on Highway Bridges Seismic Damage Assessment System for China .................................................................................1511 Liang Zong, Gang Shi, Saini Yang, Yuanqing Wang

Dynamical Testing of a Cable-Stayed Bridge Model Using Wireless Sensor Network..........................................................................................................1517 Chengyin Liu, Haijing Zhang, Xianyin He, Yukun Guo

Sensitivity Analysis of Damage Identification of Random Beam Structures to Static Load Position...............................................................................1525 Bin Huang, Furong Zhu, Ruifang Zong, Roohollah Fallahi Seresh

Structural Damage Assessment by Using Strain Frequency Response Function .....................................................................................................................1530 Zengguo Jiang, Bo Chen, Zhen Zhang

Structural Damage Identification of Municipal Bridges Using Strain Frequency Response Function.....................................................................................1535 Zengguo Jiang, Zhen Zhang, Bo Chen

Wavelet Transform based Hybrid Damage Detection Method for Structures .............1539 Xin Zhao, Wenjian Yang

Damage Assessment of a Bridge after a Barge Collision..................................................1547 Yanyan Sha, Hong Hao

Application of Compressive Sampling for Civil Shtructral Health Monitoring ..........1553 Yuequan Bao, Hui Li, Jinping Ou

Time Series Based Structural Nonlinear Damage Identification Algorithm Using ARMA/GARCH Model............................................................................1559 Liujie Chen, Ling Yu

·xxii·

Bearing Damage Identification Study of Long-Span Continuous Rigid Frame Bridge....................................................................................................................1566 Jie Niu, Fupeng Chu, Longhua Wang, Zhouhong Zong

Damage Identification of Structures with Substructural Flexibility ...............................1573 Yong Xia, Xiaoqing Zhou, Shun Weng

Improved Substructuring Method for Calculation of Eigensensitivity ..........................1581 Shun Weng, Tuoyu Jiang, Hongping Zhu, Ling Mao, Yong Xia

Monitoring Highway Traffic of Long-Span Bridges Based on WIM Data ..................1589 Zhiwei Chen, Youlin Xu, Kaiyuen Wong

A Bio-Inspired Damage Detection Approach Based on Multi-Scale Wavelet Finite Element ......................................................................................1596 Songye Zhu, Wenyu He, Weixin Ren

Damage Identification in Beam Structures Based on Displacement Statistical Moment......................................................................................................................1603 Dansheng Wang, Wei Xiang, Xiaoqiang Chen, Hongping Zhu

Damage Assessment of Axially Loaded Cracked Beam Using Bayesian Estimation Theory ....................................................................................................1609 Yaoting Zhang, Changyu Fang, Chao Ma

Guided Wave Based Methods for Damage Detection: Experimental Study on Concrete Joint ..................................................................................1615 Ying Wang, Hong Hao

Experimental Vibration Analysis for Identification of a Long Span Bridge ................1621 J. Zhang, J. Prader, F. Moon, A.E. Aktan

Numerical analysis of Fire-Damaged Columns Strengthened by Cross-Sectional Enlargement .............................................................................................1635 Lixian Liu, Qiongxian Gao, Long Lv

Experimental Evaluation on Performance of All-Steel BRBs .........................................1644 Quan Chen, Tao Li, Chunlin Wang, Shaoping Meng, Hanbin Ge

Experimental and Analytical Studies on Torsion of BRBs...............................................1651 Tao Li, Quan Chen, Chunlin Wang, Jing Wu, Hanbin Ge

科

学出版社


www.abook.cn

Smart Materials and Structures

The 12th International Symposium on Structural Engineering

NUMERICAL ANALYSIS OF FIRE RESPONSE OF SPATIAL PRE-STRESSED

STEE TRUSS IN FIRE

Xintang Wang, Hongliang Sun, Pingxin Sun, Jinyi Zhang Faculty of Architecture, Civil Engineering & Environment, Ningbo University, Ningbo 315211, P.R. China Abstract: The computational model of the spatial pre-stressed steel truss subjected to fire load is established based on the previous work about analysis of the planar pre-stressed steel structures in fire. The model provides the convenience that the cables with higher tensile strength used in the structure can be set arbitrarily and stretched for more times, but all pre-stress lose of the cables set in the structure are not considered in the computational model. Numerical analysis of fire behavior of a spatial pre-stressed stainless steel truss is analyzed, and the spatial truss has the plan size of 3400×1400. Effect of different factors on bearing capacity and fire behavior of the structure subjected to fire load is discussed based on numerical results. It is shown that fire condition, value of pre-tension, external loads and times of pre-tension have great effects on behavior and bearing capacity of the pre-stressed steel trusses in fire. The conclusions can be used as basis of fire-resistant design of the structures and in-depth analysis of fire behavior of the spatial pre-stressed steel trusses in fire. Keywords: Fire response, spatial stainless steel truss, pre-stressed, numerical analysis 1 INTRODUCTION

Pre-stressed steel structures with better mechanical behavior is a kind of special structures which has been widely used in practical engineering (WANG, 2004), but its overall performance will change greatly when fire happens because of its speciality. At present, there are only a few papers for researching on fire behavior of the pre-stressed steel structures, and most of them just analyzed some engineering structures using the software (YANG, HAN, et al., 2007). The theoretical analysis of behavior of the string supported structure in fire was finished in paper (Bai, Shi, et al., 2009).

Based on analysis of the spatial pre-stressed steel structures under room temperature in the earlier work (WANG, 2004), the computation model of the spatial pre-stressed steel trusses stretched for more times subjected to fire load is established according to the response mechanism of the structure in fire. The fire response and thermal expansion effect of the structure resulting from the change of material properties are considered in the model. The EC3 constitutive model is simplified in order to make the existing material constitutive model could be

applied easily (Li, Han, et al., 2007). Effect of different factors on behavior of the

whole structures is discussed by analysis of bearing capacity of the spatial pre-stressed stainless steel trusses in fire. The research work presented here may be used as basis of further in-depth research on fire behavior of the pre-stressed steel structures.

2 FUNDAMENTAL ASSUMPTIONS AND SIMPLIFICATION OF THE CONSTITUTIVE MODEL

The following assumptions are suggested for establishment of computational model of the pre-stressed steel structures as:

(1)Each node of the structure is suggested as ideal hinge.

(2)The tension cables can slide freely at contact point of nodes, which means the friction resistance, eccentricity and loss of prestress are neglected.

(3)The temperature of each element of the structure in fire is distributed uniformly.

The folded element is used to simulate the flexible cable to establish the computational

科学出版社


www.abook.cn

·206·

model of the pre-stressed steel structure, which is shown in Fig.1. The node number of the folded cable-element is p , and the intermediate nodes can slide freely according to the actual character of the cable element.

Figure 1. The folded cable element

The strain expression of the cable element has the formula as

0

1s s s

Bs

(1)

Where0

S is initial length of the cable element;

the row matrix s

B reflects the geometrical

characteristics of the folded cable-element, and is related to the coordinates of spatial folded nodes (Wang, 2004); s is nodal displacement matrix of the cable element.

In order to establish the computational model of the cable-prestressed steel structures subjected to fire load, the material constitutive models and the variation rules of material property parameters are needed to be determined first. For convenience of application, EC3 model is simplified with series expansion, and the final expression is obtained as follows:

TTTEEE~

)(0

(2)

1 2

/

(1 ) /

/( )

T pT

T T pT yT

T yT yT tT

tTyT

uT tT

yT uT tT tT uT

E

E E

E f

f

f

≤

≤ ≤ (3)

Where 2

2

2 2

1 22 2

2 2

( )2 ( )

( ) [1 ( ) ]

yT yT pT pT

T yT pT

yT pT yT pT

b

a cE f fb

a c

2

1

2 T

T

TE

EE

2

1 2( )]T pT T pT yT T pT PT

bE f

a c

TTT ca

bE 2

2

2

])(1[)(

2

2

22

2

22

2

pTyTpTyT

pTyT

T

ca

bff

b

ca

22

2 2

1[1 ( ) ]

2T yT pT

b

a c

)/)((TpTyTpTyT

Eca

2)( cEcbTpTyT

)(2)(

)( 2

pTyTTpTyT

pTyT

ffE

ffc

In which the related parameters take the values as

yT = 0.02,

tT = 0.02,

uT =0.2 according

to the EC3 model (Li, Han, 2007). The parameters

pTyTTffE ,, are initial elastic

modulus, yield strength and proportional limit of steel at temperature T, which can be measured by experiment.

tT is the maximum strain at

the temperature T. The quadratic function is used to substitute the original nonlinear section of EC3 model.

3 ESTABLISHMENT OF ANALYSIS MODEL OF WHOLE STRUCTURE

For analysis of the cable-prestressed spatial steel truss subjected to fire load, the initial temperature field and displacement field at room

