Download - Deep Mixing Columns - 03 Sabatini
-
7/24/2019 Deep Mixing Columns - 03 Sabatini
1/11
EVALUATIONOFDEFECTSINDEEPMIXING
METHOD
COLUMNS
USED
FOR
STRUCTURAL
FOUNDATIONSUPPORT
Paul
Sabatini
October
17,
2012
Geosyntec
Consultants
Oak
Brook,
IL
ProjectOverview
Some 6,000, 23m long DMM elementswere installed to support 124 LNG processtrain mats
During early phases of construction, coringrevealed clay inclusions and unmixed zones
Owner requested analyses andassessments to confirm a fit
for
purpose
foundation system
-
7/24/2019 Deep Mixing Columns - 03 Sabatini
2/11
Project
Overview
(Cont.) We reviewed available DMM core data
Contractors QA/QC coring program should
be increased
Site activities and investigations proposed
to provide reliable physical data
Contractor declined to implement our
investigative proposals
DMMCoreReview
-
7/24/2019 Deep Mixing Columns - 03 Sabatini
3/11
As
Built
DMM
Elements
SummaryofCoreEvaluation
Summarized results of 182 cores using
Contractor field observations of full
diameter poor quality core (PQC)
Information used to define defect scenarios
for 3D numerical modeling
Design Defect Length = 1 m
Worst Case Defect length = 2m
-
7/24/2019 Deep Mixing Columns - 03 Sabatini
4/11
Numerical
Modeling 3D finite element analyses of DMM columns
(with defects) and foundation mats
Use Zone 5 foundation with lowest DMMreplacement ratio (i.e., on the order of 0.35)because drag loads would be largest
Drag loads from longterm settlements in theUnit II clays
Seventeen (17) cases without downdrag and
five additional cases with downdrag
AnalysisModel
-
7/24/2019 Deep Mixing Columns - 03 Sabatini
5/11
Loaded
Areas
of
Mats
Checkerboard(CB)EntireMat A50Partial
A load of 50 kPa is applied over the shaded areas.
The self-weight of the mat, 18 kPa, is applied over the remainderarea of the mat.
HorizontalDistribution
of
Defect
Regions
RegionA RegionCRegionB
RegionD RegionFRegionE
-
7/24/2019 Deep Mixing Columns - 03 Sabatini
6/11
HorizontalDistributionofDefectRegions(Cont.)
RegionAModified RegionBModified2RegionBModified1
RegionFModified RegionCBOutRegionCBIn
VerticalDistribution
of
DMM
Column
Defects
2mDefectDistribution 1mDefectDistribution
0.5 m DefectUnit I
Layer
Unit II
Layer
Unit IIIA-1
Layer
+2.23 m LAT+2.73 m LAT
-0.5 m LAT0 m LAT
-6 m LAT-5 m LAT
0.5 m Defect
1.0 m Defect
Bottom of Mat
Unit I
Layer
Unit II
Layer
Unit IIIA-1
Layer
-0.5 m LAT0 m LAT
-5.5 m LAT-5 m LAT
0.5 m Defect
0.5 m Defect
Bottom of Mat
-
7/24/2019 Deep Mixing Columns - 03 Sabatini
7/11
Results
for
Perfect
DMM
Case
Soil LayersSettlement
(mm)
Layer
Thickness
(m)
Unit I 1.1 3.98
Unit II 6.7 3
Unit IIIA-1 10.4 11
Unit IIIB-1 3.2 7
Unit IIIB-2 4.3 12
Unit V 4.3 6
Unit VI 3.5 36
Sum 33.4 78.98
ResultsforA50Case
MaxSettlement=36.3mm
Min.S22=2130kPa
Max.S22=1950kPa
-
7/24/2019 Deep Mixing Columns - 03 Sabatini
8/11
Results
for
B50
Case
MaxSettlement=43.9mm
Min.S22=3210kPa
Max.S22=2970kPa
EffectsofDefects
Case Settlement
(mm)
MatStress
(%
of
Design
Allowable)
MatStress
(%
of
Ultimate)
Perfect 33.4 85 52
A50 36.3 104 64
B50 43.9 157 96
B50 Modified1 34.6 82 51
B50 Modified2 36.0 83 51
Pressure=50kPa
DefectLength=2m
-
7/24/2019 Deep Mixing Columns - 03 Sabatini
9/11
Summary
Defects
Only Even for perfect DMM elements, 85
percent of the allowable design stressesare mobilized
Partial defects on the edges or in interiorDMM island elements are not critical andresult in computed stresses very similar tothose where no defects exist
Defects are critical if they are continuous
across multiple DMM elements
Downdrag,Non
Uniform
Loading
and
DMMElementswithDefects
Address effects of downdrag settlements onmat supported by perfect DMM elements andthose with defects
Downdrag settlements result from secondarycompression settlements of the Unit II clays
Analyses are conducted for a 3m thick and 6mthick Unit II clay layer
Nonuniform (i.e., checkerboard) loading also
considered
-
7/24/2019 Deep Mixing Columns - 03 Sabatini
10/11
CriticalDowndrag Scenario
UnitILayer
UnitIILayer
0.5mLAT0mLAT
5.5mLAT5mLAT
0.5mDefect
0.5mDefect
BottomofMat
UnitIIIA1Layer
8Defects
2Defects
15cm
0cm
UnitII
5cm
DistributionofDMMColumnDefects,DefectRegionsandLoadingArea
SecondaryCompressionSettlements oftheUnitIIsoillayer
EffectsofDefectsandDowndragDefectAssessment Downdrag Assessment
C as e Se tt le me nt
(mm)
MatStress
(%ofDesign
Allowable)
MatStress
(%of
Ultimate)
Downdrag
Settlement
(cm)
Settlement
(mm)
MatStress
(%ofDesign
Allowable)
MatStress
(%of
Ultimate)
Perfect 33.4 85 52 10/5 49.4 104 64
A50Modified
1m
34.1 84 52 10/5 50.7 103 64
CBPerfect 32.6 104 64 15/5 51.1 122 75
CBInside
Modified1m
33.8 119 73 15/5 52.9 141 87
-
7/24/2019 Deep Mixing Columns - 03 Sabatini
11/11
Summary Downdrag settlements result in significant
increases in DMM and mat stresses
For Case A50Modified1m, computed mat stressesare very similar to those for the Perfect case
Even with nonuniform loading, continuous defectsup to 1m long, and the largest postulateddowndrag settlements, computed mat stresses forCase CBInsideModified1m are less than theequivalent ultimate concrete stress for the mat
ProjectSpecific
Conclusions
on
FoundationFitnessforPurpose
Foundation mats with three (3) interior DMM islandelements or more are fitfor purpose
Mats with only one interior DMM island element are mostprone to adverse performance if defects exist in thatelement
The assumption of defects (or completely unmixed zones)occurring at specific elevations and which penetrate acrossthe entire width of a DMM element is conservative
Most of the dead load for each mat has been applied andmeasured total settlements range from less than 2 mm to10 mm.