final report (balsa wood bridge design)
TRANSCRIPT
Final Report CEE 4803 Balsa Wood Bridge Design Project
December 9, 2015
Group 3 Arman Yosal Josia Tannos Maya Goldman Savannah Brooks
Table of Contents I. Introduction……………………………………………………………………….....page 1 II. Concept…………………………………………………………………………....page 23 III. Design Methods……………………………………………………………………..page 2 IV. Construction Techniques………………………………………………………….page 37 V. Testing and Performance………………………………………………………….page 78 VI. Post Test Evaluation……………………………………………………………..page 810 VII. Conclusion………………………………………………………………………....page 10 VIII. Appendix
I. Introduction This report covers the initial ideation and design, construction, testing and failure analysis
of a balsa wood truss bridge with span of forty inches, height of six inches and width of four inches and a weight of 135.9 grams. Using SAP2000 the three dimensional simple warren truss model was loaded with a thirty pound distributed force to account for a factor of safety of two. The bridge was constructed by laminating together wood pieces using super glue, included gusset plates to act as joints, and consisted of a platform to take into consideration how the bridge would be loaded during testing. The performance of the bridge exceeded expectations by holding twentyeight percent more weight than it was designed for and failed spectacularly by multiple lateral cross bracings and frame members splintering or cracking at the center of the span. The following sections will describe in more detail the design methods, construction techniques, testing and performance, and post test evaluation.
II. Concept Before construction could commence, a suitable design had to be chosen. The primary
goal for the preliminary plans was to create the most efficient system possible. This means that the bridge should be efficient in terms of cost by being conservative with material orders. It was also designed to be efficient in terms of loading capabilities. This means that the bridge was designed to withstand the 15 pounds of applied load required but not much more after that. It The goal was to create a bridge that would fail as close to this weight requirement as possible while still meeting the 0.25” maximum deflection allowance using the least amount of material necessary. In order to do this, various typical truss designs were examined. In the end, a simple warren truss design was chosen based on its overall symmetry and redundancy which would allow for uniform self weight distribution and ease of construction. After this, hand calculations were completed in order to ensure forces on each member fell below the minimum weight requirement. The minimum requirement was 15 pounds, however, in order to prevent the possible hindrances and performance issues during testing caused by possible human errors made during the construction process, a safety factor of two was applied to the system. This means that calculations for forces on each member were calculated using a load of 30 pounds, instead of 15 pounds. This was to give more leeway during performance testing. These calculations were completed for simple warren trusses that contained varying amounts of triangles. These designs were then eliminated until two possible choices remained based on the height and span constraints given as project requirements. The options for the truss was either a five triangle truss or a seven triangle truss. Ultimately, the design with seven triangles was chosen as can be seen in Figure 1a. Finally, the design was put into SAP2000 in order to create a model as close to real world system loading behaviors as possible. The 3D model created in this program was constructed for the idealized case of loading on the center lateral member. A factor of safety of two was also applied to this model. Modeling the design in SAP2000 gave more exact estimates of how the structure would react when the load is applied, allowing possible points of
weaknesses to be documented and strengthened during construction, lessening the chances of premature system failure.
III. Design Methods
An integral component of the design process after the initial design concept had been completed was planning the cross sectional area of each member in the truss. This was calculated using a form of the stress equation, and balsa wood properties found in Table 4a.The P ,A = σ maximum stress each member can hold is 0.68 ksi for compressive stress and 1.10 ksi for tensile stress. The results are shown in Table 2a. Four different crosssectional areas were used in order to build the truss bridge. All of these can be seen in Table 5a. The materials ordered then can seen in Table 1a. Without considering crosssectional area, the deflection exceeds the criteria. Deflection decreases with increased crosssectional area. So, in the final model, the crosssectional area of some members were increased to reduce the deflection. For example, member HJ in the center of the truss was designed to have a cross sectional area of ⅜” by ⅜”. This was achieved by laminating four 3/16” pieces together.
Deflection tests were also completed once all materials ordered had arrived. This was done in order to calculate the properties for each piece since balsa wood properties can vary from piece to piece. After completing the test, it was determined that the calculated modulus of elasticity was ten times greater than that the accepted one published online. These properties found through testing were not used, however. Altering these properties in SAP2000 would have altered the entire design and the materials required. As engineers, efficiency in terms of cost and time are of the utmost importance, both of which would have been sacrificed should these new values been used. These properties could have been used, if samples of the material were provided in order to complete deflection tests before material orders go out or if it were possible to reorder materials.
Another component of the truss design is lateral bracing along the top and bottom, in the shape of X’s. These lateral bracing members were placed to join each section of the frame except the center where the load was applied. These elements were added to account for torsion due to human error during the construction of the structure. However, since these were added merely to add stability should the bridge be slightly misshapen or skewed and since the cross sectional area of these pieces is moot, the smallest available crosssectioned pieces were used. This was done in order to limit the amount of weight added to the structure. Thus, the 1/16” pieces of Balsa wood were used. Another design detail to account for loading is the top center member where the load bar was placed. In order to create a strong platform to rest this bar on, three pieces of 3/16” balsa wood were placed on top of the center member to create a platform just over half an inch wide. Also, gusset plates were used at each of the joints in order to increase stability at these possible points of weakness.