temperature are denoted by 0

~T and 0 ,

respectively, and the increment of temperature of

the structure is expressed as 1

~T ,

2

~T , …,

nT~

. The temperature in state i is denoted as

i

k

kiTTT

1

0

~~~ , and the vectors of nodal

displacements and internal forces of the structure are i and iN , respectively.

The structural response equation for computation of 1

1

i of the pre-stressed steel

truss subjected to fire load is obtained by

·207·

principle of virtual work, which is shown as

1

0 1 , 1[ ] [ ]

ii i

eT s kT T ie k

K K R

(4)

Where the matrix i

eTK ][

0is the expansion

form of stiffness matrix of truss element e. The

matrix ikTsK has the expression as

T[ ] [ ] [ ]i

i sT sks kT sk sk

sk

E AK B B

S (5)

skB][ is strain matrix of folded cable-element,

and the general expression is given in reference

(Wang, 2004). i

sTE is the initial elastic modulus

of the steel cable at temperaturei

T~

, and the

following expression is used in accordance with the current results(Zhou, Li, et al., 2008).

skB][ is strain matrix of folded cable-element,

and the general expression is given in reference

(Wang, 2004). i

sTE is the initial elastic modulus

of the steel cable at temperaturei

T~

, and the

following expression is used in accordance with the current results( Zhou, Li, et al., 2008).

0sT T gsE E (6)

Where T stands for the reduction factor of

the corresponding materials of the structure,

0gsE is elastic modulus of steel cable at room

temperature. The following formula is used for the yield

strength of steel cable at high temperature.

0yT T y (7)

Where 0y and yT are yield strength at

ambient temperature and high temperature T, respectively. The expression of load vector 1, iTR which results in variation of

temperature is

1 2

, 1 , 1 , 1T i T i T iR R R

(8)

In which 1

1, iTR and 2

1, iTR are caused

by changes in temperature of truss element and cable element, respectively. The nodal load of truss element e at displacemental state i caused from the temperature increment

1, ie

T

is shown as

, 1

, 1

, 1

mT i

eT i

jT i

PP

P

(9)

Where

T

, 1 , 1mT i e emT iP T q

T

, 1jT e ejT iP T q

, 1

, 1 0

i i

T e T e i

emT i

E A Tq

, 1

, 1 0

i i

T e T e i

ejT i

E A Tq

Where the expression of transformation matrix

eT can be found in the paper (Wang,

2003). 2

1, iTR is caused by the temperature changes of cable element, and has the expression as follows:

2

, 1, 11

q

kT iT i kk

R p S

(10)

Where

, 1, 1

i i

sk sT sT sk ikT ip A E T

It is noted that i

sTE and i

sT are the value of

the elastic modulussT

E of steel cable element and

the thermal expansion coefficient sT

at

temperature iT , respectively. The expression of

sTE and

sT can be determined with

experiment. For the cable-prestressed steel structure at the

displacemental state i , the increment of displacement response of the whole structure

resulting from the temperature change of 1

~

iT

can be obtained by the governing equation as

11 1

2[ ] [ ]

ii i

T T iK K (11)

In which 1 1

0[ ] [ ] [ ] [ ]i i i i

T T s T TK K K K 1 1 1

0[ ] [ ] [ ]i i i

T T s TK K K

Where the matrix i

TK ][

0is the global stiffness

matrix of truss element and i

TsK ][ is cable

element’s at the temperature i

T~

. It should be especially noted that the relationship presented below holds true, that is

ii

i 21 (12) The relation shows that the Eq(4) and (11) are

correlated by Eq (12), although they are two equations for solving

1 and 2 ,

respectively. The overall bearing capacity of the planar

pre-stressed tubular trusses in fire will be analyzed by the modal established here. With

科学出版社


www.abook.cn

·208·

using the model presented above, the failure elements need to be determined, and failure of the compression element is judged according to the equation below

R T

c

Nf

A

≤ (13)

Where R

is resistant coefficient of steel

material and the approximate value is 1.1; T

is strength reduction factor of steel material, that is /T yT yf f ; c

is ratio of the stability

coefficient of axially loaded compression members in high temperature to that in room temperature, the value refers to paper (Li, 2007). All the parameters such as 、

R and

c for

tension member are taken as 1. All the other parameters take the values of corresponding material in the moment temperature.

4 NUMERICAL ANALYSIS OF A STAINLESS TRUSS

A simply supported stainless spatial pre-stressed steel truss is shown in Fig.2. It is noted that two cables which has the serial number of E-195 are set in the structure. The initial elastic modulus of stainless steel of the structure is 2.0×105MPa, the initial yielding strength is 205MPa; the elastic modulus of cable is 2.05×105MPa, and the tensile strength is 1570MPa.

Figure 2. The computational model of numerical analysis of the stainless steel spatial truss

Also it is noted that all the elements of the truss are stainless steel tubes, and the geometrical

dimension of them take as 16×2mm for top chords, 16×1mm for bottom chords and obligue belly bars, and 6×1mm for vertical braces, respectively. The diameter of steel cables is 4mm. The poisson’s ratio of all the materials is 0.3.

The vertical concentrated load which is denoted as P is applied to the upper nodes of the truss presented in Fig.2. For a comparative study, both one-time pretension and two-times pretension are considered for application of prestress here. For the case of one-time pretension, the load P is 400N, the pre-stress produced in the steel cable is 180MPa, and the intial displacement of the central node 21 is 6.25mm. for the case of two-times pretension, the whole load P is 550N, the pre-stress in cable is 515MPa, and the intial displacement of the central node 21 is 6.13mm and stress of E-195 is 661.57MPa.

It is noted that variation of the mechanical properties of all the materials of the structure shown in Fig.2 under fire takes from the values given in the paper (Li, 2007).

For the prestressed steel structure presented here, two fire conditions are considered. One is that the fire temperature goes up to 300℃ during 15min, and then keeps up slowly; the other is that the fire temperature goes up to 300℃ during 18min. According to analysis, the response temperature kakes the same values with the fire temperature, and the temperature of all the bottom chords is the same and 50℃ higher than that of top chords, and the belly bars have the same temperature with the bottom chords.

The response displacements of typical nodes and the response stresses of typical elements including the cable element are shown in Fig.3 for one-time pretension under the first fire condition.

(a) Displacement response

Figure 3. Response for one-time pretension under fire condition one

·209·

(b) Stress response

Figure 3. Response for one-time pretension under fire condition one (continued)

It is seen from the results shown in Fig.3 and Fig.4 that the structural response of the pre-stressed stainless steel truss are greatly affected by the way of pretension. The results show that variation of stress response of the steel cable for one-time pretension is larger than that for two-times pretension for fire condition one. It is denoted that the ultimate response displacements for the two way of pretension have the same value of 14mm for the fire condition presented above.


(b) Stress response

Figure 4. Response for two-times pretension under fire condition one

The results for fire condition two are shown in Fig.5 and Fig.6.


(b) Stress response

Figure 5. Response for one-time pretension under fire condition two


(b) Stress Response

Figure 6. Response for two-times pretension under fire condition two

科学出版社


www.abook.cn

·210·

It is seen from comparation of the results shown in Fig.3~Fig.4 and the results shown in Fig.5 ~Fig.6 that there are obvious difference for the structural response of the prestressed steel truss shown in Fig.2 for the two fire conditions presented here. The results show that the heating rate has great effect on the response displacements and the time to failure of the pre-stressed stainless steel truss.

5 CONCLUSIONS

It is concluded that variation of cable stress of the structure in fire for one-time pretension is larger than that for two-times pretension for fire condition one, for which the fire temperature goes up to the maximum value during 15min. For the fire condition one, there is obvious difference between the times to failure for the way of pretension, and the time to failure when the structure fail to bear the external loads is 18.5min for one-time pretension and 9.6min for two-times pretension. Further more, the maximum response temperature of the structure is 339℃ for one-time pretension and 240℃ for two-times pretension for the same fire condition. However, for the fire condition two that the fire temperature goes up to 300 during 18min, the time to failure presented above is 21.7min for one-time pretension and 12.7min for two-times pretension, and the maximum temperature is 253℃ for two-times pretension. The results show that the heating scheme has great effect on performance of the pre-stressed spatial stainless steel truss in fire.

ACKNOWLEDGEMENTS

The work reported here was supported by National Natural Science Foundation of China (No. 51078187).

REFERENCES

Bai, Y., Shi, Y.J., and Wang, Y.Q.(2009). Theoretical Analysis and Numerical Simulation on Behaviour Properties of Large Span Cable-supported Structures under Fire Conditions. Science in China Series E: Technological Sciences, 52:8, 2340-2349.

Li, G.Q., Han, L.H., Lou, G.B., and Jiang, S.C.(2007). Fire Resistance Design of Steel Structure and Composite Steel-Concrete Structures. Beijing: China Architecture & Building Press.

Wang, X.T. (2003). Computational Structural Mechanics and Programming. Beijing: Science Press. (in Chinese)

Wang, X.T. (2004). Standardized Optimization of the Cable-Prestressed Steel Trusses. Spatial Structures, 10 :1,16-20. (in Chinese)

Yang, Z., Han, Q.H., Jin, H., and Wang, L. (2007). Numerical Analysis of the Fire-Resistance of Prestressed Composite Space Truss System. Journal of Tianjin University, 40: 8, 1007-1012. (in Chinese)

Zhou, H.T., Li, G,Q., and Jiang, S.C.(2008). Experimental Studies on the Properties of Steel Strand at Elevated Temperatures. Journal of Sichuan University (Engineering Science Edition), 40:5, 106-110. (in Chinese)


PREDICTION OF LONG-TERM STRENGTH OF CONCRETE BASED ON

ARTIFICIAL NEURAL NETWORK

Xiaoming Yang, Dan Shi College of Civil Engineering and Architecture, Liaoning Technical University, Fuxin 123000, P.R. China

Abstract: Recently, the safety of existing civil engineering structures attracts more and more attention. The long-term strength of concrete plays a key role during the assessment of safety and durability for civil engineering structures. The strength of concrete will gradually decrease during the service of civil engineering structures. It is significant to accurately predict the strength deterioration of concrete for correctly evaluating the safety of structures. The factors affecting the long-term strength of concrete include environment type, age, climate, water cement ratio, amount of cementing material and so on. In this paper, artificial neural network with powerful mapping ability has been selected to predict the long-term strength of concrete. First, there-layer BP neural network with age, water cement ratio, amount of cementing material as input and long-term strength as output was built. Then, the neural network was trained by the samples measured in real structures and the well-trained neural network was test. From the test results, the trained neural network can accurately predict the long-term strength of concrete with the error less then 9%. Keywords: Concrete, long-term strength, artificial neural network, prediction

1 INTRODUCTION

As the development of economy, the safety of civil engineering structure has received more and more attention. During long-term service life of several civil engineering structures, they are inevitably affected by environmental erosion, materials deterioration, load and fatigue effect so that different extents of damages happened in these structures and the load carrying capacity of components in these structures decreased, as shown in the work by Ou, (2005). When extreme condition happens, the structure could collapse or get failure, which would make great loss of life and property. For ensuring the safety of civil engineering structure, the condition assessment periodically is needed. The condition assessment is that the real values of structural parameters are measured and calculated for making sure weather the structure is safe. The key point in condition assessment is how to measure the real strength of material used in the existing structure.

Reinforced concrete is most widely used in civil engineering, as shown in the work by Pan, Bian, and Yang (2011) and Cheng, and Liu

(2003). But, it is easy for the concrete to be eroded by environmental factor during its long service life. A lot of diseases, including concrete carbonation, chemical corrosion, freeze-thaw destruction, alkali-aggregate reaction and so on, will lead to decrease of concrete strength. Also, strength of material is the key parameter for calculating the load carrying capacity of structure and one of signs for ensuring the structural failure. Thus, to accurately obtain the real strength of concrete in existing structures, long-term strength of concrete, is of significance for correctly deciding the condition of structure, ensuring its safety and decreasing the cost of maintenance.

At present, the main methods for measuring the concrete strength in field are rebound method, ultrasonic method, pull-out method and core drilling method. Thought these methods have been widely used, the great cost of time and engineer is inevitable when the measured structure is large and there are many measure points in the structure. Besides, there are many locations in large-scale structure, such as the location in box girder, beam-column joints and concrete filled steel tube, can reached using the normal method while the strength of concrete in these location is more important for condition

科学出版社


www.abook.cn

·212·

assessment. So, a method which can conveniently, fast and accurately obtain the long-term strength of concrete in any location of the assessed structure is urgently needed.

Artificial neural network (ANN) is the mathematic tool widely used in various field in recent years. It physically simulates the information procession system of human brain so that it has not only normal ability of data procession but also the thinking, study and memory ability for knowledge procession. So the technology of ANN can be used to predict the long-term strength of concrete. The network is trained by the real measurement of concrete long-term strength and then the relationship between the influencing factors and long-term strength is learned. It is convenient and less labor and time to predict the long-term strength of concrete by the trained ANN.

2 ARTIFICIAL NEURAL NETWORK

The artificial neural network is one information procession system which is built by simulating and simplifying the ability, architecture and some basic properties of true brain. From the viewpoint of system, ANN is the nonlinear and dynamic system consisting of a lot of neurons with massive and perfect combination. ANN has many advantages such as robustness, parallel structure, self-learning, self-organization, adaptive, nonlinear and so on.

At present, in the practical applications of ANN, BP network (using BP training algorithm for ANN) plays an important role as shown in previous work by the authors (Lautour and Omenzetter, 2009), as shown in Fig.1. BP algorithm

Figure 1. Architecture of BP neural network

is a training algorithm of error back propagation. It is a kind of training for multilayer neural network, based on the gradient descent method under the guidance of teacher. The main idea of error back propagation algorithm is divided the

training process into two stages. In the first stage (positive process), input information from input layer to hidden layer processes and is used to calculate the actual output value of each element layer by layer. And in the second stage (reverse process), if it fails to obtain the expected output value in the output layer, it will calculate the difference value between actual and expected output value (error) to adjust the weighted value.

BP algorithm works according to the following stages.

(1) The calculation formula of the output

node’s output lO

1) The input of input nodes： jx 2) The output of hidden nodes：

i ij j ij

y f x (1a)

In this formula, the connected weighted value is ij , and the node threshold is i

3) The output of the output node：

l ij i li

O f T y (1b)

In this formula, the connected weighted value is ijT , and the node threshold is l

(2) The modified formula of output layer (between hidden nodes and output node)

1) The desired output of the output node: it 2) Error control The error of all samples is E:

1

P

kk

E E

(2)

In this formula, kE is one sample’s error:

2

1

1

2

nk k

k l ll

E t O

(3)

In this formula, P is the number of samples, n is the number of output nodes.

3) Error formula： 1l l l l lt O O O (4a)

4) Modified weighted value：

1li li l iT k T k y (4b)

In this formula, k is the number of iterations. 5) Modified threshold：

1l l lk k (4c)

(3) The modified formula of the nodes in hidden layer (between input nodes and the hidden nodes)

1) Error formula：

1i i i l lil

y y T (5a)

·213·

2) Modified weighted value：

1ij ij i jk k x (5b)

3) Modified threshold：

1i i ik k (5c)

Although the application of BP algorithm is very widely, it has some questions, such as falling into local minima easily, converging slowly and so on. Thus, many improved algorithms are applied, including the self-adaptive and variable step-size algorithm and -M algorithm.

3 INFLUENT FACTORS ON LONG-TERM STRENGTH OF CONCRETE

The influent factors on long-term strength of concrete need to be selected as input of ANN for its usage. These influent factors include environment type, age of concrete, water cement ratio, climate condition, amount of cementing material and so on. The environment obviously affects the long-term strength of concrete. When concrete service in harsh environment (for example coast), many types of disease will happen and lead to rapid decrease of concrete strength. Nevertheless, concrete indoor will keep its long-term strength quite well. In theory, the strength of concrete will increases with time and the speed of increase is fast at beginning, then becomes slow and stops at last. In fact the strength of concrete will decrease after it reaches maximum. The water cement ratio is the key parameter for calculating mixture ratio and is also the key factor for concrete strength. The water cement ratio obviously affects the long-term strength. The climate condition is used to consider of time of freeze-thaw cycles which badly deteriorates the concrete strength. The time of freeze-thaw cycles can be calculated according to positive and negative of temperature in one year. The more the time of freeze-thaw cycles is, the steeper the long-term strength of concrete decreases. As far as the amount of cement material, it helps increase the strength of concrete. The above influence factors have different extent effect on the long-term strength of concrete and the effect can’t be described using exact mathematic relationship.

ANN with nonlinear mapping ability is suitable to study the relationship between the long-term strength and these influence factors.

4 CASE STUDY

4.1 Architecture of ANN

In order to verify the feasibility of the prediction of concrete long-term strength by ANN, the existing measured data of the concrete strength are used to train and test the ANN. Three-layer BP neural network with one input layer, one output layer and one hidden layer was built firstly. In input layer there are three nodes separately representing water cement ratio, age and compound replacement ratio. There is one node in output layer, representing the strength of concrete. The number of nodes in hidden layer is directly related to the accuracy of the prediction of the neural network, and now there is still no accurate and effective method to make sure the nodes’ number in hidden layer directly. The common method is that the range of nodes’ number in hidden layer is decided depending on experience and then the final number of nodes in hidden layer is confirmed by trial training.

4.2 Training of ANN

Before prediction, ANN should be trained by measured strength of concrete, which help ANN gain the exact mapping relationship between the long-term strength of concrete and each influencing factors. Generally, training samples of neural network come from the measured date of the concrete under different conditions. In this paper, 50 groups of training samples and 10 groups of testing samples are selected from previous work by the authors (Nie, Wang and Li, 2008; Okan and Cengiz, 2008; Su, 1996; Zhao and Liu, 2005). Table 1 only shows ten groups of train samples due to the limitation of the paper length. L-M algorithm is selected to training the neural network and the target error is 0.05. After 15 iterations, the training of the ANN is finished.

科

学出版社


www.abook.cn

·214·

Table 1. Part of training samples of neural network

Sample

Water

cement

ratio

Compound

replacement

ratio

Age(day)

Measured

value of

strength (MPa)

1 0.80 0.40 365 72.0

2 0.90 0.44 28 55.9

3 0.50 0 60 52.8

4 0.50 0.30 90 77.6

5 0.64 0.22 28 55.7

6 0.50 0.30 365 93.4

7 0.96 0.50 180 65.5

8 0.48 0 365 72.5

9 0.54 0.17 28 41.5

10 0.64 0.22 90 56.3

4.3 Test of ANN

After training of the ANN, it is tested by the selected testing samples. Table 2 shows the 10 testing samples and testing results. From table 2 it has been found that all test results for 10 samples are correct except for sample 5 which has large error. For the other samples, the relative errors of prediction are less than 9%, which can satisfy the requirement of real civil engineering. Thus, it is feasible to predict the long-term strength of concrete by artificial neural network.

Table 2. Testing samples and testing result

Samp

le

Water

cement

ratio

Compound

replacement

ratio

Age

(day)

Measured

value of

strength

(MPa)

Predicted

value

(MPa)

Relative

error

1 0.32 0.09 60 76.0 78.21 2.90%

2 0.5 0 120 61.8 61.31 0.79%

3 0.64 0.22 28 46.6 47.69 2.34%

4 0.31 0.12 60 80.2 77.61 3.23%

5 0.47 0.14 28 42.1 56.55 34.32%

6 0.64 0.22 60 53.1 55.90 5.27%

7 0.28 0.1 28 65.2 66.50 1.99%

8 0.41 0.15 365 94.5 93.91 0.62%

9 0.35 0.1 60 71.2 77.10 8.28%

10 0.32 0.13 28 69.4 65.27 5.95%

5 CONCLUSIONS

The long-term strength of concrete is one of the most important parameters during the condition assessment for civil engineering structures. It is significant to accurately and rapidly predict the long-term strength of concrete for correctly evaluating the safety of structures, ensuring the safe use of structures, and decreasing the maintenance costs of structures. In this paper, the method of artificial neural network, which has powerful nonlinear mapping ability, is selected to predict the long-term strength of concrete by training and testing the neural network using the samples from the measured value of long-term strength. Testing results show that trained neural network can predict the long-term strength of concrete effectively, and it is feasible to assess the condition of reinforced concrete structure using the developed method.

ACKNOWLEDGEMENTS

The financial support from National Natural Science Foundation of China under grant 51008148 is greatly acknowledged.

REFERENCES

Cheng, Y.H., and Liu, B. (2003). Present and Tendency of the Research on Durability of Reinforced Concrete Structure. Journal of Northeastern University, 24:6, 600-605.

Lautour, de OR, and Omenzetter, P. (2009). Prediction of seismic-induced structural damage using artificial neural networks. Engineering Structures, 31, 600-606.

Nie, F.Y., Wang, T.Z., and Li, L.X. (2008). Study and Application on Evaluation for Later Age Strength of Concrete. Ash Comprehensive Utilization, :1, 40-42.

Okan, K., Harun, T., and Cengiz, D.A. (2008). An Artificial Neural Network Approach for Prediction of Long-term Strength Properties of Steel Fiber Reinforced Concrete Containing Fly Ash. Journal of Zhejiang University Science A. 9:11, 1514-1523.

Ou, J.P. (2005). Research and Practice of Smart Sensor Networks and Health Monitoring for Civil Infrastructures in Mainland

·215·

China. Science Foundation in China, 19:1, 8-12.

Pan, H.k., Bian, Y.D., and Yang, L.H. (2011). Reliability Analysis of Reinforced Concrete Structure Based on Durability and Deterioration Grade. Journal of Building Structures, 32:1, 105-109.

Su, Z. (1996). Confecting and Application of C60 Pumping Concrete Containing Fly Ash. Concrete, 2, 31-36.

Zhao, S.L., and Liu, Y. (2005). Performance Prediction of Commercial Concrete Based on RBF Neural Network. Computer Engineering, 31:18, 36-37.

科

学出版社


www.abook.cn


EFFECT OF PVA FIBER ON THE FREEZING AND THAWING BEHAVIOR OF

CONCRETE

Fei Xu1, Ju Chen1, Weiliang Jin1, Jihua Zhu2 1 Institute of Structural Engineering, Zhejiang University, Hangzhou 310058, P.R. China 2 College of Civil Engineering, Shenzhen University, Shenzhen, 518061, R.P. China

Abstract: In this study, the effect of polyvinyl alcohol (PVA) fiber on the concrete damage under freeze-thaw condition was investigated. Experimental investigations were conduct to measure the workability, compressive strength, bending strength, total charge passed and frost resistance of PVA fiber concrete in fresh water and NaCl solution freeze-thaw environment. For frost resistance, the relative dynamic modulus of elasticity and the relative mass change of specimens at each cycle were measured. In addition, the surface deterioration of concrete specimens was also observed. Specimens having different volume content of fiber were tested in order to find the optimization PVA fiber content under freeze-thaw condition. Six series of specimens with PVA fiber volume content of 0%, 0.1%, 0.25%, 0.5%, 0.75% and 1.0% were tested. Keywords: Concrete, durability, fiber, freezing and thawing

1 INTRODUCTION

Frost damage to concrete has long been the thorny issue maintaining in cold region especially that having freeze-thaw temperature. With wide application of fiber reinforced concrete, the frost damage of fiber concrete has drawn more and more attention. Significant research efforts have gone to improve and certify durability and ductility behavior of fiber reinforced concrete. Previous research shows the desirable performance of fiber reinforced concrete on tensile stress (Li et al. 2001) and strains (Lin and Li 1997) behavior as well as fatigue resistance (Li and Zhang 2002; Kenser et al. 2003). It is known that incorporation of polypropylene fiber in concrete gain satisfying durability when concrete exposing to freeze-thaw cycles (Morgan et al. 1992). Studies (Sahmaran and Li 2007; Sahmaran et al. 2009) on the behavior of polyvinyl alcohol(PVA)-ECC subjected to freeze-thaw cycles also show excellent frost resistance. However, research also reported that polypropylene fiber does not significantly change the frost resistance of concrete (Karahan and Atis 2011). Contradiction indicates the effect of fiber incorporated into

concrete on frost damage is to be explored. PVA fiber as a new kind of fiber has desirable

characteristics of high elasticity modulus, high tensile strength, chemical stability, which has been successfully applied into the cementitious composite as a kind of reinforcement. However, large quantities of tests including durability studies have focused on the properties of ECC, but few research works so far has been reported on the frost resistance of PVA fiber concrete. Therefore, the influence of different fiber contents and different freeze-thaw environments on the frost damage resistance of PVA fiber concrete is investigated in this study. This is achieved by measuring the workability, compressive strength, bending strength, total charge passed (ASTM C1202-94, 1994) and frost resistance of PVA reinforced concrete in fresh water and NaCl solution freeze-thaw environment. In addition, fiber concrete costs more from using of fibers which is more expensive per weight unit when compared with plain concrete. Therefore these requires to find optimization PVA fiber content in concrete to achieve excellent properties as well as achieving minimized cost/benefits ratio. This study also attempts to find optimization PVA fiber content in concrete for frost resistance.

·217·

2 EXPERIMENTAL PROCEDURE

2.1 Materials

The materials used in the production of concrete mixture were a Chinese standard (GB175-92) 425# Portland cement (which has standard compressive strength of 42.5 MPa at the age of 28 days), natural river sand with fineness modulus of 2.36, and crushed stone with maximum and minimum grain size of 16 and 5 mm respectively, water, RECS15 PVA fiber, and a naphthalene-based high efficiency water reducer admixture (FDN). The physical and mechanical properties of the Poly Vinyl Alcohol (PVA) fiber used are list in Table 1.

2.2 Mixture Composition and Preparation

Standard concrete mixture (C40) with the water/cement ratio of 0.43 by mass was used in this investigation. In order to maintain good workability of fresh mixture, special fiber content matrices with adjusted FDN dosages are used. Since the FDN is a non-air- entrainment admixture, dosage is considered to have minor effect on the frost durability of each mixture. Mixture ingredients are presented in Table 2. The test specimens are labeled such that the fiber content could be identified from the label. For example, the labels “PCF-0” and “PCF-0.1” define the specimens with 0% and 0.1 % volume content PVA fiber respectively. In total six series of specimens having different fiber volume contents, namely PCF-0, PCF-0.1, PCF-0.25, PCF-0.5, PCF-0.75 and PCF-1 with the corresponding fiber volume contents of 0%, 0.1%, 0.25%, 0.5%, 0.75% and 1.0%, were tested as shown in Table 2.

The mixing procedure of plain and fiber reinforced composition is different attempting to promoting workability of fresh mixture. Cement, fine and coarse aggregate were dry mixed for 45s in a concrete truck mixer. For fiber concrete,

the dry mixing was prolonged to 90s. After slow addition of water and FDN, mixing was continued for a period of 120s (180s for fiber concrete). The additional mixing activated the polycarboxylate molecules and improved their orientation which resulted in a reduced initial workability loss (Felekoglu et al. 2009). Last, the freshly mixed fiber concrete was fed into specimen moulds and vibrated on a concrete platform vibrator. The specimens were demolded 24 hours after casting and cured in a lime saturated water tank for 28 days at room temperature (23±2℃) with relative humidity of 100%.

2.3 Specimens and Testing Programs

Concrete prisms with size of 100 × 100 ×400mm were prepared for freeze-thaw cycling test in accordance with ASTM C666 Procedure A (2008). 28 days after casting, the specimens were moved into freeze-thaw chamber. Each series of specimens were subjected to two different freeze-thaw environments, water and 5 % (by mass) sodium chloride solution (NaCl) respectively. Three prisms were tested repeatedly and the average value was used for each test data presented. The temperature of concrete specimens was controlled by a sensor embedded in the center of a standard concrete specimen, ranged from -17.0 ± 2 to 8.0 ± 2℃, and one cycle approximately lasted 3 hours. The deterioration of specimens during freeze-thaw cycles was assessed by the relative mass change (RMC) and the relative dynamic modulus of elasticity (RDME). The RMC is the ratio of the mass change (current value minus initial value) after certain cycles to the initial value. The RDME, namely the ratio of the DME value after certain cycles to the initial value, was measured according to ASTM C597 (2009). According to the criteria of test procedure, the specimen fails if its RDME drops to 60% or less or its loss of mass exceeds 5.0%, and the failure cycle numbers of specimens were also recorded.

Table 1. Physical and mechanical properties of RECS15 PVA fiber

Configurations Color Specific Gravity

Length Tensile Strength

(MPa) Chemical Stability

Absorption

Resin-bundled chopped White 1.3 8 mm 1600 Stable Minimal

Table 2. Ingredients and characters of concrete

Mixture ingredients PCF-0 PCF-0.1 PCF-0.25 PCF-0.5 PCF-0.75 PCF-1 Cement ( kg/m3 ) 512 512 512 512 512 512 Water ( kg/m3 ) 220 220 220 220 220 220 Sand ( kg/m3 ) 567 567 567 567 567 567

Aggregate ( kg/m3 ) 1054 1054 1054 1054 1054 1054 FND ( % by mass of cement) 0 0.1 0.25 0.5 0.75 1.0

PVA fiber ( kg/m3 ) 0 1.3 3.25 6.5 9.75 13 Air content ( vol. % ) 0.395 0.665 0.79 0.81 0.9 0.92

Compressive strength at 28 days ( MPa ) 46.06 50.74 51.74 56.13 49.75 46.40 科

学出版社


www.abook.cn

·218·

3 RESULTS AND DISCUSSION

3.1 Effect of PVA Fiber on the Frost Resistance of Concrete

The RDEM of concrete with different fiber volume content in fresh water and 5% NaCl solution at different freeze-thaw cycles are presented in Fig.1. The ultimate freeze-thaw cycles of concrete specimen series PCF-0, PCF-0.1, PCF-0.25, PCF-0.5, PCF-0.75 and PCF-1 in fresh water were 75, 100, 250, 150, 125 and 125 cycles, respectively. The ultimate freeze-thaw cycles of concrete specimen series PCF-0, PCF-0.1, PCF-0.25, PCF-0.5, PCF-0.75 and PCF-1 in NaCl solution were 75, 100, 225, 150, 125 and 150 cycles, respectively. Plain concrete, PFC-0 series, failed most quickly after 75 freeze-thaw cycles both in fresh water and in NaCl solution, with RDME dropping blew 60%. It is shown that PVA fiber is able to improve the frost resistance of concrete. However, the frost resistance of concrete does not always increase as the fiber volume content increases. Specimen series PCF-0.25 have best frost resistance in both fresh water and NaCl solution.

The RMC results are presented in Fig.2. It is shown that the mass of specimens increase when freeze-thaw cycle increases in fresh water, which might due to the crack propagation and water absorption, as shown in Fig.2(a). On the other hand, all specimens kept losing weight in NaCl solution with the increasing freeze-thaw cycle except for series PCF-1, as shown in Fig.2(b).

Surface deterioration of specimens is shown in Fig.3 and Fig.4. PVA fiber is able to greatly enhance the surface deterioration resistance of the specimens, and it is shown that the larger fiber content, the better surface scaling resistance. This may contribute to the increment of air content with the increasing fiber volume content as shown in Table 2, since air entrainment has significant effect on scaling resistance of concrete (Pigeon et al. 1996). In addition, these randomly distributed fibers in concrete may also prevent mortar layer from splitting away. These two reasons may explain why fiber concrete has much better surface scaling resistance than plain concrete.

Fig.1 and Fig.2 indicate the frost resistance of concrete does not always increase as the increment of fiber content. This may be explained that incorporation of PVA fiber has both positive and negative effects on the concrete

Figure 1. RDME of concrete subjected to freeze-thaw cycles

Figure 2. RMC of concrete subjected to freeze-thaw cycles

·219·

subject to freeze-thaw cycles. The positive effects are that fiber in concrete can help defer crack propagation (Sahamaran et al., 2009) which will effectively enhance the frost resistance of concrete. Moreover, it is widely approved that large quantity of voids inside the concrete can considerably defer two basic frost durability problems: internal cracking and surface scaling (Pigeon et al., 1996). These are proved by test results of improved compressive and bending strength, and air content with increased fiber content compared with plain concrete. However the negative effects should not be ignored. It should be noted that the interface between PVA fiber and concrete matrix is also an internal defect site, which has negative effect on frost resistance.

3.2 Effect of NaCl Solution on the Deterioration of Concrete Exposed to Freeze-thaw Cycles

That existence of NaCl in freeze-thaw cycles will cause severely scaling on plain concrete has been widely approved. It is shown that surfaces of fiber concrete were also severely scaled in NaCl solution compared with those in fresh water, seen in Fig.3-ig.5 provides a comparison of RDME and RMC of each series of specimens in fresh water and NaCl solution. The mass of specimens in NaCl solution were reduced while the mass of specimens in fresh water were added under freeze-thaw cycles. Comparison indicates that the RDME deteriorations of specimens in fresh water and NaCl solution are almost the same, which means that NaCl has little effect on the RDME deterioration. In addition, test results indicate that the maximum freeze-thaw cycles for specimens in NaCl solution and fresh water are similar.

Figure 3. Surface scaling of specimens in fresh water

Figure 4. Surface scaling of specimens in NaCl solution

科学出版社


www.abook.cn

·220·

Figure 4. Surface scaling of specimens in NaCl solution(continued)

Figure 5. RDME and RMC of concrete subjected to freeze-thaw cycles

4 CONCLUSIONS

(1) Compared with plain concrete, the compressive strength, bending strength, air content and resistance of Cl- iron permeability of PVA fiber concrete are increased. However, the workability of concrete is decreased.

·221·

(2) PVA fiber effectively enhances the frost resistance of concrete both in fresh water and NaCl solution. It is shown that specimens having 0.25% volume content PVA fiber has the best frost resistance both in fresh water and NaCl solution. In addition, PVA fiber greatly enhances the surface scaling resistance of concrete in NaCl solution. PVA fiber concrete has almost the same number of freeze-thaw cycles in both fresh water and NaCl solution.

ACKNOWLEDGEMENTS

The authors are grateful to the finical support from the National Natural Science Foundation of China (513000-N10901), Natural Science Foundation of Zhejiang Province of China (Y1100063) and Shenzhen Key Lab on Durability of Civil Engineering (SZDCCE 10-18).

REFERENCES

ASTM C1202-94 (1994). Standard test method for electrical indication of concrete ability to resist chloride ion penetration.

ASTM C597 (2009). Standard test method for pulse velocity though concrete.

ASTM C666 (2008). Standard test method for resistance of concrete to rapid freezing and thawing.

Felekoglu B., Tosun K. and Baradan B. (2009). Effect of Fiber Type and Matrix Structure on the Mechanical Performance of Self-compacting

Micro-concrete Composites, Cement and Concrete Research, 39:11, 1023-1032.

Karahan O., and Atis C.D. (2011). The Durability Properties of Polypropylene Fiber Reinforced Fly Ash Concrete. Materials & Design, 32:2, 1044-1049.

Kenser, K., Billington, S. L. and Douglas, K. S. (2003). Cyclic Response of Highly Ductile Fiber-reinforced Cement-based Composites. ACI Materials J., 100:5, 381-390.

Li, V. C., Wang, S. and Wu, C. (2001). Tensile Strain-hardening Behavior of Polyvinyl Alcohol Engineered Cementitious Composite (PVA-ECC). ACI Materials J., 98:6, 483-492.

Li, V. C. and Zhang, J. (2002). Monotonic and Fatigue Performance in Bending of Fiber-reinforced Engineered Cementitious Composite in Overlay System. Cement and Concrete Research, 32:3, 15-23.

Lin, Z. and Li, V. C. (1997). Crack Bridging in Fiber Reinforces Cementitious Composites with Slip-hardening Interfaces. J. Mechanics and Physics of Solids, 45:5, 63-787.

Morgan, D.R., Mcaskill, N., Curette, G. G. and Malhotra, V. M. (1992). Evaluation of Polypropylene Fiber Reinforced High-volume Fly Ash Shotcrete. ACI Mater J., 89:2, 169-177.

Pigeon, M., Marchand, J. and Pleau, R. (1996). Frost Resistance Concrete. Construction and Building Materials, 10:5, 399-448.

Sahamaran M., Lachemi M. and Li V.C. (2009). Assessing the Durability of Engineered Cementitious Composites under Freezing and Thawing Cycles. Journal of ASTM International, 6:7, paper ID JAI102406.

Sahmaran M. and Li V.C. (2007). De-icing Salt Scaling Resistance of Mechanically Loaded Engineered Cementitious Composites. Cement and Concrete Research, 37:7, 1035-1046.

科

学出版社


www.abook.cn


STUDY ON THE EFFECT OF NATURAL CLIMATE ON THE CORROSION OF

REBAR IN CONCRETE

Jianhua Jiang College of Civil and Transportation Engineering, Hohai University, Nanjing 210098, P. R. China Abstract: Natural climate is an important factor that affects the durability of reinforced concrete (RC) structures, and therefore must be taken into account in the degradation models for durability of RC structures. In order to investigate the development of corrosion rate of the reinforcing rebar in concrete under natural climate environment, the long-term test on the corrosion of the rebar in concrete was carried out under the sheltered condition in natural climate; at the same time, the test on the response of the micro-environment of concrete corresponding to the natural climate environment was also conducted in order to analyze the effect of natural climate on corrosion of rebar in concrete. The results show that, the corrosion rate of the rebar in concrete under natural climate environment is fluctuant and time-dependent; the changes in environmental condition of the micro-environment of concrete lag behind that of natural climate, especially in relative humidity. It is concluded that the corrosion of the rebar in concrete is directly affected by the micro-environment of concrete that depends on the random fluctuations of natural climate, and temperature is the primary factor that influences the corrosion of rebar in concrete under sheltered condition of normal atmospheric environment. Keywords: Natural climate, micro-environment, corrosion rate, temperature, relative humidity 1 INTRODUCTION

Natural climate environment is the real environment of RC structure, and is of randomness, uniqueness and complexity due to the effects of geographical location, meteorological factors, etc. For RC structures located in different regions, their service lives are usually different because of the variation in natural climate environment, even though these structures are identical. Therefore, the researches on durability of RC structures under natural climate environment are of great significance.

Compared with the constant artificial climate environment, the natural climate environment has obvious stochastic fluctuation, which would cause the fluctuations in the micro-environment of concrete. Therefore, the effects of natural climate condition must be considered during the study on the durability of RC structure. Song et al. (2006) presented the equivalent diffusion coefficient model of carbon dioxide in cracked concrete considering the effects of pore water saturation and temperature of concrete. After taking account of the effects of concrete’s temperature , relative humidity, etc on the

chloride diffusion coefficient, a calculation model of chloride penetrating into concrete based on finite difference is established by Shi and Luo (2004). Lopez and Gonzalez (1993) and Enevoldsen et al. (1994) analyzed the influence of pore water saturation, resistivity, and internal relative humidity of concrete on the corrosion rate of rebar and suggested that a critical pore water saturation or relative humidity initiates corrosion of rebar. Song and Liu (2000) established the quantitative relationship between the anode-cathode area ratio of corroded rebar and the pore water saturation in carbonated concrete based on experimental study. Liu and Weyers (1998) developed a prediction model for the time-varying corrosion rate of rebar, which considered the influence of chloride-ion concentration, temperature, and resistance of concrete based on long-term exposure test under natural climate environment.

In order to accurately establish the prediction model for the degradation of RC structure by considering the effects of climatic condition, the regularity and mechanism about the influence of climate environment on the deterioration must be found out first, and the response relationships of temperature and relative humidity between in

·223·

natural climate environment and in the micro-environment of concrete should be further clarified.

The development of corrosion rate of rebar in concrete under sheltered conditions in the natural climate environment was studied through a long-term test, and the mechanism was also analyzed, in this paper. At the same time, the response relationship between the micro- environment of concrete and the natural climate environment was studied through the long-term monitoring of environmental parameters (i.e. temperature and relative humidity) in both concrete and natural climate, and the effects and mechanism of natural climate environment on the corrosion rate of rebar in concrete were further analyzed.

2 EXPERIMENT ON CORROSION RATE OF REBAR IN CONCRETE UNDER NATURAL CLIMATE ENVIRONMENT

2.1 Specimen Design

The concrete grade of specimen is C25, and the quality proportion of cement, sand, gravel and water is 1, 1.6, 2.96 and 0.54 respectively in the mixture. Cement is type 42.5 ordinary Portland cement based on Chinese standards; the fine aggregate is river sand with the fineness modulus of 2.7; the coarse aggregate is gravel with size of 5~15 mm, and its main constituents are limestone and trachyte; the mixing water is ordinary tap water.

The dimension of the specimens is 100 mm×150 mm×250 mm, shown in Figure 1. A hole is preformed using plastic pipe in the specimen when pouring concrete, in order to embed the temperature-humidity probe. The diameter of hole is 20 mm, and the distance from the concrete upper surface to the preformed hole is 25 mm. At the same time, in order to study the effect of the micro-environment of concrete on the corrosion rate of the rebar, a HRB335 steel bar and a stainless steel bar with diameter of 14 mm are arranged in the concrete specimen, and the thickness of concrete cover is 25mm. The chloride salt (sodium chloride) accounted for 5% of cement mass was admixed into the concrete to simulate the steel corrosion induced by chloride ion. The serial number of specimens is NCE-S1 and NCE-S2, respectively.

Figure 1. Specimen of steel corrosion in concrete under natural climate environment (unit: mm)

2.2 Test Method

The specimens were cured for 28 days in the standard curing condition after pouring and demolding, and then the initial relative humidity in concrete was controlled to about 80% corresponding to the room temperature through the proper treatment, in order to make the relative humidity response obvious under natural climate environment. After that, the concrete surfaces except for the concrete cover were sealed with plastic film and paraffin, and the temperature-humidity probe (see Figure 2) was embedded into the preformed hole using plastic pipe and sealant. The test was carried out in the outdoor exposure test site, and the specimens were under sheltered condition in the natural climate environment. The corrosion rate (corrosion current density) of rebar in concrete was periodically measured using the electrochemical test system (see Figure 3) during the test process, and the temperature and relative humidity values were simultaneously recorded via the probe.

Probe Register

Figure 2. Rotronic temperature-humidity sensor

Figure 3. Setup for measuring corrosion rate

科学出版社


www.abook.cn

·224·

3 EXPERIMENTAL RESULTS AND ANALYSIS

The test cycle is more than one year. The measurement parameters include the corrosion current density of the rebar and the corresponding internal temperature and relative humidity in concrete.

3.1 Corrosion Rate of Rebar

According to the fundamental theory of metal corrosion, the corrosion current density can be taken as the representative value of corrosion rate of rebar. Figure 4 shows the measured results of the corrosion current density of the rebar in the concrete specimens under natural climate environment.

Figure 4. The long-term data of corrosion rate of rebar in concrete

It can be seen from the Fig.4 that the corrosion rate of rebar in concrete has obvious time-variation under natural climate environment, and specific as follows: (1) the corrosion rate shows a general trend of time-varying mode caused by the development of rust layer; (2) the corrosion rate shows fluctuation in detail due to the effects of climate environment. In short, the time-varying process of corrosion rate of rebar in concrete under natural climate environment is different from that under constant artificial climate environment, which is mainly because the natural climate environmental condition is not consistent.

3.2 Temperature and Relative Humidity

Figure 5 shows the evolutions of temperature and relative humidity in both natural environment and the micro-environment of concrete in the specimen NCE-S1 under sheltered conditions in the natural climate environment.

(a) Temperature

(b) Relative humidity

Figure 5. Temperature and relative humidity both in concrete’s micro-environment and in natural environment (12-14 means Dec.14 in the abscissa axis)

The test results in Figure 5 show that, (1) the temperature and relative humidity in concrete are both in the state of fluctuation under natural climate environment; (2) the evolution of internal temperature in concrete is basically consistent to that of ambient temperature, but the fluctuation of the former is slightly smaller than that of the latter; (3) the fluctuation of internal relative humidity in concrete is significantly smaller than that of ambient humidity.

In a word, the temperature response is significantly faster than the relative humidity response between concrete and natural environment. It is because the temperature and relative humidity response of concrete is essentially a heat transfer and a mass transfer process respectively, and the heat transfer of concrete is far superior to mass transfer.

4 ANALYSIS ON EFFECTS OF CLIMATE ENVIRONMENT ON CORROSION RATE OF REBAR

According to the time-varying mode of steel corrosion rate in concrete under the constant

·225·

artificial climate environment, the corrosion of rebar is still in the stable period before the corrosion cracking in this experiment (Yuan et al., 2010). However, the experimental results indicate that the development of corrosion rate of rebar does not exhibit the stable period after the decline, but is in a state of rise with fluctuation under the natural climate environment. The reasons are as follows.

4.1 Effect of Temperature in Concrete

Figure 6 shows the evolutions of the temperature and the corrosion rate of the rebar in concrete under the sheltered conditions in the natural climate environment.

(a) NCE-S1

(b) NCE-S2

Figure 6. Corrosion rate of the rebar in concrete and temperature in a function of time (12-14 means Dec. 14 in the abscissa axis)

It can be seen from Figure 6 that the change of the corrosion rate of the rebar is relatively consistent with that of the temperature according to the increase of time. This is because that the corrosion rate of rebar in concrete is not only affected by the development of rust layer, but also affected by the fluctuation of temperature in concrete simultaneously. Therefore, the corrosion rate of rebar fluctuates in all stages and doesn’t exhibit a clearly stable period.

The steel corrosion in concrete is an electrochemical reaction process. According to the basic law of chemical kinetics, the rate equation of chemical reaction is as follows (Logan, 1996).

1 1M Nkc c . (1)

where Mc , Nc are the concentrations of

reactants; 1 , 1 are the orders of reaction;

k is the reaction rate coefficient, and the relationship between k and the temperature can be expressed as the Arrhenius equation, namely Eq.2.

aE

RTk Ae

. (2) where A is the pre-exponential factor or frequency factor; aE is the activation energy; R

is the gas constant; T is the absolute temperature.

aE and A can be regarded as constants within

certain temperature range. Eq.1 and Eq.2 show that the effect of

temperature on the corrosion rate of rebar in concrete can be attributed to the influence of temperature on the coefficient k , and approximately obeys the exponential function.

4. 2 Effect of Pore Water Saturation in Concrete

The fluctuation of relative humidity in concrete’s micro-environment, caused by the natural climate, is really the change in the moisture content or pore water saturation of concrete. In the case of varying temperature, the relative humidity cannot reflect the moisture content of concrete alone. Therefore, the pore water saturation is used to analyze the effect of moisture content of concrete on the corrosion rate of rebar.

The pore water saturation of concrete is calculated on the basis of the temperature and relative humidity in concrete’s micro-environment, and the specific calculation method will be described in another paper. Figure 7 shows the evolutions of the pore water saturation of concrete and the corrosion rate of rebar under the sheltered conditions in the natural climate environment.

It can be seen from Figure 7 that the effect of the pore water saturation of concrete on the corrosion rate of rebar seems to be not clear. In fact, the corrosion rate of rebar should increase when the pore water saturation increases because reducing the resistivity of concrete can strengthen the path of ions between the cathode and anode of steel corrosion. It is admitted that, however, the short supply of oxygen in concrete due to pore water saturation could reduce the corrosion rate of rebar. In the test, the pore water saturation of concrete has not reached the extent of inhibiting the supply of oxygen needed for steel corrosion. In this regard, Figure 7 does not

科学出版社


www.abook.cn

·226·

(a) NCE-S1

(b) NCE-S2

Figure 7. Pore water saturation and steel corrosion rate in concrete (12-14 means Dec. 14 in the abscissa axis)

display the effect of pore water saturation itself, but demonstrates the comprehensive effects of the pore water saturation, temperature and the development of rust layer on the corrosion rate of rebar in concrete. The effect of the pore water saturation is concealed or counteracted because the influences of temperature and rust layer are more prominent.

5 CONCLUSIONS

The following concluding remarks have been drawn from this study:

(1) The corrosion rate of rebar in concrete is of time-variation and significant fluctuation under the natural climate environment, therefore its time-varying process is different from that of corroded steel in concrete under the constant artificial climate environment in some extent due to the effects of concrete’s micro-environment.

(2) The temperature and pore water saturation in concrete’s micro-environment have significant effect on corrosion rate of rebar. The higher temperature in the concrete, the higher corrosion

rate of rebar, and their relationship obeys index function proximately. Under the normal atmospheric environment, the higher pore water saturation of concrete, the bigger corrosion rate of rebar. The concrete’s micro-environmental temperature is the primary factor that affects the corrosion rate of rebar in concrete under the sheltered conditions in the natural climate environment.

ACKNOWLEDGEMENTS

The author wishes to acknowledge the financial support from the Fundamental Research Funds for the Central Universities (No: 2012B02614).

REFERENCES

Enevoldsen, J.N., Hansson, C.M., and Hope, B.B. (1994). The Influence of Internal Relative Humidity on The Rate of Corrosion of Steel Embedded in Concrete and Mortar. Cement and Concrete Research, 24:7, 1373- 1382.

Liu, T., and Weyers, R.W. (1998). Modeling the Dynamic Corrosion Process in Chloride Contaminated Concrete Structures. Cement and Concrete Research, 28:3, 365-379.

Logan, S.R. (1996). Fundamentals of Chemical Kinetics. London: Longman, England.

Lopez, W., and Gonzalez, J.A. (1993). Influence of The Degree of Pore Saturation on The Resistivity of Concrete and The Corrosion Rate of Steel Reinforcement. Cement and Concrete Research, 23:2, 368-376.

Shi, Y.H., and Luo, G. (2004). A Finite Difference Calculating Model of Chloride Penetrating into Concrete under Various Factors. Industrial Construction, 34:5, 58-61.7-10. (in Chinese)

Song, H.W., Kwon, S.J., Byun, K.J., and Park, C.K. (2006). Predicting Carbonation in Early-aged Cracked Concrete. Cement and Concrete Research, 36:5, 979-989.

Song, X.B., and Liu, X.L. (2000). Experiment Research on Corrosion of Reinforcement in Concrete through Cathode-to-Anode Area Ratio. ACI Materials Journal, 97: 2, 148-155.

Yuan, Y.S., Jiang, J.H., and Peng, T. (2010). Corrosion Process of Steel Bar in Concrete in Full Lifetime. ACI Materials Journal, 107:6, 562-567.


THE INFLUENCE OF STEEL REINFORCEMENT ON CREEP AND

SHRINKAGE OF CONCRETE COLUMNS

Tiejun Liu, Leilei Guo, Dujian Zou Shenzhen Graduate School, Harbin Institute of Technology, Shenzhen 518055, P.R. China Abstract: It becomes increasingly important for estimating creep and shrinkage accurately as resultes of the increasing high-rise reinforced concrete buildings. Steel reinforcements have a significant impact on the creep and shrinkage of concrete, which makes the estimation more complex and difficult. In this paper, the restraint effect of steel reinforcements was analyzed and the formulas to calculate time-dependent deformation and stress of axially compressed columns were derived. A case study was done to discuss the time dependent creep and shrinkage of columns according to CEB-FIP1990 code and GL2000 code. The deformation of columns with different reinforcement ratios was calculated according to GL2000 code. It is shown that, due to the existence of steel reinforcements, the 50 years deformation was decreased by about 10%-40%. To evaluate the validity of the theoretical derivation, the creep and shrinkage of RC column specimens with different reinforcement ratios was experimental investigated. The experimental resultes indicated that deformation reduction ratio induced by steel reinforcement was higher than the calculated values. Keywords: Creep, shrinkage, reinforcement ratio, restraint effect 1 INTRODUCTION

The shrinkage and creep are inherent characteristic of concrete material. The gravity load bearing elements of concrete buildigns are subjected to shringkage and creep during the construction process and service life, and the influence of shrinkage and creep on performance of structures becomes increasing significant with the increasing of high-rise buildings. A lot of shrinkage and creep prediction models have been constructed, such as the ACI series, CEB series, B3 and GL2000. These prediction models are generally derived based on large numbers of test datas which mostly foucesed on one-dimensional plain concrete specimens. It should be noted that the current prediction models are not applicable for RC member as steel has an impact on the shrinkage and creep of concrete. A simple method to deal with the influence of reinforcement is transforming the area of steel to certain area of concrete. But this method neglects the influence of stress loss. Actually the stress of concrete decreases along the loading time due to the existence of steel and the stress loss of steel will also influence the development of time-dependent derformation of concrete. This problem can be solved by dividing the calculation time into many

segments. The deformation and stress of every time interval is calculated with the iterative method.

Column is one of most important gravity load bearing elements in a structure. It is significant to evaluate their time-dependent derformation accurate. In this paper, the restraint effect of steel reinforcements is analyzed and the formulas to calculate time-dependent deformation and stress of axial compresses columns were derived. Furthermore, an experiment was carried out to investigate validity of the theoretical derivation.

2 FORMULATION DERIVATIONS

The formulas to calculate time-dependent deformation and stress of axially compressed column are derived as follows.

Firstly, three fundamental assumptions should be satisfied.

1) Plane sections remain plane all the time; 2) No slip occurs at the steel– concrete

interfaces; 3) The creep is linear and the strains of

shrinkage and creep are independent and meet superposition principle.

The working-time is divided into large 科

学出版社


www.abook.cn

·228·

numbers of calculation intervals. Considering an external time-invariant force N acts on the column at time t(0).

The instantaneous strain of concrete is equal to the strain of steel according to plane-section assumption.

, (0) , (0)c t s t (1)

The external load is withstanded by concrete and steel：

, (0) , (0)c t s tN N N (2)

Where Nc,t(0) =the force that concrete withstands at time t(0), Ns,t(0) =the force that steel withstands at time t(0).

The stress of concrete and steel can be deduced from Eq.1 and Eq.2.

, (0)

0

, (0)

1c t

s

c t

N

EA

E

(3)

, (0)

0 , (0)

, (0)

1

ss t

sc t

c t

NE

EA E

E

(4)

Where A0 =cross-sectional area of the column, Es =modulus of steel, Ec,t(0) =modulus of concrete at time t(0), ρ =reinforcement ratio.

The strain of concrete and steel can also be deduced as follows,

, (0) , (0)

0 , (0)

, (0)

1c t s t

sc t

c t

N

EA E

E

(5)

The development of deformation from time t(i) to time t(i+1) is derived below,

If the steel does not exist, the strain of concrete should be,

, (0)

1

, (0)

1,

1 , ( )

,

, ( )

(0), ( 1) (0), ( )

( ), ( 1) ( ), ( )

( ), ( 1) ( ), ( 1)

c t

i

c t

ic j

j c t j

c i

sh

c t i

t t i t t iE

t j t i t j t iE

t i t i t i t iE

Where (t(j),t(i)) =creep coefficient from time

t(j) to time t(i), ,c i =stress decrement of

concrete from time t(j-1) to time t(j), Ec,t(i) =modulus of concrete at time t(j), sh (t(i),t(i+1))

=shrinkage from time t(i) to time t(i+1). Actually as the influence of steel, the

deformation of concrete decreases and the stress also has a decrement. Assuming the actual deformation of concrete is , 1c i . The force

decrement of concrete is shown as follows,

, 1 , ( ) 1 , 1 0( ) (1 )c i c t i i c iN E A (7)

Assuming the actual deformation of steel is

, 1s i . The force increment of steel is,

, 1 , 1 0s i s s iN E A (8)

The deformation of concrete is always consistent with that of steel which can be expressed as follows,

, 1 , 1c i s i (9)

The force decrement of concrete is equal to the force increment of concrete,

, 1 , 1c i s iN N (10)

The strain increment of concrete and steel at time t(i+1) can be deduced from Eq.9 and Eq.10,

, 1 , 1 1

, (0)

1

1c i s i i

s

c t

E

E

(11)

The strain at time t(i+1) can be deduced,

, ( 1) , ( 1) , ( 1) , 1c t i s t c t i c i (12)

The stress at time t(i+1) is also deduced,

, ( 1) , ( ) , ( ) 1 , 1( )c t i c t i c t i i c iE (13)

, ( 1) , ( ) , 1s t i s t i s s iE (14)

The strain and stress of concrete and steel at any time can be calculated.

3 CASE STUDY

A 500 × 500 mm2 square column is 3 m high and the reinforcement ratio is 1.82%. The column is subjected to an axial load of 1250 kN. The column is loaded on the 28th day. The humidity is 60%. The calculation step is shown as Table 1.

Table 1. The calculation step

Age 28d-90d 90d-365d 1y-3y 3y-10y 10y-50y

Step 3d 7d 21d 60d 180d

The deformation development of columns is shown in Figure 1. 50 years later the strains of RC columns are decreased to 71.9% and 77.1% respectively when compared to plain columns according to GL2000 model and CEB-FIP 1990 model.

The development of stress of concrete is shown in Figure 3. The stresses of concrete are decreased to 34.7% and 49.2% of the initial stress according to GL2000 model and

(6)

·229·

CEB-FIP1990 model. The existence of steel has a significant impact on stress distribution on column section.

Figure 1. Development of strain of column

Figure 2. Development of stress of concrete

The development of stress of steel is shown in Figure 3. The stresses of steel are increased to 5.7 times and 5.2 times of the initial stress according to GL2000 model and CEB-FIP1990 model. The increased stress may lead to the yield of the steel.

Figure 3. Development of stress of steel

The deformation of columns with different reinforcement ratios is calculated according to GL2000 model. As shown in Figure 4, the deformations of concrete are decreased by 12.5%, 20.4%, 27.1%, 32.8%, 37.8%, 42.1% respectively compared to the plain column when the reinforcement ratios are 0.5%, 1%, 1.5%, 2%, 2.5%, 3%.

Figure 4. Deformation of columns with different reinforcement ratios

4 EXPERIMENT

An experiment is carried out to investigate the validity of the theory. There are five groups of specimens with reinforcement ratios being 0, 0.5%, 0.9%, 1.4%, 2% respectively. Each group has two specimens and the size is 150 mm × 150 mm × 300 mm. The concrete mixture was case in mold and then compacted by a vibrating table. The specimens were moist cured for 7 days. Then they are put in a room environment with relative humidity being 75%. High precision vibrating wire strain gage was used to measure the strain and the embedded temperature sensor can prevent the specimens from the influence of site temperature. The specimens were subjected to an axial load of 103.5kN. The datas are shown in Figure 5.

Figure 5. Deformation of specimens with different reinforcement ratio 科

学出版社


www.abook.cn

·230·

It is shown that with the increase of reinforcement ratio, the restraint effect trends to increase. The deformation reduction ratio is defined as the deformation of specimens with different reinforcement ratio divided by the deformation of the plain specimen. All the rates of deformation reduction are listed in Table 2 and compared with the theoretical values calculated with CEB1990 model and GL2000 model.

Table 2. The deformation reduction rate

Reinforcement ratio

Test data CEB-FIP1990 GL2000

0 1 1 1 0.5% 88.0% 90.7% 92.0% 0.9% 83.0% 84.2% 86.3% 1.4% 74.5% 77.1% 80.0% 2.0% 68.3% 69.8% 73.4%

The deformation reduction rates are 88%, 83%, 74.5%, 68.3% respectively corresponding to reinforcement ratio being 0.5%, 0.9%, 1.4%, 2%. It is shown that the influence of reinforcement on shringake and creep of concrete is more significant than numerical simulation.

5 CONCLUSIONS

1) It should be noted that the reinforcement has a significant restraint effect on the time-dependent deformation of column. With the increase of reinforcement ratio, the restrain effect enhances. The stress of steel is increased to several times which may lead to the failure of the member or structure.

2) The experiment also show a consistant restraint effect on the deformation of specimens due to reinforcement. The deformation were decreased to 88%, 83%, 74.5%, 68.3% respectively corresponding to reinforcement ratio being 0.5%,

0.9%, 1.4%, 2%. Generally the deformation reduction rate according to different reinforcement ratio is close to the calculated values of prediction models.

ACKNOWLEDGMENTS

This work is support by National Nature Science Foundation of China (Grant NO. 51178154, 50938001), Basic Research Program of Shenzhen Science and Technology Plan (Grant NO. JC201005260163A).

REFERENCES

Comite euro-international du beton. (1990). CEB-FIP MODEL CODE. Wiltshire Redwood Books.

Fu, X.Y., and Gao, H. (2008). Analysis on Shrinkage and Creep for Reinforced Concrete Column. Journal of Building Structures, S1,191-199.

Goel, R., Kumar, R., and Paul, D.K. (2007). Comparative Study of Various Creep and Shrinkage Prediction Models for Concrete. Journal of Materials in Civil Engineering， 19:3,249-260.

Pan, Z.F., Lü, Z.T., and Fu, C.C. (2011). Experimental Study on Creep and Shrinkage of High-Strength Plain Concrete and Reinforced Concrete. Advances in Structural Engineering, 14:2, 235-246.

Sun, C. (2010). Research and Application of Long-term Creep and Shrinkage Effects on Reinfored Concrete Strcture. Doctor’s Theses of Harbin Institute of Technology,43-49.


STATIC BEHAVIOUR OF FLOATING PIERS UNDER UNILATERAL LOADS

Yujun Qi, Fubin Zhang, Weiqing Liu College of Civil Engineering, Nanjing University of Technology, Nanjing 211816, P.R. China Abstract: The rotation angle of floating pier under unilateral load should be limited to small range to ensuring safety. First, the design guidelines in China for the structural system of floating piers are proposed. Second, a mechanical model is built, and based on the hydrostatic theory a mechanical equation is achieved to obtaining exact solution of tilting angle for a pontoon with rectangular section. Then, the applicable condition of approximate solution is studied. And then, based on the exact solution, the effects of various parameters on the tilt of pontoon are studied. The results show that the greater width smaller height and lower center of gravity position of pontoon is advantageous to the static stability of the floating piers. Keywords: Floating piers, stability, unilateral loads, tilting angle, hydrostatic theory 1 INTRODUCTION

Floating piers are kinds of piers which are built in many ports of the world, because they have various advantages compared to land based structures, e.g., they are easily adaptable to large fluctuations in the mean water level, they are mobile and can be relocated relatively easily, they offer less obstruction to water circulation and fish migration, and so on (Ken-ichi Uzaki, 2011).

Usually, the structural system of floating piers consists of a number of floating bodies, called pontoons which are connected to each other by hinges. And pontoons are connected to piles by a guide frame, rigidly attached to the pontoon, and having rollers in its sides. Thus, free vertical movement of pontoon along the pile is allowed, when loading on the pontoons or varying water levels. During an offset loading on the pier, a rotation of potoons about its transverse axis can also generates, because of the gap between the pile and guide frame. Obviously，the rotation angle of potoons should be designed so that they have sufficient capacity to sustain these loads.

Due to the special structural formation of floating piers and mobility of the pontoons, they have high flexibility. This makes both of their static response and dynamic response different from that of fixed piers. The dynamic response is investigated byHsien Hua Lee (2007), Azadeh Mostofi (2012), Zahra Tajali (2011) and Subrata K. Chakrabarti (2006). These results offered useful references for the floating-piers design,

but these methods of calculations are hard to be applied for their difficult. An approximate method can be used for estimating the rotation angle of potoons under the vertical offset loads for the floating piers are similar to those of ships.

In this paper, first the design guidelines of floating piers are investigated. Then based on the concept of buoyancy, the accurately solving formula of rotation angle for potoons with rectangular section are studied. Finally, use the accurately solving formula, the static behaviour of floating piers under vertical offset loads is investigated.

2 DESIGN GUIDELINES

At present, there is no design code of floating piers in China, for floating piers starts relatively late. Based on the project experience and actual situation in China, and reference foreign codes, “design: small craft berthing faciltiies” and “Guidelines for design of marinas”, etc. the design guidelines for the structural system are proposed as following (He Wenqin, 2004).

(1) Width of walkway Main pier: 2.0-3.0m Finger pier: 1.0-1.5m (2) Design loads For the main pier, design live loading shall be 科

学出版社


www.abook.cn

·232·

2000-3000Pa, concentrated live load shall be 1600N.

For the finger pier, design live loading shall be 1000-2000Pa, concentrated live load shall be also 1600N.

(3) Static stability The inclination of both main pier and finger

pier shall be no greater than 1/10, when the vertical loads loaded on the half of walkway.

3 STATIC STABILITY OF SINGLE- PONTOON WITH RECTANGULAR SECTION

3.1 Problem Description

A typical floating pier is conposed of main pier and finger piers which have continuous pontoons with rectangular section. The deck is connection and the pontoons are disconnected at the joints of main pier and finger piers. The distributed load q is loading on the half of the walkway as shown in Fig.1.

Figure 1. Unilateral load (a) on the main pier and (b) on the finger pier

3.2 Modling and Calculating

3.2.1 Basic assumption

During this study, following basic assumptions are employed.

1) Pontoon is rigid body. It means that the effet of elastic deformation on the liquid pressure is neglected.

2) Piers are independent. The interaction between main pier and finger pier is neglected because pontoons are disconnected at the joints of main pier and finger piers.

3) Restraint of pile is ignored, because of the gap between the pile and guide frame.

4) Distributed loads are equivalent to concentrated loads.

3.2.2 Modling

Based on basic assumptions, the geometry of the problem is shown in Fig.2. The pontoon is idealized as two-dimensional, and Cartesian coordinates are employed and fixed in the pontoon. The z-axis is directed vertically upwards from an origin on the middle point of the bottom of the pontton. The pontoon is W×H×L (width×height×length), and is of draft h0 under dead load. A mass m loaded on the middle point of the top of the pontton, the draft of pontoon is denoted by h as shown in Fig.2. And then the mass m moved to the position of distance e from point (0, H) along the x-axis on the top of pontoon, the rotation angle of pontoon is , and the increasing amount of length that is in the side of pontoon under water level is h calld tilting height.

Figure 2. Sketch of mechanical model (a) when loaded on the (0, H) and (b) when loaded on the (e, H)

3.2.3 Analysis and calculat

The hydrostatic pressure p1 and p2 at lower left corner and lower right corner respectively can be written as:

1 ( )cosp g h h (1)

2 ( )cosp g h h (2) Where is dency of water and g is acceleration of gravity.

The F1, F2 and F3 are resultant forces of hydrostatic pressure on the left side, the right side and the bottom of pontoon respectively, and can be written as:

1 1

1( )

2F p h h L (3)

2 2

1( )

2F p h h L (4)

3 1 2

1( )

2F p p WL (5)

The action points of F1, F2 and F3 are denoted by (x1, y1), (x2, y2) and (x3, y3) respectively, and can be written as:

·233·

1 1

1( , ) ( , ( ))

2 3

Wx y h h (6)

2 2

1( , ) ( , ( ))

2 3

Wx y h h (7)

3 3( , ) ( ,0)6

hWx y

h

(8)

The and h should be satisfying the following geometric relation:

tan 2 /h W (9) The moments balance equation about the

origin O as shown in Fig.2, can be written as: 1 1 0 0sin sin cosGF y G y P H P e

2 2 3 3F y F x (10)

Put Eqs. (1) - (9) into the Eq.(10), the equation (11) is obtained as following:

3 2 0 01 1 2( ) (

3 6GG y

gL h gW LW

2 2)( ) 0

HgLh P h Pe

W (11)

Thus, h can be obtained, because there is just one unknown h in this simple cubic Eq (11).

The obtained h should be satisfying the following Eq. (12), to prevent that the deck is immersed in water.

h h H ≤ (12) Then, put h into Eq. (9), the rotation angle

of pontoon can be obtained. Furthermore, the first item in Eq.(11) can be

ignored when h is smaller value, and the h can be immediate obtained as following. This is an approximate solution.

)22

6

1( 2002

W

HPgLh

W

yGLgW

Peh

G

(13)

3.3 Result and Discussions

A parametric study was carried out to investigate the static behaviour of the floating pier to the various overturning moment (induced by eccentric load P and eccentricity e) and structural parameters. Value ranges of these parameters are as following.

(1) Size parameters The width W and height H of pontoon are

called size parameters. The value ranges of width W and height H are 1.0~3.0m and 0.5-2.4m respectively, and the value of H/W was

less than 1.0. In addition, the length L was a constant 1.0m, for two-dimensional problem.

(2) Moment parameters The overturning moments were induced by

eccentric load P which is 250~3600N and eccentricity e which is 0.1-1.2m.

(3) Dead weight parameters The dead weight parameters including initial

draft h0 and center of gravity position yG0 determine the initial floating state of pontoons. All of the value ranges of h0/H and yG0/H was 0.1-0.9 in 0.1 increments.

3.3.1 Accuracy of approximate solutions

The accuracy of approximate solution was investigated in comparison to that of exact solution as shown in the Fig.3. The results show that the variation of tilting height h with parameters is same for the results of Eq. (12) and Eq. (14), and the error increased with increase of tilting height h.

(a)

(b)

Figure 3. Comparition between approximate solutions and exact solutions at W=1.0m, P=500N and (a) yG0/H=0.2 and (b) yG0/H=0.5

科

学出版社


www.abook.cn

·234·

(a)

(b)

Figure 4. Relative errors- tilting angle curve for (a) W=0.5m and (b) W=2.4m

3.3.2 Effect of size parameters

The width W is important sise parameters, for it has direct influence on the buoyancy of pontoon when the length L is constant. Fig.5 shows the tilting angle vs. width W of pontoon curves, and shows that the tilt angle of pontoon decreased with the increase of width W of pontoon under the same overturning moments. Therefore, the static stability can be improved when the width W of pontoon is increased.

Figure 5. Effect of width W on tilting angle at a constandt H of 0.5m (case a: P=250N, e=0.5m, h0/H=0.1; case b: P=250N, e=0.5m, h0/H=0.2; case c: P=750N, e=0.2m, h0/H=0.1; case d: P=750N, e=0.2m, h0/H=0.2)

When the width W of pontoon is same, the tilt angle of pontoon increased with the increase of height H of pontoon, as shown in the Fig.6, because of the increase of overturning moments with the increase of height H of pontoon. Therefore, greater height H of pontoon is disadvantageous to the static stability of the pontoon.

Figure 6. Effect of height H on tilting angle at a constandt W of 1.0m (case a: P=250N, e=0.1m, h0/H=0.1; case b: P=250N, e=0.2m, h0/H=0.1)

3.3.3 Effect of moment parameters

The inclination of pontoons is induced by the overturning moments. An initial overturning moment M0 is difined by product of eccentric load P times the eccentricity e. it represents the overturning moments when just put the eccentric load P on the pontoon and the rotation of pontoon is not beginning. The overturning moment under the hydrostatic equilibrium is denoted by MP, and it is unknown befor solving the Eqs. (11) or (13). The effect of moment parameters on the tilting height h can be studied by the initial overturning moment M0.

Fig.7 to Fig.9 show the effect of tilting height h on initial overturning moment M0 for both H/W=1/2 and H/W=1/4.5 conditions. The figure clearly shows that increase in initial overturning moment M0 increases the tilting height h for given conditions. Furthermore, for H/W=1/4.5 condition, the tilting height h is nearly at the same initial overturning moment M0. And, for H/W=1/2 condition, the tilting height h is great difference at the same initial overturning moment M0 and increase in eccentric load P increases the difference of tilting height h. For instance, when the value of the eccentric load P from 250N to 1250N (meanwhile, the value of the eccentricity e from 0.5m to 0.1m), the tilting

·235·

height h increases by about 110.8% at a constant initial overturning moment M0 of 125N·m.

Figure 7. Effect of initial overtuming moment M0 On the tilting height h at W=1.0m and H=0.5m (case a: e=0.1m and P is from 0 N to 1500N; case b: P=250N and e is from 0 m to 0.5m; case c: e=0.2m and P is from 0 N to 1000N; case d: P=500N and e is from 0 m to 0.5m)

Figure 8. Effect of initial overtuming moment M0 On the tilting height h at W=2.4m and H=1.2m (case a: P=1800N and e is from 0m to 1.2m; case b: P=3600N and e is from 0m to 0.6m)

The reasons for the above phenomenon can be finding from the overturning moment MP. the overturning moment MP, as shown Fig.2, can be weitten as:

sin cosPM P H P e (14)

Substituting Eq. (9) and M0=Pe in Eq. (14), we obtain

0cos ( 2 )P

HM M P h

W (15)

Thus, the effet of the second item in Eq. (15) increased with the increase of value of H/W and

eccentric load P.

Figure 9. Effect of initial overtuming moment M0 On the tilting height h at W=1.0m and H=0.5m (case a: P=3000N and e is from 0 m to 0.9m; case b: P=3600N and e is from 0 m to 0.6m; case c: e=0.6m and P is from 0 N to 36000N)

3.3.4 Effect of dead weight parameters

The dead weight parameters include the initial draft h0 and center of gravity position yG0. Fig.10 and Fig.11 show the effects of initial draft h0 and

center of gravity position yG0 on tilting angle respectively.

Fig.10 clearly shows that increase in value of yG0/H increases the tilting angle for given conditions and the increasing rate increased with the increase of initial draft h0.

Fig.11 shows that there is a peak on each curve of tilting angle on h0/H at different position, respectively. The reasons for the above phenomenon can be finded from by the Eq. (13). The denominator in Eq. (13) can be weitten as a quadratic function:

CBhAhhf 0200 )( (16)

Where, A, B and C are coefficients, and gLA , )(2 0GP yhgLB

W

HPLgWgLhC P

2

6

1 22

Where, hP is the draft caused by the eccentric load P, and can be weitten as:

gWL

PhP

(17)

The equation of symmetry axis of (16) is

PG hyh 00 (18)

科

学出版社


www.abook.cn

·236·

Figure 10. Effect of ratio of yG0/H On the tilting angle at W=1.0m and H=0.5m (case a: h0/H=0.1; case b: h0/H=0.2; case c: h0/H=0.3; case d: h0/H=0.4)

Figure 11. Effect of ratio of h0/H On the tilting angle at W=1.0m and H=0.5m (case a: yG0/H=0.2; case b: yG 0/H=0.4; case c: yG0/H=0.5; case d: yG 0/H=0.6)

When yG0-hp≤0, there are a minimum value and a maximum value for the quadratic function (16) and Eq. (13) at h0=0, respectively. When yG0-hp﹥0, there are a minimum value and a maximum value for the quadratic function (16) and Eq. (13) at h0= yG0-hp, respectively.

4 CONCLUSIONS

This paper investigates the static behaviours of the floating piers composed of pontoons with rectangular section on the unilateral load. Based

on the hydrostatic theory, a mechanical equation is achieved to obtaining exact solution of tilting angle. It is proved that proposed approximate solution in this paper is closed to exact solution when tilting angle is less than 15 degrees. This study further reveals that both increase in width of pontoon and decrease in height of pontoon decrease the tilting angle. The study shows that the overturning moment and initial overturning moment have great difference for greater height-width ratio of pontoon, even greater than 110%. The study further shows that increase in center of gravity position increases the tilting angle for given conditions and the increasing rate increased with the increase of initial draft. Thus suitably parameters shall be selected so as to increase static stability of foating piers on unilateral load.

REFERENCES

Azadeh Mostofi, Khosrow Bargi. (2012). New Concept in Analysis of Floating Piers for Shipberthing Impact. Marine Structures, 58-70.

He Wenqin. (2004). On the Design of Marina. Port & Waterway Engineering, 61-64. (in Chinese)

Hsien Hua Lee, Liang-Yin Chen, Wen-Kai Weng, S.-W Shyue. (2007). The Prediction of the Dynamic and Structural Motions of a Floating-pier System in Waves, Ocean Engineering, 1044-1059.

Ken-ichi Uzaki, Yoshito Ikehata and Nobuhiro Matsunaga. (2011). Performance of the Wave Energy Dissipation of a Floating Breakwater with Truss Structures and the Quantification of Transmission Coefficients.

Subrata K. Chakrabarti, Partha Chakrabarti and M. Sri Krishna. (2006). Design, Construction, and Installation of a Floating Caisson Used as a Bridge Pier. Journal of Waterway, Port, Coastal, and Ocean Engineering, 143-156.

Zahra Tajali, Mehdi Shafieefar. (2011). Hydrodynamic Analysis of Multi-body Floating Piers under Wave Action. Ocean Engineering, 1925-1933.


A THERMODYNAMICAL CONSISTENT MICROPLANE PLASTIC-DAMAGE

MODEL FOR CONCRETE

Jianying Wu, Shilang Xu Department of Civil Engineering, Zhejiang University, Hangzhou 310058, P.R. China Abstract: Microplane theory is a general approach to construct complex three-dimensional material constitutive relations from the simpler one- or two- dimensional behavior on generic planes with predefined spatial orientations. Though microplane theory has been successfully applied to many engineering materials, it is still challenging to reconcile the conflict between the thermodynamic consistency and the numerical capability, e.g. in reflecting the deviatoric-induced dilatancy typical in concrete like geomaterials. In this work, we present a new strain-space microplane plastic-damage model for concrete, with (major) symmetric material stiffness so that the reported thermodynamical inconsistency is eliminated. Moreover, the internal frictions are explicitly considered on each microplane under compression by using a non-associated evolution law for the tangential behavior, and then the deviatoric-induced dilatancy can be well captured. An additional advantage is that the hysteretic loops during unloading/reloading cycles are also satisfactorily reproduced. Several typical numerical simulations are presented to validate the above conclusions. Keywords: Concrete, microplane model, plastic-damage, dilatancy 1 INTRODUCTION

Materials generally contain a multitude of defects in the form of microvoids. Upon the initial state it is reasonable to assume that the distribution of microvoids is homogeneous in an initially isotropic material. During the loading history, the old microvoids as well as the newly initiated ones may evolve into the so-called microcracks which undergo irreversible growth mainly “in the direction perpendicular to the maximum tensile strain or stress” (Krajcinovic and Fonseka, 1981). The microcracks evolution leads to the microstructure changing from initial isotropy to anisotropy (i.e. the so-called damage induced anisotropy). The inelastic behavior of isotropic materials is nowadays well-understood in either the plasticity (Chen, 1994) or damage mechanics (Krajcinovic, 2003). On the contrary, although large quantities of relevant researches have been carried on, the modeling of damage induced anisotropy is far more difficult and poses great challenges to both the scientific and engineering communities.

As either a pure macroscopic or a mesoscopic damage model is difficult to model the damage induced anisotropy. It is desirable to combine the advantages of both types of approaches. Microplane theory is such a general concept to

construct complex three-dimensional material constitutive relations from the simpler one- or two-dimensional behavior on generic planes with predefined spatial orientations. Though microplane theory has been successfully applied to many engineering materials, mainly attributed to the noteworthy contributions from Bazant and cowokers (Bazant and Oh, 1983; Bazant and Prat, 1988; Carol et al., 1992; Carol and Bazant, 1997; Wu, 2009), its refinement still remains an open issue. One of the most challenging problem of microplane theory is how to reconcile the conflicting between thermodynamic consistency and numerical capabilities (Carol et al., 2001), e.g. in the modelling of the deviatoric induced dilatancy in concrete like geomaterials.

In this work, we present a new microplane plastic-damage model for concrete, with major symmetric material stiffness so that the reported thermodynamical inconsistency is effectively eliminated. Moreover, the internal frictions are explicitly considered on each microplane under compression by using a non-associated evolution law for the tangential behavior, and then the shear-induced dilatancy can be well captured. An additional advantage is that the hysteretic loops during unloading/reloading cycles can also be satisfactorily reproduced. Several numerical examples of concrete are also presented.

科学出版社


www.abook.cn

·238·

2 GENERAL FORMULATION

2.1 Strain Space Plastic-damage Model

Similar to the strain space plasticity model (Han and Chen, 1986), the stress tensor s and its rate s are expressed as

p cre co , (1) where e : : E is the elastic stress with E being the material stiffness tensor, and p is the plastic stress; co : : E is the continuum stress rate that would be obtained if the microcracks did not evolve, and the cracking stress rate cr represents relaxations of the incremental elastic behavior due to the microcracks evolution, constiting of two parts, i.e. the degradation part d : : E and the plastic component p . In 1-D case the above definitions of the corresponding stress rates are illustrated in Fig.1.

Figure 1. 1-D illustration of the continuum, degradation and cracking stress rates

Due to microcracks evolution, concrete material exhibits both stiffness degradation and plastic deformations. Therefore, despite of the analogy to the strain space plasticity model (Han and Chen, 1986), Eq.(1) is characterized with a variable secant stiffness E , and the evolution laws for the stiffness and the plastic stress rate have to be postulated. However, this would be a rather challenging task for a classical tensor invariant-based macroscopic constitutive model (e.g. plasticity, damage mechanics, or their combinations) is used. Comparatively, it is much easier to overcome this problem within the framework of microplane theory. The key point of microplane theor is that the macroscopic properties can be determined as an integral or summation of all contributions from predefined microplanes over possible spatial orientations.

2.2 Microplane Plastic-damage Model

Within the framework of microplane theory, the normal, volumetric, deviatoric and tangential strains on each microplane are resolved as

N : , : N VV (2)

D T: , : D T (3)

where the projection operatorsN ,V ,D andT are expressed in terms of the unit normal vector n of the microplane

1,

3 N n n V I (4)

1,

3 ID n n I T n n n n (5)

for the second-order and symmetric fourth-order identity tensors I and I , respectively. According to the principle of virtual work the stiffness E is expressed as an integral form

0 0 0 T

V V D T

3

2d

D TE E E

V V D D T TE (6)

where is the unit hemisphere surface; the integrity variables V

, D and T

characterize the

damage of the microplane volumetric, deviatoric and tangential elastic moduli, 0

VE , 0DE and 0

TE .

Upon initial virgin state, i.e. V D T 1.0 , the

stiffness E should be identical with the elasticity tensor of the material, so that

0 0 0 0 0 0 0

V D T(1 2 ), (1 )E E E E E (7) for the second-order and symmetric fourth-order identity tensors I and I, respectively. Note that the stiffness tensor (6) is of major symmetry which is thermodynamically consistent. Therefore, the continuum stress rate co in Eq. (1) is expressed as

co co T co

V D T

co 3

2 ds s V D T s (8)

where the microplane volumetric, deviatoric and tangential continuum stress rates are given in the incremental elastic form

co 0 co 0 co 0

V V V V D D D D T T T T, ,s E e s E e E s e (9)

Similarly, the crack stress rate cr in Eq. (1) can be expressed as

T

V V D D T

3

2

cr dg g g V D T (10)

where the microplane Lagrangian multipliers , with the subscripts V, D or T , satisfy the Kuhn-Tucker conditions

0, 0, 0f f ≤ ≥ (11)

where g and f are the microplane potential and yield functions, respectively, all dependent on the microplane strains.

2.3 Yield and Potential Functions

In this work the following simple microplane

·239·

yield functions are assumed

0f e ≤ (12)

A loading state corresponds to the actual values of the strain-like internal variables (the cracking thresholds) being larger than the maximum values reached so far. In other words, the cracking stresses evolve only under loading states while they remain constant for unloading and reloading stage. Once crack evolves, the cracking thresholds are updated accordingly.

The internal friction on a rough plane plays an important role on the nonlinear behavior of concrete in compression. It is this mechanism that causes the shear induced dilatancy and leads to non-associated evolution laws. In this work, it is assumed that the microplane volumetric and deviatoric evolution laws are associated while the tangential behavior is non-associative. In other word, the microplane potential functions are expressed as

V V D D,g f g f (13)

T T N T( )g H e (14) where is the frictional coefficient reflecting the shear-induced dilatancy; the Heaviside function

( )H is defined as ( ) 1H x if 0x and ( ) 0H x if 0x≤ , introduced to guarantee that the internal

frictions are only activated on those microplanes under compression. Therefore, the stress rate (10) becomes

cr cr cr cr

V D T N

3

2

cr dT V D T N (15)

where the microplane stresses cr

V , cr

D , cr

N and cr

T

are given by cr cr

V V D D,V DH e H e (16) cr

N N T T T T T( ) ,H H e H e (17) with length Te of the microplane tangential strain

vector Te and the hardening/softening functions

H .

2.4 Hardening/softening Functions

In this work the hardening/softening functions are assumed as follows

0H e h E (18) where 0E represents the secant modulus of the

microplane stress-strain curve and h is defined

as h e . In this work, the following

functions are adopted for

V 1 V

V V

V 1 V 1 V

1exp | | 0( )

1 | | | | 0

p

p q

e a ee

e a e a e

≥ (19)

D 1 D

D D

D 2 D

1

2

exp | | 0( )

exp | | 0

p

p

e a ee

e a e

≥ (20)

3

T T T 3 V( ) exp | | (p

e e a e (21) where , ,a b p and q are parameters controlling the

volumetric compressive behavior; 1 2,p p and 3p

are exponential parameters affecting the curve shape; 1 2,a a and 3a are parameters controlling the

peak strength of the microplane stresses; is an parameter to reflect the lateral confinement; is the McAuley brackets, defined as max( ,0)x x . Usually the following parameters are applicable for most concrete: a = 0.005, b = 0.225, p = 0.25

and q = 2.25; 1p = 1.0, 2 3p p = 1.5.

To describe the nonlinear behaviour due to damage evolution and plastic flows, we adopt the following simple rule: the volumetric and deviatoric tensile behavior is assumed to be elastic damage, while the volumetric, deviatoric compressive behavor and the tangential one are assumed to be plastic.

More specifically, during the volumetric and deviatoric unload/reloading state the tensile and compressive virgin curves act as envelopes and straight lines are assumed with a certain slope. The compressive behavior is assumed to be elastoplastic, that is, the unloading branches are assumed to have the initial slopes 0

VE and 0DE ,

i.e. V D 1.0 , and the tensile parts of the

diagrams shift accordingly to the points where the unloading compressive branches intersect the strain axis. In constrast, the unloading/reloading behavior in tension is assumed to follow a secant slope between the maximum point reached in the tensile curve and the origin of that curve, i.e. V V and D D . The above rules which

were also adopted in Carol et al. (1992), are shown in Figs.2(a) and 2(b).

For the tangential behavior, a different rule is assumed for the unloading/reloading branches. That is, the virgin loading curve is used as an envelope, and unloads with an initial stiffness 0

TE .

During unloading when the strain axis is reached zero stress is assumed. For the reloading branch, a straight line from the reloading point (the end point of the unloading) to the starting point of unloading is followed. Thus a reloading branch with a slope smaller than the initial stiffness 0

TE is

used herein. In this way, a very simple loop can 科

学出版社


www.abook.cn

·240·

be obtained at the microplane level, which is sufficient for a reasonable unloading/reloading hysteretic loop at the macroscopic level; see Fig.2(c) for the illustration.

(a) Microplane volumetric behavior

(b) Microplane deviatoric behavior

(c) Microplane tangential behavior

Figure 2. Microplane stress-strain relations

2.5 Loading/unloading/reloading Rules

As different microplane moduli are adopted to determine the stress, under general triaxial stressstrain states the criterion that discriminate the loading/unloading/reloading states should be introduced. In this work the following simple criteria (see

Fig.2 for the details) are adopted

p p max

p min

tensile loading : 0, 0

compressive loading : 0, 0

unloading/reloading: otherwise

e e e e e

e e e e

≥ ≥

(22)

for the volumetric and deviatoric behavior, and max

T T

min max

T T T

min max

T T T

loading : 0

unloading : 0,

reloading: 0,

e e

e e e e

e e e e

≥

(23)

for the tangential behaviour. Here, e referes to either of the volumetric strain Ve or the deviatoric

one De , with pe being the corresponding plastic

strains; maxe and maxe denote the maximum and

minimum values of the microplane strains ever reached so far.

3 NUMERICAL EXAMPLES

In the first example the uniaxial compression test performed by van Mier (1984) is simulated with the following parameters: 0E 25GPa, 0 0.18, 1a 5.0×10-4, 2a 5.0×10-3, 3a 1.5×10-3

and 0.0. In Fig.3 the predicted curve of axial stress vs. volume change is shown. As we can clearly see, the shear-induced dilatancy can be capatured fairly well.

Figure 3. Hysteretic loops under uniaxial cyclic compression

The next example corresponds to the cyclic behavior of a concrete specimen subjected to a cyclic uniaxial compression. In this example, the uniaxial strain xx cycles at 0.0015, 0.0024, 0.0035, 0.0045 and finally increases to the value

·241·

0.006. The predicted results of stress-strain curves are shown in Fig.4, with the following parameters: 0E 26.8GPa, 0 0.18, 1a 7.0×10-5,

2a 2.0×10-3, 3a 1.7×10-3 and 0.0. Remarkably, owing to the possible combination of loading/unloading/reloading states on each microplane, both the stress-strain curve and the hysteresis loops can be predicted realistically.

Figure 4. Hysteretic loops under uniaxial cyclic compression

In this final example we consider the behavior of concrete under biaxial compression, with the parameters 0E 30GPa, 1a 5.0×10-5, 2a 1.7×

10-3, 3a 1.5×10-3 and 0.88. Numerical

simulations uner plane stress state with different ratios between principle stresses are performed. The predicted biaxial strength envelope is shown in Fig.5. The simulation result agrees with the test data (Kupfer et al. 1969) fairly well.

Figure 5. Comparison of strength envelop with biaxial compressive test by Kupfer et al. (1969)