IV. Construction Techniques First the materials were gathered and counted to ensure all necessary items were
purchased and collected. See Table 1a. of the appendix for material order. Tsquares (metal rulers shaped like a T) and XActo knives were used to cut all the member pieces to the correct lengths based on the dimensions of the bridge design seen in Figure 1. The amount of each piece with specified length and crosssectional area were tabulated in Table 3a. and used during this cutting process to keep track of how many of each kind were needed. In order to keep the cuts consistent, the members were made slightly longer so that the length was a mark on the ruler as opposed to in between two marks.
a. Front view of truss
(b) Side view of truss
Figure 1. Dimensions of truss members based on design concept and methods and minimum requirements of 36” length by 4” width by 5” height
After all the pieces were cut, they were laminated together to make the necessary
configuration. For example, to make member AB ¼” X ¼”, four ⅛” pieces of length 5.7” were glued together. This was done for every member twice to create the two frames and these were taped together and labeled for easy access when connecting all the pieces together at a later time. Next the gusset plates were cut to 1.5”X1” except at the connection joints F,H, and J where the dimensions were cut to 1.5”X1.5” to account for the larger members. Two gusset plates on either side of the connection and for two frames resulted in cutting sixty total. Finally, small pieces were cut and glued to the ends of each member which was connected to a member thicker than itself, so that at the connection all members were touching the gusset plates.
Image 1. Laminated piece on small member (red circle) to increase height (green line) and create
connection to gusset plate Before construction all the pieces were laid out in their proper positions and there was a
discussion about how to form the connections so that the proper angle for each triangle was maintained. A handdrawn sketch was used as a template, and the pieces were placed on the template. The first corner, A, was the starting point. Each connection was formed by glueing the members to the gusset plate so the triangle matched the template, and the frame was moved each time to form the next triangle. This method, however, was flawed since the template did not take into account the thickness of the members or the actual connection size, nor were the lines perfectly straight. It was also difficult to gauge the linearity of the overall structure as it grew because each triangle was formed relative to the template instead of the overall frame. To combat most of these issues, a CAD drawing was printed on a plotter to actual size, which replaced the handdrawn template.
Image 2. Laminated members organized in shape of truss
Image 3. Hand drawn template method Image 4. CAD template method
While the first truss frame was being finished, the next frame was created by aligning the pieces on top of the existing frame and then glueing the members to the gusset plate, starting at the corner and working left to right towards the other corner. This method was chosen for schedule efficiency and because after the first frame was complete it was more important for the two sides to be exact copies of each other than for the new frame to follow the template (symmetry).
Image 5. Second truss built on top of first truss
Next, the two frames had to be connected with lateral bracing and crossbracing. In order for the two frames to remain vertical and at the correct distance apart when glueing the lateral bracing members to either side, tongue depressors hole punched at either end and clamped together with a screw and nuts were secured along the length of the bridge as seen in Image 6.
Image 6. Tongue depressor system used to stabilize truss frames when adding lateral bracing Once group members confirmed the width between frames at both ends of bridge were equal and the frames were perpendicular to the table (not leaning or rotating), a lateral member was glued on either end and placed between the frames, with a gusset plate at the top of the ⅛” members. The 3/16” members were simply glued to the frame without gusset plates.
Image 7. Gusset plates on ⅛” lateral members Image 8. Connection for 3/16” members The crossbracing pieces of size 1/16” were cut and glued on an individual basis after the whole bridge was built. These pieces were glued to the truss frame without gusset plates.
Image 9. Crossbracing top view
V. Testing and Performance
First, the supports of the bridge at each end were placed on the table. The supports in contact with the table must not exceed 0.5 inches according to project specifications. One shim was used on the right back support (based on Image 10 configuration). The bridge was tested by placing a steel metal bar on the platform at the top center of the bridge. Then, a chain was used to lift the bucket filled with loads. The loading started with ten pounds and weight was added until the bridge failed. The configuration of the bridge and the loading is shown in Image 10.
Image 10. Loading the bridge
The bridge was expected to have a minimum height of five inches, minimum width of four inches and minimum length of 36 inches. As shown in Table 1, the bridge met the height, width and length requirements. Another specification was that the bridge was expected to deflect no more than 0.25 inches. When a load of 15 pounds was applied at the center of the bridge, the deflection was almost zero.