4 CONCLUSIONS

The proposed microplane plastic-damage model is not only thermodynamically consistent, but also is able to capture the internal frictions. The numerical results demonstrate that the typical behavior of concrete in compression, e.g. the shear-induced dilatancy, the unloading/reloading hysteretic loops, and the strength enhancement due to lateral confinements, etc., can be well captured.

ACKNOWLEDGEMENTS

The first author (J.Y. Wu) would like to acknowledge the support from the National Natural Science Foundation of China (51008130), the State Key Laboratory of Subtropical Building Science (2008ZA10) and the Fundamental Research Funds for the Central Universities (2012FZA4017).

REFERENCES

Bazant, Z. P. and Oh, B.-H. (1983). Microplane Model for Fracture Analysis of Concrete Structures. Proc. Symp. Interact. Non-Nucl. Munitious Struct., US Air Force Academy, Springs. Co.: 49-55.

Bazant, Z.P., Prat, P.C. (1988). Microplane Model for Brittle-plastic Material. I: theory and II: Verification. J. Eng. Mech. ASCE, 114 : 10, 1672-1699.

Carol, I. and Bazant, Z.P. (1997). Damage and Plasticity in Microplane Theory. International Journal of Solids and Structures, 34: 29, 3807-3835.

Carol, I., Jirasek, M. and Bazant, Z.P. (2001). A Thermodynamically Consistent Approach to Microplane Theory. Part I. Free energy and consistent microplane stresses. International Journal of Solids and Structures, 38: 2921-2931.

Carol, I., Prat, P.C. and Bazant, Z.P. (1992). New Explicit Microplane Model for Concrete: Theoretical Aspects and Numerical Implementation. International Journal of Solids and Structures, 29: 9, 1173-1191.

Chen, W.F. (1994). Constitutive Equations for Engineering Materials: Plasticity and Modeling. Elsevier, Amsterdam. 科

学出版社


www.abook.cn

·242·

Han, D.J. and Chen, W.F. (1986). Strain-space Plasticity Formulation for Hardening-softening Materials with Elastoplastic Coupling. International Journal of Solids and Structures, 22: 935-950.

Krajcinovic, D. (2003). Damage Mechanics. Elsevier B.V., Netherlands.

Krajcinovic, D., Fonseka, G.U. (1981). The Continuous Damage Theory of Brittle Materials. J. Appl. Mech. 48, 809-824.

Kupfer, H., Hilsdorf, H.K. and Rusch, H. (1969). Behavior of Concrete under Biaxial Stresses. ACI Journal, 66: 656-666.

van Mier, J.G.M. (1984). Strain-softening of Concrete under Multiaxial Loading Conditions. Ph.D. Dissertation, University of Eindhoven, The Netherlands.

Wu, J.Y. (2009). An Alternative Approach to Microplane Theory. Mechanics of Materials, 41: 87-105.


SMART AGGREGATES USED FOR SEISMIC STRESS MONITORING IN CONCRETE STRUCTURES

Shuang Hou, Haibin Zhang, Jinping Ou Civil Engineering, Dalian University of Technology, Dalian 116024, P.R. China Abstract: A smart aggregate(SA) using PZT material as the sensing element for compressive seismic stress monitoring is developed in this paper. The sensor consists of a piece of PZT sheet sandwiched by a pair of stone blocks through epoxy. Investigation of the sensors performance involves of the loading scheme in which the alternating stress superimposed on the compressive pre-stress are applied on the smart aggreagate by the servohydraulic test machine. The polarization process of the selected PZT material under pre-stress is investigated to decide the waiting time before calibration. The degradation of piezoelectric coefficient of PZT under pre-stress in range of 0-24MPa is evaluated. Keywords: Smart aggregates, PZT, compressive stress, degradation 1 INTRODUCTION

One of the major obstacles in the research of seismic collapse mechanism for concrete building structures is the lack of reliable method to monitor the interal stress in concrete during the collapse stage.Based on the evaluation of the stress wave propogatoin in concrete, a cement-based smart aggregate sensor (SA sensor for abbreviation in this paper) was proposed by Song for damage identification in concrete structures(Song et al. 2005), and it has also been used in concrete early strength monitoring(Gu et al. 2006), and over height vehicle-bridge collision monitoring(Song et al. 2007). This monitoring method could can only qualitatively evaluate the structure state, while the stress history of the structure can not to be obtained during the damage process. Utilzing the direct piezoelectric effect, the SA sensor is used in trafic load measurement on the road(YANG et al. 2005),(Li et al. 2006). PZT and PVDF in the form of sheet were utilized for low strain measurement (Jayant and Inderjit 2000). However, the application of the SA sensor in relatively high stress/strain is not found so far.

In the sensor application for PZT material, the piezoelectric response is linear and reversible when subjected to small stress(ANSI and IEEE 1987). However under relatively high stress, the

irreversible changes in the material, which takes a finite time, will cause hysteresis. Alguero(Alguero et al. 2001) found that for a soft PZT(P-5H), its piezoelectric coefficient varies with both the level of the pre-stress and the stress amplitude. For the selected PZT material, the piezoelectric coefficient drops by 50% when subjected to pre-stress of 30MPa, and increases by 11% per MPa of the stress amplitude. The significant degradation in piezoelectric coefficient for soft PZT when subjected pre-stress were also found by other researchers(Calderon-Moreno 2001; G et al. 2001; Yimnirun et al. 2006).

In this paper, a smart aggregate using PZT material for seismic compressive stress monitoring is developed. Its application in monitoring the relatively high stress is evaluated through tests. The loading scheme of alternating stress superimposing on static pre-stress was performed though servohydraulic machine for the new sensor’s calibration.

2 SENSING PRINCIPLE

Under small field conditions, the constitutive relations for a piezoelectric material are(ANSI

科学出版社


www.abook.cn

·244·

and IEEE 1987)

d

i ij j im mD e E d (1) c E

k jk j km md E s (2)

which can be rewritten as d

c E

D Ee d

d s

(3)

where a vector D of size (3×1) is the electric displacement (Coulomb/m2), is the strain vector (6×1), E is the applied electric field vector (3×1) (Volt/m), and m is the stress

vector (6×1) (N/m2). The piezoelectric constants

are the dielectric permittivity ije of size (3×3)

(Farad/m), the piezoelectric coefficients dimd (3×6) and c

jkd (6×3) (Coulomb/N or

m/Volt), and the elastic compliance Ekms of size

(6×6) (m2/N). For a patch of piezoelectric material (see Fig.1), the poling direction, which is usually along the thickness, is denoted as the 3-axis. The 1-axis and 2-axis are in the plane of the patch. In the case of a stress sensor, where the applied external electric field is zero, Eq.3 becomes

1

2

1 15

3

2 24

4

3 31 32 33

5

6

0 0 0 0 0

0 0 0 0 0

0 0 0

D d

D d

D d d d

(4)

where, 1 , 2 and 3 are the normal stress

in the 1, 2, and 3 direction respectively, 4 ,

5 , and 6 are the shear stresses in the 2-3,

1-3, and 2-3 planes respectively, and the coefficients 31d , 32d , and 33d relate the normal

stress in the 1, 2, and 3 directions respectively to a field along the poling direction, 3D . The

coefficients 15d and 24d relate the shear

stress in the 1-3 plane to the field 1D and the

shear stress in the 2-3 plane to the field 2D .

The electric displacement can be related to the generated charge with relation to

1

1 2 3 2

3

d

d

d

A

q D D D A

A

(5)

where 1dA , 2dA , and 3dA are the components

of the electrode area in the 2-3, 1-3, and 1-2 planes respectively. When the PZT is used to measure the uniaxial compressive stress, it is generally in the thin form of patches with its two faces coated with thin electrode layers with the applied stress in the 3 direction. Since the areas of 1A and 2A is much smaller than that of 3A ,

Eq.5 can be rewritten as

3 3q D A (6)

Substituting Eq.4 in Eq.6 gives 3

3 31

i ii

q d A

(7)

The charege q can be converted to the voltage V by charge-voltage converter in relation of

qV

C (8)

where C is the capacitance the feedback capacitor of the converter.

If the sensitivity of the SA sensor is defines as

3S

VS

(9)

where 3S is the stress on the SA sensor

which is in the 3 direction of the PZT sheet. From Eq.7 to 9 the sensitivity can be related to the piezoelectric coefficient of the PZT sheet by

33

31

ii

i

AS d

C

(10)

where i is the stress ratio expressed as

ii

， 1,2,3i (11)

where 3S is the stress on the SA sensor in the

3 direction.

Figure 1. PZT sheet

·245·

3 SENSOR DESIGN AND FABRICATION

Figure 2 demonstrates the structures of the proposed sensor. The sensor consists of a piece of PZT sheet connected with a piece of two-wire cable and a pair of stone blocks. The commercially-available soft PZT ceramic referred as P-5H with major composition of Pb(TiZr)O3 and featuring high sensitivity was chosen. The PZT’s properties measured by the supplier are listed in Table 1. Although the hard PZT is reportedly more resistant to degradation in piezoelectric coefficient(Calderon-Moreno 2001; G et al. 2001; Yimnirun et al. 2006), the soft PZT will render the senor with potential as an actuator in active sensing. The stone used here is the marble which has the relatively low hardness to allow manually cutting. The young’s modulus of the marble is 51.5GPa which is comparable with the concrete. The Young’s modulus of epoxy used is 2.5GPa, as supplied by the manufacture. The size of the PZT sheet is 15mm×15mm and its thickness is 0.3mm. The PZT sheet was connected to the two-wire cable on its two sides at the position close to the edge by welding. The marble stone was cut into the size of 25×25×12 mm. Then they were aligned together along the side of 12mm long and were held tightly. Along one central line in the interface between the two marble block, a hole with diameter of 3mm and depth of 10mm was pored to accommodate the connected cable and the welding points on the PZT sheet. Then the PZT sheet is sandwiched by the marble blocks though epoxy. After the epoxy was cured, the thickness of the epoxy layer is measured to be about 0.45mm.

Figure 2. Structure of the proposed smart aggregate

Table 1. Typical properties of the selected PZT P-5H

Young’s modulus(GPa) 46 Density(kg/m3) 7.5 d31(pC/N) 186 d33(pC/N) 670 d15(pC/N) 660

From the design and the fabrication techneque, it is evident that the newly developed sensor has two advantages over the cement-based SA sensor when used for the seismic stress monitoring. The first one is its stability in performance. For the commonly-used concrete, the compressive strenth is generally not higher than 60MPa. In this range, both the hard and the soft PZT are in elastic stage(Calderon-Moreno 2001), so are the marble block and the epoxy. Therefore as a whole the sensor works in the linear range when subject to seismic load. However for the cement-based SA sensor, nonliear behavior will occur due to the nonlinear characteristics of cement. The other advantage is its consistance in fabrication quality. When fabricaing the cement-based SA sensor, the PZT sheet is embeded in the cement mortar in a small mold which is then vibrated for the consilidation of the mortar. In this process, the PZT sheet position is subject to change, reducing the consistance in its sensitivity.

The sensitivity of developed the sensor has also been calculated theoretically. The finite element analysis (by ABAQUS 6.10) was carried out to investigate the stress distribution in the PZT sheet when an uniform stress of 10MPa is applied on the SA sensor along the direction of the PZT’s thickness. The stress distributions in the three directions in the mid interface of the SA sensor are shown in Fig.3. Due to the existence of the hole in the SA sensor, the stress in the PZT becomes nonuniform. The stress ratio 1 , 2 and 3 on PZT sheet to the stress on the sensor are 0.324, 0.289 and 1.22, respectively. Here the average stress on the PZT is used. The charge generated by stress 3 can be calculated to account for 88 percent of the total charge, indicating that the sensor is less sensitive to the stress in the direction 1 and 2. For the case when feedback capacitance of the charge amplifier is set to 200nF, the sensor’s sensitivity can be calculated by Eq.10 to be 1.05V/MPa.

For the general charge amplifier with maximum charge input of 1.0×106pC, by Eq.7, the amplitude range of stress for the sensor can be calculated to be 5.2MPa. In this amplitude range, the sesnor is able to monitor the seismic response in structures when there is no significant damage. To monitor the stage of concrete crushing, the amplitude range of the sensing system should be extended by using the charge-to-voltage converter with larger maximum input charge.

科学出版社


www.abook.cn

·246·

(a)

(b)

(c)

Figure 3. Stress distribution in the mid interface of SA sensor in the: (a) 1 direction, (b) 2 direction and (c) 3 direction

4 EXPERIMENTAL RESULTS

4.1 Experimental Setup

The test configuration is shown in Fig.5. An alternating stress superimposed on compressive pre-stress is applied on the SA sensor by the servohydraulic test machine(MTS 810). The generated charge signal will be converted to voltage signal by charge amplifier(HK9301, Hengke Tech. Co., China). The voltage signals then were measured by the data acquisition

system(NI4472 module), which also measures the load signal applied by the servohydraulic test machine. Fig.6 shows the outlook of the test system.

Figure 4. Test configuration for voltage and stress measurement

Figure 5. Outlook of the test system

4.2 Depolarization Process Under Pre-stress

The calibration of the SA sensor could start only after this depolarization process has finished. The time dependence of the process for the PZT used in this paper is investigated by evaluating the sensitivity variation with pre-stress. Measurement was performed on a pair of identical SA sensors, denoted as SA1 and SA2, which were aligned together and applied with two levels of pre-stresses, 4.8MPa and 14.4MPa for 30 minutes. Meanwhile, a series of alternating compressive stress were superimposed on the pre-stress for measuring the sensor’s sensitivity. The two levels of pre-stress are corresponding to the low and high axial loads that might occur in concrete building structures. The loading scheme is shown in Fig.7. As an example, Fig.8 shows the applied alternating stress during 994-1006s, which consists of a group of sine wave stress with ever-increasing amplitudes up to 2.4MPa, and frequency of 3MPa. It can be seen that the responses of these two sensors are quite close. The SA output voltage against the applied load is also drawn in Fig.9, indicating the existence of slight hysteresis. This could result from the nonlinear piezoelectric behavior of the PZT under pre-stress and the frequency characteristics of charge amplifier used in this paper.

For simplicity, the amplitudes ratio of the

·247·

alternating stress to the SA output voltage in each loading cycle was calculated as the sensor’s sensitivity(see Fig.10), which also equals the average slope of the voltage versus stress curve. The sensitivities of the two sensors are 0.63 and 0.60V/MPa, respectively. The amplitudes of the sensor output voltage and the amplitudes of the stress keeps linear in the range of 0-2.25MPa. This finding defers greatly from the result of the literature(Alguero et al. 2001) which shows the piezoelectric coefficient linearly increases by 10% per MPa for the soft PZT-5H in the amplitude range of 0-2.5MPa.

Figure 6. Load history

Figure 7. Amplitudes of SA output vs. applied stress during 694-1006s

Fig.11 shows the sensitivity variation of the SA sensors at the two stress levels of 4.8MPa and 14.4MPa compared with their initial values that are assumed to be the value measured at the time of 0.1 minute. It can be found that during the 30 minutes the depolarization process is fast in the beginning and then become slow. At the stress level of 14.4MPa, the depolarization completed within 15 minutes. While at the stress level of 4,8MPa, it seems to take longer time. At this stress level, the sensitivity dropped by 3% during the first 15 minutes and dropped only 2% in the consecutive 15minutes. Thus it can be concluded that for concrete building structures

with pre-stress of interest, 15 minutes is the proper time for exhausting the most part of the depolarization process. In the following tests, 15 minutes is used as the waiting time before the calibration test starts.

Figure 8. Sensitivity variation of SA vs. holding

4.3 Influence of Pre-stress

Arranged in cascades in step of 4.8MPa within 4.8-24MPa, the uniaxial compressive pre-stress was applied on the aligned sensors of SA1 and SA2 (see Fig.12). In each step the stress was held for 15 minutes and at the end of the step a group of ever-increasing alternating stress was superimposed on the pre-stress for evaluating the sensors’ sensitivities. Fig.13 shows an example of the alternating stress which is the same as depicted previously. The sensitivities of the SA sensors at each load level are drawn in Fig.14. It can be seen that for the sensors, even at the pre-stress level of 24MPa, theirs sensitivities drop no more than 3%. This means that for the SA sensor the influence of the pre-stress on the piezoelectric coefficient is negligible. This finding is very appealing since the initial static load differs greatly in building structures according to location and it is difficult to be identified accurately in advance.

Figure 9. Applied stress for evaluating the influence of pre-stress time 科

学出版社


www.abook.cn

·248·

Figure 10. SA sensitivity vs. pre-stress

5 CONCLUSIONS

From the work carried out in this paper, the following conclusions can be drawn. The structure of the smart aggregarte and fabrication technique renders the proposed SA sensor with higher stability in performance and higher consistence in its fabrication quality in comparison with the cement-based SA sensor. When embedded in building structures, the initial static stress won’t influence the sensor’s sensitivity. As presented in the tests, under pre-stress of up to 24MPa, the sensor’s sensitivity drops no more than 3%. With load amplitude lower than 2.25MPa, nonliearity in its piezoelectric response is not observed.

REFERENCES

Alguero, M., Cheng, B. L., Guiu, F., Reece, M. J., Poole, M., and Alford, N. (2001). Degradation of the d33 piezoelectric coefficient for PZT ceramics under static and cyclic compressive

loading. Journal of the European Ceramic Society, 21:10-11, 1437-1440.

ANSI, and IEEE. (1987). IEEE Standard on Piezoelectricity., ANSI/IEEE std.

Calderon-Moreno, J. M. (2001). Stress induced domain switching of PZT in compression tests. Materials Science and Engineering A, 315:1-2, 227-230.

G, Y., Liu, S., Ren, W., and Mukherjee, B. K. (2001). Effects of uniaxial stress on the piezoelectric, Dielectric, and mechanical properties of lead zirconate titanate piezoceramics. Ferroelectrics, 262, 207-212.

Gu, H., Song, G., Dhonde, H., Mo, Y. L., and Yan, S. (2006). Concrete early-age strength monitoring using embedded piezoelectric transducers. Smart Materials and Structures, 15:6, 1837.

Jayant, S., and Inderjit, C. (2000). Fundamental Understanding of Piezoelectric Strain Sensors. Journal of Intelligent Material Systems and Structures, 11, 246 - 257.

Li, Z., Yang, X., and Li, Z. (2006). Application of Cement-Based Piezoelectric Sensors for Monitoring Traffic Flows. Journal of Transportation Engineering, 132:7, 565-573.

Song, G., Gu, H., Mo, Y. L., Hsu, T., Dhonde, H., and Zhu, R. R. H. Health monitoring of a concrete structure using piezoceramic materials., 108-119.

Song, G., Olmi, C., and Gu, H. (2007). An overheight vehicle-bridge collision monitoring system using piezoelectric transducers. Smart Materials and Structures, 16:2, 462-468.

YANG, X., Li, Z., DING, Y., and LI, Z. (2005). Test on Sensor Effect of Cement Matrix Piezoelectric Composite. TRANSACTIONS OF TIANJIN UNIVERSITY, 11:2, 133-136.

Yimnirun, R., Laosiritaworn, Y., and Wongsaenmai, S. (2006). Effect of uniaxial compressive pre-stress on ferroelectric properties of soft PZT ceramics. Journal of Physics D: Applied Physics, 39:4, 759.


REINFORCEMENT OF DAMAGED FRAME JOINT WITH CARBON FIBER

Zhengfei Chang 1,2, Zhiyong Yang 1, Jun Hu1,Gang Wang1,2 1College of civil engineering and architecture, WHUT, Wuhan 430070, P.R.China 2Shenzhen Metallurgical Engineering CO.LTD, Shenzhen 518054, P.R.China

Abstract: Most structure is damaged in earthquake areas, and research on seismic performance of damaged RC frame joints is limited. The study will take load in end of column with force-displacement mixed control on the condition of changing axial compression ratio by Quasi-static test of cyclic loading. It focuses on the analysis and comparison of indexes which have important effect on the seismic performance, such as hysteretic behavior, strength degradation, stiffness degradation, ductility, energy dissipation capacity, failure modes between undamaged joint and column-damaged joint after reinforcement with carbon fiber, and obtains the effect of axial compression ratio for frame joint strengthened with carbon fiber. Research results will be applied to reinforcement of damaged structure, has a broad business value. Keywords: Reinforcement, damaged, frame joint, carbon fiber

1 INTRODUCTION

The joints of the reinforced concrete frame are such important parts in the structure, which connect the beams and columns transferring and distributing the internal forces, mediating the deformation of the components, and assuring the integrity of the structure, and they are important elements in the frame structures. Several passed earthquakes have proved the weakness of the nodes that if the nodes are destroyed, the whole structure would collapse. China is a country with frequent earthquakes that there are a lot of insufficient seismic performance of reinforced concrete frame structures.In order to reduce earthquake damage and loss, it needs to repair and retrofit the current building, espically retrofitting the frame joints.Research of reinforcement of reinforced concrete has an economic and social significance.

At present, research on seismic performance of damaged RC frame joints is limited both at home and abroad. Most of these studies have only 2-3 specimens, and fewer factors are taking into account with undamaged joint which cause the data discrete. Study on seismic performance of reinforced damaged node is a blank stage. It will take load in end of eleven columns with force-displacement mixed control on the

condition of changing axial compression ratio by Quasi-static test of cyclic loading. It focuses on the analysis and comparison of indexes which have important effect on the seismic performance, such as hysteretic behavior, strength degradation, stiffness degradation, ductility, energy dissipation capacity, failure modes between undamaged joint and column-damaged joint after reinforcement with carbon fiber, and obtains the effect of axial compression ratio for frame joint strengthened with carbon fiber.

2 TEST PROGRAMME

2.1 Specimen Design

SJ1 is undamaged specimen, SJ1 is damaged specimen.

Figure 1. Specimen 科

学出版社


www.abook.cn

·250·

Table 1. Sample parameters

Specimen number Dimensions(mm)

SJ1-SJ2 Beam:200×200 Column:300×300

Table 2. Reinforcement of specimen

Beam

reinforcement

Column

reinforcement

Beam(Column)

Stirrup

Up:2B16

Behind:2B18

6B16

A6@100

Figure 2. Reinforcement of beam-column joints

2.2 Loading System

Constant invariant vertical load is imposed on end of column in advance by hydraulic jack which is unchanged during the test process.For horizontal loading, the mixed loading of the control force and displacement control method is used in the test process as Fig.3.

Figure 3. Loading system

2.3 Carbon Fiber Reinforcement

Put a layer of carbon fiber cloth in the corner of

beam-column joints, which extends along the column of 300mm and the beam 500mm. Put two layers of carbon fiber cloth on core area of nodes in horizontal and vertical direction, and extends horizontally beam 200mm, and extends vertically column 300mm. Put cloth in beam-column range with width of 100 u-shaped hoops @100 as Fig.4.

Figure 4. Carbon reinforcement

3 TEST RESULT AND DISCUSSION

3.1 Hysteresis Curve and Skeleton Curve

From Fig.5 and Fig.6, Reinforcement of specimen’s curve in load control stage pinch phenomenon more clear, but in loaded late shrink phenomenon gradually is improve, which is due damage of specimen, though it is reinforced with carbon fiber, carbon fiber is elastic material which cannot play role in early. When test is to late load stage, carbon fiber cloth began play role, and made delay back curve became full. In the same case there is a clear role on improving hysteresis curve of axial compression ratio of pinching shrinkage phenomena. Direct hysteresis curve of reinforced specimen undamaged saturation level is superior to reinforcement of damaged specimen.

(a) SJ1 Figure 5. Hysteresis curve of reinforcement

·251·

(b) SJ2

Figure 5. Hysteresis curve of reinforcement (continued)

(a) SJ1

(b) SJ2

Figure 6. Skeleton curve of reinforcement

3.2 Ductility

Ductility is the structure or components in bearing capacity of deformation capacity without the significant drop in. Ductility is generally represented by ductility coefficient,

structure of ductility coefficient is an important indicator of measuring seismic performance of structures. In this test, the use of displacement ductility of ductility coefficient is to measure the specimens, that end-column limit of displacement and yielding displacement expressed as ratios of ends of cylinder displacement ductility coefficient:

/u y

Table 3. Ductility

Specimen number Ductility

SJ1 5

SJ2 4

Ductility of undamaged joint is better than damaged joint about 25% after reinforcement.

3.3 Energy Dissipation Capacity

Energy dissipation coefficient is to measure the energy capacity of the specimen.

ABC ADC

OBE ODF

S SE

S S

Hysteretic curve is surrounded by SABC and the SADC area, SOBE, and SODF represents the area of the triangle shown in Fig.7.

Figure 7. Calculation diagrams

(a) SJ1

Figure 8. Calculation diagrams 科

学出版社


www.abook.cn

·252·

(b) SJ2

Figure 8. Calculation diagrams.(continued)

Table 4. Energy dissipation capacity

Specimen number Energy dissipation capacity SJ1 3.2 SJ2 2.31

Energy dissipation capacity of undamaged specimen is significantly better the damaged one after reinforcement about 38%.

3.4 Crack and Damage

When load is up to 120kN, which is recorded as yielding load and horizontal displacement value is 23.89 mm as yielding displacement. Changing to displacement control, it continues to increase load, end-column appears the plastic hinge. After the end of the test, the surface of carbon fiber cloth brush, the presence of carbon fiber

(a) SJ1

(b) SJ2

Figure 9. Crack and damage

cloth improves the core properties of shear capacity Failure mode of SJ1 is belongs to column and core damage.damage.

Early in the load, hysteresis curve is a linear relationship, when the delay loop is not obvious. When load is up to 80kN, which is recorded as yielding load and horizontal displacement value is 19.83mm as yielding displacement. Changing to displacement control, it continues to increase load, core and cylinder end cracks is suppressed, crack shift to beam as beam of plastic hinge appears. After the end of the test, the surface of carbon fiber cloth brush, it is found that the core of cross inclined crack before damage occurs. Due to the presence of carbon fiber cloth, it improves the core area and end-column resistance to damage. Failure mode of SJ2 is belongs to beam damage.

4 CONCLUSIONS

Undamaged specimen is superior to damaged one in yield load, maximum load, and ultimate load, strength degradation, stiffness degradation, and energy dissipation capacity after reinforcement, but the late is superior to the former in yielding displacement limit, maximum displacement. There is little difference between ductility.

The damage is more obvious impact on the carbon fiber reinforcement, it reduces reinforcement of the bearing capacity of specimens at all stage, and all phases of displacement have increased. It is weakening the strength of stiffness degradation and energy dissipation capacity.

REFERENCES

Chris P. Pantelides. (2008). Seismic Rehabilitation of Reinforced Concrete Frame Interior. Beam- Column Joints with FRP Composites. Journal Of Composites For Construction, 435-445.

M J Shannag, M.A.Alhassan. (2005). Seismic Upgrade of Interior Beam-Column Subassemblages with High-Performance Fiber-Reinforced Concrete Jacket. ACI Structural Journal, 102:131-139.

·253·

Saleh.H.Alsayed1, Tarek H. Almusallam. (2010). Seismic Rehabilitation of Corner RC Beam- Column Joints Using CFRP Composites. Journal Of Composites For Construction., 681-692.

Saptarshi Sasmal, Balthasar Novák. (2011). Numerical analysis of fiber composite - steel

plate upgraded beam - column subassemblages under cyclic loading. Composite Structures, 599-610.

Seyed. S. Mahini. (2010). Strength and ductility of FRP web-bonded RC beams for the assessment of retrofitted beam-column joints. Composite Structures, 1325-1332.

科

学出版社


www.abook.cn


STUDY ON DISTRIBUTION OF STEEL CORROSION PRODUCTS IN

CONCRETE STRUCTURES

Yingyao Wu, Yuxi Zhao, Weiliang Jin ¹ Institute of Structural Engineering, Zhejiang University, Hangzhou 310058, P.R. China

Abstract: This study investigated a reinforced concrete specimen that had deteriorated in an artificial environment for two years. Samples were made from the specimen and the steel/concrete interface and corrosion-induced cracks were observed by SEM operated in BSE mode and analysed by EDX. Rust distributions at steel/concrete interface and in adjacent concrete were observed and the mechanism was discussed. The thickness of corrosion-filled paste (CP) and its variation with the growth of steel corrosion reveal that the penetration of corrosion products into concrete and the forming of corrosion layer proceed simultaneously. Before concrete surface cracking during the corrosion process, the rust will neither fill the cracks nor penetrate the concrete adjacent to cracks. Based on the observations, a two-stage model, instead of the three-stage model, is proposed to describe the concrete cracking process induced by steel corrosion. Keywords: Steel reinforced concrete, SEM, rust, interfaces, cracks

1 INTRODUCTION

Corrosion of steel in concrete is the major cause of the durability problem of reinforced structures. The corrosion of steel produces a pressure on the surrounding concrete as the corrosion product is about 2-6 times the volume of the original steel, and eventually causes cracks at steel/concrete interface. Once the cracks penetrate the concrete cover, corrosion of the steel is accelerated, leading to serious damages like spalling of the concrete cover.

The three-stage model proposed by Liu and Weyers (1998) has been accepted widely to quantitatively describe the concrete-cracking process induced by steel corrosion. In this model, the process can be taken as the three stages, i.e. 1) the stage for rust to fill the porous zone in adjacent concrete, 2) the stage for rust to create expansive pressure on the surrounding concrete, and 3) the stage for rust to fill the cracks.

The second stage has been widely studied both in experiment research and analytical models. For the first stage, the previous studies by the authors (Asami and Kikuchi, 2003; Duffóet al., 2004; Chitty et al., 2005; Care et al., 2008; Jaffer and Hansson, 2009; Wong et al., 2010; Michel et al., 2011; Zhao et al., 2012) have observed the area of concrete around the rebar penetrated by corrosion products, so called ‘corrosion-filled paste (CP)’ according to the the

work by Wong et al. (2010). However, more quantitatively experimental research is needed to study the thickness of CP and its variation with the growth of steel corrosion. As for the third stage, although the rust is assumed to fill the corrosion-induced cracks in the three-stage model, the experimental verification of this assumption is, in fact, not sufficient yet.

In this study, a reinforced concrete specimen that had deteriorated in an artificial environment for two years was investigated. The rust distributions at the steel/concrete interface and corrosion-induced cracks were observed using a scanning electron microscope (SEM). This study can reveal the progress and mechanism of concrete cracking induced by steel corrosion and thus verifies the three-stage model proposed by Liu and Weyers (1998).

2 EXPERIMENTAL PROGRAM

2.1 Reinforced Concrete Specimen

A reinforced concrete specimen with the dimensions shown in Fig.1 was used in this study. The specimen contained three ribbed bars with a nominal diameter of 16 mm. The rebars were used as received, and no efforts were made to remove the existing millscale. A low cover

·273·

depth of 20 mm was adopted to reduce the time for corrosion initiation.

Figure 1. Schematic of the reinforced concrete specimen (dimensions are in mm)

The mixture proportions of the concrete are

shown in Table 1. The concrete contained a commercial calcium nitrite based corrosion inhibitor. The 28-day compressive strength of the concrete was 56 MPa as measured on 150mm cubes.

Table 1. The mixture composition of concrete specimens (kg/m3)

Cement 126

Blast-furnace slag 168

Fly ash 126

Sand 735

Aggregate 1068

Water (l/m3) 145

Ratio of water/binder 0.345

Water-reducing admixtures 5.04

Corrosion inhibitor 8.4

2.2 Curing and Exposure History

The specimen was covered in damp hessian immediately after casting and sprayed with water once every day for two weeks. After curing, the side and bottom surfaces of the panel were surface treated with epoxy resin and a polyurethane coating to ensure that chloride predominantly penetrated through the top cover, with minimal penetration through the other faces of the panel. The concrete panel was then subjected to alternate wetting and drying cycles. Each test cycle lasted for 3 days and consisted of spraying the blocks with 3.53 wt. % sodium chloride solution for 4 hours and subsequently allowing them to dry at 40 °C for the remainder of the time to complete one full cycle.

After 2 years of exposure, longitudinal cracks were observed on the top face of the specimen, running approximately parallel to the reinforcing bar, as shown in Fig.2. The specimen was then removed from the environmental chamber for further testing.

Figure 2. Schematic diagrams of the cut specimens

2.3 Sample Preparation

The specimen was cast into a low-viscosity epoxy resin to minimise any artificial damage during the sample preparation process. Then, it was carefully cut to extract the corner and middle rebars, as shown in Fig.2. The cut panels were labelled L, M and R, representing the sections containing the left-corner, middle and right-corner rebar, respectively.

Each panel was then sectioned sequentially to produce a series of 10 mm-thick cross-sectional slices. Cutting was carried out using an abrasive cutter and diamond blade suitable for hard brittle materials. An example of a slice is shown in Fig.3(a). Slices from panel L and R with different corrosion and cracking degree were chosen for this study. The locations of the slices are indicated in Fig.2. The samples were further prepared for the SEM observation by trimming down slices to smaller blocks including the complete steel/concrete interface and the cracks (Fig.3(b)). The samples were all polished using a lapping and polishing machine. To prevent further corrosion, these samples were kept in a dry environment (relative humidity less than 30 %) before observation.

(a) (b)

Figure 3. Sample preparations. (a) Slice L-9. (b) Sample for SEM

2.4 Observation and Measurements

A scanning electron microscope (Hitachi 科

学出版社


www.abook.cn

·274·

S3400N) operated in the backscattered electron (BSE) mode was used to study the steel/concrete interface and cracks. The energy spectrum analysis (EDS) analysis was applied to study the content of Fe and O in the observation areas.

(1) Steel/concrete interfaces The steel/concrete interfaces were observed

using SEM in BSE mode. The iron Oxides accumulated at steel/concrete interface was investigated and the contents of Fe and O were analysed by EDS. Approximately 80 points evenly distributed around the rebar were observed to cover the entire steel/concrete interface. The thicknesses of corrosion layer (CL) and corrosion-filled paste (CP) were measured at every point observed.

(2) Rust distributions in corrosion-induced cracks

The corrosion-induced cracks were observed using SEM in BSE mode. The content of Fe was analysed by EDS to investigate rust distributions in cracks.

3 RESULTS

3.1 Rust Distributions at Steel/concrete Interfaces

Fig.4(a) shows a BSE image at a steel/concrete interface, which shows the most typical rust distributions at the steel/concrete interface. The line distribution of Fe and O across the steel/concrete interface was analysed by EDS along an analytical line as shown in Figure 4(a). In Fig.4(b), the horizontal axis represents the distance from the starting point of the analytical line, while the vertical axis is the counts of photoelectrons per second, reflecting the contents of Fe and O respectively. The EDS analysis indicates five regions along the analytical line, i.e. concrete (C), corrosion-filled paste (CP), millscale (MS), corrosion layer (CL) and steel (S) respectively. It can be noted that the steel corrosion products accumulate between the steel and the millscale (inside the millscale as CL) as well as penetrate into adjacent concrete (outside the millscale as CP). The formation of these five regions is discussed below.

In the presence of chloride ions in concrete, reaction between ferrous ions (Fe2+) formed in the anodic region (steel) and hydroxyl (OH-) formed in the cathodic region (concrete) is believed to involve soluble intermediary ‘green complexes’ according to the previous study by Sagoe-Crentsil and Glasser (1993), which will

break down once meet OH- and undergo further oxidation to form solid precipitates. The possible reactions during steel corrosion are shown below (by assuming the ‘green complexes’ to be ferrous chloride (FeCl2) according to the study by Shi et al., (2009)): Anode: Fe - 2e- Fe2+

Fe2+ + 2Cl- FeCl2 Cathode: O2 + 2H2O + 4e- 4OH- Overall: 2Fe + 4Cl- + O2 + 2H2O 2Fe(OH)2 + 4Cl-

(a)

(b)

Figure 4. Rust distributions at steel/concrete interface. (a) BSE image and an analytical line across the interface. (b) The distributions of Fe and O analyzed by EDS along the analytical line

The millscale of a rebar embedded in concrete was found to be porous in the work by Jaffer and Hansson (2009), allowing soluble species involved in the above reaction to move through it from anodic to cathodic regions or the opposite way. The ions reaction and migration process can be simplified and represented in Fig.5. In Scenario 1, hydroxyl (OH-) formed in concrete penetrates through the millscale into the surface of steel. When OH- meets FeCl2, FeCl2 breaks down and further oxidate to Fe(OH)2. This scenario briefly describes the process that the steel corrosion accumulates between steel and millscale, forming CL. While in Scenario 2, as shown in Fig.5, the soluble intermediary ‘green complexes’ FeCl2 moves through the

·275·

porous millscale away from the steel surface to the adjacent concrete, then FeCl2 and OH- reacts and forms Fe(OH)2, which is the same reaction as in Scenario 1. The precipitates Fe(OH)2 fill in the porous cement paste, result in the formation of CP.

Figure 5. Schematic of ions migration and reaction during steel corrosion in the presence of chloride ions in concrete

It can be found that the millscale preserved well after the growth of the corrosion layer. This is because iron (Fe) in steel is more easily to be oxidized than ferric ion in millscale under low oxygen supplementary in concrete. Therefore, generally speaking, the millscale of the rebar is hardly to be further oxidized before surface cracking of the concrete cover.

3.2 Distribution of Corrosion-filled Paste (CP)

Samples L-9, R-5 and R-7 spanning a range of degrees of corrosion and damage were analysed using SEM in BSE mode. Thicknesses of the corrosion layer (CL) and the corrosion-filled paste (CP) were measured at approximate 80 points evenly distributed around the rebar to cover the entire perimeter.

Although the data collected by the above observation and measurement are scattered, the statistical result of data from all samples observed displays some regularity, by grading these scattered data into several groups as shown in Fig.6. The abscissa in Fig.6 illustrates the range of CL in each group, while the ordinate is the average thickness of CP in each group. The result reveals that the average thickness of CP increases along with the growth of CL, this tendency does not change until the thickness of CL exceeds 300 μm.

Figure 6. Average thickness of corrosion-filled paste (CP) for different thickness of corrosion layer (CL)

Such a regularity found above is different with the assumption of the three-stage model proposed by Liu and Weyers (1998), which assumes that the corrosion products will never create expansive pressure on the surrounding concrete before porous zone around the steel/concrete interface is fully filled. The result in this study, however, reveals that maybe the penetration of corrosion products into the porous zone of concrete and the forming of corrosion layer at steel/concrete interface, in fact, proceed simultaneously after the initiation of steel corrosion. The mechanism of this fact has been discussed in Section 3.1.

It should be recognized that, as samples involved in this study are limited, so to further confirm the conclusion in this section, more experiment should be carried out with much more samples spanning a larger range of degrees of corrosion taken into account.

3.3 Rust Distribution in Cracks

In the previous study by the authors (Zhao et al., 2012), the rust distributed in corrosion-induced cracks in electro-chemically corroded reinforced concrete specimens was observed by SEM. The observation reveals that rust does not fill the corrosion-induced cracks. However, in electron- chemically accelerated corrosion experiment, the high speed of steel corrosion may result in a lack of time for corrosion products to fill cracks. The specimen investigated in this study, by contrast, had deteriorated in an artificial environment for two years, thus presenting a more factual rust distribution in corrosion-induced cracks.

Fig.7(a) illustrates the BSE images of rust distribution in corrosion-induced cracks obtained from sample L-9 in which the cracks penetrate concrete cover, while Fig.8(a) shows the situation of sample R-7 with only the inner cracks. It can be seen from Fig.7(a) that the 科

学出版社


www.abook.cn

·276·

concrete at both sides of the crack are brighter which corresponds with the peaks of content of Fe in the EDS analysis in Fig.7(b), indicating that corrosion products penetrate concrete adjacent to cracks but the crack itself is not filled. This phenomenon is the same as the observation reported in the previous study by Zhao et al., (2012). For the situation of the inner crack in sample R-7, the BSE image in Figure 8a and the EDS analysis in Figure 8b show no rust filling in the crack and no rust lining the surface of the cracks either, as the concentration of Fe keeps low along the entire analysis line across the inner crack.

(a)

(b)

Figure 7. Rust distributed in a crack penetrating concrete cover in sample L-9. (a) BSE image and an analytical line across the crack. (b) The distribution of Fe analyzed by EDS along the analytical line

Comparing Fig.7 with Fig.8, it can be seen that the rust will not penetrate the concrete at sides of a crack until the crack reaches the concrete surface. In fact, only after the crack penetrates concrete cover during the acceleration corrosion process, the outer solution can ingress into the crack, then some corrosion products

dissolve in the solution and are carried away from the rebar by the solution, lining and penetrating the edges of cracks.

(a)

(b)

Figure 8. Rust distributed in an inner crack in sample R-7. (a) BSE image and an analytical line across the crack. (b) The distribution of Fe analyzed by EDS along the analytical line

According to the discussions above, it can be reckoned that before concrete surface cracking, the rust will neither fill the cracks nor penetrate the concrete adjacent to cracks. Therefore, the filling of corrosion-induced cracks with rust corresponding to Stage 3 in the three-stage model proposed by Liu and Weyers (1998) needs not to be considered in the corrosion-induced concrete surface cracking model. Two-stage mode, i.e. the stage for rust to fill the porous zone around the steel/concrete interface and the stage for rust to create expansive pressure on the surrounding concrete, is proposed to predict the corrosion-induced concrete surface cracking. What’s more, in this two-stage model, the forming of the penetration of corrosion products into the porous zone of concrete and the forming of corrosion layer at steel/concrete interface should be considered simultaneously at the initiation of steel corrosion.

·277·

4 CONCLUSIONS

This study investigates a reinforced concrete specimen that has deteriorated in an artificial environment for two years. Rust distributions at steel/concrete interface and corrosion-induced cracks are observed by SEM operated in BSE mode. The following conclusions have been drawn from this study:

1) Steel corrosion in the presence of chloride ions in concrete is a complex progress. The steel corrosion products accumulate between the steel and the millscale (inside the millscale as CL) as well as penetrate into adjecent concrete (outside the millscale as CP), due to ions migration and reaction.

2) The result in this study shows that maybe the penetration of corrosion products into the porous zone of concrete and the forming of corrosion layer at steel/concrete interface proceed simultaneously after the initiation of steel corrosion.

3) Before concrete surface cracking during the corrosion process, the rust will neither fill the cracks nor penetrate the concrete adjacent to cracks. Therefore, the filling of corrosion- induced cracks with rust needs not to be considered in the corrosion-induced concrete surface cracking model.

4) Two-stage model, instead of the three- stage model, is proposed to predict the corrosion-induced concrete surface cracking. What’s more, the stage for rust to fill the porous zone around the steel/concrete interface and the stage for rust to create expansive pressure on the surrounding concrete should be considered simultaneously but not separately.

ACKNOWLEDGEMENTS

Financial support from the Fundamental Research Funds for the Central Universities of China with Grand No. 2012FZA4018 is gratefully acknowledged.

REFERENCES

Alexander Michel, Brad J. Pease, Mette R. Geiker, Henrik Stang and John Forbes Olesen.

(2011). Monitoring Reinforcement Corrosion and Corrosion-induced Cracking Using Non-destructive X-ray Attenuation Measurements. Cement and Concrete Research, 41: 11, 1085-1094.

G.S. Duffó, W. Morris I. Raspini and C. Saragovi. (2004). A Study of Steel Rebars Embedded In Concrete During 65 years. Corrosion Science, 46: 9, 2143-2157.

H.S. Wong, Y.X. Zhao, A.R. Karimi, N.R. Buenfeld and W.L. Jin. (2010). On The Penetration of Corrosion Products From Reinforcing Steel Into Concrete Due to Chloride-induced Corrosion. Corrosion Science, 52: 7, 2469-2480.

K. Asami and M. Kikuchi. (2003). In-Depth Distribution of Rusts on a Plain Carbon Steel and Weathering Steels Exposed to Coastal–industrial Atmosphere for 17 years. Corrosion Science, 45: 11, 2671-2688.

K.K. Sagoe-Crentsil and F.P. Glasser. (1993). ‘Green Corrosion Products’, Iron Solubility and the Role of Chloride in the Corrosion of Steel at High pH. Cement and concrete research, 23: 4, 785-791.

S. Care, Q. T. Nguyen, V. L’Hostis and Y. Berthaud. (2008). Mechanical Properties of the Rust Layer Induced by Impressed Current Method in Reinforced Motar. Cement and Concrete Research, 38: 8-9, 1079-1091.

S.J. Jaffer and C.M. Hansson. (2009). Chloride-induced Corrosion Products of Steel in Cracked-concrete Subjecte to Different Loading Conditions. Cement and Concrete Research, 39: 2, 116-125.

Shi Hui-sheng, Guo Xiao-lu and Zhang He. (2009). Influence of Chloride Anion on the Corrosion of Steel Bar in Concrete. Cement technology, 5, 21-25. (in Chinese)

W.J. Chitty, P. Dillmann, V. L Hostis and C. Lombard. (2005). Long-term Corrosion Resistance of Metallic Reinforcements in Concrete—a Study of Corrosion Mechanisms Based on Archaeological Artefacts. Corrosion Science, 47: 6, 1555-1581.

Y.P. Liu and R.E. Weyers. (1998). Modeling the Time-to-corrosion Cracking in Chloride Contaminated Reinforced Concrete Structures. ACI Materials Journal, 95: 6, 675-681.

Yuxi Zhao, Jiang Yu, Yingyao Wu and Weiliang Jin. (2012). Critical Thickness of Rust Layer at Inner and out Surface Cracking of Concrete Cover in Reinforced Concrete Structures. Corrosion Science. 59: 7, 316-323.

科

学出版社


www.abook.cn


INFLUENCES OF OBSTRUCTION ACTION OF REINFORCING BAR ON

CHLORIDE DIFFUSION AND CORROSION INITIATION

Chengming Lan1, Jie Yuan2, Hui Li3 ¹ National Center for Materials Service Safety, University of Science and Technology Beijing, Beijing 100083, P.R. China ² School of Transportation Science and Engineering, Harbin Institute of Technology, Harbin 150090, P.R. China 3 Research Cener of Structural Monitoring and Control, School of Civil Engineering, Harbin Institute of Technology, Harbin 150090, P.R. China Abstract: Chloride-induced corrosion of reinforcing steel causes reinforced concrete structures performance deterioration over time in marine environment. Few researches concern the influences of obstruction action of reinforcing bar on chloride diffusivity in concrete and the prediction results about corrosion initiation or service life may be non-conservative. In this paper, the influencing factors of chloride diffusion coefficient and chloride binding isotherms in concrete are reviewed comprehensively. The two-dimensional FEM models considering insulation boundary of bar for chloride diffusion are built. Chloride concentration field and contour at corner of square cross-section and in front of reinforcing bar are present. Finally, the influences of boundary conditions, dc (concrete cover to reinforcing bar diameter ratio) and obstruction action of reinforcing bars on corrosion initiation predictions are discussed thoroughly. Keywords: Chloride diffusion, corrosion, service life, obstruction action, binding isotherm 1 INTRODUCTION

Large infrastructures including bridge and offshore structures have been increasingly built in many countries. These structures are often exposed to severe marine environments. However, reinforced concrete structures suffer from corrosion of reinforcing bar due to chloride ingress (Oh and Jang, 2003). Corrosion damage leads to future structural distress due to the loss of the reinforcing bar cross-sectional area as well as loss of bond along the steel-concrete interface, concrete cracking and delamination.

The durability is a major concern in the design of concrete structures that are exposed to marine environment. Chloride ingress implies a complex interaction between physical and chemical processes. However, under various assumptions, this phenomenon can be simplified to a diffusion problem (Tuutti, 1982 a). Acturally, the chloride penetration into concrete is influenced by many parameters such as temperature, humidity, types of cement, and mixture proportion. Chloride binding also has an important effect on chloride penetration and hence, on the time to corrosion initiation. The

conventional diffusion analysis has neglected the existence of reinforcing bars in concrete. Furthermore, the prediction results about corrosion initiation or service life may be non-conservative and this may be dangerous for RC structures. The penetrated chloride profiles in concrete, however, may be greatly influenced by the existence of steel bars because the chloride diffusion cannot occur through the reinforcing bars and chlorides are accumulated in front of reinforcing bars.

With this context, chloride diffusion model, influencing factors of chloride diffusion coefficient and chloride binding isotherms in concrete are reviewed, firstly. Two-dimensional FEM models for chloride diffusion considering insulation boundary of reinforcing bar and concrete are built to simulate the concentration profiles in front of reinforcing bars. The obstraction actions of reinforcing bar on chloride diffusion are illustrated by lots of simulated results. Finally, the influences of boundary conditions, dc and obstruction action on corrosion initiation predictions are discussed thoroughly. Based on the simulation results, many valuable conclusions are derived to guide

·255·

the actual design for reinforced concrete durability.

2 CHLORIDE DIFFUSION

The penetration of chloride ions into concrete is a complex process involving differenet transport mechanisms such ionic diffusion and convection (mainly in form of capillary suction). Frequently, diffusion is considered the main transport process of chlorides into concrete. Diffusion denotes the motion of chloride ions within the pore solution caused by their concentration gradient.

2.1 Diffusion Law

This process is often described by Fick’s 2st law, and is expressed as:

*f fc

C CD

t x x

(1)

with

* cc

b

e f

11

DD

C

C

(2)

where *

cD is the apparent diffusion coefficient

(m2/s) and b fC C is the “binding capacity”

of the concrete binder (m3 of pore solution / m3 of concrete) as defined by Nilsson et al. (1994),

cD is the effective diffusion coefficient when

the concentration is expressed in kilograms per cubic meter of concrete (m2/s), e is the

evaporable water content (m3 evaporable water/m3 concrete), and fC is the free chloride

concentration (kg/m3 of pore solution) at depth x (m), bC is the concentration of bound

chlorides (kg/m3 of concrete). The total, bound, and free chloride concentrations in concrete are related by

t b e fC C C . (3)

2.2 Diffusion Coefficient

The apparent diffusion coefficient *

cD is

obtained from a reference value of diffusivity which is a characteristic of concrete, 0D , by

consideration of relative humidity, h , temperature, T , age t , and free chloride concentration, fC , the general form of cD can

be expressed as shown in Eq. (4):

c 0 1 2 3 4 f( ) ( ) ( ) ( )D D f h f T f t f C (4)

where 0D is referred to as the reference

chloride diffusivity when all influencing factors assume values of unity, 1( )f h denotes the

influence of the relative humidity, 2 ( )f T

denotes the influence of temperature, 3 ( )f t

represents the influence of the age of concrete,

4 f( )f C indicate the influence of the free

chloride content. The 0D for a concrete cured

for 28 days can be obtained from its water to cement ratio ( w c ) of concrete by using the

equation recommended by American Concrete Institute (ACI) Committee 365 service life prediction model as (Bentz and Thomas, 2001) (see Figure 1)

0

2.40 12.06

0 28 days 10 w c

tD D . (5)

The results from Zhao (2001) and test results obtained by authors of this paper are also shown in Fig.1. It can be seen that ACI Committee 365 model gives a good prediction result.

Figure 1. ACI Committee 365 service life prediction model

According to Bažant and Najjar (1971), 1( )f h takes the form

14

1

1( ) 1

1 c

hf h

h

(6)

where h is the actual pore relative humidity, and ch is the humidity at which cD drops

halfway between its maximum and minimum values. Bažant and Najjar (1971) reported that

ch remains constant for defferent concretes or

cement pastes – i.e., 0.75ch .

Temperature has a significant effect on the rate of diffusion in more than one respect. First of all, temperature changes the adsorption heat,

科学出版社


www.abook.cn

·256·

and temperature increases the frequency of thermal vibrations of the diffusion, which can be taken into account by the Arrhenius’ law (Xi and Bažant, 1999)

2

ref

1 1( ) exp

Ef T

R T T

(7)

where refT is the reference temperature at

which 0D is determined (=296 ºK) and T is

the actual absolute temperature in concrete (ºK), R is the gas constant ( 8.314R J/(mol ºK)), and E is the activation energy of the chloride diffusion process (kJ/mol). E has been found to depend on water to cement ratio w c and on

the cement type (Page et al., 1981), as a value of 44.6 for OPC and 0.5w c (Xi and Bažant,

1999). According to Martín-Pérez et al. (2001), 3 ( )f t

takes the form

ref3 ( )

mt

f tt

(8)

where reft is the time of exposure at which 0D

has been evaluated ( ref 28t days), t is the

acuual time of exposure in days and m is the age reduction factor, dependent on the concrete mix proportions. In this paper, m is assumed to be 0.35 (Val, 2007).

The fourth factor, 4 f( )f C is to account for

the dependence of the chloride diffusivity on the free chloride concentration

4 f ion f( ) 1 ( )mf C k C (9)

where ionk and m are constants, ion 8.333k

and 0.5m (Xi and Bažant 1999).

3 BINDING ISOTHERM

Sandberg (1999) found that the amount of bound chloride increases as the concentration of hydroxide ions in pore solution decreases. The relationship between free and total chloride in concrete with a chloride and hydroxide ion gradient is almost linear. The relationship between free and bound ions over a range of chloride concentrations at a given temperature are known as the chloride binding isotherms. Three types of binding isotherm (i.e., linear, Langmuir, Freundlich binding isotherm) have been proposed to describe the relationship,

which are decribed in the following sections. No binding:

b 0C , b

f

0C

C

, *

c cD D . (10)

Linear isotherm: Tuutti (1982 b) proposed a linear binding

isotherm, which can be expressed as follows:

b fC C , b

f

C

C

, * cc

e1

DD

(11)

where is the slope of the line. Although several experimental works have reported that the relationship between bound and free chlorides is non-linear (Sergi et al., 1992; Nilsson et al., 1994; Tang and Nilsson, 1993; Tritthart, 1989), long-term exposure tests results also give a linear binding isotherm for various cements (Mohammed and Hamada 2003).

Langmuir isotherm: The Langmuir isotherm, derived from

physical chemistry, is assuming monolayer adsorption, which explains that the slope of the isotherm curve at high concentrations approaches zero. It is of the following form:

f b2b

f f f

*

c

2

e f

, 1 1

1

1

c

C CC

C C C

DD

C

(12)

where and are constants, which vary with the binder composition. These coefficients are obtained by non-linear curve-fitting of the experimental data and have no physical meanings (see Fig.2). Tang and Nilsson (1993) have stated that the relationship between bound and free chlorides is best described by the Langmuir isotherm when the concentration level of chloride in the pore solution is less than 1.773kg/m3.

Freundlich isotherm: The Freundlich binding isotherm can be

expressed as follows:

1bb f f

f

*

c1

f

e

,

1

1

c

CC C C

C

DD

C

(13)

where and are binding constants. Tang and Nilsson (1993) suggested that monolayer adsorption occurs at low concentrations (which is described better by Langmuir isotherm), but that adsorption becomes more complex at concentrations higher than 0.05M and is

·257·

described better by the Freundlich isotherm. The difference between the Freundlich and Langmuir isotherms is their behavior at high concentration.

Figure 2. Idealized binding isotherms for a concrete with 40% slag replacement level and water to binder ratio = 0.3 for two different exposure conditions: (a) s 0.5C M and (b)

s 2.5C M (Martín-Pérez et al., 2000)

4 NUMERICAL MODELS

Considering the dependence of the chloride diffusivity on various influencing variables as well as the time-varying boundary conditions, the nonlinear partial differential equation, Eq. (1), is solved by finite element software Comsol Mutiphysics. Fig.3 shows the cross-sections of the studied reinforced concrete components and the positions of reinforcing bars. It can be noted that chlorides come from two directions at the corner of square cross-section.

Two different boundary conditions were considered in the calculations: 0.5 and 2.5M chloride ions concentration solutions. The former condition simulates complete submersion

in sea water, whereas the latter simulates conditions more characteristic of marine structures in the splash zone or bridge decks exposed to de-icing salts (Martín-Pérez et al., 2000). Initial conditions were: zero chloride content within the concrete componets; actual pore relative humidity 0.75h . Values of the other parameters used in the analysis are: actual absolute temperature 296T ºK, concrete evaporable water content e 8% , and

reinforcing bar diameter 25d mm. It was assumed that the cement content was 450 kg/m3 of concrete, and the reference diffusion coefficient 12

0 5.105 10D m2/s (based on

Eq.(5)) for 0.32w c . For the free chloride

concentration was large, the Frundlich isother was used to express the binding capacity of concrete, and the values of the parameters showed in Fig.2 (b) were used in analysis.

Figure 3. Cross-section of reinforced concrete members and positions of reinforcing bars

5 RESULTS ANALYSIS

The free chloride concentrations contour in square cross-section for s 2.5C M and 50t

years are plotted in Fig.4. It can be seen that the free chloride concentrations increased significantly at the frontier of reinforcing bar towards boundary, especially at the corner of square cross-section which were influenced by two-directional diffusion. For free chloride cannot pass through reforcing bar, chloride gathered at frontier of reinforcing bar and then diffused around reforcing bar in concrete.

科

学出版社


www.abook.cn

·258·

Figure 4. Distribution and contour of free chloride concentrations in cross-section

Corrosion initiation predictions were obtained by assuming that the corrosion initiation of a reinforced concrete structure subjected to chloride ingress corresponds to the period until reinforcing steel bar depassivation. The free concentration threshold used was 0.09% by mass of cementitious materials, a lowerbound reported in Glass and Buenfeld (1995). The 0.09% by mass corresponds to 5.0 kg/m3 of pore solution based on the assumption of a 450 kg/m3 binder content and an evaporable water content of 8% by volume.

Fig.5 shows free chloride concentration evolve as increasing time with and without reinforcing bar obstruction action at point A, B and C in Fig.3 which were influenced by chloride come from two directions. It can be seen from Fig.5 that frontal chloride ions reach point B earlier than point A, for it was a little closer from boundary to point B than to point A. The free chloride concentrations increase rapidly at point A and concentrations are much larger than that at point B. For s 2.5C M, the free

chloride concentrations at point B reach the threshold firstly, but for s 0.5C M, the free

chloride concentrations at point A reach the threshold.

The free chloride concentrations at point C were influenced by chloride from one direction. The free chloride concentration at point B was larger than that at point C under the same condition

(a) s 2.5C M

(b) s 0.5C M

Figure5. Free chloride concentrations evolution at point A, B and C with and without obstruction action of reinforcing bars

and as the time increases, the influence became more significant. Also, it can be seen that as the time increases, the obstruction influence of reinforcing bar became more significant with the same boundary concentrations. As the surface chloride concentrations increase, the obstruction actions of reinforcing bars become indistinctive.

The corrosion initiation predictions of square cross-section at point C for different c d were

shown in Fig.6. For s 2.5C M, corrosion

initiation time decreased about 19% considering reinforcing bar obstruction with dc from 1.2

to 2.8 (see Fig.6(a)). There were little influences of dc on the rate of change of corrosion

initiation considering reinforcing bar obstruction or not at high chloride concentrations. For

s 0.5C M, corrosion initiation time decreased

from 40.61% to 34.10% considering reinforcing bar obstruction with dc from 1.2 to 2.8 (see

Fig.6(b)). It can be seen that the lower the chloride concentration was, the more visible the influence of reinforcing bar obstruction was. In

·259·

addition, the influence of dc on corrosion

initiation was more significant at low chloride concentration.

(a) s 2.5C M

(b) s 0.5C M

Figure 6. Corrosion initiation predictions of square cross-section at point C for different

dc

6 CONCLUSIONS

The influences of obstruction action of reinforcing bar on chloride diffusivity and corrosion initiation were investigated. The following conclusions were obtained from this case study:

(1) ACI Committee 365 model for reference chloride diffusion coefficient gives a good prediction result in comparison with the actural test results.

(2) Chlorides were accumulated in front of reinforcing bars in concrete and the free chloride concentrations increased in front of reinforcing bars, especially for bars at corner of square cross-section which were influenced by two dimensional chlrodies. The reinforcing bars at corner in square cross-section would be corroded

at first. (3) The obstruction action of reinforcing bars

on chloride diffusion was significant, especially for lower surface chloride concentrations. As the surface chloride concentrations increased, the obstruction actions of reinforcing bars became indistinctive. The influence of dc on corrosion initiation was more significant at low chloride concentration.

ACKNOWLEDGEMENTS

This study was financially supported by NSFC projects (Grant Nos. 50525823, 50538020, 50278029, 50974125, 51008302 and 51125017) and the National Fundamental Research Projects (Grant Nos. 2010CB226804 and 2011CB013604).

REFERENCES

Bažant, Z. P. and Najjar, L.J. (1971). Dying of concrete as a nonlinear diffusion problem. Cement and Concrete Research, 1: 461-473.

Bentz, E.C., and Thomas, M.D.A. (2001). Life-365: Service life prediction model and computer program for predicting the service life and life-cycle costs of reinforced concrete exposed to chlorides. American Concrete Institute, Cleveland, 7-11.

Glass, G.K., and Buenfeld, N.R. (1995). Chloride Threshold Levels for Corrosion Induced Deterioration of Steel in Concrete. In: Nilsson, L. O. Ollivier, J.P. (Eds.), Chloride Penetration Into Concrete. RILEM, Paris, 429-440.

Martín-Pérez, B., Pantazopoulou, S.J., and Thomas, M.D.A. (2001). Numerical Solution of Mass Transport Equations in Concrete Structures. Computers and Structures, 79: 13, 1251-1264.

Martín-Pérez, B., Zibara, H., Hooton, R.D., and Thomas, M.D.A. (2000). A Study of the Effect of Chloride Binding on Service Life Predictions. Cement and Concrete Research, 30: 8, 1215-1223.

Mohammed, T.U., and Hamada, H. (2003). Relationship between Free Chloride and Total Chloride Contents in Concrete. Cement and Concrete Research, 33: 9, 1487-1490.

Nilsson, L.O., Massat, M., and Tang, L. (1994). The Effect of Non-linear Chloride Binding on the Prediction of Chloride Penetration into Concrete Structures. In: V.M. Malhotra (Ed.), 科

学出版社


www.abook.cn

·260·

Durability of Concrete, ACI, Detroit, 469-486.

Oh, B.H. and Jang, B.S. (2003). Chloride Diffusion Analysis of Concrete Structures Considering Effects of Reinforcements. ACI Materials Journal, 100:2, 143-149.

Page, C.L., Short, N.R., and El Tarras, A. (1981). Diffusion of Chloride Ions in Hardened Cement Paste. Cement and Concrete Research, 11: 3, 395-406.

Sandberg, P. (1999). Studies of Chloride Binding in Concrete Exposed in a Marine Environment. Cement and Concrete Research, 29: 4, 473-477.

Sergi, W., Yu, S.W., and Page, C.L. (1992). Diffusion of Chloride and Hydroxyl Ions in Cementitious Materials Exposed to a Saline Environment. Magazine of Concrete Research, 44: 158, 63-69.

Tang, L., and Nilsson, L.O. (1993). Chloride binding Capacity and Binding Isotherms of OPC Paste and Mortars. Cement and Concrete Research, 23: 2, 247-253.

Tritthart, J. (1989). Chloride Binding in Cement

II. The Influence of the Hydroxide Concentration in the Pore Solution of Hardened Cement Paste on Chloride Binding. Cement and Concrete Research, 19: 5, 683-691.

Tuutti, K. (1982 a). Analysis of Pore Solution Squeezed Out of Cement Paste and Mortar. Nordic Concrete Research, 1:25. 1-16.

Tuutti, K. (1982 b). Corrosion of steel in Concrete. Swedish Cement and Concrete Research Institute. Stockholm.

Val, D.V. (2007). Factors Affecting Life-cycle Cost Analysis of RC Structures in Chloride Contaminated Environments. ASCE Journal of Infrastructure Systems, 13: 2, 135-143.

Xi, Y.P., and Bažant, Z. P. (1999). Modeling Chloride Penetration in Saturated Concrete. ASCE Journal of Materials in Civil Engineering, 11: 1, 58-65.

Zhao, S.C. (2001). Reliability Based Assessment and test research of durability of reinforced concrete structures. Ph. D. Dissertation. Dalian University of Technology. China.


EXPERIMENTAL AND NUMERICAL STUDY ON SEISMIC BEHAVIOR OF

CIRCULAR STEEL-CONCRETE-CFRP-CONCRET SOLID HYBRID

COLUMNS

Yongjun Liu, Dong Wang, Yu Wang

School of Civil Engineering, Shenyang Jianzhu University, Shenyang 110168, P. R. China Abstract: New circular Steel-Concrete-CFRP-Concrete hybrid column (referred as SCCC hybrid column or SCCC column for brevity), suggested by first author, consists of outer steel tube, inner FRP tube, annular concrete between two tubes, and core concrete encased in CFRP tube. In this paper, the seismic behavior of SCCC column and the effects of changes of CFRP tube’s diameter on seismic behavior of SCCC columns are investigated by both tests and finite element analysis. Three 260 mm diameter and 780 mm long SCCC columns were tested under constant axial load and reversed cyclic lateral load in the laboratory. In addition, finite element analysis using ABAQUS are conducted on the three SCCC columns under identical load. The results of finite element simulation are coincident with that of the tests. As the tube diameters of SCCC columns changed from 130mm to 160mm and 190mm, the value of ductility coefficient increased 3.40% and 15.50%. All the results demonstrate that the SCCC columns have excellent strength, ductility, energy absorption capacity, and the diameter of CFRP tube can influence the seismic performance of SCCC columns significantly. This study provides a basis for further research of SCCC column seismic behavior, and shall prompt the engineering apply of SCCC column. Keywords: SCCC column, seismic behavior, CFRP tube, numerical simulation 1 INTRODUCTION

Carbon Fiber Reinforced Polymer (CFRP) composite materials have experienced a continuous increase of use in civil and structural engineering around the world in the last decades. One particularly successful use of CFRP composite materials in structural engineering is repair and rehabilitation of existing reinforced concrete structures. With the rapid advances in practical applications and scientific researches, the use of CFRP composite materials in new construction has received more and more attention.

In recent years, several CFRP confining concrete columns, for example, CFRP tube encased concrete column, CFRP-concrete-steel double-skin column, and Concrete Filled CFRP-Steel Tube have been suggested for new constructions (Mirmiran et al., 1998; Teng et al., 2004; Zhao et al., 2005). Research results have shown that CFRP tubes are effective at improving stiffness, ductility, and load carrying capacity of concrete columns because CFRP tubes place the concrete under triaxial stress state and increase both the ultimate strength and

strain of concrete (Zhu et al., 2006; Teng et al., 2007; Gu et al.,2006).

In all above-mentioned existing hybrid columns, concrete is confined by outer CFRP tubes, thus, many challenges are brought to structural engineers. One of these challenges is the understanding and prediction of their fire behavior. It is well known that fire hazard cannot be avoided in offshore drilling platforms, high-rise building, etc. Preliminary studies indicate that the performance of CFRP under fire conditions is well below that of traditional materials (Ji et al., 2008; Liu et al., 2008). The mechanical properties of CFRP materials and resins tend to degrade very quickly after these materials reach their glass-transition temperature, Tg. As a result, the CFRP tubes will lose their confinement effect and the concrete columns may collapse.

How to improve the fire performance of structural members containing CFRP materials with less or even without fire protection is a key issue that needs to be addressed for the widespread application of CFRP. Under this circumstance, a new type of circular hybrid Steel-Concrete-CFRP-Concrete (SCCC) column 科

学出版社


www.abook.cn

·262·

(see Fig.1.) was suggested by authors for new construction to overcome shortcomings of existing hybrid columns and then to achieve more fire-resistant structures.

Figure 1. Cross section of SCCC column

In SCCC columns, the three constituent

materials are optimally combined to achieve several advantages than FRP tube directly exposed columns. Different from CFRP tube directly exposed columns, the most prominent feature of SCCC column is that the CFRP tube is protected by annular concrete. When a SCCC column is exposed to a fire, the temperature of CFRP may not exceed the glass-transition temperature of polymer matrix. Thus, CFRP tube may confine core concrete during the whole course of fires without need for additional fire protections. In addition, thanks to outer steel tube, the annular concrete can’t fall and spall during fires and is more effective and durable than such protection system as coating and plaster board.

As a kind of new structural member, many aspects of structural properties of SCCC column need to be studied. So far, some research works have been carried out to study static behavior under compression load and fire resistant behavior of SCCC column (Liu et al., 2010, Li, 2011). However, fairly limited work is about seismic, anti-impact, and anti-explosion property of SCCC columns. In this paper, the seismic behavior of SCCC column and the effects of changes of CFRP tube’s diameter on seismic behavior of SCCC columns are investigated by both tests and finite element analysis.

2 EXPERIMENTAL INVESTIGATION

2.1 Test Specimens

Three test specimens of SCCC column were made. Fig.2 illustrates the geometry and sizes of test specimens. They had a 350mm circular cross section and with a CFRP tube of 130mm,

160mm, or 190mm diameter respectively. The steel tubes were made of Q235 steel plates with a thickness of 2mm. The CFRP tubes of all columns had 3 plies of carbon CFRP sheets, the thickness of which is about 0.7mm. The fibers of FRP tubes were oriented in the hoop direction for full utilization of tubes as confinement reinforcement. Each column had two thick steel flange plates welded on top and bottom, through which it was fixed to the laboratory strong floor and sliding plate for testing.

Figure 2. Specimen geometric dimension

2.2 Material Properties

The steel tubes were obtained from Q235 steel plates with 500MPa yield strength. The stress-strain relationship was established by performing three tensile coupon tests. The concrete was prepared at the Structural Laboratory of Shenyang Jianzhu University. The mix consisted of normal portland cement and crushed limestone with 5 mm maximum size. The water cement ratio was 0.44 and the resulting concrete strength was 34.7MPa on the day of column testing. The capacity of the composite material was established to be approximately 1967MPa. 2.3 Test Setup, Instrumentation, and Loading Program Each column was tested under a constant axial compression and incrementally increasing lateral deformation reversals, simulating seismic loading. A 2000kN oil jack was used to apply constant axial compression throughout the test. Meanwhile, a 1000kN capacity servo-computer controlled MTS hydraulic actuator was placed horizontally for the application of lateral deformation reversals. Fig.3 illustrates the test setup.

·263·

(a) Photograph

(b)Schematic view

Figure 3. Test setup

2.4 Loading Regime

Each specimen was first axially loaded to 300kN, and then the specimens were then subjected to a reverse cyclic lateral displacement history applied in a number of incremental steps. The tests were performed using the displacement control loading regime, in which the initial lateral displacement Δ=6mm, then the lateral displacement was increased 0.5Δ in each step. Fig.4 depicts the lateral loading regime for SCCC column tests.

Figure 4. Loading regime

3 TEST RESULTS

3.1 Observed Behavior

For specimen SCCC-130, in which the diameter of CFRP tube is 130mm, when the lateral displacement reached 18mm, Bulge appeared at the steel tube 120mm away from the of the top of flange, which mean the steel tube started to yield. When the lateral displacement reached to 27mm, the bulge was quite obvious. When the lateral displacement increased to 36mm, the steel tube cracked along the welding line. Even so, the SCCC column could continue to work due to existing of the CFRP tube. Fig.5 shows the failure mode of specimen SCCC-130.

Figure 5. Failure mode of specimen

3.2 Hysteresis Curves

Hysteresis loops of all the specimens under the effect of low cyclic reciprocating load are shown in Fig.6, from which we can find: (1) As the increasing of CFRP tube diameter, the fuller hysteresis loop is, which means better seismic behavior. (2) When reaching the peak load, with increasing of cycle index, the bearing capacity reduces slowly, which shows an improvement in both energy dissipation ability and ductility.

(a) SCCC-130

Figure 6. Hysteresis curves 科

学出版社


www.abook.cn

·264·

(b) SCCC-160

(c) SCCC-190

Figure 6. Hysteresis curves(continued)

3.3 Skeleton Curves

Skeleton curves of the three SCCC column specimens are shown in Fig.7, from which we can find that the skeleton curves are in the shape of ‘S’, which means that specimens experience three stages of elastic, plastic and ultimate failure under the effect of low cyclic reciprocating load.

Figure 7. Skeleton curves

3.4 Ductility Coefficient

Displacement ductility coefficient of each

specimen can be calculated from skeleton curves according to following formula /u y , in

which u is the ultimate displacement; y is

the yield displacement. The ductility coefficient of SCCC-130, SCCC-160, and SCCC 190 equals 5.8, 6.0, and 6.7 respectively. As a result, following conclusion can be drawn that the ductility coefficient increases as the diameter of CFRP tube increase.

4 NUMERICAL INVESTIGATION

4.1 Element Type

The CFRP tube and steel tube of SCCC column were modeled using the general purpose shell element S4R in the commercially available software package ABAQUS, and concrete was modeled using 3D solid C3D8R element. Fig.8 shows finite element meshes of the four parts of a SCCC column.

(a) Core concrete (b) CFRP tube

(c) Annular concrete (d)Steel tube

Figure 8. Finite element meshes

·265·

4.2 Numerical Results

Fig.9 shows the calculated hysteresis curves using ABAQUS, and Fig.10 shows the comparison of experimental and numerical hysteresis curves. From Fig.9 and Fig.10, we can find that numerical hysteresis curves are similar with the experimental ones. The reason that there are some non-significant differences among experimental and numerical hysteresis curves may be due to the drawback of constitutive model of concrete.

(a) SCCC-130

(b) SCCC-160

(c) SCCC-190

Figure 9. Calculated hysteresis curves

(a) SCCC-130

(b) SCCC-160

(c) SCCC-190

Figure 10. Comparison of experimental and numerical hysteresis curves

5 CONCLUSIONS

Experimental and numerical study were undertaken to investigate the seismic behavior of a SCCC columns, with emphasis on the effects of changes of CFRP tube’s diameter on seismic behavior of SCCC columns. The following conclusions can be drawn:

科学出版社


www.abook.cn

·266·

(1)SCCC column suggested by first author has excellent strength, ductility, energy absorption capacity, and the diameter of CFRP tube is an important parameter that can influence the seismic performance of a SCCC column. (2)The general purpose software package ABAQUS can be used for further study of seismic behavior of a SCCC columns partially replacing model test.

ACKNOWLEDGEMENT

The authors are very appreciated to National Key Technology R&D Program in the 12th Five Year Plan of P. R. China under grant 2012BAJ11B02.

REFERENCES

Gu W, Zhao Y, Shang D. (2006). Load Carrying Capacity of Concrete Filled CFRP-steel Tubes under Axial Compression. Engineering Mechanics, 23:1, 149-153.

Ji G, Li G, Li X, Pang S, Randy J. (2008). Experimental Study of FRP Tube Encased Concrete Cylinders Exposed to Fire. Composite Structures, 85:2, 149-154.

Li G. (2011). Study on Mechanical Properties of Circular Steel-concrete-CFRP-concrete Solid Composite Short Columns under Axial Compression. Shenyang: Shenyang Jianzhu

University. Liu Y, Jia J. (2008). Parametric Study of

Temperature Distribution in CFRP Strengthened Beam-slab Assembly under Fire Conditions. Proceedings of the 5th International Conference on Advanced Composite Materials in Bridges and Structures, Manitoba, Canada: 114.

Liu Y, Tang Y, Wang Q. (2010). New Circular Hybrid Steel-concrete-CFRP-concrete Column and Its Fire Behaviour. Concrete, 32:1, 37-39.

Mirmiran A, Shahawy M, Samaan M. (1998). Effect of Column Parameters on FRP-confined Concrete. Journal of Composites for Construction, 2:4, 175-185.

Teng J, Yu T, Wong Y, Dong S. (2007). Hybrid FRP–concrete–steel Tubular Columns: Concept and behavior. Construction and Building Materials, 21(4): 846-854.

Teng J, Yu T, Wong Y. (2004). Behavior of Hybrid FRP-concrete-steel Double-skin Tubular columns. Proceedings of the Second International Conference on FRP Composites in Civil Engineering, Adelaide, Australia: 811-818.

Zhao Y, Gu W, Xu J, Zhang H. (2005). The Strength of Concrete Filled CFRP-steel Tubes under Axial Compression. Proceedings of the Fifteenth International Offshore and Polar Engineering Conference, Seoul, Korea: 19-24.

Zhu Z, Ahmad I, Mirmiran A. (2006). Seismic Performance of Concrete-filled FRP Tube Columns for Bridge Substructure. Journal of Bridge Engineering, 11:3, 359-370.


SHEAR BEHAVIOR OF RC BEMS CONTAINING CORRODED STEEL BARS

Xin Xue1, Seki Hiroshi2, Yu Song1 ¹ School of Architecture and Civil Engineering, Xiamen University, Xiamen 361005, P.R. China ² Department of Civil and Environmental Engineering, School of creative Science and Engineering, Waseda University, Japan Abstract: The influence of corrosion of steel bars (longitudinal bars and stirrups) on the shear behaviour of RC beams has been examined by experience. Experimental results indicate that, if corrosion level represented by maximum sectional-area loss is less than 35%, the corrosion of stirrups has little effect on the shear load carrying mechanism of RC beams and the shear capacity can be predicted accurately using modified truss theory. On the other hand, when the longitudinal bars corrode, the diagonal shear cracks shift to the loading point and cause a transition of load-carrying mechanism. Keywords: Corrosion, steel bars, shear behaviour, load-carrying mechanism, mximum sectional-area loss 1 INTRODUCTION

Unlike flexural failures, reinforced concrete shear failures are relatively brittle and, particularly for members without stirrups, can occur without warning. Corrosion of embedded reinforcement is one of the most critical consequences of aging RC members subjected to chloride-induced corrosion damage and concrete carbonation (Hong, 2006). To prevent these RC members from shear failure, there is a need to develop a tool that can evaluate and predict the shear behaviour during their service life.

The influence of steel bar corrosion on shear behavior has not been widely investigated. As steel bars corrode, the following effects result: sectional area loss, deterioration in bond effects, and the generation of corrosion cracks (JSCE 2006). About the influence of the corrosion of longitudinal bars, numerous published literature (Sato et al., 2003; Matsuo et al., 2004; Hashimoto et al., 2003., Xue et al., 2009) have pointed out that, when the longitudinal bars corrode, the load-carrying mechanism shift from beam mechanism to arch mechanism and the shear capacity increase. Most of the previous studies, however, have focused on the corrosion level of the longitudinal bars, while ignoring the influence of shear span. Shear span is generally thought to play a very important role in the shear behavior of RC beams. It is agreed that sectional area loss will cause decreases in the stirrup shear

resistance. About the influence of bond strength deterioration on shear behaviour, however, due to a variety of factors such as inavoiable corrosion of longitudinal bars, variation in stirrup style and etc., published literature have different opinions, expecially (Xu et al., 2004; Oshita et al., 2006; Morigawa et al., 2007; Christoper et al., 2006; Zhao et al., 2008).

Given the above mentioned background, this paper make a experimental investigation on the shear behaviour of RC beams containing corroded steel bars, and made efforts to develop a proper method that can evaluate the shear capacity accurately.

2 EXPERIMENT DETAILS

2.1 Specimen Details

Specimen details are shown in Figure 1. The specimen sections were designed as 12cm in wide and 24cm in heigtht. To prevent the specimens from failing in flexure, high strength screw-type steel bars were used for the longitudinal bars. Round bars φ6 (SR235 in Japanese specification) were used for the stirrups with close-type. Spacing of stirrups is designed as 120mm, eqavalent to 0.39% of steel ratio. Mechanical properties of the reinforced steel

科学出版社


www.abook.cn

·279·

bars are shown in Table 2. To avoid premature anchorage failure due to steel bar corrosion, the two ends of the longitudinal bars were fixed to the steel plates placed at two sides of the specimens using nut. Since the longitudinal bas and stirrups are individually subjected to accelerated corrosion tests (will be described later), insulatin job was done to the contact area between the longitudinal bars and the stirrups to prevent electric current flow through each other.

Specifications of the specimens are shown in Table 1. Experimental parameters were designed as corrosion target (longitudinal bars or stirrups), shear span to effective depth ratio (a/d), and

corrosion levels. Specimens of No.7-No.11 were for the investigation on the effect of stirrup corrosion specimens, and remains are for the investigation of the effect of longitudinal bar corrosion. The specimens can be divided into several series with the same a/d, and each series containing one specimen and several corrosion specimens. The mix proportion of concrete is specified in Table 3. The specimens had been moist cured for one week and then dry cured for at leat 28 days since conrete placement. Load tests were carried out soon after the accerlerate corrosion tests. 28 day compressive strength of specimens was also specified in Table 1.

Table 1. Experimen details

Longitudinal bars Massloss (%)

Stirrups Mass loss(%) No. a/d 2) Specimens 1)

cf (N/mm2)

Corrosion target 4)

PAM PPM PAM PPM

PEst5) (KN)

No.1 B(2.0)-ms 38.0 — 0.0 0.0 0.0 0.0 163.5

No.2 B(2.0)-m1s 39.1 7.8 15.1 1.0 1.6 178.6

No.3

2.0

B(2.0)-m2s 39.3 ○

9.1 15.0 3.0 4.9 206.1

No.4 B(2.6)-ms 33.1 — 0.0 0.0 0.0 0.0 148.0

No.5 B(2.6)-m1s 35.1 3.1 5.0 3.4 5.6 154.8

No.6 B(2.6)-m2s 35.9 ○

18.4 32.0 17.8 29.2 164.2

No.7 B(2.6)-ms1 34.2 0.6 0.9 5.6 6.1 144.0

No.8 B(2.6)-ms2 33.9 1.0 1.5 8.3 11.2 148.2

No.9 B(2.6)-ms3 34.6 1.5 2.3 14.0 20.9 138.9

No.10 B(2.6)-ms4 35.1 1.7 3.0 16.3 27.6 140.8

No.11

2.6

B(2.6)-ms5 34.9

△

0.8 0.6 19.3 34.2 138.9

No.12 B(3.2)-ms 35.2 — 0.0 0.0 0.0 0.0 141.1

No.13 B(3.2)-m1s 34.9 5.0 8.0 2.9 4.8 141.0

No.14

3.2

B(3.2)-m2s 33.3 ○

6.0 8.5 1.5 2.5 143.4

1) B(2.6)-ms1

a/d

ms : Sound specimensm1s : Longitudina lbar corrosion level:1;No stirrup corrosionms1 : Stirrup corrosion level:1; No longitudinal barcorrosion

2) a/d : Shear span to effective depth ratio 3) cf : Compressive strength of concrete 4) Corrosion target : — No corrosion; ○ Longitudinal bars; △ Stirrups 5) Ptest : Load at failure

Table 2. Steel bar properties

Steel bars Specification Sectional

area Yield

strength

Longitudinal D19 USD685A 287 706

Stirrups j6 SR235 28 300

Series a(mm) b(mm) a/d S1(mm) S2(mm)

B(2.0)-ms 440 1120 2.0 80

B(2.6)-ms 580 1400 2.6 100

B(3.2)-ms 700 1640 3.2

120

100

Table 3. Mix proportion

Unit(kg/m3) W/C(%)

s/a(%) Cement Water

Fine aggregat

Coarse aggregate

Admixture

48.1 51.8 387 186 881 842 5.8

Figure 1. Specimen details

men details

·280·

2.2 Accelerate Corrosion Test

Electrochemical corrosion shown in Figure 2 is performed to moddeling steel bar corrosion in a short period. The specimens were immersed in 3% NaCl solution for about 24 hours and then the corrosion target (longitudinal bars or stirrups) and copper plate placed outside the specimens were connected to the anode and cathode of a constant current generator respetively. The corrosion levels were achived by controlling the corrosion test period. Average mass loss of steel bars due to corrosion can be calculated from amount corrosion time using Faraday’s law. After load tests, the steel bars in concrete were taken out to investigate their corrosion state. Firstly, the concrete debris and some corrosion products were roughly removed using a sand blast, and then steel bars were put in a 10% ammonium hydrogen citrate solution for 24 hours to achieve a thorough cleaning.

Figure 2. Electrochemical corrosion tests

Two indexes of percentage average mass loss (PAM) and percentage maximum local mass loss (PMM) were used to nvestigate the average corrosion sate and the maximum corrosion state of steel bars. The PAM of the stirrups can be calculated using Eq.1.

/ 100PAM w w (1) Where Δw is the average mass loss of the corroded stirrups, and w is the mass of the original stirrups.

With regard to PMM, for longitudinal bars, the value can be achieved by evaluating the critical sectional area loss of the longitudinal bars. At a most critical corrosion section, the diameter of two mutually perpendicular directions were measured using a calliper and the residual cross-sectional area could be calculated making assume that residual section is elliptical one. For stirrups, the value of PMM can be achieved by a tension test. After the evaluation of PAM, the side-leg portions were cut from the corroded stirrups and then subjected to a tension test. the PMM could be calculated by Eq.2.

( ) / 100y y yPMM f f f (2)

Where fy is the yield strength of the original

stirrups, f 'y is the yield strength of the side-leg test piece.

2.3 Load Test

The specimens were subjected to load tests under simply supported conditions, as shown in Figure 3. The load was applied at mid span with a displacement increment of 0.2 mm/minute. Application of the load was monitored using a load cell, and the deflection at the mid-span was measured using displacement transducers placed under the specimens. During the tests, the width of diagonal cracks was measured using a visual crack comparator.

Figure 3. Load tests

3 EXPERIMENT RESULTS

3.1 Corrosion State of Steel Bars

Rust stains and cracks along the steel bars subjected to accelerate corrosion tests were observed on the surface, as shown in Figure 4. These cracks were thought to be induced by the expansion of corrosion products.

The relationship between PAM and PPM is shown in Figure 5. Regarding PPM/PAM ratio, the value of stirrups is less than that of the longitudinal bars; this is because that the critical corroded section of stirrups was always at the corner, which is not contained in side-leg portions.

3.2 Results of Specimens with Corroded Stirrups

For specimens No.7~No.11 whose stirrups were subjected to accelerated corrosion, corrosion is also confirmed along longitudinal bars. The PAM of each specimen, however, is less than 2%. Since this value is such a small one that the mechanical performance and bond behaviour of longitudinal bars hardly change, and therefore the change in shear behaviour can be exclusively attributed to the corrosion of stirrups.

All the specimens failed in shear compression due to the crushing of concrete near loading

科学出版社


www.abook.cn

·281·

point. Load-displacement behaviour is shown in Figure 5. The load capacity decrease with the increasing corrosion level of stirrups, the reduction in the shear strength carried by stirrups due to the crossion area loss can be considered to be responsible for this result. The crack pattern at failure and the opening behaviour of critical diagonal crack are shown in Figure 6 and Figure 7 respectively. In Figure 6, the cracks due to corrosion product expansion are also illustrated using dotted line, and the heavy line refers to the critical diagonal crack which results in the failure of specimen. The number of diagonal cracks in corroded specimens is less than that of the sound specimens. The deterioration in bond strength of bond strength between stirrups and the sourrounding concrete is assumed to account for this. Howerer, there is hardly any difference between the corroded specimens and the sound specimen in the location and inclined degree of the critical diagonal crack. From the crack pattern, it can be observed that the stirrups intersecting the critical diagonal crack did not change in corroded specimens.

Figure 4. PMM-PAM relationships

Figure 5. Load-displacement behaviour

Figure 6. Crack pattern at failure

Figure 7. Crack opening behavior

In accordance with the modified truss theory widdly used for the evaluation of shear behavior of RC beams, the critical diagonal crack behaviour plays a very important role in shear behaviour of RC beams, by which the load carrying mechanism of RC beams is determined. From above describtion and discussion, it can be expected that the stirrup corrosion have slight influence on load-carrying mechanism of RC beams. Therefore, the Eq.1 basing on modified truss can also be applied to predict the shear strength of RC beams containing corroded stirrups.

u c sV V V (3)

Where Vu is shear strength of RC beams, Vc is the shear strength carried by concrete, Vs and is shear strength carried by stirrups. It should be noted that the critical sectional loss of stirrups must be took into account when Vs is calculated.

Vc - PPM relationship is shown in Figure 8. Vc did not change irrespetive of stirrup corrosion state.

·282·

Figure 8. Vc-PMM relationships

From experimental resuls, it is confirmed that stirrup corrosion just decrease the shear strength carried by stirrups, did not affect the load carrying mechanism, and thus did not change the shear strength by concrete. Taking into account the critical sectional loss of stirrups, the shear strength can be predicted accurely using modified truss theory.

3.3 Results of Specimens with Corroded Longitudinal Bars

All the specimens failed in shear compression. The crack pattern at failure is shown in Figure 9. The flexural cracks perpendicular to the longitudinal bars decreased with the increase of corrosion level of longitudinal bars. The deterioration in bond strength is assumed to account for this. It can be observed from the crack pattern that the critical diagonal crack shiftted progressly to the loading point and the area of compression zone of concrete above the critical diagonal crack increased as the longitudinal bar corrosion level increased. It can be expected that the longitudinal bar corrosion caused transition of load-carrying mechanism of RC beams. When the corrosion level increased, tie-arch mechanism became stronger in the corroded specimens.

(a) a/d=2.0

(b) a/d=2.6

(c) a/d=3.2

Figure 9. Crack pattern at failure

Figure 10. Vc-PAM relationships

More than 2% of PAM due to corrison is also confirmed in most of the stirrups. However, from the describtion in 3.2, the stirrup corrosion did not affect load-carrying mechanism. The transition of load-crrying mechanism can be attributed absolutely to the longitudinal bar corrosion. When the bond strength between longidutinal bars and the surrounding concrete deteriorated due to corrosion, Bernoulli-Euler theory can not be aplicalbe and the tensile strength arising in longitudinal bars was transferred directly to the support, and thus enhanced the tie-arch mechanism. The transition of load-carrying mechanism can have two

科学出版社


www.abook.cn

·283·

opposing effects on the entire shear strength. The enhancement of tie-arch mechanism caused the increase in the shear strength Vc carried by concrete. On the other hand, since the critical diagonal crack shifted to load point and became steeper, the number of stirrups intersected the critical diagonal crack decreased, and therefore resulted in the reduction in shear strength Vs sustained by stirrups. In this case, the role of stirrups placed near the support is to prevent the propagation of bond crack and enhance flexural rigidity of longitudinal bars, and thus made the tie-arch mechanism stronger. It can be expected that the entire shear strength vary with the longitudinal bar corrosion level and steel ratio of stirrups. In this paper, since the beneficial effect of increase of Vc due to the enhasement of tie-arch mechanism seemed to surpass the adverse effect of reduction in Vs, the shear strength increase with the corrosion level of longitudinal bars.

Variation of Vu with PAM of longitudinal bars is shown in Figure 10. It can be observed that the extent to which the shear strength increase vary with the shear span to effective depth a/d. The shorter the shear span is, the more increase in the shear capacity.

4 CONCLUSIONS

The shear behaviour of RC beams containing corroded steel bars (longitudinal bars and stirrups) was investigated, and with the scope of this papere, some conclusions can be drawn as following:

1) Stirrup corrosion has little effect on the load-carrying mechanism of RC beams. If the PMM is less than 35%, wih the consideration of sectional area loss of stirrups, the shear capacity of RC beams containing corroded stirrups can be predicted using modified truss theory.

2) Longitudinal bar corrosion may cause load-carrying mechanism transition. The derioration in bond strength of longitudinal bars results in the strengthen of tie-arch mechanism, and thus increase the shear capacity of RC beams.

3) The transition of load-carrying mechanism due to longitudinal bar corrosion is influenced by the shear span to effective depth of RC beams.

REFERENCES

Christoper, H. and William, C.F. (2006). Tests of Reinforced Concrete Beams with Corrosion-damaged Stirrups. ACI Structural Journal, 103:1, 133-141.

Hashimoto, K., Morikawa, H. and Kobayasi, H. (2003). Evaluation of Shear Behavior of RC Beams Taking into Consideration of Reinforcement Corrosion. Proceedings of the Japan Concrete Institute, 25:2, 1009-1014. (in Japanese)

Hong, N. (2006). Architectural Corrosion and Sustainable Development. Industrial Construction, 36：3，76-79. (in Chinese)

Matsuo, T., Sakai, L., Matsumura, T. and Kanetsu, N. (2004). Studies on the Shear Load Carrying Mechanism of RC Beams with Corroded Reinforcement. Concrete Journal, 15:2, 69-77. (in Japanese)

Morigawa, H., Kasamatsu, T. and Kobayashi, H. (2007). Assessment of Shear Performance of RC Members Taking Account the Deteriorated Bond Effects Between Concrete and Reinforcement. Proceedings of the Construction Research Institute Foundation, 49, 1-8. (in Japanese)

Oshita, H., Sawai, K., Hatano, Y. and Inoue, S. (2006). Effects of Expansion of Concrete, Bond and Anchorage Characteristics of Shear Reinforcement on Load Carrying Behavior of RC Beams Members. Proceedings of the JSCE Kansai-Chapter Annual Meeting, V-27-V-28. (in Japanese)

Sato, Y., Yamamoto, T., Hattori, A. and Miyagawa, Y. (2003). Influence of Stirrups and Longitudinal Barscorrosion on Shear Behavior of RC Members. Proceedings of the Japan Concrete Institute, 25:1, 821-826. (in Japanese)

Xu, S. H. and Nie, D. (2004). The Shear Behavior of Corroded Simply Supported Reinforced Concrete Beams. Journal of Building Structures，25:5, 98-104. (in Chinese)

Xue, X., Seki, H., and Hiromori, S. (2009). Influence of Reinforcing Bar Corrosion on Shear Behavior of RC Beams. Doboku Gakkai Ronbunshuu E，65:2, 161-177. (in Japanese)

Zhao, Y.X., Jing and W. L.X. (2008). Analysis on Shearing Capacity of Concrete Beams with Corroded Stirrups. Journal of Zhejiang University (Engineering Science), 42:1, 19-24. (in Chinese)


PRELIMINARY STUDY ON FINITE ELEMENT SIMULATION OF STEEL-

ASPHALT COMPOSITE ISOLATION LAYER

Fei Yao1, Jing Shao1, Shouping Shang2 ¹ College of Civil and Transportation Engineering, Hohai University, Nanjing 210098, P.R. China ² College of Civil Engineering, Hunan University, Changsha 410082, P.R. China Abstract: Finite element model of reinforced - asphalt composite isolation layer which considering influence of the asphalt ointment and brick piers is presented on the basis of shaking table test in this paper. Subroutine UBEAM is programmed to simulate the nonlinear characters of the steel bars under strong earthquakes; Subroutine USPRNG（nonlinear spring） is programmed to simulate the nonlinear interaction between the brick pier and the upper beam in the earthquake. Based on this calculation model, modes of each specimen of the shaking table test were calculated, a 3D model time history analysis is taken to simulate each shaking table test conditions. The calculated results are compared to the experimental results to verify the accuracy of the calculation model. Keywords: Seismic isolation, finite element, nonlinear, stiffness, damping 1 INTRODUCTION

Steel- asphalt compound isolation layer is a new isolation device with broad prospects for its excellent seismic isolation effects, high safety reserve, resistance to vertical pull force and overturning, low cost, etc. It has been applied in pilot project after its damping effect has been experimentally verified, but research on the working mechanism and computational model is still very limited. At present, simplified calculation model only consider vertical bearing capacity and horizontal stiffness of steel bars on elastic stage is adopt to analyze and design the isolation layer. But, shaking table test results show that the steel- asphalt compound isolation layer is a complex nonlinear system. When input acceleration is large, influence of the asphalt ointment and brick piers to the dynamic characteristics of the isolation layer can not be ignored. So， more precise calculating method such as numerical simulation method is needed.

In this paper, finite element model of reinforced - asphalt composite isolation layer considering influence of the asphalt ointment and brick piers is presented on the basis of shaking table test. Subroutine UBEAM is programmed to simulate the nonlinear characters of the steel bars under strong earthquakes; Subroutine USPRNG（nonlinear spring） is programmed to simulate the nonlinear

interaction between the brick pier and the upper beam in the earthquake. Based on this calculation model, modes of each specimen of the shaking table test were calculated, a 3D model time history analysis is taken to simulate each shaking table test conditions. The calculated results are compared to the experimental results to verify the accuracy of the calculation model.

2 ESTABLISHMENT OF THE MODEL

When the relative displacement of the isolation layer occurs because of the reduction of height, it is the deformation caused by the geometrical conditions rather than external forces. In that case, solid elements are better able to simulate the specimen in the 3-D displacement and deformation. But if solid elements are used to simulate each vertical steel bar, the number of elements is too much in order to meet the computing requirements for accuracy; otherwise the results are not accurate in the calculation of the flexural rigidity. This can result in the number of elements is too great to make the calculation difficult to carry out. Besides, steel- asphalt compound isolation layer is a complex nonlinear system consists of a variety of 科

学出版社


www.abook.cn

·268·

different physical properties of materials. If these materials are all simulated by solid elements, it will related to the contact problem of multiple objects, and unpredictable problems will appear in the calculation, such as: the contact object penetrate each other, the calculation is difficult to converge, etc.

A set of steel- asphalt compound isolation layer modeling method is summarized in this paper, considering the role and influence of each component of the isolation layer. Finite element analysis model is shown in Fig.1.

Figure 1. Schematic of finite element analysis model

2.1 Simulation of Vertical Steel Bar

In order to reduce the number of elements and improve computational efficiency, No. 52, 2-D beam element is chosen to simulate the vertical bar. It has a very big advantage in analyzing the axial force, bending moment, shear force of the rod-shaped structure. Compared to using solid elements, the number of elements is reduced by more than an order of magnitude. But it can not solve the deformation problem caused by the geometric conditions, that is the highly reduction when there is relative displacement in isolation layer. This deformation problem is the direct cause of the redistribution of internal forces in the isolation layer, and its importance can not be ignored. Therefore, the nonlinear spring user subroutine USPRNG is used to simulate vertical and horizontal load redistribution phenomenon because the height of the isolation layer reduced. Also, to solve the reinforced nonlinear problems, the bilinear model is adopted in nonlinear subroutine UBEAM.

2.2 Simulation of Beams and Mass Blocks

The boundary conditions imposed by the lower beam is mainly played the role of fixed reinforced lower end of the isolation layer, therefore no need for direct modeling in the finite element model, just need to fixed degree of freedom of the bottom of the steel bars. When

Modal analysis, fixed X, Y, Z three directions displacement and X, Y, Z three directions of rotation; when time history analysis, fixed X, and Z displacement of the two directions and X, Y, Z three directions of rotation, and imposed the boundary conditions of acceleration in the Y direction.

Upper beam and the mass does not affect the mechanical properties of isolation layer as the main components, in order to simplify the model, the upper beam and mass are simulated by No. 52 elements. The upper beam is mainly played the role of fixed steel bars and delivering upper weight, so set the stiffness very big. Beam elements which simulation the mass blocks has also been given a large stiffness, mass density derived from the conversion.

2.3 Simulation of Asphalt Ointment

Asphalt ointment located between the brick piers is for anti-rust, while the asphalt ointment between the top of the brick piers and bottom of upper beam has a certain influence on the mechanical performance of the isolation layer. This effect is mainly reflected in: transmission of normal pressure between the beams and brick piers, increased damping of the isolation system when the relative displacement between the upper beam and brick pier occurs. Considering this feature of asphalt ointment, non-linear springs in MARC are adopted to simulate asphalt ointment. Through programming subroutine USPRNG, setting stiffness and damping of the non-linear spring to simulate the character of the asphalt. Compared to directly using 3-D element to simulate asphalt, it not only simplifies the model, greatly reduce the computation time, but also makes the purpose of calculated explicit, and much more convenient to adjust the calculation parameters.

2.4 Simulation of Brick Piers

The brick piers improved reliability and stability of the isolation layer in case of large earthquakes. Its main role is: when the relative displacement of the isolation layer is large, sharing part of the vertical load with vertical steel bars; at the same time, friction between it and the upper beam occurs. This friction is related to the relative displacement of the isolation layer. The larger the relative displacement, the more the vertical loads brick pier get, so the larger the normal pressure and the larger the friction.

Horizontal stiffness of brick piers is much larger than which of the vertical steel bars. In the whole dynamic response of the isolation layer, dots fixed six degree of freedom are used to

·269·

simulate the brick piers. Through setting the non-linear spring between the dots and the upper beam, subroutine USPRNG is programmed to simulate the role of the brick piers.

2.5 User Subroutine

MSC.Marc has lots of user subroutines and powerful features, stress-strain relationship of materials or structures can be self-defined. Two subroutines are adopted in this paper: the beam element subroutine UBEAM and the nonlinear spring subroutine USPRNG.

2.5.1 Subroutine UBEAM

In this paper, the nonlinear elastic beam is adopted. In the elements there are four generalized stress components F ,

xM , yM , T

and four generalized strain components , xK ,

yK and . Their relationship as Eq.1:

11

22

33

44

x x

y y

DF

DM K

DM K

DT

(1)

Where F , xM ,

yM , T are the axial force,

moment in the X direction, moment in the Y direction, torque.

ijD are elements in

cross-sectional stiffness properties matrix. For the nonlinear elastic beam, the generalized stress and generalized strain relationship is nonlinear, the stiffness matrix changes as the strain changes.

To vertical steel bars of the isolation layer, the main consideration in the earthquake case is the nonlinear stage of the material. In this paper, a simplified reinforced stress - strain curve is used, i.e. the stiffness of the two-stage.

2.5.2 Subroutine USPRNG

In MARC user subroutine USPRNG can define non-linear springs, dampers and foundation stiffness. The constant can be modified during loading process. In this paper, a nonlinear spring (including vertical and horizontal) to simulate the interaction between the brick piers and the upper beam, the nonlinear spring force and damping can be defined.

In the case of large earthquake, larger relative displacement occurs in the isolation layer. There is nonlinear interaction between the brick piers and the upper beam, including the vertical support that brick piers give to the upper beam and horizontal friction between them. Due to the

presence of asphalt ointment, the damping of the isolation layer can be increased when there is relative motion between brick piers and the upper beam.

The interaction between the brick piers and on the beam is directly related to the relative displacement of the isolation layer. So when determine the interaction in the process to vibration, we must first obtain the relative displacement of the isolation layer at this moment. By calling NODVAR subroutine to extract the node results in the MARC database, including the displacement, velocity, acceleration and other parameters of each calculated increment as the judgment basis, calculate the value of the vertical and horizontal interaction, and then substitute into the next incremental step to calculate.

3 VERIFICATION OF THE MODEL

3.1 Modal Analysis

Modal analysis results of the isolation layer specimen with structural dynamics method and finite element model results are compared and listed in Table 1.

Table 1. Baseband frequency calculated with two methods (Hz)

No. Structural dynamics method

finite elements method

Error（%）

GZC200-6-1 1.88 1.97 4.7 GZC300-6-1 1.02 1.07 4.9 GZC200-6-2 1.88 1.97 4.7 GZC200-8-2 3.33 3.50 5.1

It can be seen from Table 1, calculated

baseband frequency of the finite element modal analysis is slightly larger than structural dynamics, about 5%. This is because the structural dynamics calculation method only considers the horizontal stiffness of the vertical steel bars, the finite element model considering the horizontal stiffness of the asphalt ointment.

3.2 Acceleration Response Analysis

Due to the limited accuracy of the shaking table test equipment, computer input acceleration time history waveform and the acceleration time history waveform of the shaking table actually output is not exactly the same, so compare the curve of the experiment and calculation does not 科

学出版社


www.abook.cn

·270·

make sense. But the rationality of the finite element model can be verified by comparing the two acceleration amplitude. Acceleration amplitude of the tests and calculation are listed in Table 2.

Table 2. Acceleration amplitude of the tests and calculation (g)

No. tests calculation Error（%）

200El0.1 0.50 0.46 8.0 200El0.2 0.46 0.45 2.2 200El0.3 0.42 0.44 4.8 200El0.4 0.42 0.43 2.4 200Taft0.1 0.53 0.54 1.8 200Taft0.2 0.63 0.53 15.8 200Taft0.3 0.45 0.48 6.7 200Taft0.4 0.40 0.38 5.0 300El0.1 0.48 0.44 8.3 300El0.2 0.41 0.41 0.0 300El0.3 0.39 0.38 2.5 300Taft0.1 0.66 0.54 18.1 300Taft0.2 0.49 0.51 4.1 300Taft0.3 0.42 0.46 9.5 200-6-2El0.1 0.66 0.71 7.6 200-6-2El0.2 0.69 0.70 1.4 200-6-2El0.3 0.64 0.70 9.4 200-6-2El0.4 0.61 0.65 6.6 200-6-2El0.5 0.57 0.60 5.3 200-6-2El1.0 0.55 0.55 0.0 200-6-2El1.7 0.48 0.52 8.3 200-6-2Taft0.1 0.72 0.81 12.5 200-6-2Taft0.2 0.77 0.79 2.7 200-6-2Taft0.3 0.74 0.74 0.0 200-6-2Taft0.4 0.69 0.61 11.6 200-6-2Taft0.5 0.71 0.61 14.1 200-6-2Taft1.1 0.61 0.58 4.9 200-6-2Taft1.7 0.48 0.47 2.1 200-8-2El0.1 0.85 0.79 7.1 200-8-2El0.2 0.64 0.75 17.2 200-8-2El0.3 0.70 0.72 2.9 200-8-2El0.5 0.73 0.71 2.7 200-8-2El0.6 0.71 0.69 2.8 200-8-2El1.7 0.70 0.67 4.3 200-8-2Taft0.1 0.94 0.88 6.4 200-8-2Taft0.2 0.80 0.84 5.0 200-8-2Taft0.3 0.55 0.70 27.3 200-8-2Taft0.4 0.64 0.67 4.7 200-8-2Taft0.6 0.60 0.61 1.7 200-8-2Taft1.5 0.37 0.41 8.1

It can be seen from Table 2: error percentage

between calculated and experimental values is less than 20%; in 40 conditions, the error percentage of less than 10% accounted for 33, calculation results are reasonable. There are several conditions with relative large error,

because of the limited accuracy the shaking table test, input acceleration waveform cannot always consistent; there will be some differences between each specimen due to construction reasons; isolation layer is a complex non-linear systems with many factors affect, the idealized computer models cannot simulate a variety of random; we also observed in the experiments, the mechanical properties of the isolation layer has relationship with the loading history, after a number of vibration especially vibration with large acceleration, asphalt ointment softens and extrusion, brick pier partial damaged. These phenomenons will have some impact to force performance in follow conditions. But as a new type of isolation device, its performances manifested in the experiments and the calculations are very satisfactory.

4 CONCLUSIONS

In this paper, based on early research on shaking table tests of steel-asphalt compound isolation layer, nonlinear finite element model is established using large finite element the software MSC.MARC. Subroutine UBEAM and USPRNG are programmed to simulate the material nonlinear of steel bars and nonlinear interaction between the brick pier and the upper beam. Using this model, modes of each specimen of the shaking table test were calculated, a 3D model time history analysis is taken to simulate each shaking table test conditions. Compared with the experimental results, the results show that this model can calculate the natural frequency of the isolation layer specimen accurately, can simulate the mechanism and dynamic response of the isolation layer specimen, and can calculate performance parameter, i.e. the acceleration amplitude, of the isolation layer specimen isolation reasonably. However, as the number of test data are not enough, the computational model needs more experimental data support further improvement.

ACKNOWLEDGEMENTS

Supported by “the Fundamental Research Funds for the Central Universities” (2012B02214). This support is gratefully acknowledged.

·271·

REFERENCES

Shang, S.P., and Yao, F. (2011). Study on Structure of a New Type of Composite Isolation Layer and Its Shaking Table Test. China Civil Engineering Journal, 44:2, 36-41. (in Chinese)

Shang, S.P., Gao, Z.Y. (2011). Research on

Isolation Properties of Composite Isolation Pier by Shaking Table Test. Journal of Earthquake Engineering and Engineering Vibration, 31:6, 117-122. (in Chinese)

Shang, S.P., Huang, Q.T. (2012). Full-Scale Experimental Research on Steel Asphalt Isolation Pier. Journal of Building Structures, 33:3, 132-139. (in Chinese)

科

学出版社


www.abook.cn


COMPARITIVE ANALYSIS ON Z-DIRECTION QUALITIES OF STRUCTURE

STEELS AMONG DIFFERENT CODES

Yuanqing Wang, Yuanyuan Zhang, Yongjiu Shi Key Laboratory of Structural Engineering and Vibration of China Education Ministry Department of Civil Engineering, Tsinghua University, Beijing 100084, P.R. China Abstract: Steel is not isotropic material. For plate steel, especially for thick plate steel, because of melting and rolling process, the properties through three dimensions are different, among which the one through Z-direction is the worst. The most important index to measure the Z-direction properties is section shrinkage on Z-direction simple tension test. Sulfur content, Z-direction impact toughness and fracture toughness are the other important indexes measuring Z-qualities. Z-direction qualities are important while plate steel is bearing load through thickness position. Bad Z-direction qualities may result in lamellar tearing. Since few codes around the world make full accounts of Z-quality analysis and control, and more thicker plate steel is being applied nowadays, further research on the Z-direction qualities must be carried on and the codes be more rationally revised. In this article, different codes on Z-direction qualities are compared and analyzed. Existing retrieval reveals that the euro code EN1993-1-10:2005, German regulation Dast 014-81, Japanese code WES 3008, Chinese code GB5313-85 and Chinese procedure book of melting JGJ 81-2002 have detailed indexes with the Z-direction values analysis of various situations. As a conclusion, the effective thickness of the welding seam, the system and procedure of melting, the shrinkage due to the thickness of the material thickness, the remote constraint by the method of the joints and the preheating process to the steel are the main five different aspects to analyze the Z-direction values of different structures. In corresponding to the five different aspects, many codes have different means to choose different class of steel according to the Z-values of the structures applied to the project. At last, upon these codes, some measures are taken to prevent from the Z-direction tearing and other damages. Future research can be taken referring to the existing codes in order to precise the parameters or add up new impact factors to the codes. Keywords: Thick plate steel, Z-direction properties, lamellar tearing, section shrinkage 1 INTRODUCTION

The indexes which can be used to evaluate Z-direction qualities are basic mechanical properties, chemical proposition of the steel, Z-direction tension thrinkage ψz and Z-direction impact strength. Among these indexes the most important are chemical proposition of the steel and Z-direction tension thrinkage ψz. In this paper, Euro code, Japanese code, Chinese code and some other codes are consulted and compared. The procedures to assaying different situations and choosing different quality steel are exactly the same: to evaluate the Z-direction indexes first by add up the Z-direction values which depend on different aspects of actual material and conformation, and then choose different classes of Z-direction reinforced steel by the classification of the

Z-direction values of the actual situation. The types of Z-direction indexes to evaluate the sensitivity are slightly different, while the numerical values of them are also different in different codes. For Z-direction qualities, National code GB5313-85 is applied in China for Z-direction properties of thick plate steel.

2 A BRIEF INTRODUCTION IN Z-DIRECTION QUANTITIES OF DIFFERENT CODES

In British Standard BS5950-1:2000(Sructual use of steelwork in building), it is pointed out that to prevent the steel from Z-direction lamellar tearing, connection details, melting methods and

·285·

procedures, the high Z-direction quality steel’s option should be concernd.

In Euro code the Z-direction qualities are defined as this: after Z-direction test on sampling material according to Euro code EN 10002, expressed the thrinkage as percentage, and the numerical values can represent the weaken of the Z-direction qualities.

It is indicated in Australia code AS4100 that the reasons influence the Z-direction quality and cause Z-direction lamellar tearing are as these: the improper connection of melting, the through thickness high stress during thrinkage and the through thickness ductility reduction resulting from discontinuous of the material.

In Canadian code, it is emphased that in filet melt and T-joint melt conditions etc. measures should be taken to prevent or reduce the constrained stress through the thickness. When the bad condition joint is inevitable, resorts should be applied to prevent from Z-direction damage.

In Japanese code JIS G3199-1992, it is stated that, of the hot rolling plate steel, the properties through thickness direction is weaker that the rolling direction or wideness direction. To improve the qualities through thickness, measures should be taken in manufacture process.

In Chinese code, it is stated that the Z-direction qualities is defined by the thrinkage of simple tension through thickness.

It is clear that there are more precise and extent details and more comprehensive applications in different thuicknesses and different strenghth in Euro code than in Chinese code or Japanese code. Firstly the Euro code of Z-direction qualities can be applied in not only plate steels, but also in flat steel and section steels. Secondly the Euro code can be applied in thicknesses from 15mm to 400mm while Japanese code and Chinese code can be applied from 15mm to 150mm. At last the Euro code can be applied in strength up to 960MPa, while Japanes and chinsese code up to 500MPa.

3 REQUIRMENTS AND CLASSI- FICATIONS OF Z-DIRECTION PROPERTIES IMPROVED STEEL IN DIFFERENT CODES

3.1 Summary

In Euro code EN10164-2004(steel products with improved deformation properties perpendicular to the surface of the product) there is special

standard for the qualities needed of the Z-direction quality steel and the classification needed of the Z-direction quality needed steel. There are two factors needed for the Z-direction improved steels, say the reduction of area (section thrinkage) and ultrasonic testing of the materials. After this Euro code designed, the requirments to the Z-direction indexes of some other European countries shall point to this code. In Japanese code JIS G3199-1992 and Chinese code GB5313-85, the qualities needed for the Z-direction properties are Z-direction recuction, sulphur percentage of the materials and ultrasonic testing of the materials.

3.2 Requirements of Reduction

For the reduction of the material, Euro code, Japanese code and Chinese code have exactly same requirements, as described in Table 1.

Table 1. Requirements of Z-direction reduction of Euro code, Japanese code and Chinese code

Reduction of area in % Quality class Minimum average

of three tests Minimum of individual value

Z15 15 10 Z25 25 15 Z35 35 25

3.3 Requirements of Ultrasonic Testing

It is prescribed in Euro code that plate steels should be detected according to EN10160, while section steels according to EN10306. Among the EU countries there are special ultrasonic testing codes in UK, Germany and France. For UK, BS5996-1993(Specification for Acceptance levels for internal imperfections in steel late, strip and wide flats, based on ultrasonic testing), for Germany, SEL072:1998, and for France, NFA04-305 March1983 are applied for ultrasonic testing. The testing methods vary from manual detecting, assisted manual detecting, semi-automatic detecting to automatic detecting. As regards to Japanese and Chinese code, plate steel offered by supplier according to codes of through thickness properties must be detected by ultrasonic testing. The ultrasonic testing methods should be negotiated and coincide with codes. For Japan, it is declared in JIS G3136-05(Rolled Steels for Building Structure) that for SN400C, SN490C and other steels with Z-direction qualities requirements, the grade of ultrasonic testing must be no less than JIS G 0901. For China, according to GB19879-2005, 科

学出版社


www.abook.cn

·286·

steel witn Z-direction qualities requirements should be tested piece by piece. Testing methods should coincide with code GB/T 2970 and the grades shall be noticed in the contract.

3.4 Requirements of Sulphur Content

In existing search, only Japanese code and Chinese code have specified sulphur content limitations, but the limitations are slightly different, as described in Table 2.

Table 2. Requirements of sulphur content according to Japanese code and Chinese code

Limitations of sulphur content by % Quality class Chinese code

GB5313-85 Japanese code JIS G3199-1992

Z15 0.010 0.010 Z25 0.007 0.008 Z35 0.005 0.006

4 REQUIREMENTS AND CLASSI- FICATIONS TO THE Z-DIRECTION QUALITIES IMPROVED STEEL OF DIFFERENT CODES

The available design Z-direction quality value for the special joint and material value should be accounted before choosing the Z-direction improved steel. The Z-direction quality values are made certain with some equations. The equations vary from different codes. The types and values of different Z-direction indexes to evaluate the sensitivity are a slightly different in different codes.

4.1 Euro Code for Z-direction Quality Value Indexes

It is stated in Euro code EN1993-1-10:2005 that lamellar tearing may be neglected if the following condition is satisfied, as to Eq.1.

ZEd ≤ZRd (1) Where: ZEd represents the required design Z-value. ZRd represents the available design Z-value for the material. To accurate the design value ZEd, there is an equation as Eq.2.

ZEd= Za + Zb + Zc+ Zd + Ze (2) Where: Za represents the criteria of weld depth which affects the target value of ZEd. The effective

weld depth aeff is as Figure 1.

Figure 1. Effective weld depth aeff

Zb represents the criteria of shape and position of welds and connections affecting the target value of ZEd. Zc represents the effect of material thickness affecting the target value of ZEd. Zd represents the effect of remot restraint affecting the target value of ZEd. Ze represents the influence of preheating which affects the target value of ZEd. There are detaild tables with each numerical data of these factors in DAST 014-81.

4.2 Germany Code for Z-direction Quality Value Indexes

For the target value of Z direction qualities, it is cleared in German code DAST 014-81. Define LTR as the sensitivity indexes of lamellar tearing. LTR can be added up with the factors named INF(X), and then the choice of materials with different Z-direction qualities can be determined by the summaried value of LTR. Like Euro code, INF(X) has some different factors as follows: the effective thickness of weld seam, the shape and position of welds, the thickness of the material, the extent of constrain of the joint and the preheating environment. There are detaild tables with each numerical data of these factors in DAST 014-81.

4.3 Japanese Code for Z-direction Quality Value Indexes

Japan has an exactly same table in WES 3008(Specification for through thickness characteristics of steel plate and wide flat) with Euro code for the lamellar tearing sensitiveness, as in Eq.3.

ZEd= Za + Zb + Zc+ Zd + Ze (3) The difference is there is no preheating factor ini this equation. There are detaild tables with each numerical data of these factors in WES 3008.

4.4 Japanese Code for Z-direction Quality Value Index

There is no detailed code to classify the Z-direction sensitiveness, but in the Chinese

·287·

regulations of construction manual of structure steels, the German code of DAST 014-81 is cited to specify the sensitiveness of actual joint and material.

5 COMPARISON OF CODES IN THE SELECTION OF Z-DIRECTION IMPROVED STEEL

In different codes, the target value of Z-direction sensitiveness is at first determined. And the Z-direction improved steel which to be used in high sensitiveness structure must be examined and classified, and at last the selection of the material is made by cover the danger of high sensitiveness structures with high quality material.

5.1 Euro Code for Choosing the Right Class Material

In Euro code, the classes of steel are specified according to code EN10164-2004, and then the selection of steel is determind with code EN1993-1-10-2005. The process is as follows: Firstly, before selecting the class of Z-direction property improved steels, the consequences for the damage while lamellar tearing accurs must be considered. Secondly, after the class of steel strength is selected, the Z-direction property of steel must be specified, for example as Z15, Z25 andZ35. After the steels installed or connected into structures, lamellar tearing must be detected to make clear whether lamellar tearing has accured. Thirdly, the factors which may affect Z-direction properties shall be considered to prevent lamellar tearing. Fourthly, the sensitiveness can be judged with the ductility of through thickness direction according to code EN10164. The grade of this ductility can be classified by Z-direction property class. At last, after the Z-direction property indexes are graded with table of ZEd, the using of Z-direction property improved steel can be chosen with some corresponding codes. In the chapter of material of EN 1993-1-1:2005, the details needed for Z-direction propertie improved steel are specified. And in the following three codes the table for selecting Z-direction properties

improved steel is listed: EN1993-2:2006(steel bridges), EN1993-1-1(design of steel structures) and EN1993-6(crane supporting sructures). The regulations are the same as Table 3.

Table 3. Choice of quality class according to EN10164

Target value ZEd Quality class Zed≤10 —

10≤Zed≤20 Z15 20≤Zed≤30 Z25

30≤Zed Z35

5.2 Euro Code for Choosing the Right Class Material

The selection of Z-direction properties improved steels is regulated by code JIS G3199-1992 after the indexes of Z-direction property sensitiveness is cleared. For some special steel such as SN400C and SN490C, Z-direction properties are clearly announced. The average ductility of 3 samples must be no less than 25%, while the lowest ductility of a single sample must be no less than 15%. For some other steel, Z-direction properties can be offered as annex properties of the material.

5.3 Chinese Code for Selecting the Right Class Material

It is regulated in some Chinese codes that the Z-direction properties improved steel is needed while there are structures which require highly improved through thickness properties. But there are no clear indexes to estimate the sensitiveness of sturctures and the urgency to need Z-direction properties improved steel. Actually, the German code DAST014-81 is cited in procedure book of melting JGJ 81-2002 for the sensitiveness of lamellar tearing. In Chinese code, requirements of Z-direction properties improved steel are as annex to the normal steel attributes, such as 4C-Z25.

5.4 Other Countries

In Austrilian codes, American codes and Canadian codes there are no specific regulations about selecting Z-direction properties improved steels. It is only mentioned that through thickness properties must be considered to avoid lamellar tearing when in structures sensitive to lamellar tearing.

科

学出版社


www.abook.cn

·288·

6 MEANS TO IMPROVE Z-DIRECTION PROPERTIES AND PERVERTING FROM LAMELLAR TEARING

In different codes, there are some tips to improve Z-direction properties and prevent from lamellar tearing. In Euro code EN1993-1-10-2005, these aspects are considered to prevent from lamellar tearing: the criticality of the location because of acturally tensile stress and the extent of redundancy, the strain through the thickness direction because of badly made connection, the patterns of joint details or welding details and the chemical properties of the material which be traversely stressed. In Australian the instructs to lamellar tearing can be found in paper Walsh, P. Use of the Australian Standard for Concrete Structures, CSIRO, Inkata Press, 1998. In Japan, the means to improve through thickness properties during manufacture process are mentiond in paper ‘SN Steels-Production Methods and Properties’. It is bring forward that to reduce the detrimental defects of steel, to ameliorate the central segregation can improve through thickness properties. In China, means to improve Z-direction properties are put forwad in procedure book of melting JGJ 81-2002: to control the Sulfur content, to adjust the connection design and to apply suitable melting craft.

7 MEANS OF TESTING METHODS OF Z-DIRECTION PROPERTIES OF DIFFERENT CODES.

Only Euro code, Japanese code and Chinese code are compared in this paper.

7.1 Test Unit

Each consignment must be divided in to test units.

In Euro code EN10164-2005, for flat products of class Z15, Z25 and Z35 are based on the sulphur content by ladle analysis, as given in Table 4.

Table 4. Test unit for flat products

Test unit for

S>0.005%a S≤0.005% Quality

class Parent plate

or coilb Max.40 tc Castd

Z15 If agreed Xe x

Z25 x - Xe

Z35 x - Xe

a Ladle analysis. b Coil applies to wide strip, narrow strip and slit strip. c Or part there of of products. d Products with the same heat treatment. e Unless otherwise agreed at the time of the order. See Option 3.

For sections, the test unit shall consist of

material from the same cast and same heat treatment with at most of max 40t or part thereof. For Japanese and Chinese codes, Z15 steel shall be tested by units which consist of same cast and same heat treatment, while Z25 and Z35 steel shall be tested one piece by one.

7.2 Location and Quantity of Sampling

It is all regulated in Euro code, Japanese code and Chinese code that one sample contain six test pieces to be machined shall be taken out of each test unit, from which 3 for testing, and the other 3 reserved for supple testing. For Euro code, the flat steel and section steels are clearly accounted of sampling method but for Japanese code and Chinese code only plate steels are clearly accounted.

7.3 Test Pieces with or Not with Eextension Piecies

It is clearly regulated in Euro code that in which situation the extension pieces shall be applied in the test, as in Table 5.

Table 5. test pieces with or notwith extension pieces

Product thickness t With or not with extension pieces

15≤t≤20mm mandatory 20<t≤80mm optional t>80mm Not acceptable

It is regulated in Japanese code and Chinese code that when the length of product thickness is not enough to support a test sample, extension pieces can be applied.

·289·

7.4 Judgements about Testing

In Euro code, Japanese code and Chinese code, there are legals to determin whether the steel tests meets the satisfaction of steel class. Details can be found in EN10164-2005, GB5313-85 and WES 3008.

8 CONCLUSIONS

The differences and resemblance of Z-direction indexes, the test method of Z-direction properties improved steels and the measures to take avoiding Z-direction distruction among different codes are systemeticly compared in this paper. From above we can conclude that most countries nowadays don’t have so much specific details with Z-direction properties evaluating and Z-direction properties improved steels classifying. O nly Euro code and Japanese code has detailed indexes grading and choosing Z-direction properties improved steels. Therefor some times the steels are wasted when we overestimate the sensitiveness of lamellar tearing, while sometimes the structure isdesigned dangerously when the dangerous situation is underestimated. For the progress of thick plate steels, the Z-direction properties should be clealy and reasonably anylised with experimental and theoriotical method. Then the reasonable assessment can be applied into production.

ACKNOWLEDGMENTS

The writers gratefully acknowledge the support for this work, which was funded by the Natural Science Foundation of China(Grant No. 51178224 and No. 50778102) .

REFERENCES

ANSI/AISC 360-05. An American National Standard. Specification for Structure Steel Buildings.

AS4100-1998.Australian Standard, Steel structures. AS4100Suppl1999. Steel structures-commentary

(supplement to AS 4100-1998). Bonade, R; Mueller, P; Spatig, P. (2008).

Fracture toughness behavior in the ductile-brittle

transition region of the tempered martensitic Eurofer97 steel: Experiments and modelling. Engineering Fracture Mechanics. 75:13, 3985-4000.

BS5950-1:2000. Structural use of steelwork in building.

BS5996-1993. Acceptance levels for internal imperfection steel plate, strip and wide flats, based on ultrasonic testing.

CAN/CSA-S16-01. A National Standard of Canada. Limit states design of steel structures.

DAST 014-81. EN 10136:2002. Iron and steel-Ultrasonic

testing of H beams with parallel flanges and IPE beams.

EN1011-2. recommendations for arc welding of Ferrite steel.

EN10160-1999. Ultrasonic test for steel plates thicker than 6mm.

EN10164-2005. Steel products with improved deformation properties perpendicular to the surface of the product.

EN1993-1-1.Design of steel structures. Part1-1:General rules and rules for buildings.

EN1993-1-10:2005. Design of steel structures, Part 1-10: Material toughness and Through-thickness properties.

EN1993-2:2006. Design of steel structures, Part 2: Steel bridges.

EN1993-6. Design of steel structures. Part6 Crane supporting structures.

GB/T 2970-2004. Thicker steel plates—Method for ultrasonic inspection. (in Chinese).

GB19879-2005. Steel plate for building structure. (in Chinese).

GB5313-85. Steel plate with through thickness characteristics. (in Chinese).

JIS G3136:2005. Rolled steels for building structure.

JIS G3199-1992. Specification for through-thickness characteristics of steel plate wide flat.

Manual of steel construction. China steel construction society.

NFA04-305 March1983. IRON AND STEEL PRODUCTS. Ultrasonic reflection testing of plate not less than 6mm thick, Definition of quality-Test method.

SEL072-1998. Ultrasonically tested heavy plate, Technical delivery specifications.

Seta Ichiro. (1996). SN Steels-Production Methods and Properties. Japan Welding Society. 65:3,237-239. (in Japanese)

WES 3008. Specification for through thickness characteristics of steel plate and wide flat.

科

学出版社


www.abook.cn


RESEARCH ON IMPACT TOUGHNESS OF Q460C HIGH-STRENGTH

STEEL AT LOW TEMPERATURE

Yuanqing Wang1, Yun Lin2, Yannian Zhang2, Gang Shi1, Yongjiu Shi1

¹ Key Laboratory of Structural Engineering and Vibration of China Education Ministry, Department of Civil Engineering, Tsinghua University, Beijing 100084, P.R. China ² School of Civil Engineering, Shenyang Jianzhu University, Shenyang 110168, P.R. China Abstract: In order to obtain the influence degree for toughness of high strength steel, impact toughness tests of Q460C steel with 14 mm thickness as a new type of high strength structural steel at low temperature were conducted in this paper. SANS type pendulum impact testing machine as well as a low temperature incubator were employed for the experimental study. What’s more, the paper reported the results to identify the effects of low-temperature flexibility, plate thickness and material strength on the plastic deformation capacity governed by brittle fracture. Moreover, analyze the scanning electronic microscope of fracture surfaces of the specimen at different temperature points. In addition, the charpy impact power level was compared with that of Q345 steel with 60 mm, 90 mm, 120 mm and 150 mm thickness respectively. And the results demonstrated that: the impact toughness of Q460C steel declined with the descending temperature; The influence of steel strength on low temperature brittleness of Q460C steel was less apparent than that of plate thickness on that of Q345 steel.The fracture surfaces of impact test specimens of Q460C steel, which impact broke at-20 ha℃ d already finished the transition form ductile to brittle, with a large of notable brittle characteristics. Meanwhile, Boltzmann function fit analysis got -11.1 as the brittle ductile transition temperature ℃for Q460C steel. Test results showed that low temperature brittleness of Q460C steel were obvious, which should be paid sufficient attention to. Keywords: Charpy impact toughness, high strength steel, low temperature, impact energy, toughness-brittleness transition temperature 1 INTRODUCTION

High strength steel plate is more and more widely applied in construction industry for the past few years, such as 690 MPa high strength steel being used in Australian Sydney World Square's conversion layer, 690 MPa and 650 MPa high strength steel in the Star City hotel located in the central area, 460 MPa high strength steel in the key position of the “bird’s nest” project, 600 MPa steel I-column in Japanese Yokohama’s Landmark Tower building and other large-span house structures and high-rise as shown in the work by Pocock (2006).

There is large area of cold region inland springing up the high strength steel structure, whereas low temperature prompts the toughness of steel structural members to reduce. Furthermore the cold brittleness is not ignored. Moreover Q460C steel, the high strength construction steel was used for the charpy

impact test at low temperature in this paper. Not only to get tendency of the impact energy value AKV with temperature changing, but also to obtain the toughness-brittleness transition temperature. What’s more, in order to investigate the influencing level of intensity as well as thickness on impact energy values at low temperature, compare the AKV value of Q460C steel with that of U71Mn rail steel and Q345 steel’s.

2 OVERVIEW OF IMPACT TEST

2.1 Material and Method

According to GB/T 229-2007 Metallic Materials Charpy Pendulum Impact Test Method, which is test standards in Chinese, impact test of Q460C steel with 14mm thickness was carried out at five temperature points containing respectively

·291·

20℃, 0℃, －20℃, －40℃and－60℃. Select three samples at each temperature point in order to get longitudinal impact energy value AKV with the changes of low temperature. The experimental results may be useful to the structural design considering cold brittleness using the high-strength steel. And Q460C steel, as high strength construction-structure steel in Chinese steel standards, whose technical conditions met GB / T 1591-2008 High-strength Low-alloy Structural Steel, was delivered at controlled rolling status. Moreover its carbon equivalent was equal to 0.468% and the main chemical composition was shown in Table 1. And detailed size of impact specimen was shown in Figure 1 (a), the machining impact specimen in Figure 1 (b), and overall specimens after impact test in Figure 1 (c).Make sure that the direction of specimen’s gap and steel rolling direction were consistent.

Table 1. Chemical composition of Q460C steel (%)

C Mn Si P S Cu 0.190 1.510 0.220 0.013 0.004 0.007

(a) Detailed size

(b) Impact specimen

(c) The specimens after impact test

Figure 1. Impact specimen of Q460C steel

2.2 Equipment

Impact test was taken in the incubator to cool the impact specimens through steaming a mixture of liquid nitrogen and air .And test performed in the Strength Mechanical Laboratory of the School of Aerospace of Tsinghua University, which had a full set of low temperature test equipment, and “SANS” type pendulum impact testing machine as shown in Figure 2 .

Figure 2. SANS type pendulum impact testing machine

3 EXPERIMENTAL RESULTS

3.1 Trend of Impact Toughness of Q460C Steel with Low Temperature Changing

The results of the charpy impact test of Q460C high strength steel varying with temperature were drawn as the curve of impact absorption energy and temperature (K-T chart), which was shown in Figure 3.

Figure 3. Impact toughness of Q460C with temperature

Figure 3 displayed that the impact energy values of Q460C steel reduced obviously with the decrease of temperature, which indicated Q460C steel had obviously cold brittleness. The average value of the AKV at －60℃ comparing with that at 20℃ was reduced by 62.1%. It was implied that Q460C steel is quite sensitive to low temperature.

Besides, the test value at －20℃ and below － 20 ℃ temperature point didn’t meet the requirement of GB/T 19879-2005 Steel Plates for Building Structure Standard, which regulates the longitudinal impact energy value of Q460 steel at 0℃, －20℃ or －40℃ should not be less than 34 J. Therefore precaution of the low temperature brittle fracture should be considered

科学出版社


www.abook.cn

·292·

when designing.

3.2 Compared with U71Mn Rail Steel and Q345 Structure Steel

Yield strength value of U71Mn steel rail was suggested as 460 MPa as shown in the work by Zhang et al. (2006), according to domestic yield strength data of rail steel after hot rolling and referring to the latest provision of Russian and American rail standards. Further the results of comparative analysis of low temperature impact power value of Q460C steel and U71Mn rail steel at five low temperature points in previous work by the authors (Wang et al., 2007) were obtained, which were drawn as Figure 4.

Figure 4. Impact toughness of Q460C steel compared with that of U71Mn steel

Figure 4 can clearly show characteristics of impact energy-temperature curve of Q460C steel and U71Mn rail steel. The average AKV values of Q460C steel were greater than that of U71Mn rail steel range from 0℃ to －60℃, which indicated the low temperature fracture-toughness of structure steel was better than that of rolled rail steel at the same yield strength.

The fracture-toughness of steel was not only related with its thickness but alse with its strength. Subsequently, the low temperature impact power value of Q460C steel at 14 mm thickness was analyzed by comparing with that of Q345 steel at 60 mm thickness, 90 mm thickness, 120 mm thickness and 150 mm thickness respectively in previous work by the authors (Wang et al., 2010), which were shown in Figure 5.

Figure 5 displayed that when tempareture fell off from 20℃ to －20℃, the AKV values of Q460C steel at 14 mm thickness were less than that of Q345 steel at 150 mm, 120 mm, 90 mm and 60 mm thickness in order.

Figure 5. Compared with AKV value of Q345 steel at 60mm, 90mm, 120mm and 150mm thickness respectively

However, from －20℃ to －60℃, the AKV value of Q460C steel was rising and gradually higher than that of Q345 steel with different thicknesses. At －40℃, its value was between that of Q345 steel of the four thicknesses, but at －60℃, its value was over all that of Q345 steel of the four thicknesses. The above results showed that both the thickness and strength of structure steel affected its low temperature impact toughness. When tempareture brought down below －20℃, it was more significant to influence on the cold brittleness for thickness of Q345 steel than that of strength of Q460 steel.

4 TRANSITION TEMPERATURE

Ductile-brittle transition temperature is usually acted as important criterion of prevention from fracture in engineering. But general data of experiment is largely discrete; therefore choice of right function to fitting the curve of impact energy with temperature is reasonable. There is a smaller residual and better correlation when using Boltzmann function to carry out the regression analysis in the form Eq.1 as shown in the work by Zhao (2004).

0

1 22( )/d.

1 ekv T x x

A AA A

(1)

In Eq.1: Akv represented impact energy, T as temperature, A2 (J) as up-platform energy, A1 (J) as down-platform energy, x0 ( ℃ ) as ductile-brittle transition temperature and dx (℃) as range of transition temperature, of which the smaller the dx, the more easily the materials are changed from plastic to brittle.

·293·

The regression analysis of AKV value of Q460C steel and that of U71Mn rail steel in previous work by the authors (Wang et al., 2007) were carried out by the Boltzmann function fitting as shown in Figure 6. Furthermore Boltzmann fitting parameters of Q460C steel were shown in Table 3, and comparison of the fitting parameters of the AKV value of Q460C steel with that of Q345 steel with different thicknesses in previous work by the authors (Wang et al., 2010) using Boltzmann function was shown in Figure 7 .

Figure 6. Boltzmann fitting parameters of Q460C and that of U71Mn steel rail

Table 3. Boltzmann fitting parameters

A1 A2 x0 dx

24.4 58.8 －11.1 3.3

Figure 7. Boltzmann fitting parameters of Q460C and that of different thicknesses of Q345

1) Figure 6 indicated the regression coefficient of the AKV value of Q460C steel by Boltzmann fitting was more than 97.3%. And

ductile-brittle transition temperature of Q460C steel was below that of U71Mn rail steel, and the range of transition temperature of the Q460C steel was more than that of U71Mn rail steel. It was illustrated that fracture toughness of Q460C structure steel was better than that of U71Mn rail steel of the same strength grade at low temperature.

2) Table 3 and Figure 7 indicated that comparing with Q345 steel at 60 mm, 90 mm, 120 mm and 150 mm thicknesses respectively, Q460C steel had the lowest up-platform energy and the highest ductile-brittle transition temperature. Therefore Q460 steel was relatively more sensitive to low temperature.

5 SEM ANALYSIS OF FRACTURE

The impact specimen fractures of Q460C steel were observed with scanning electron microscope as shown in Figure 8, which were micro-morphology of the specimen fractures ruptured at 20℃, 0℃, －20℃, －40℃ and －60℃ respectively. And the photographs were taken near the center part of the fracture with amplification of 1000 times.

1) Figure 8 form (a) to (e) showed that the dimple and tearing ridge of fracture were reducing with the decrease of temperature. The fracture morphology presented microscopic hole gathered type and the fractures had dimples with different sizes while fracturing at 20℃, 0℃ or －20℃. But the dimples varied from big to small in size, deep to shallow in depth and long to short in length with temperature decreasing. Overall, it was major of toughness deformation, and no obvious brittle features.

2) Figure 8 (d) and (e) showed that not too much macro-plastic deformation came out in the morphologies. And that cleavage steps and tongue shapes can be observed obviously, and the fracture was crystalline with many strong reflect-light aspects; consequently the plastic deformation was minimal. Additionally, the fracture had little fiber zone but major of radiation zone, which mostly presented the mechanism of cleavage fracture, having significant property of brittle fracture while rupturing at －60℃.

科

学出版社


www.abook.cn

·294·

(a) 20℃ (b)0℃

(c) －20℃ (d) －40℃

(e) －60℃

Figure 8. Morphology of impact fractures of Q460C specimens

6 CONCLUSIONS

The impact energy values of Q460C steel were reducing obviously and its impact toughness deteriorated with the temperature decreasing.

1) Generally, the fracture toughness of structure steel was better than that of the rail steel with the same strength level at low temperature.

2) The impact energy value of Q460C steel with 14 mm thickness didn’t meet the requirement of 34J in the standards, which still need to improve in chemical composition and rolling technology.

3) Using the Boltzmann function fitting to the analysis of impact energy value of Q460C steel had a good correlation. The fitting parameters

showed that the down or up platform energy were small, and Q460C steel was easier to present brittle transition at － 11 ℃ low temperature.

4) The fracture toughness of structure steel was not only related with the thickness but also with strength to some extent. When temperature fell below －20℃ , it was more significant influence of the thickness of Q345 steel than that of the strength of Q460 steel on cold brittleness.

In general, Q460C steel has obvious cold brittleness features, whereupon it is necessary to carry out again the amount analysis of the toughness at low temperature, which should be taken enough attention.

ACKNOWLEDGEMENTS

This work was financially supported by the National Natural Science Foundation (51178244 and 50778102).

REFERENCES

Pocock, G. (2006). High Strength Steel Use in Australia, Japan and the US. The Structural Engineer, 11, 27-30.

Wang, Y.Q., Hu, Z.W., Shi, Y.J., Zhou, H., and Chen, H.(2010).Testing the Impact Resistance of Thick Structural Steel Plate at Low Temperatures. Journal of Harbin Engineering University, 31:9, 1179-1184. (in Chinese).

Wang, Y.Q., Xi, W., and Shi, Y.J. (2007). Experimental Study on the Impact Toughness of Rail Steel at Low Temperature. Journal of Tsinghua University (Sci &Tech), 47:9, 1414-1417. (in Chinese).

Zhang, Y.H., Chen, C.Y., and Zhou, Q.Y. (2006). Research on the Index Value of Yield Strength of Rail Steel. Railway Engineering, 3, 92-94. (in Chinese).

Zhao, J.P., Zhang, X.M., and Shen, S.M. (2004). On the Method of Data Processing for Ductile-brittle Transition Temperature. Petro-chemical Equipment, 33:4, 29-32.


THE IMPACT OF CRACKS ON DIFFERENT AGE OF CONCRETE

SMALL-SIZED HOLLOW BLOCK MASONRY

Yin Wang, Zhijian Shi, Yadi Hu College of Urban and Rural Construction, Agricultural University of Hebei ，Baoding 071000,P.R. China Abstract: In the latter part of concrete small-sized hollow block masonry cracks after using many reasons, including the design factors that are not in place, the material factors, construction factors, climate, natural conditions, factors, etc., in which the age of c concrete small-sized hollow block length of cracks in the masonry of the impact is different. Cracks in the masonry structure will cause adverse effects, affecting the use of construction and durability, this paper field test, to identify non-instars of concrete small-sized hollow block masonry cracks on impact. To reduce or prevent age of concrete hollow block the effects of crack, to strictly control the construction of block wall on the age, age as long as possible, especially in areas influenced by temperature, such as wall, the top wall, should pay attention to blocks of age, in a possible, concrete hollow block age should be greater than 42 days, the best age to block more than 56 days is better. Keywords: Concrete small-sized hollow block, crack, age of concrete block 1 INTRODUCTION

Cracks in the late after use of the concrete small-sized hollow block masonry for many reasons, of which the design is not in place, material factors, construction factors, climate, natural conditions, factors, including age concrete small hollow block of the length of the masonry cracks .

Block forming the conservation, the shrinkage rate of 0.35~0.50mm/m than clay brick wall contraction .Shrinkage will produce considerable stress,when the masonry mortar strength, cohesion, and certain areas of mortar joint is not full, the wall there will be shrinkage cracks . In particular, the rush period, the conservation of wet block on the wall, natural shrinkage, wet block mortar joint around the hairline cracks.

“Concrete small-sized hollow block building technical regulations”JGJ/T14-2004, “Technical specification, masonry structure constructed detail”02G01-2,to prevent cracks also made provisions effectively prevent and reduce cracks played a role.

2 THE AGE OF THE CONCRETE SMALL HOLLOW BLOCK CRACKS

In order to test the age of concrete small hollow

block masonry with manufacturers consultation, not the age of the block in the block manufacturer to cover production houses, masonry masonry management space plane shown in Figure 1 :

Production management space is a single layer, housing bay 3300mm deep into 4200mm, the height is 3200mm .

In-situ concrete roof, concrete strength C20, 60 thick polystyrene board roof insulation, the outside corner and three core columns, T- walls have four core columns, core pillars anchored into the ground beam and floor, there is no ring beam, the basis of strip foundation, each room has a 900mm door, leaving the stack 200mm, 1500mm windows a center set, the external walls of face brick, interior wall plaster walls, brush the white latex paint . Roof hipped color steel roof .

In order to test a unified environment, the block is so arranged that : used for external walls is the age of 84 days of the block, ② ③ axis transverse wall the age of 21 days of the block, ④, ⑤ The axis of the transverse wall with age for the 28-day block, ⑥, ⑦ axis crosswall with the age of 42 days block, ⑧, ⑨ axis of the cross wall of the age of 56 days of block . All the blocks using the MU10, masonry mortar to adopt Mb5 mixed mortar .

Housing built after a year and a half, the cracks generated by the housing survey, 科

学出版社


www.abook.cn

·296·

different wall vertical crack width as shown in Table 1

Figure 1. Production management space level map

Table 1. Crosswall cracks of different width (mm)

crosswall

Crosswall

of ①⑩

axis

Crosswall

of②③

axis

Crosswall

of④⑤

axis

Crosswall

of⑥⑦

axis

Crosswall

of⑧⑨

axis

width of

crack vague 0.5~0.8 0.1~0.2 vague invisible

It can be seen from Table 1, concrete

small-sized hollow block masonry wall, the size of the cracks is indeed the age of the block length of cracks ② The ⑨ axis intermediate transverse wall can be seen, the age the longer the period, the smaller the cracks, age over 28 days, produce a very small cracks, when the age reaches 42 days, cracks were visible when the age reaches 56 days, no cracks apparent .

Table 1 shows the concrete small hollow block production later in the process of formation strength, concrete block may also occur shrinkage deformation, if the short block of the age of the block has not yet fully shrinkage deformation masonry masonry masonry masonry in a row from the subsequent shrinkage deformation of cracks, the age of the shorter block performance is more obvious, the longer the age of block masonry masonry cracks more smaller. ① and ⑩ shaft external wall cracks are

barely visible, and ⑥ ⑦ axis transverse wall interior wall similar reason, ①, the ⑩ shaft external wall is not just the cracks and to shrinkage deformation of the block, due to the external walls insulation measures, cracks in the wall and indoor and outdoor temperature difference is related to both work together to make wall cracks

3 IMPROVEMENTS

Not the age of the concrete small hollow block cracks is different, the smaller the age of the longer block masonry cracks in the wall, on the contrary, the greater the cracks . Therefore, during the construction of small concrete hollow block masonry, approach the age of the concrete small hollow block of control.

In order to guarantee the quality of the concrete small hollow block masonry, test analysis comparison of several measures :

(1) Strict control of the small hollow concrete block, age, approach the block must have a product certificate .

(2) Age-site construction of concrete small hollow block should be greater than 42 days, and

·297·

the best selection of concrete small hollow block age greater than 56 days .

(3) Generally small hollow concrete block factory age of 28 days, during construction preparation conditions of the site license, the block should be completed before ordering, early approach, moisture, rain at the scene to take measures to ensure that the block on wall masonry of age more than 42 days is appropriate .

(4) Such as facades, and the top wall masonry affected by temperature of the wall, try to use the age of greater than 56 days of the block, to ensure the quality of the masonry .

(5) Possible after the completion of the main masonry, post-renovation try to stagger the time interval, so that the concrete small hollow block as far as possible complete the shrinkage deformation .

4 RECOMMENDATIONS

To reduce or prevent the cracks of different ages of concrete small hollow block, to strictly

control the block wall construction on age, age as long as possible, especially by the temperature greater impact on areas such as external walls, the top-level wall, should be noted that the block of the age of possible concrete small hollow block age should be greater than 42 days, it is best used to block better with age greater than 56 days .

REFERENCES

Gao Xiaoping, the Jingui real, Sanchi Cheung .02 G01, masonry structure construction detail. Beijing, China Planning Press, 2003 .

The Sun cyanide Ping, Tangdai, new, Yan Xi .JGJ/T14-2004 Concrete Small Hollow Brickwork Construction Technical Regulations. Beijing, China Building Industry Press, 2004 .

Zhang Guan Fu, Yong-Feng Liu . (2000). Of the crack of the Concrete Small Hollow Block Construction Design. New wall Materials and Construction, 1:35-37.

科

学出版社


www.abook.cn


INNOVATIVELY DESIGNED EXTERNALLY FRP REINFORCED EXPANSIVE

CONCRETE BEAMS

Qi Cao1, Zhongguo John Ma 2 ¹ Assistant Professor, School of Civil Engineering, Faculty of Infrastructure Engineering, Dalian University of Technology, Dalian 116024, P.R. China ² Associate Professor, Department of Civil and Environmental Engineering, The University of Tennessee, Knoxville, TN, 37996, USA Abstract: The major cause of cracking in bridge decks, concrete pavements, as well as slabs on grade is restrained shrinkage of the concrete. Expansive concrete (EC) and fiber-reinforced polymer (FRP) are explored to develop an innovative structural system as one possible method of delaying concrete cracking and eliminating steel corrosion. A series of “coffee can” tests were carried out to measure and compare the expansion of concrete from two candidate materials. The selected EC candidate mix was then optimized to get the maximum expansion as well as a decent concrete strength. The optimized expansive concrete mix was used to make FRP- expansive concrete beams. Both glass FRP (GFRP) sheets and carbon FRP (CFRP) sheets were used for comparison. Initial cracking loads and ultimate loads were compared for the proposed structural beams. Keywords: FRP, expansive concrete, lime-system, prestress 1 INTRODUCTION

The major cause of cracking in bridge decks, concrete pavements, as well as slabs on grade is due to restrained shrinkage of the concrete and warping stress. One possible method of eliminating the cracking is to use expansive concrete. Cusick and Kesler (1977) conducted a series of studies on the behavior of ettringite-system expansive concrete used in bridge decks. It was documented that bridge decks with expansive concrete did not crack or had only a few cracks compared with ones with Type I cement concrete. Russell (1978) undertook a study to investigate effects of type of cement, type of aggregate, percentage and position of reinforcement, slab thickness, and curing conditions on the expansion and subsequent shrinkage of concrete made with expansive concrete. It was reported that heavily reinforced slabs had less expansion than lightly reinforced slabs. Phillips et al. (1997) found that expansive concrete minimized shrinkage cracking of concrete based on the evaluation of practical application of bridge decks with expansive concrete. After those early studies on ettringite-system cement discussed above, Russell et al. (2002) studied lime-system expansive concrete mixes to

develop an expansion between 0.03% and 0.1% while keeping a minimum concrete strength of 27.6 MPa (4000 psi). While Russell et al. (2002) has used lime-system cement up to 10%, no work has been reported using lime-system cement above 10% to study the expansion behavior of expansive concrete. Tests of using lime-system cement more than 10% were conducted in this study. It was believed to be necessary to achieve higher expansion to overcompensate the shrinkage and get pre-stressing effect from “external” reinforcement. Generally, expansive concrete is defined as a concrete that is made of expansive cement that expands to an amount equal to or greater than the following shrinkage. When concrete is restrained by reinforcement, compression stress will be generated in the concrete and this could offset tensile stress caused by shrinkage (ACI Committee 223, 1998; ASTM C 845, 2004). ACI 223 (1998) characterizes typical length change properties over time for both conventional and expansive concrete as shown in Fig.1. It indicates that the final expansion of expansive concrete can be designed to be greater than the anticipated shrinkage, resulting in a residual pre-stressing of the reinforcement in the concrete after the initial shrinkage has occurred.

·299·

Figure 1. Typicl length change characteristics of expansive concrete and Portland cement concrete (ACI223 1998)

There are two ways to produce expansive concrete. One is to use expansive type cement, also called shrinkage compensating cement, such as type K, M and S. These three types of expansive cement differ by the amount of aluminate compound they contain. The other is to add on expansive cement component to the cement. In essence, the property of Type K cement differs from those of regular Portland cement only with respect to percentage of sulfates and aluminates. The other kind of expansive cement is lime-system cement. Accordingly, there are two mechanisms of achieving expansive concrete with an added component. The first one is called “ettringite crystal” development which occurs with ettringite-system cement. The second is calcium hydroxide platelet crystals development that occurs with lime-system cement. For ettringite crystal development, the introduction of water to the aluminate compound in the cement creates a hydrothermal solution that forms a precipitate: an expansive, needle-like, amorphous mineral called ettringite. It has a specific gravity of 1.7 and is comprised mainly of water. It is the formation of ettringite that causes the expansion of expansive concrete made with expansive cement. There is one concern about so called “delayed ettringite formation (DEF)”. It is believed to be one cause of early concrete deterioration (Moffat, 2005). Moffat (2005) pointed out that heat is required for the formation of ettringite. In fact, the more heat, the greater the initial expansion is. The expansion of lime-system expansive concrete obtained through the formation of calcium hydroxide platelets rather than ettringite is an advantage of lime-system because it removes the

possibility of DEF formation. The objectives of this study were to develop the expansive concrete mixes and types of FRP composites required to produce the proposed externally FRP reinforced expansive concrete system and to demonstrate the structural efficiency of the proposed system. Both carbon fibers composites and glass fibers composites were considered and compared. The initial cracking load of the proposed system was determined from structural tests.

2 EXPERIMENT PROGRAM

2.1 Materials

2.1.1 Expansive concrete materials

As stated earlier, ettringite-system cement and lime-system cement are the most commonly available and used cement to make expansive concrete. Both ettringite-system expansive concrete and lime-system expansive concrete were chosen to do the expansion test. These two different expansive cements adding at the same dosage for expansive concrete were compared in this study to find out which can give a higher total expansion. The expansion behaviors during this time period were also compared. Their typical mix designs were obtained. The same admixture ratio was selected as 19% of Type I cement for both mix designs. Table 1 and Table 2 show the mix designs for ettringite-system expansive concrete and lime-system expansive concrete, respectively.

2.1.2 Coffee can test

For the expansive concrete expansion test method, ASTM C 878/C 878M (2003) gives a standard test method for determination of expansion of shrinkage-compensating concrete. This method applies for the specimen that is internally restrained by a threaded rod affixed to end plates. ASTM C 157/ C157M (2006) presents a test method to determine the length change in concrete. The concrete specimens are stored in the lime-saturated water and their lengths measured by a comparator at any age to get the length change at that time period. It should be pointed out that ASTM C 157/ C157M (2006) length change specimen in this method is

科

学出版社


www.abook.cn

·300·

Table 1. Mix design for ettringite-system EC

Mixture Proportions

Portland Cement Type I/II, kg/m3

(lb/yd3) 279(470)

Ettringite-system Cement, (19%), kg/m3 (lb/yd3)

53(90)

Fine Aggregate, (ASTM C-33 Sand), kg/m3 (lb/yd3)

649(1095)

Coarse Aggregate, (ASTM C-33 #57 Stone), kg/m3 (lb/yd3)

1067(1800)

Water, L/m3 (gallon/yd3) 203(41)

Water Reducer, L/m3 (oz/yd3) 0.929(24)

Table 2. Mix design for lime-system EC

Mixture Proportions, kg/m3 (lb/yd3)

Portland Cement Type I/II 310(523)

Lime-system Cement, (19%) 59(99)

Fine Aggregate, (ASTM C-33 Sand) 713(1203)

Coarse Aggregate (3/8’’), (ASTM C-33 #8 Stone)

876(1478)

Coarse Aggregate (3/4’’), (ASTM C-33 #67 Stone)

214(361)

Water 166(280)

Water Reducer, L/100kg (oz/100lb cementitious material)

1.3(20)

under no restraint. In the proposed system, however, expansive concrete is restrained by the “external” FRP reinforcement. Both ASTM methods (ASTM C 878/C 878M, 2003; ASTM C 157/ C157M, 2006) are not appropriate for the restrained condition. Thus, the “coffee can” test method is used to measure the expansion of expansive concrete. By using this method, values of expansion rate could also be obtained roughly. It is believed that the coffee can method is adequate to compare and screen two different expansive concrete sources. The disadvantage of using the coffee can test method is that it is not able to capture the reading of shrinkage after the expansion happens. In the coffee can test method, two dial gauges were placed at the quadrant points at the location in the middle depth of the can and were used to measure the diameter change

over the time period of expansion. The dial gauges were fixed on metal stands which were fixed on the table. The gauges were set in position perpendicular to the can’s circular surface. Test instrument set up is shown in Fig.2. The environment condition was at a temperature of 75 and a relative humidity of 64%.

Figure 2. “Coffee can” test setup

Seven coffee can tests were conducted. Two of them were conducted for selection of expansive concrete sources and five of them were tested for optimization of expansive concrete mixes. The original diameter of the can was measured using a clamp and digital calipers. Once everything was set, the freshly made concrete was poured into the coffee can and filled up to within 25 mm (1 in.) to the top. It needs to be assured that the dial gauges have some initial readings. The optimum location for measurement is at the lowest 1/4 to 1/3 of the can height. Measurements were observed hourly during the first 10-12 hours. After eight hours, water was added on top of the can with 25 mm (1 in.) depth. Measurements were continued daily at 24 hours after casting the specimen. Then readings were recorded for about seven days after which the expansion has terminated. Three cylinders from each mix were tested for compressive strength at seven days based on ASTM C 39/C 39 M (2005). The total expansion of the two mixes over seven days is shown in Fig.3. From Fig.3, we can see that ettringite-system mix achieves the maximum expansion at about six days with the value of 0.06%, while lime-system mix reaches the maximum expansion at 11 hours after casting with the value of 0.15%. By comparison, lime-system mix gives a higher value of expansion compared to ettringite-system mix along with a higher initial increasing rate. This summarizes that lime-system expansive concrete. Mix shows a higher total expansion and is chosen for the next step of the test.

·301·

Figure 3. Comparisons of total expansion-time curves of ettringite-system EC and lime-system EC

2.1.3 Optimization of lime-system expansive concrete mixes

A series of concrete mixes, compressive strength and expansion tests were conducted to seek an optimal mix that can achieve maximum expansion while achieving concrete compressive strength at 27.6 MPa (4000 psi) or more. In the first step, the concrete mix shown in Table 2 was used to make concrete and test the compressive strength and expansion. It was found that the balance between the need for strength and need for expansion was not attained. In the second step, a mix specification for high performance concrete (HPC) was modified to seek the balance point. The cementitious materials were replaced at the same proportion with lime-system cement. Table 3 shows the final expansive concrete mix used for the following beam test. Testing results are shown in Fig.4. It can be seen that the balance point is close to 20% dosage, but the strength is not desirable at this point. At 15% lime-system cement dosage, concrete strength is about 41.4 MPa (6000 psi) which exceeds 27.6 MPa (4000 psi), and the maximum expansion is 0.25%, which could compensate for the shrinkage of concrete. Therefore the expansive concrete mix with 15%

Table 3. Mix design for high performance lime-system EC

Mixture Proportions, kg/m3 (lb/yd3) Portland Cement Type I/II 541(913) Lime-system cement, (15%) 116(196) Fine Aggregate 937(1580) Coarse Aggregate, (1/4’’) 398(672) Fly ash 155(261) Silica fume 77(130) Water 178(300) High range water reducer 26(44.2)

lime-system cement is selected for the next step test.

Figure 4. Effect of lime-system cement dosage on compressive strength and maximum expansion of EC (Note: 1 MPa=0.145 ksi)

2.1.4 Fabrication of FRP sheets

In this research, both glass fiber-reinforced polymer (GFRP) sheets and carbon fiber-reinforced polymer (CFRP) sheets were fabricated by a hand lay-up process in the laboratory. The materials for making GFRP are bi-axial glass fiber, unsaturated polyester (Isophthalic) and hardeners (MEKP). Based on GFRP specimen design suggested by material supplier, the weight of the resin was twice the weight of the fiber and a 10% extra was used to compensate for the loss of resin due to fabrication. The weight of hardener was 1% of the resin and it was mixed well with resin before application. For the hand lay-up process, a thick glass plate was prepared on a flat surface, and a mold release agent was treated on the surface of the mold. The mold dimension was 533 mm (21 in.) long by 152 mm (6 in.) wide and 152 mm (6 in.) high. Then glass fibers were laid up by fiber orientation on the four sides of the mold. Unsaturated polyester resin was mixed with hardeners in glass container by stirring with a glass rod. The mixed resin system was used to wet out fibers and bond the fibers together. The mold was released after three days and the GFRP box samples were trimmed and cured at room temperature. Two specimens were made and used for testing for each of three-layer and five-layer GFRP. Carbon fiber and epoxy resin as well as resin hardener were used to fabricate CFRP specimen. The procedures to fabricate CFRP are basically similar as those for GFRP. The molds for CFRP fabrication were “internal” molds instead of “external” molds as used in GFRP fabrication. 科

学出版社


www.abook.cn

·302·

The fiber was cut first according to the beam dimension and then the molds were placed with plastic cover. The resin system for CFRP fabrication was epoxy resin mixed well with hardener. The mixing ratio of epoxy to hardener is 100:34.5 by weight according to the material provider. The mixing was done in five minutes at full speed of the drill mixing system. Then the mixed resin was applied to the carbon fiber by squeezing the fiber with a roller for evenly distributing the resin on the fiber. Once the fiber was saturated with resin, the mold was wrapped with the pre-made fiber reinforced polymer. The specimens were kept at room temperature and humidity for 24 hours until releasing the molds out. After that, the specimens were cured for seven days before pouring concrete. Two specimens were made and used for testing for each of one-layer and two-layer CFRP.

2.2 Beam Tests

2.2.1 Testing procedures

Table 4 lists the experiment matrix of FRP/concrete beams. Four beams were tested for each mix and each type of FRP. Fig.5 shows the instrumentation of the beam specimen. The top and bottom layers are FRP layers. The top view and bottom view are the same. Concrete was made and poured into the FRP specimens. After casting, the specimens were cured by covering the top surface with wet burlaps (ASTM C 192, 2007). Based on the curing scheme, the burlaps were kept wet constantly during first 11 days at a room temperature of 75ºF and a relative humidity of 76%. After that, the burlaps were removed and room temperature and relative humidity were changed to 84ºF and 61% respectively. The strain data were collected for 28 days. After 28 days, the room conditions were changed back to the original with a temperature of 75ºF and a relative humidity of 76%.

Table 4. Matrix of FRP/concrete beam tests

Reinforcement type Reinforcement ratio HPC EC 3-layer 1 1

GFRP 5-layer 1 1 1-layer 1 1

CFRP 2-layer 1 1

Fig.5 shows the four point bending test setup (ASTM C 78, 2008). The span of the beams was 457 mm (18 in.) and the distance between the

loads was 152 mm (6 in.). The simply supported beams were loaded at a rate of 0.254mm/min (0.01 in. / min).The beams were instrumented to record load, deflection and strain. White paint was sprayed on both sides of the beam to facilitate observing the concrete crack’s development during the loading procedure. Cylindrical concrete specimens were tested for compressive strength according to ASTM C 39/C 39M.

Figure 5. Beam bending test setup

2.2.2 Initial cracking load and ultimate loads

The beam test results are summarized in Table 5. It shows the initial cracking load and ultimate load for the eight tested beams. The cracking values indicate that the differences between HPC and expansive concrete decrease as the axial stiffness of FRP reinforcement increases. Table 5 also shows that the ultimate loads of CFRP specimens are higher than those of GFRP specimens. For GFRP specimens, 5-layer specimens show a higher ultimate load than 3-layer specimens. For CFRP specimens, 2-layer specimens show a higher ultimate load than 1-layer specimens. For the same FRP type and layer, it shows that the expansive concrete specimens have higher ultimate loads than HPC specimens.

Table 5. Summary of the bending test results

Specimen Initial cracking load

(kN) Ultimate load (kN)

HG1 59.2 148.6 EG1 42.3 178.8 HG2 56.5 219.7 EG2 42.3 229.5 HC1 62.3 217.5 EC1 55.2 294.5 HC2 72.5 321.2 EC2 71.6 384.8

Note: 1kN=0.225kip. Note: H: HPC, S: EC, G: GFRP, C: CFRP, the number after letter indicates reinforcement ratio.

·303·

Fig.6 shows the effect of the axial stiffness of FRP reinforcement on the cracking load (Pcr) between HPC and EC specimens. Different concrete strength (f 'c) and cross section dimensions (b, h) were adjusted. The cracking load at “EA=0” was calculated based on tensile cracking of plain concrete beam at modulus of rupture (fr) of concrete. We can see that, as FRP reinforcement stiffness increases, the cracking loads of both EC and HPC increase. It is also shown that EC has a higher cracking capacity than HPC at all tested stiffness points. This is due to expansion of expansive concrete and confinement of FRP reinforcement. Expansive concrete specimens have a residual pre-stressing (“P/A”) effect which HPC specimens do not have. P/A increases as the FRP reinforcement stiffness increases and it helps to delay the cracking and increases Pcr. This demonstrates that the pre-stressing effect increases as the axial stiffness of the FRP reinforcement increases.

Figure 6. Effect of the axial stiffness of FRP reinforcement on Pcr (Note: 1 kN=0.225 kip)

3 CONCLUSIONS

Based on the experimental investigation carried out in this paper, the following conclusions are made:

(1) The proposed hybrid FRP reinforced expansive concrete system shows a good perspective to delay concrete cracking and improve the ultimate capacity.

(2) Lime-system expansive concrete mix shows a higher expansion than ettringite-system expansive concrete mix. The mix using 15% lime-system cement replacement of Type I cement gives the desired strength and expansion in balance.

(3) Cracking load results indicate that the prestressing effect (P/A) for expansive concrete specimens increases as the axial stiffness of the FRP reinforcement increases.

(4) Based on the testing tests, at the same

axial stiffness of FRP reinforcement, expansive concrete specimens show higher cracking resistance and ultimate load capacity.

ACKNOWLEDGMENTS

The financial support to the second author provided by the National Science Foundation – NSF CAREER program (CMS – 0550899) is gratefully acknowledged. Material donations received for specimen fabrication from International Admixtures Incorporated (IAI), Fyfe Company and LLC, CTS Cement Manufacturing Corporation are deeply appreciated. Communications with Robert Gulyas of BASF Construction Chemicals, LLC on the historical development of expansive concrete are also acknowledged.

REFERENCES

ACI Committee 223. (1998). Standard Practice for the Use of Shrinkage-Compensating Concrete. American Concrete Institute, Farmington Hills, MI.

ASTM C 157/C157M. (2006). Standard Test Method for Length Change of Hardened Hydraulic-Cement Mortar and Concrete. ASTM Annual Book of Standards, Volume 04.02, American Society of Testing and Materials, West Conshohocken, PA.

ASTM C 192. (2007). Standard Practice for Making and Curing Concrete Test Specimens in the Laboratory. ASTM Annual Book of Standards, Volume 04.02, American Society of Testing and Materials, West Conshohocken, PA.

ASTM C 39/C39M. (2005). Standard Test Method for Compressive Strength of Cylindrical Concrete Specimens. ASTM Annual Book of Standards, Volume 04.02, American Society of Testing and Materials, West Conshohocken, PA.

ASTM C 78. (2008). Standard Test Method for Flexural Strength of Concrete (Using Simple Beam with Third-Point Loading. ASTM Annual Book of Standards, Volume 04.02, American Society of Testing and Materials, West Conshohocken, PA.

ASTM C 845. (2004). Standard Specification for Hydraulic Expansive Cement. ASTM Annual Book of Standards, Volume 04.02, American 科

学出版社


www.abook.cn

·304·

Society of Testing and Materials, West Conshohocken, PA.

ASTM C 878/C 878M. (2003). Standard Test Method for Restrained Expansion of Shrinkage-Compensating Concrete. ASTM Annual Book of Standards, Volume 04.03, American Society of Testing and Materials, West Conshohocken, PA. Cusick, R.W., and Kesler, C.E. (1977). Behavior of Shrinkage-Compensating Concretes Suitable for Use in Bridge Decks. T & AM Report No 372, Final Summary Report, University of Illinois, 1977, 1-28.

Moffat, B. S. (2005). Shrinkage-Compensating Concrete: An investigative Study. Metropolis

& beyond proceedings of the 2005 Structures Congress and the 2005 Forensic Engineering Symposium, New York, NY, 1033-1043.

Phillips, M. V., Ramey, G. E., Pittman, D. W. (1997). Bridge Deck Construction Using Type K Cement. Journal of Bridge Engineering, 2:4, 176-182.

Russell, H. G., Stadler, R. A., and Gelhardt, H. G.(2002). Shrinkage-Compensating Concrete Made with an Expansive Component. Concrete International, 24:8, 107-111.

Russell,H.G. (1978). Performance of Shrinkage-Compensating Concretes in Slabs (RD057.01D), Portland Cement Association, 1978.


BEHAVIOR OF SERC SHORT COLUMNS WITH H-ROLLED STEEL UNDER

AXIAL COMPRESSION

Chenxia Wang 1,2, Ziyan Guo 2 ¹ Department of Civil Engineering, Nanjing University of Aeronautics and Astronotics, Nanjing 210016, P.R. China ² Architecture and Civil School, UST Inner Mongolia, 014010, P.R. China Abstract: A new style for encased steel concrete composite structure is presented, which is called Encased H-rolled Steel Concrete composite columns. According to the full range analysis about this kind column under axial compression, the factors affecting the ultimate bearing capacity are analyzed. Then contrasting SERC short columns on short columns of steel reinforced concrete, the advantages of the SERC columns are recognized. Keywords: Encased H-rolled steel concrete, short-columns, full range analysis, ultimate load capacity 1 INTRODUCTION

Steel Encased Reinforced Concrete（SERC）is one form of steel concrete composite structure. Early encased steel structure mainly including encased angle steel. Then, the steel tube concrete composite structure has a wide range of applications and development. In recent years, The SERC structure applications to China’s residential construction of steel structure. Foreign says encased H-rolled steel concrete composite structure for Partially Encased Composite Columns. This SERC structure mainly

Figure 1. H-rolled steel of SERC column section form 1.H-rolled steel 2. Longitudinal reinforcement 3.Stirrups 4.concrete 5.studs 6.steel tie bar 7.fillet welds8.the point of weld

made of H-rolled steel(welding or hot-rolled), longitudinal reinforcement, stirrup, stud, steel tie and so on. Its section has various forms, As shown in Fig.1 shows.From the study in foreign countries and see. This section of the component have more research is America, Canada and some European countries. Some of them has been patented technology. Domestic studies in this field mainly made of H-rolled steel(welding or hot-rolled), longitudinal reinforcement, stirrup, stud, steel tie and so on. Its section has various forms, As shown in Fig.1 shows. Some of them has been patented technology. Domestic studies in this fieldis less, Xi’an university of architecture and technology to this form of beam component had research. This beam component is more trouble for site construction. Usually, complete the concrete of beam steel ribs before assembly, then hoisting; literature sauthor used on both sides of the casting: the first pouring one side of concrete, wait for it reached the early strength. Then pouring in the other side of the concrete. Use this way to solve the test problems. This paper studies the Encased H-rolled Steel composite columns and the main consideration to construction of factors. In the actual application of structure column for more feasible. It can form framework with beam steel and used to build high-rise and super-tall buildings. Especially, if this form of column in the steel structure housing, not only can reduce the

科学出版社


www.abook.cn

·306·

components section area in pure steel structure house but also increase effective use area. Design of components in this paper and foreign research is different, Foreign only do hot H-rolled steel and welding thin-walled H-rolled steel research. Design of H-rolled steel in this paper for ordinary welding H-rolled steel and don’t happen local buckling phenomenon like welding thin-walled H-rolled steel.

2 BASIC ASSUMPTIONS

(1) Component under pressure from start until destruction, section always keeps the plane deformation.

(2) Between steel and concrete without slippage and always to keep working together before destruction.

(3) Material of stress-strain curve known. (4) Steel don’t happen local buckling. In the

test, flange and plate’s flakiness ratio in H-rolled steel satisfy the requirements of the Steel Structure Standard.

3 MATERIAL CONSTITUTIVE RELATIONSHIP

3.1 Reinforced and Steel Constitutive Relationship

In this paper use completely elastic-plastic’s Double line model, Apply to long flow’s low strength steel. Model’s Mathematical expression as follows:

s y ≤ s s sE (1)

,y s s h ≤ ≤ s yf (2)

3.2 Core Concrete Constitutive Relationship

Although the concrete only three directions stress by constraint from steel. But in the test design, another direction with the steel tie to concrete has certain restriction. Suppose ensure

that concrete in three mono-axial compressive states, between tie bar and plate core concrete stress-strain relationship curves use improved Kent-Park model. As shown in Figure 2 shows.

OB: ck ≤ , 2

2c

c c

kfk k

(3)

OB c cuk ≤ ≤

1c m ckf Z k (4)

Figure 2. Restrictive concrete stress-strain curvel

4 SPECIMEN DESIGN

When the analysis, design the three groups of specimens, were Consider contain steel rate, steel tie spacing, concrete strength to component performance influence. Specimen’s length is 800mm, Steel tie diameter is 10mm, Yield strength is 235 N/mm2, reinforced and steel Elastic modulus 6 22.06 10 N/mmssE . The

compressive strength of concrete cube take standard, Concrete cover depth is 25mm, chart1-chart3 is the parameters of specimen and calculation SERC The axial compressive limit bearing capacity of short column result.

Table 1. The first group of specimens（crtain steel rate different）

Specimen Numbers SERC1-1 SERC1-2 SERC1-3

The strength of concrete level C30 C30 C30

H-rolled steel ection size (mm）

( w fB H t t ) 200×200×6×6 200×200×6×8 200×200×6×10

·307·

(continued)


Contain steel rate%

Configuration tie situation

Calculation Nu（kN）

NuCorresponding strain c

8.82

10@100

1503.5

0.0022

10.76

10@100

1671.5

0.0022

12.70

10@100

1839.5

0.0022

Table 2. The second group of specimens（steel tie spacing different）




( w fB H t t ) 200×200×6×8 200×200×6×8 200×200×6×8

Contain steel rate%




10.76

10@100

1671.5

0.0022

10.76

10@150

1648.6

0.0021

10.76

10@200

1637.6

0.0021

Table 3. The third group of specimens（The strength of concrete ranks different）




( B H t tw f ) 200×200×6×6 200×200×6×6 200×200×6×6

Contain steel rate%




8.82

10@100

1503.5

0.0022

8.82

10@100

2085.6

0.0021

8.82

10@100

2446.1

0.0021

5 CONFIGURATION H-ROLLED STEEL SERC COMBINATION COMPONENT STRUCTURE PERFORMANCE ANALYSIS

5.1 Load-deformation Relation Curves

This paper use strength superposition principle, Through nonlinear analysis program sure SERC Axial compression of short column ultimate bearing capacity, Analysis the influence of bearing capacity factors and draw the load-deformation relation curves, As shown in Fig.3-Fig.5 shows:

Figure 3. The steel ratio is not the same as the load of-the strain curve

Concrete strain

科

学出版社


www.abook.cn

·308·

Figure 4. Steel tie the load is not the same as the distance-strain curve

Figure 5. The strength of concrete is not the same as the level of load and strain curve

5.2 Result Analysis

Through the analysis of Table 1 to the data in Table 3, Find the first group specimen reach the limit bearing capacity corresponding strain

c

is 0.0022; Find the second group specimen reach the limit bearing capacity corresponding strain

c , With the improvement of Steel tie spacing is

0.0022,0.0021,0.0021,show the trend of decrease. Find the third group specimen reach the limit bearing capacity corresponding strain c , With

the improvement of strength grade is 0.0022, 0.0021, 0.0021, show the trend of decrease too. From load-strain curve (Fig.3-Fig.5) found, Contain steel rate and the strength of concrete level are the main factors to affect the SERC axial compressive limit bearing capacity of short column, With the steel rate increase, Limit bearing capacity increased; The spacing between steel tie to limit bearing capacity influence is very small. With The spacing between steel tie become small, the curve of the late more gentle. This shows that the ductility of the components

increases. The trend of the present anatomizes with c .

The Fig.5 know, The strength of concrete level increase, The bearing capacity of the pillars increase too. But Later curve are steeper then normal concrete. In other words, The ductility of the pillars down, The strength of concrete low component, Its ductility are better then High strength concrete.

The Fig.3 found, although with the increase of contain the steel ratio, Limit bearing capacity of column become increase too. But The curve of the late basic parallel. That is the ductility of the pillars did not improve, this and the first group specimen reach the limit bearing capacity corresponding strain c is 0.0022 the coincide.

Through the above analysis, can see such as the contain steel ratio will not affect the ductility of the component, If you want to improve the ductility of the components, only increase the steel ratio is not feasible, should by reasonable allocation of steel tie ways come true.

6 CONCLUSIONS

From the above analysis can be seen, SERC axial compression short column has the following advantages:

(1)High bearing capacity, It can be used in The main stress components of high-rise and super high-rise building. H-steel casting concrete to strengthen the steel structure stability, stiffness and strength, If structural measures is properly that can ensure steel and concrete can work together until the damage. Application of high strength concrete can greatly improve the bearing capacity of the SERC pillars, But ductility is lower than ordinary strength concrete.

(2)The steel ratio and the strength of concrete level is influence the SERC axial compressive limit bearing capacity of short column the main factors. With the steel rate increase and improve the strength of concrete, the ultimate bearing capacity increased; As in the steel flange reasonable allocation steel tie, so ductility is superior to a steel reinforced concrete structures.

(3)Convenient installation, Beam-column joints is easy to deal with, Shorten the construction period.

(4)Fire resistance performance is better than pure steel structure. Bare outer surface area of the steel than pure steel structure decreased more than half, Greatly improves the fire resistance of

·309·

components. At the same time in a follow-up study deserves

further investigation of have the following: (1) This procedure is only considered Steel tie

volume stirrup ratio to constraint of concrete strength improve influence, Without considering the steel to restriction of concrete effect.

(2) The Constraints of concrete stress-strain curve used Suitable for high strength concrete improvement the Kent-Park model, whether siutable for ordinary concrete need further discussion.

(3) Because in the actual production of component will generate residual stress, rsidual stress of component of the bearing capacity of just how much, also need to be further discussed.

REFERENCES

Ministry of Construction department of Personnel and education etc. (2005). MB Light steel structure residential building

system. Beijing: China Architecture Industrial press.

O. S. Bursi, R. Zandonini.(2006). Seismic Behavior of a 3D Full-Scale Steel-Concrete Composite Moment Resisting Frame Structure. Composite Construction in Steel and Concrete , 641-652.

R. Tremblay, B. Massicotte. (1998). Experimental study on the behavior of partially encased composite columns made with light welded H steel shapes under compressive axial loads. Structural Stability Research Council 1998 Annual Technical Session & Meeting. Atlanta, Georgia, 195-204.

R. Vincent. (2000). Design and application of partially encased non-compact composite columns for high-rise buildings. Proceedings of Composite Construction IV, Engineering Foundation, Banff, Canada, 854-864.

ZHAO Qiao-rong.(2004). Outsourcing steel reinforced concrete beam properties research. Xi’an: Xi’an university of architecture and technology Master’s degree thesis.

科

学出版社


www.abook.cn


HYDRATION OF ULTRAFINE AND ORDINARY PORTLAND CEMENT AT

EARLY AGES

Jikai Zhou1,Yuanyuan Yan1,Xudong Chen1, Minquan Zhou2, Xiuming Jiao2 1College of Civil and Transportation Engineering, Hohai University, Nanjing, Jiangsu, 210098, P.R. China 2Tianhuangping Pumped Storage Power Co. Ltd. , Anji 313302, P.R. China Abstract: Experimental results of a quantitative of the hydration of ultrafine and ordinary Portland cement pastes were presented in this paper. The degree of hydration of cement pastes was determined based on the determination of non-evaporable content. The purpose of this study was to explore the mechanism of the effects of particle size distribution on hydration kinetics of cement. The data obtained from test indicates that at each age before 7 days, the non-evaporable water content of the ultrafine cement paste was higher than that of the ordinary cement paste. The fineness has a great effect on early 7 days hydration; however, the view that increased fineness improves hydration at later ages is not correct. In the later age, the rate of hydration slows down and diffusion of silicate ions through a lays of existing products becomes the rate-limiting process. The porosity of ultrafine cement paste is lower than of ordinary cement paste. The difference in porosity of two cement paste is most pronounced at early ages before 14 days. A general reduction in porosity with increasing in age is observed from test results. The ultrafine cement consistently produces higher gel/space ratios at all seven testing ages. At an age of 1 day, the ordinary Portland cement achieves 79% of its fine counterpart. Keywords: Ultrafine cement, cement paste, non-evaporable water, degree of hydration, porosity, gel/space ratio 1 INTRODUCTION

Cement producers over the years have developed several new types of binders which either have extended the area of usage of cement or have demonstrated technical and ecological advantages (Uchikawa, 1994; Sarkar et al., 2001; Marek Gawlicki et al., 2010). The ultrafine cement is attracting attention as the filler for repairing deteriorated parts and cracks in concrete structures （Persson, 1996）. Besides this, particle size distribution-controlled ultrafine cement is used as the bonding material of insulation with metal fittings in power-transition line （Neithalath et al ., 2009).

In the recent past, an ultrafine cement was used to improve oil production in nearly 200 wells in the Provost area of eastern Alberta, Canada (Sarkar et al., 2001). Kaufmann et al. (2004) reported the use of ultrafine slag cement and ultrafine Portland cement for tertiary grouting of a control shaft in the Warm Springs Dam in Sacramento. Nearly 5700 ft3 of the grout was placed in the water-filled voids and cracks in this operation. A water shutoff system was a

developed using a silicate-based polymer slurry, followed by an ultrafine cement.

The hydration of cement-based materials has been extensively investigated. Various reviews (De Schutter et al., 1995; Alina Badanoiu et al., 2011; L’Krajci et al., 2010; Dabic et al., 2000; Schindler et al., 2005; Meinhard et al., 2008) have summarized the kinetic and mechanistic arguments advanced to explain experimentally hydration phenomena observed. However, as Brown (1989) has pointed out for Portland cement, elucidation of hydration kinetics is complicated by the particle size distribution of the material. To understand the factors controlling the mechanisms of cement hydration, the factors controlling its hydration kinetics must also be understood. It is well-known that hydration kinetics and particle fineness are related (Bentz et al., 2008). However, this relationship is at best qualitatively understood. With the exception of a study by Bentz et al. (1999) of the hydration of stirred, high water-to-cement ratios cement suspensions, the effect of particle size distribution on the kinetics of cement hydration seems to have been ignored.

Experimental results of a quantitative of the

·311·

hydration of ultrafine and ordinary Portland cement pastes were presented in this paper. The degree of hydration of cement pastes was determined based on the determination of non-evaporable water content. The purpose of this study was to explore the mechanism of the effects of particle size distribution on hydration kinetics of cement.

2 EXPERIMENTAL PROGRAM

The test program consisted of the following measurements: (a) particle size distribution; and (b) thermogravimetric weight loss measurement.

The chemical composition of the ultrafine cement with 98% particles finer than 10 μm, which was the starting material, is compared with that of a Type P 52.5 cement in Table 1. ⅡFig.1 shows the particle size distributions of ultrafine (Fig.1 (a)) and ordinary cements (Fig.1 (b)) measured by laser diffraction.

Table 1. Chemical composition of ultrafine and ordinary cement (mass (%))

Oxide SiO2 Al2O3 Fe2O3 CaO MgO Na2O K2O TiO2

ultrafine 19.36 4.89 1.70 63.98 1.07 0.12 0.70 0.24

Ordinary 21.69 5.01 3.46 66.37 0.88 0 0.65 0

(a) Ultrafine cement

(b) Ordinary cement

Figure 1. Measured differential particle size distributions for the two cements

2.1 Sample Preparation

Cement paste samples were mixed using a paddle mixer following a procedure suggested by Williams et al. (1999), which is similar to ASTM C-305. Cement and water were initially mixed at 140 rpm for 30s, followed by a pause for 1 min before mixing for another 2.5 min at 285 rpm. Cement paste adhering to the sides of the mixing bowl was scraped and the entire mixture was mixed for another 2.5 min at 285 rpm. The pastes were mixed in a mechanical mixer with 5-1 capacity, and cast in a 70.7×70.7×70.7-mm steel mould. Cement paste samples with the water-to-cement ratio equal to 0.45, were evaluated in the test program. The steel moulds were removed 24 h after mixing, and the cubes were cured in water at 25 .℃

2.2 Thermogravimetry

Thermogravimetry (TG) has been used frequently for studying the hydration of Portland cements because the hydration reaction between the mixing water and cement can be reversed when the hardened cement paste is subjected to high temperature (Voigt et al., 2004; Wang et al., 2011). Samples of approximately 150 mg were taken by crushing cubes of cement paste. The samples were heated up to a temperature of 1050 accord℃ ing to a defined heating regime. During the heating process, the sample loses weight due to the evaporation of the free and the decomposition of the chemically combined water. To distinguish between evaporable and non-evaporable water, the temperature was first increased to 105 , held at that value for 2 hours, ℃and then increased further to 1050 . The ℃heating rate was 10 /min. The result of the TG ℃is the development of the weight loss of the sample during heating. The heating regime for one of the tested cement pastes is given in Fig.2.

Figure 2. Heating regime of cement paste measured after 36 h by thermogravimetry

科学出版社


www.abook.cn

·312·

Two parameters, the amount of the total water and the non-evaporable water in the cement paste, can be derived directly from TG measurements. The difference between the initial sample weight m0 and the weight at 1050 ℃(m1050) yields the amount of the total water held in the cement paste at the time of testing. The amount of the total water needs to be corrected by the weight loss of the dry cement powder at 1050 with respect to its initial mas℃ s (L0/1050). The amount of non-evaporable water is calculated from the weight loss measured between 105℃ and 1050 corrected by the loss ℃of ignition of the dry cement powder at 1050 ℃with respect to its mass at 105 (℃ L105/1050). The amount of the two types of water is expressed in gram per gram of original cement. The equations for calculating the content of the total water (wt/c) and non-evaporable water per gram of origin cement (wt/c) are given in Eq. (1) and (2), respectively. If both quantities are known, the amount of the evaporable water per gram of original cement can be calculated with Eq. (3). The calculation of the non-evaporable water from the weight loss of cement paste after heating was first derived by Powers and Brownyard (1947).

0

0 / 1050

1050

1 1tw m

Lc m

(1)

105105 /1050

1050

1 1nw mL

c m (2)

e t nw w w

c c c (3)

The amount of non-evaporable water is commonly considered as an approximation of the chemically bound water, although much water from the interlayer spaces (which is by definition part of the chemically bound water) is lost during the drying process (Lam et al., 2000; Zhou et al., 2011). Thus, by relating the content of non-evaporable water at a certain t to that for complete hydration, the degree of hydration can be calculated (Eq. (4)). The content of non-evaporable water for complete hydration

/n completew c was determined based on the

maximum water content of the clinker phases of the used cement as can be found in the literature (Sanahuja et al., 2007). The value obtained from this calculation ( /n completew c = 0.236 and 0.227 for

ultrafine and ordinary cement, respectively) correspond well with that given by Copeland (1956), as typical for fully hydrated Portland cement pastes ( /n completew c = 0.23).

/n n

complete

w w

c c

(4)


3.1 Non-evaporable Water /degree of Hydration

The results of the determination of non-evaporabel water content are shown in Fig.3. At each age before 7 days, the non-evaporabel water content of the ultrafine cement paste was higher than that of the ordinary cement paste. However, at ages after 7 days, there is essentially no difference in the values of non-evaporable water content between the ultrafine and ordinary cement paste with the same water-to-cement ratio 0.45.

Figure 3. Non-evaporable water content relative to the cement content

Based on the data of non-evaporable water content, Results for the hydration degree are shown in Fig.4. It can be seen that the fineness has a great effect on early 7 days hydration; however, the view that increased fineness improves hydration at later ages is not correct. This may have caused by various reasons such as the amount of water required for setting and/or pre-hydration problems of that ultrafine cement. Since the experiments were performed by using the constant water-to-cement ratio, proper compaction might have not provided as the particle size of ultrafine cement is fine. On the other hand, the pre-hydration might have commenced according to the LOI (loss on ignition) values of ultrafine cement in Table 1. Very fine particles of ultrafine cement may also cause deterioration in rheology and uniformity of distribution. These very fine particles may have quickly hydrated because of high surface

·313·

area in the certain parts of the cast, which would have caused some weak zones in the whole. And in the later age, the rate of hydration slows down and diffusion of silicate ions through a lays of existing products becomes the rate-limiting process.

Figure 4. Degree of hydration of cement pastes

3.2 Porosity

The porosity influences directly many properties of cement-based materials, and is one the important factors to consider when evaluating the influences on materials’ properties (Wang et al., 1999; Pann et al., 2003; Kondraivendhan et al., 2010; Zhou et al., 2011). The use of Mercury Intrusion Porosimetry (MIP) to determine the porosities of cement pastes is quite common (Cook et al., 1999; Kumar et al., 2003). However, in this study, the porosities ( )t , were

determined using an expression based on Power and Brownyard (2008)’s model for hydration:

cement cement

cement

( / ) 1.15 0.06

1 ( / )

w ct

w c

(5)

where cement is the specific gravity of cement. In

this study, the specific gravity of ultrafine and ordinary Portland cement is 2.94 and 3.13 respectively. In subsequent discussion, t

refers to water-filled porosity in cement paste. Fig.5 depicts the variation in porosity with

time for ultrafine and ordinary cement pastes. As can been seen from Fig.5, the porosity of ultrafine cement paste is lower than of ordinary cement paste. The difference in porosity of two cement paste is most pronounced at early ages before 14 days. The hydration of ultrafine cement is accelerated by fine particle, and this

result in the increase of C-S-H gel which fills the pores of cement paste. A general reduction in porosity with increasing in age is observed, confirming that the cement paste becomes denser with proceeding hydrating of cement particles.

Figure 5. Porosity of cement pastes

3.3 Gel/space Ratio

It is well-know that the compressive strength of concrete depends on the gel/space ratio determined from the degree of cement hydration and water-to-cement ratio (Bentz et al., 1999). A gel/space ratio is defined as the ratio of the volumes of the hydrated cement to the sum of the volumes of the hydrated cement and of the capillary pores (Brown et al., 1989). For cement pastes, assuming that 1 ml of hydrated cement occupies 2.06 ml, the gel/space ratio is given by:

2.06

/c c

pc

c c

vx

v w c

(6)

where pcx is the gel/space ratio of the cement

paste, cv is the specific volume of anhydrous

cement, c is the degree of hydration of cement,

and /w c is the original water-to-cement ratio. The value of cv is equal to 0.341 and 0.319 for

ultrafine and ordinary cement respectively. The calculated results are shown in Fig.6. As

shown in Fig.6, the ultrafine cement consistently produces higher gel/space ratios at all seven testing ages. At an age of 1 day, the ordinary Portland cement achieves 79% of its fine counterpart. These increased gel/ratios at any given age are likely due to their increased reactivity and the smaller inter-particle spacing.

科

学出版社


www.abook.cn

·314·

Figure 6. Gel/space ratio of cement pastes

4 CONCLUSIONS

Experimental results of a quantitative of the hydration of ultrafine and ordinary Portland cement pastes were presented in this paper. The main conclusions are as follows:

(1) At each age before 7 days, the non-evaporable water content of the ultrafine cement paste was higher than that of the ordinary cement paste. However, at ages after 7 days, there is essentially no difference in the values of non-evaporable water content between the ultrafine and ordinary cement paste with the same water-to-cement ratio 0.45.

(2) The fineness has a great effect on early 7 days hydration; however, the view that increased fineness improves hydration at later ages is not correct. In the later age, the rate of hydration slows down and diffusion of silicate ions through a lays of existing products becomes the rate-limiting process.

(3) The porosity of ultrafine cement paste is lower than of ordinary cement paste. The difference in porosity of two cement paste is most pronounced at early ages before 14 days.

(4) The ultrafine cement consistently produces higher gel/space ratios at all seven testing ages. At an age of 1 day, the ordinary Portland cement achieves 79% of its fine counterpart.

ACKNOWLEDGEMENTS

The authors are grateful to the National Natural Science Foundation of China (No. 50979032), the Technical Project State Grid Xin Yuan Company Limited (THPKJ-2011-2), the Fundamental Research Funds for the Central Universities (Grant No. 2011B11047) and the

Research Innovation Program for College Graduates of Jiangsu Province for the financial supports.

REFERENCES

Alina Badanoiu, Jenica Paceagiu, and Georgeta Voicu (2011). Hydration and Hardening Processes of Portland Cements Obtained From Clinkers Mineralized with Fluoride and Oxides. Journal of Thermal Analysis and Calorimetry, 103,879-888.

Bentz, D.P., Garboczi, E.J., Haecker, C.J., and Jensen, O.M. (1999). Effects of Cement Particle Size Distribution on Performance Properties of Portland Cement-based Materials. Cement and Concrete Research, 29, 1663-1671.

Bentz, D.P., Sant, G., and Weiss, J. (2008). Early-age Properties of Cement-based Materials. I: Influence of Cement Fineness. ASCE Journal of Material Civial Engineering, 20:7, 502-508.

Brown, P.W. (1989). Effects of Particle Size Distribution on the Kinetics of Hydration of Tricalcium Silicate. Journal of American Ceramic Society, 72:10, 1829-1832.

Cook, R.A. and Hover, H.C.(1999). Mercury Porosimetry of Hardened Cement Pastes. Cement and Concrete Research, 29, 933-943.

Copeland, L.E. (1956). Specific Volume of Evaporable Water in Hardened Portland Cement Pastes. Journal of American Concrete Insititue, 27:8, 863-874.

Dabic, P., Krstulovic. R., and Rusic. D.(2000). A New Approch in Mathematical Modeling of Cement Hydration Development. Cement and Concrete Research, 30, 1017-1021.

De Schutter, G. and Taerwe, L. (1995). General Hydration Model for Portland Cement and Blast Furnace Slag Cement. Cement and Concrete Research, 25:3, 593-604.

Kaufmann, J., Winnefeld, F., and Hesselbarth, D. (2004). Effect of the Addition of Ultrafine Cement and Short Fiber Reinforcement on Shrinkage, Rheological and Mechanical Properties of Portland Cement Pastes. Cement and Concrete Research, 26 ,541-549.

Kondraivendhan, B. and Bhattacharjee, B. (2010). Effect of Age and Water-cement Ratio on Size and Dispersion of Pores in Ordinary Portland Cement Paste. ACI Materials Journal, 107:2, 147-154.

Kumar, R. and Bhattacharjee, B. (2003). Study on Some Factors Affecting the Results in the Use of MIP Method in Concrete Research.

·315·

Cement and Concrete Research, 33:3, 417-424.

L’Krajci, S.C., Moujmdar, M., Kuliffayova, I., and Janotka (2010). Microstructure of Portland Cement Mortar Amended by Burnt Kaolin Sand. Journal of Thermal Analysis and Calorimetry, 100, 779-787.

Lam, L., Wong, Y.L., and Poon, C.S. (2000). Degree of Hydration and Gel/space Ratio of High-volume Fly Ash/cement Systems. Cement and Concrete Research, 30, 747-756.

Marek Gawlicki, Wiesława Nocun-Wczelik and Łukasz Bak. (2010). Calorimetry in the Studies of Cement Hydration: Setting and Hardening of Portland Cement–calcium Aluminate Cement Mixtures. Journal of Thermal Analysis and Calorimetry, 100, 571-576.

Meinhard, K. and Lackner, R.(2008). Multi-phase Hydration Model for Prediction of Hydration-heat Release of Blended Cements. Cement and Concrete Research, 38, 794-802.

Neithalath, N., Persun, J., and Hossain. A.(2009). Hydration in High-performance Cementitious Systems Containing Vitreous Calcium Aluminosilicate or Silica Fume. Cement and Concrete Research, 39, 473-481.

Pann, K.S., Tsong, Y., Chao, W.T., and Lin, T.D. (2003). New Strength Model Based on Water-cement Ratio and Capillary Porosity. ACI Materials Journal, 100:4, 311-318.

Persson, B.(1996). Hydration and Strength of High Performance Concrete. Advanced Cement Based Materials, 3, 107-123.

Powers, T.C. and Brownyard, T.L. (1947). Studies of the Physical Properties of Hardened Portland Cement Paste. Journal of Thermal Analysis and Calorimetry, 43:2, 669-680.

Sanahuja, J., Dormeiux, L., and Chanvillar, G. (2007). Modeling Elasticity of a Hydrating Cement Paste. Cement and Concrete Research,

37, 1427-1439. Sarkar, L.S. and Wheeler, J. (2001). Important

Properties of an Ultrafine Cement – Part I. Cement and Concrete Research, 31, 119-123.

Schindler, A.K. and Folliard, K.J. (2005). Heat of Hydration Models for Cementitious Materials. ACI Materials Journal, 102:1, 24-33.

Uchikawa, H. (1994). Ultra-fine Cements for Special Applications. Advanced Cement Based Materials, 1, 150-154.

Voigt, T. and Shah, S.P. (2004). Properties of Early-age Portland Cement Mortar Monitored with Shear Wave Reflection Method. ACI Materials Journal, 101:6, 473-482.

Wang, A., Zhang, C., and Zhang, N. (1999). The Theoretic Analysis of the Influence of the Particle Size Distribution of Cement System on the Property of Cement. Cement and Concrete Research, 29, 1721-1726.

Wang, X. and Subramaniam, K.V. (2011). Ultrasonic Monitoring of Capillary Porosity and Elastic Properties in Hydrating Cement Paste. Cement and Concrete Composites, 33, 389-401.

Williams, D., Saak, A., and Jennings H.M. (1999). The Influence of Mixing on the Rheology of Fresh Cement Paste. Cement and Concrete Research, 29:9, 1491-1496.

Zhou, J., Chen, X., Wu, L. and Kan, X. (2011). Influence of Free Water Content on the Compressive Mechanical Behavior of Cement Mortar under High Strain Rate. Academy Proceedings in Engineering Sciences, 36:3, 357-369.

Zhou, J., Chen, X., Zhang, J., and Wang, Y. (2011). Internal Relative Humidity Distribution in Concrete Considering Self-desiccation at Early Ages. International Journal of the Physical Sciences, 6:7, 1604-1610.

科

学出版社


www.abook.cn


EFFECT OF AGGREGATE ON CREEP BEHAVIOR OF SELF-COMPACTING CONCRETE

Surong Luo *, Pengfei Chao College of Civil Engineering, Fuzhou University, Fuzhou 350108, P.R. China

Abstract: In this study the creep behaviour of self-compacting concrete, in which slump flow was larger than 550mm, containing fly ash were experimental investigated and compared with those of vibrated normal concrete. Based on the experimental results from creep tests of sealed and unsealed specimens, the effects of different binder /aggregate ratios and sand ratios on creep behaviour of harden self-compacting concrete was investigated. The results indicated that the ratio of aggregate-to-concrete, especially the content coarse aggregate greatly influenced creep of SCC. The creep rate of SCC investigated was greater than that of normal concrete at early age but decreased with age. Keywords: Self-compacting concrete (SCC), creep, binder /aggregate ratios, sand ratios

1 INTRODUCTION

The introduction of new admixtures and cementitous materials has allowed the production of high performance concrete (HPC), which has high strength, good flow ability, and high durability. Self-compacting concrete (SCC) is a new kind of HPC defined as a concrete that has excellent deformability and high resistance to segregate, and can be filled with a heavily reinforced area without applying vibration. Recently, due to the structural constrains, social and economical reasons, SCC seems to be a very promising materials used more and more in civil engineering.

Creep is defined as a time-dependent increase of strain under sustained load. Creep affects serviceability, durability, stress distribution and long-term reliability of concrete structures. As a kind of concrete, creep also happen to SCC. The strength of SCC is provided by the aggregate binding by the paste at hardened state, while the workability of SCC is provided by the binding paste at fresh state. Therefore, SCC is quite different from normal concrete in mixture proportions and applied materials, especially the limited presence of aggregate. All the special characteristics in SCC mixing may have important influence on the creep behavior. However,

few papers dealing with the creep behavior of SCC was found. Therefore, the purpose of this research was to investigate the creep of SCC with different content and composition of aggregate.

2 EXPERIMENTAL

2.1 Raw Materials and Concrete Mixes Used

The fine aggregate used was sand from Ming river , and the fineness modulus of sand was 2.6. The coarse aggregate was crushed granite. All the aggregate was washed by clean water and dried by sun curing before concrete placing. Fig.1 shows the grading curves of fine and coarse aggregate.

The cementitious materials used were ordinary portland cement (OPC) with the level of 42.5 and the stair fly ash(FA). Cement was produced by Lianshi cement factory in Fujian province. The fly ash was produced by the Songyu electric power plant of Xiamen. The chemical composition of cement and FA are given respectively in Table 1. The naphthalene-based UNF-5 high effective superplasticizer with water reduced rate of 20%-25% was used.

·317·

Figure 1. Grading curves of fine and coarse aggregate

All of the mix proportions of SCC having fixed water/binder(W/B) ratio of 0.30, and a constant FA/OPC (F/A)ratio of 0.8, were determined after making some adjustment on the basis of trial mixes according to general advice from the correlative specification. All of the

fresh concrete are tested by slump cone and L-type box immediately after placing to make sure the requirements of workability of SCC are fulfilled. As shown in Table 2, the binder/aggregate ratios (B/A) in the SCC mixtures, those with same sand ratio of 0.49, SC1,SC2 and SC3 were 0.28,0.31and 0.34 respectively. The sand ratios (S%)of SC4,SC3 and SC5, with the same of the content of total aggregate, were 46%,49% and 52 % respectively.

The flow ability and workability of fresh concrete are also shown in Table 2, such as the slump flow value and H2/H1, which were measured by slump test and L-type box respectively, indicating that the requirements of workability were achieved. The mechanical properties, such as compressive strength of standard cube (150mm) and prism (100×100× 400mm) specimens are also listed in Table 2.

Table 1. The chemical compositions of cementitious materials /(wt%)

Binder CaO SiO2 Al2O3 Fe2O3 SO3 MgO K2O TiO2 MnO P2O3 LOSS OPC 64.09 20.43 5.34 3.96 3.55 1.09 1.00 0.29 0.10 0.01 0.70 FA 6.28 48.33 34.79 5.96 0.83 — 0.89 1.68 1.08 0.10 1.38

Table 2. Mixture proportions(kg/m3)

Binder Aggregate 28 days

Strength(MPa) Specific number

Cement Fly ash Water

Fine Coarse

Super- plasticizer

B/A S(%)Flow value

(mm) L box

(H2/H1) cube prism

SC1 288 192 145 834 868 6.5 0.28 49% 575 0.71 47.7 36.7 SC2 312 208 156 814 848 7.5 0.31 49% 610 0.75 51.0 38.2 SC3 330 220 165 800 832 8 0.34 49% 680 0.9 52 39 SC4 330 220 155 751 881 7.6 0.34 46% 670 0.85 43.8 34.7 SC5 330 220 165 848 783 5.9 0.34 52% 640 0.8 40.5 33.0 NC 525 0 200 660 1040 0 0.29 39% — — 38.9 33.1

2.2 Method

Prism specimens of 100×100×400mm were used for creep and parallel shrinkage tests. The specimens for the SCC were cast by pouring from a scoop without vibration, while the normal reference mix were compacted on a vibrating table. All of the specimens were removed from the molds not <24h after molding and were cured in a moist condition at a temperature range of standard environment with temperature of 20±3 until the test age of 28 days. To evaluate ℃the basic creep, half of specimens were double-layered.

studies were done on shrinkage. The strain of specimens was measured by dial gages. The total strain in the creep studies was reduced by the shrinkage strain in order to obtain the creep strain. wrapped by epoxy resin and tinfoil to prevent movement of water during the test period. The

tests were conducted in the similar environment at 20±3 and the relative humidity was held at ℃60±3%. The creep specimens were paced in traditional compressive loading devices. The loading was 30% of the current compressive strength at 28 days. By pre-loading method, the elastic strain was eliminated. Meanwhile, parallel.


Results of the creep and shrinkage strain measurement on the SCC and NC specimens are all shown in Fig.2. It is shown that all the curves have a similar shape, and the creep (real line) and shrinkage (dash line) increase rapidly at the beginning of loading period, and the rate of creep and shrinkage decrease with the age.

科学出版社


www.abook.cn

·318·

unsealed creep sealed creep unsealed shrinkage sealed shrinkage

Figure 2. Creep and shrinkage strain of each mixture

As is well known, the current researches and prediction models for creep are restricted to the service stress range (or up to about 40% current compressive strength) 0, in which creep is assumed to be linearly dependent on the applied stress. Therefore, for comparison of experimental data in different studies, the results of creep tests must be standardized as specific creep. The specific creep function is the total creep strain per unit of the applied stress, and

defines as follow: 0 0

0

0

( , ) ( , )( , )

( )total shrinkaget t t t

C t tt

(1)

In Eq.(1), 0( , )C t t is specific creep. 0t is the loading time. 0( , )total t t is the total strain measured in creep test. 0( , )shrinkage t t is the relative shrinkage strain. 0( )t is the stress of the concrete caused by the applied load.

Fig.3 shows calculated results of the sealed

·319·

and unsealed specific creep for three binder/aggregate levels. Fig.4 shows the test results of specific creep according to different sand ratios. It is can be seen that creep of SCC increases steeply up to nearly one month, and then increases smoothly with age. And the total creep (unsealed) of SCC is obviously greater than basic creep (sealed). From the comparison in Fig.3, although the compressive strength increases, it is clear that decreasing the B/A level increases the total specific creep. And the basic creep distinguishes are not so marked with the B/A. It is also shown in Fig.4 that both the total creep and basic creep of SCC increase with the sand ratios. It means that the aggregate, especially the coarse aggregate has made the positive influence on the both total creep and basic creep of SCC. The reason is that the aggregate especially coarse part can be treated as the rigid skeleton without creeping, which embedded in the matrix of creeping harder cement mortar. Therefore the more coarse aggregate causes the less creep.

Figure 3. Specific creep with different B/A

Figure 4. Specific creep with different sand ratios

Fig.5 shows the comparison for creep between SC5 and NC ,with nearly the same compressive strength: 40.5MPa and 38.9MPa. As shown in comparison, the basic creep of SCC tends to be always smaller than NC because of the more compacting structure. On the other hand, it is observed that at early age, nearly first 15-20 days in loading period, total creep of SCC increases more rapidly than NC. But with the increasing age the rate of total creep of SCC decreases more than NC. It is because that the FA/OPC of SCC is higher than NC, and the hydrated rate of FA is lower than OPC. Thus the FA in SCC can supply more strength at later age for preventing creep.

Figure 5. Comparison between SC5 and NC

4 CONCLUSIONS

1) All SCC mixtures showed much more total creep than basic creep. The total creep of SCC was very dependent on water movement.

2) Aggregate did not affect the general shape of strain-time curves of creep and shrinkage specimens, however, the proportion of the aggregate increased, the total specific creep decreased while basic creep tended to be the similar although the strength increased.

3) When the content of total aggregate and binder in concrete were held constant, both of the total creep and basic creep decreased as the level of coarse aggregate proportion increased.

4) A SCC mixture showed a more rapid increase at the early age of total creep than the associated normal concrete, and the difference declined with the age. In spite of the more coarse aggregate contained in NC, the basic creep of normal concrete was always greater than SCC because of the more compacting structure of SCC.

科

学出版社


www.abook.cn

·320·

ACKNOWLEDGEMENTS

This work was supported by the National Natural Science Funds of China (No.50978060), the Natural Science Funds of Fujian Province of China (No. 2009H0027).

REFERENCES

CCES 02-2004. (2005). Guide to Design and Construction of Self-compacting Concrete. Beijing: China Architecture and Building Press.

Granger L P, Bazant Z P. (1993). Effect of composition on basic creep of concrete and cement paste. Journal of Engineering Mechanics, 121(11)：1691-1703.

Lee Y, Yi S T, et al. (2006). Evaluation of a basic creep model with respect to autogenous shrinkage. Cement and Concrete Research, 36:1268-1278.

Luo S R, Zheng J. L, Wang G. J. (2005). Mechanical Property of Self-compacting Concrete and its Engineering Application. Engineering Mechanics, 22(1): 164-169. (in Chinese)

Nevill A M. Properties of Concrete. New York：John Wiley & Sons. Inc.1996.

Ozawa K, Mackawa K, Okmura H. (1989). High Performance Concrete Cased on the Durability of Concrete, Proceedings of the 2nd East Asia-Pacific Conference on Structural Engineering and Construction, 1:445-456.

Pan Z H, Nakamura H, Wee T H .(2006). Shrinkage Behavior of High Performance Concrete at Different Elevated Temperatures under Different Sealing Conditions. Journal of Wuhan University of Technology(Materials Science Edition). 21(3):138-141.

Poppe M, Schutter D. (2005). Creep and shrinkage of self-compacting concrete. Proceedings of the First International Symposium on Design, Performance and Use of Self-consolidating Concrete, Changsha, 329-326.


APPLIED STUDY ON THE INORGANIC THERMAL INSULATION MATERIAL

IN CONCRETE

Zhao Lin , Zhu Li , Wenjing Wang , Xiaoqing Bai , Yu Zhang , Gang Ma

College of Architecture and Civil Engineering, Taiyuan University of Technology, Taiyuan 030024, P.R.China Abstract: For the poor thermal conductivity of ordinary concrete and fire-resistant performance , durability and other disadvantages of the organic building insulation material, a new type of insulating concrete—thermal insulation glazed hollow beads concrete composed of inorganic insulation materials was proposed. The results indicate that the pre-wetting processing of inorganic heat preservation material and optimization of mixing procedure is able to reduce the slump loss of concrete. The addition of add the glassed hollow beads or expanded perlite can improve the heat preservation and insulation performance of concrete under the premise of strength. Keywords: Inorganic thermal insulation material, glassed hollow beads, concrete, pre-wetting, thermal conductivity 1 INTRODUCTION

The thermal conductivity of ordinary concrete is as high as 1.74 W / (m·k), far higher than the heat-insulating materials (materials whose thermal conductivity is lower than 0.23 W / (m·k)), its thermal insulation performance is poor. In order to achieve the effect of heat preservation and realizing national energy-saving standards, heat preservation building materials are usually used on the external surface or internal surface of building structure. In the insulation engineering practice, the paste or daub construction of the insulation material is conducted after the formation of the main structure of buildings. This approach not only increases the construction process and extends the construction period, increases the construction cost, and handicaps the stress and safety of structure. This theory in Building energy conservation and wall thermal insulation by the authors(Gu, Xie and Chen,2006).

With the status of building energy efficiency in our country becomes more and more important, disadvantages—fire-hazardous, trivial construction and so on of the traditional way of exterior wall thermal insulation are emerged. So it is important significance to study on an inorganic insulation material.

Inorganic heat preservation material is mainly refers to the particles, which are formed through special technical processing and production

processing after the volcanic rocks (perlite or pine resin rock) are crushed and screened into ore. With excellent thermal insulation performance, it is light and porous in the internal. According to the different production process, inorganic heat preservation materials are usually divided into expanded perlite and glazed hollow beads.

Concrete, researched in this paper, is a new concrete with low thermal conductivity, high strength. It has a set of load-bearing and insulation performance. The thermal conductivity is much lower than ordinary concrete on condition of meeting the required strength. As structure material of concrete, the heat preservation concrete, has the comprehensive performance —seismic, economy and saving-energy, and it is convenient to design, construction, and self-insulation of the structure can be realized. Engineering effect of building structure and building energy saving can be realized simply when the main structure construction are completed.

But also a lot of cost in insulating engineering design and construction can be saved. Especially the stress of structure can be more reasonable, avoiding the quality-safety accident that brought out by craze, damage—because of insecure combination between the main structure and thermal insulation layer. The research on thermal insulation glazed hollow beads concrete is put forward based on above foundations. 科

学出版社


www.abook.cn

·322·

2 THE STUDY ON INORGANIC HEAT PRESERVATION CONCRETE

Ordinary concrete, one of the most commonly used material in the current civil engineering, is a composite materials made of raw materials such as cement, aggregate, water and other materials. It was made from appropriate proportion. In “Civil engineering materials”, Zhao(2004) had said so.

Light weight aggregate concrete is a concrete made of raw materials such as cement, ceramics, pottery, and so on. Its dry density of apparent is less than 1950 kg/m³. Because the advantages of light weight, good durability, thermal insulation, lightweight aggregate concrete is paid more and more attention in recent years. This sentence from “Mix ratio design and pumping technology of super-light aggregate concrete” by Wang and Zhou,etc(2008).

The inorganic heat preservation concrete studied in this paper studies is made of raw materials such as cement, inorganic heat preservation material, sand, gravel, and so on. Its strength is equal to ordinary concrete, the thermal insulation performance is better than lightweight aggregate concrete, and the thermal conductivity can be less than 0.2 W / (m·k). In this article, through comparing the influence on strength, thermal insulation performance, the work performance of different inorganic heat preservation material, the inorganic heat preservation material suitable for heat preservation concrete would be find out.

3 RAW MATERIALS AND THE TEAT METHOD IN THE EXPERIMENT

3.1 Raw Materials in The Experiment

Cement: PO 42.5 ordinary Portland cement produced by Shanxi JiGang cement plant.

Stone: gravel, particle size 5-12 mm, bulk density 1630 kg/m3, produced in Qingxu Shanxi.

Sand: medium sand, fineness modulus 1.6-2.2, bulk density 1500 kg/m3, produced in Qingxu Shanxi.

Water: clean tap water. Water reducing agent: Polycarboxylic Water

Reducer, concentration of 40%, produced in Taiyuan Shanxi.

Inorganic heat preservation material. Gazed hollow beads: Product, dry apparent

density of 110 kg/m³, produced in Taiyuan. Perlite: Products, dry apparent density of 70

kg/m ³, produced in Xinyang Henan.

3.2 Experimental Method

3.2.1 Initial mix proportion of concrete

In past research based on experiment ,the experiment use the initial mix proportion as follows (Not adding inorganic thermal insulation material), This mixture 28d Cubic Anti-pressure is 49 MPa, See Table 1.

Table 1. Initial mix proportion of concrete

Cement Stones Sand Water Water

Reducing Agent

480 1130 730 218 1.2

3.2.2 The mixed quantity of inorganic thermal insulation material

The mixed quantity of inorganic thermal insulation material has significant influence to the thermal conductivity and compressive strength of concrete, on the basis of determining the initial mix proportion of concrete, add two kinds of inorganic thermal insulation material (glazed hollow beads and expanded perlite) as variables to concrete, according to the initial volume of the concrete (100%,130%,150%) for add, draw indexes of the thermal conductivity, 28d cubic anti-pressure , slump degree in different mixed quantity of inorganic thermal insulation material.

3.2.3 The amount of admixture

Using the high performance admixture mixed by Research Group, the amount of admixture is 10% of the initial cement.

3.2.4 Mix proportion of experiment

The mix proportion of 6 groups in Table 2，Group A by adding glazed hollow beads expanded perlite ,group B by adding expanded perlite.

Table 2. Mix proportion of experiment

NO. C ST SA WRA AD ITIM A1 432 1130 730 1.2(5.184) 48 100%(110) A2 432 1130 730 1.2(5.184) 48 130%(143) A3 432 1130 730 1.2(5.184) 48 150%(165) B1 432 1130 730 1.2(5.184) 48 100%(70) B2 432 1130 730 1.2(5.184) 48 130%(91) B3 432 1130 730 1.2(5.184) 48 150%(105)

Note: C: Cement(kg/m³), ST: Stones(kg/m³), SA: Sand(kg/m³), WRA: Water Reducing Agent%(kg/m³), AD: Admixtures(kg/m³),

ITIM: Inorganic thermal insulation material %（kg/m³）

·323·

3.2.5 Determination of water absorption of inorganic thermal insulation material

Using vacuum filtration method to measure the water absorption of glazed hollow beads or expanded perlites used in experiments,in previous work by the authors(Li et al,2008) the specific method is as follows:

1) Wrapped the test sample by cloth and put it into the container attached to vacuum filtration device, start the vacuum pump and test a one-minute filtration to siphoned the air from the glazed hollow beads or expanded perlites.

2) Keep vacuum pump open, and open water injection valve of the container to inlet water, carry out a 20min filtration to make the glazed hollow beads or expanded perlites absorbent water until saturation.

3) Removed the test sample and filtered the surface water by filter paper, and tested the quality changes of specimen.

3.3 Formability and Performance Test Methods

Pre-wet process for dry inorganic insulation material should carry out 1h absorbent according to the measured water absorption, using direct immersion method to achieve the required pre-wet time. Filtered surface water of Inorganic insulation material with filter paper, then carry on the concrete mixing. The total mixing water consumption is the net amount of water (water-cement ratio is 0.45) on the basis of the 1h water absorption of the inorganic thermal insulation material, mixing sequence is as follows:

1) Pre-wet inorganic thermal insulation material.

2) Join Water Reducing Agent to required water.

3) Mix inorganic thermal insulation material and admixtures with 1/3 of water 1 min.

4) Mix Sand and Stones with one-third of water 1.5 min.

5) Mix cement and surplus water 2.5 min. Forming 150 mm × 150 mm × 150 mm specimens 6 piece of each different mix proportion, curing in standard condition and measuring its 28d Cubic Anti-pressure; Forming 300 mm × 300 mm × 30 mm specimens 3 piece, curing in standard condition 28d, drying 48h in the temperature of 80 ℃,measure its weight; measuring eight point of 4 sides using vernier calipers, taking the average value, Using DRP-5W made by the architectural instrument

factory of Tianjin measuring the thermal conductivity; measuring slump degree and slump degree after 1h of different mix proportion; Forming 100 mm × 100 mm × 300 mm specimens 3 pieces, used to test the elastic modulus under static pressure; Forming 100 mm × 100 mm × 50 mm 3 piece of specimens, used to test the electric flux of concrete.


4.1 Water Absorption of Inorganic Thermal Insulation Material

Using vacuum filtration method to measure the water absorption of inorganic thermal insulation material, testing each group of sample 3 times, then take the average value. From the Table 3 can see, the water absorption of glazed hollow beads is far lower than that of expanded perlite.

This is mainly determined by the production techniques, expanded perlite is by high temperature flame heating inflation the sand, its surface structure is open. Adding the area of inflation in production techniques of expanded perlite, that is glazed hollow, raising the temperature to 1200 ℃ until it is inflation, melting surface, forming the closed pore structure, so the water absorption of glazed hollow is far lower than the expanded perlite. As shown in the work by Li and Zhang etc(2006) and Zheng and Zhang (2005).

Table 3. Water absorption of glazed hollow beads and expanded perlite

sequence number Glazed hollow

beads Expanded

perlite 1 21% 97% 2 25% 98% 3 20% 102%

Water absorption 22% 99%

4.2 Cube Compressive Strength, Thermal Conductivity and Slump

Table 4 shows that with the increase of inorganic insulation material content, the thermal conductivity and the cube compressive strength of concrete will reduce. In the two testing groups of A2 and B2,when the content of inorganic insulation material is 130%, both the concrete made by glazed hollow beads and expanded perlite showed good compressive strength and thermal conductivity.

科

学出版社


www.abook.cn

·324·

Table 4. Testing results of properties

NO. W

(kg/m³) A

(kg) S

(mm) OHS (mm)

CCS (MPa)

TC (W/(m•k))

DAD(kg/m³)

A1 218 20.2 180 170 40.8 0.026 2230A2 218 30.0 193 175 37.6 0.021 2090A3 218 33.5 200 182 33.2 0.020 1970B1 218 65.3 228 205 34.7 0.035 2050B2 218 87.5 235 208 30.9 0.031 1930B3 218 99.8 258 214 27.6 0.029 1760

Note: W: Water, A: 1h water absorption of inorganic insulation material, S: Slump, OHS: One hour slump, CCS: 28d cube compressive strength, TC: Thermal conductivity. DAD: Dry apparent density

In Fig.1,to pre-wet process before the mixing of glazed hollow beads and expanded perlite can effectively resolve the problem that the concrete loss of their working issue by inorganic insulation material segregation and water absorption, it can be seen the uniform distribution of lightweight aggregate on the concrete damaged cross-section from Fig.1. By the comparison of A1-A3 and B1~B3,the greater water absorption of inorganic insulation material, the lower strength of concrete, when the B3 specimen content is 150%,The 1h slump losses of up to 17%.Analysis of the reasons by the following two:

1) Expanded perlite particles with high water absorption, low intensity, pre-wet process is difficult to fully absorbent.

2) The expanded perlite particles are damaged in the mixing process, resulting in water absorption increases. The concrete made by expanded perlite as inorganic insulation material Still exists the phenomenon of water absorption in the process of Stillness 1h ,resulting in a great loss of 1h slump.

Figure 1. Destructive fracture of inorganic insulating concrete

The other hand, three groups of A1 - A3 specimens are better than B1 - B3 in their compressive strength, thermal conductivity and slump. The reason is:

1) Measured by the Issue group, the thermal conductivity of glazed hollow bead particles

(0.025 ~ 0.035) is lower than the expanded perlite particles (0.045 to 0.060), The thermal conductivity of A1 ~ A3 specimen are lower than B1~ B3.

2) The dry density of glazed hollow beads is 110kg / m³ ,which is higher than the expanded perlite particles’ 70kg / m³, and it’s better than the expanded perlite in intensity, so that the compressive strength of A1 ~ A3 specimen is higher than the B1 ~B3.

3) The water absorption of expanded perlite is much higher than glazed hollow beads, and expanded perlites exist breakage in mixing process, it leads to the direct consequence is that concrete’s water-cement ratio is too large, and compressive strength is too low.

4.3 Concrete Static Pressure Elastic Modulus

By measuring the modulus of static compression elasticity of concrete as shown in Fig.2, we draw some results as shown in Table 5.

Figure 2. The testing of concrete static pressure elastic modulus

Table 5. Concrete static compression modulus

NO. The elastic modulus under static

pressure (GPa) A1 27.8 A2 24.9 A3 22.1 B1 26.5 B2 23.2 B3 20.8

The research shows that the more the mixed

volume rate of the inorganic heat preservation material, the smaller the modulus of static compression elasticity of concrete.

Based on the design code of concrete structure(GB50010-2002), the elastic modulus of A1 specimens is lower than that of the C20 ordinary concrete (28.0 GPa), much lower than that of the C40 ordinary concrete (32.5 GPa). Results of analysis as follows:

·325·

1) It has shown that the dry density of concrete could be reduced to 1800-1900kg/m³ after mixing with inorganic heat preservation material, the smaller the density is, the lower the elastic modulus is.

2) In the inorganic heat preservation materials, the bulk density of glazed hollow beads (110kg/m³) is bigger than that of the expanded perlite (70kg/m³). The density of the concrete sample with glazed hollow beads is bigger than sample with expanded perlite, so its elastic modulus is higher.

4.4 The Testing of Chloride Permeability

The properties of permeability of chloride ions reflect the degree of resist the external medium to internal invading ability. It is an important index of the durability of concrete.

In the past research, the electric flux of ordinary concrete without inorganic heat preservation material is 870 C, compared with the data in Table 6, it shows that the concrete mixed with inorganic heat preservation material has better chlorine anion penetration resistance, whose electric flux is lower. The main findings are:

Table 6. Test results of 28d chloride permeability

NO. 28d Electric flux (C) A1 580 A2 465 A3 382 B1 615 B2 524 B3 423

1) By mixing inorganic heat preservation

material, the microstructure of the concrete becomes more reinforced, so its ability to counter chloride diffusion is improved.

2) The density of concrete mixed with glassed hollow beads was different from that of concrete mixed with expanded perlite, which mixed with glassed hollow beads is denser than it mixed with expanded perlite, the results shows that the electric flux of group A is lower than group B.

5 CONCLUSIONS

This experiment shows that the concrete by

mixed with inorganic heat preservation material could improve the thermal insulating properties, and the contrast experiment shows the glassed hollow beads is more suitable to the concrete than expanding pearlite. It’s shown that glassed hollow beads could greatly reduce coefficient of thermal conductivity of concrete under the premise of strength. Using the characteristics of lower elastic modulus can improve seismic resistance performance of the reinforced concrete structure.

In the past the excessive slump loss in concrete mixed with inorganic heat preservation material is a hard problem. Through the pre-wetting processing and the improvement of mixing process, the experiment solves the inorganic heat preservation material applied on concrete.

REFERENCES

Fangran Zhao.(2004). Civil Engineering Materials.Tongji University press.

Fazhou Wang, Yufei Zhou, Jun Wang,etc.(2008). Mix Ratio Design and Pumping Technology of Super-light Aggregate Concrete. Construction Technology, 9:109-111.

GB50010-2002, Design Code of Concrete Structure. China Architecture and Building Press.

Qiao Li, Yongshun Hong, Lin Zhang. (2008). Analysis The Similarities and Differences between Vacuum Filtration Method and Filter Paper Method to Test Water Retention of Mortar. Chemical Building Material, 2: 31-32.

Tianshu. Gu, Lianyu Xie, Ge Chen.(2006). Building Energy Conservation and Wall Thermal Insulation. Engineering Mechanics, 23(II):167-184.

Xiuhua Zheng, Baosheng Zhang.(2005). Effect of Pre-wetted Shale Ceramsite on Strength and Frost-resistance of Lightweight Aggregate Concrete. Journal of the Chinese Ceramic Society, 6: 758-762.

Zhu Li, Zeping Zhang, Yuanzhen Liu.(2006). The Importance of Energy Efficiency in Buildings and Introduction of a New Technique. Engineering Mechanics, 23(II):141-149.

科

学出版社


www.abook.cn

proceedings of the twelfth international symposium … · international symposium on structural...

Documents