Table 1. Measured values during testing
Height 6 in
Width 4 in
Length 40 in
Weight of bridge 135.6 g
Weight of load at failure 53.5 lb
VI. Post Test Evaluation
Image 11 shows the failed segment of the bridge. The bridge failed at a load of 53.5 lb. It can be seen that the bridge underwent shear failure at the gusset plate. Image 12 and Image 13 show the shear failure observed on the bridge. There were no members that broke because of axial forces. This shows that the member crosssectional area is large enough to withstand the axial stress caused by the load. The lateral members mostly remained intact across the span. This shows that the lateral members used were strong enough to withstand the torsional load due to the unsymmetrical shape and load offset. It is shown that the wood splintered at the connection at point H and I. The member was not able to withstand the shear force caused by the load.
The goal of this project was to design a bridge that can withstand 15 lb load located at the midspan of the bridge. This design was created by accounting for a factor of safety of two. The crosssectional area of the members were calculated so that the deflection is not more than 0.25”. Based on the test result, the ultimate strength and elastic modulus used in the SAP analysis were lower than the actual value. In the design, the amount of material was added to meet the deflection requirement. At first, by using smaller crosssectional area for each member, the designed was able to hold up to 30 lbs. However, the initial design showed that the deflection was 0.70 inches, which is more than the maximum deflection. In order to reduce the deflection, the members must have larger crosssectional area because the deflection of truss is inversely proportional to the crosssectional area of the members. During the test, the bridge does not deflect significantly (the deflection is close to zero) and it can be concluded that the bridge was over designed. This over design was due to lack of understanding of the balsa wood properties. The balsa wood used in this project were stiffer than the assumed elastic modulus and therefore a more efficient design could have been used. Moreover, by increasing the crosssectional area of each member in the truss, the weight of the structure also increased.
The span and height of this bridge exceeds the minimum requirement specified. This truss bridge spanned 40 inches while the requirement only specified a span of 36 inches. This truss bridge had a height of 6 inches while the requirement only specified a height of 5 inches. This extra span and height induced more weight in the bridge. This happened because the design plan did not account for extra length in the connection and the thickness of the member. This additional length and weight increase the weight of the bridge.
Image 11. Failed Segment of the Bridge
Image 12. Failed Gusset Plate
Image 13. Shear Failure at the Horizontal Component.
VII. Conclusion
This design was able to withstand fifteen pounds of load located at the midspan of the bridge. In the design process, a factor of safety of two was used to account for human errors, defects in materials and other factors that may affect the performance of the bridge. After the test, the bridge was able to carry 53.5 lb before failure. Moreover, when a load of fifteen pounds was applied, the bridge did not deflect significantly (the deflection was almost zero). These indicate that the bridge was over designed because it could withstand more than the expected load and the actual deflection was smaller than the expected deflection. This is caused by underestimation of elastic modulus of the balsa wood. The expected length and height of the bridge was thirtysix inches and five inches respectively. However, the actual length was forty inches and the actual height was six inches respectively. This is caused by the thickness of the material and the space between members at the connections that were not considered in the design process. This additional length and height increased the weight of the bridge. Overall, this design could be improved by using a more accurate elastic modulus and accounting for additional length and height at the connection as well as the thickness of the members.
VIII. Appendix
Table 1a. Material order for balsa wood
Size Amount
1/16” 5
⅛” 31
3/16” 4
Plate 1
Table 2a. The force in each member with factor of safety of two based on SAP2000 analysis
Members Force (lb) Direction of
Force Minimum CrossSectional Area
(in 2 ) Configuration of
Member
AB 8.46 Compression 1.24E02 1/4" x 1/4"
AC 3.9 Tension 3.54E03 1/4" x 1/4"
BC 8.46 Tension 7.66E03 1/8" x 1/8"
BD 7.8 Compression 1.14E02 1/4" x 1/4"
CD 8.46 Compression 1.24E02 1/4" x 1/4"
CE 11.72 Tension 1.06E02 1/4" x 1/4"
DE 8.46 Tension 7.66E03 1/8" x 1/8"
DF 15.62 Compression 2.29E02 1/4" x 1/4"
EF 8.46 Compression 1.24E02 1/4" x 1/4"
EG 19.52 Tension 1.77E02 1/4" x 1/4"
FG 8.46 Tension 7.66E03 1/4" x 1/8"
FH 23.42 Compression 3.43E02 3/8" x 3/8"
GH 8.46 Compression 1.24E02 1/4" x 1/4"
GI 27.32 Tension 2.48E02 1/4" x 1/4"
HI 8.46 Compression 1.24E02 1/4" x 1/4"
HJ 23.42 Compression 3.43E02 3/8" x 3/8"
IJ 8.46 Tension 7.66E03 1/4" x 1/8"
IK 19.52 Tension 1.77E02 1/4" x 1/4"
JK 8.46 Compression 1.24E02 1/4" x 1/4"
JL 15.62 Compression 2.29E02 1/4" x 1/4"
KL 8.46 Tension 7.66E03 1/8" x 1/8"
KM 11.72 Tension 1.06E02 1/4" x 1/4"
LM 8.46 Compression 1.24E02 1/4" x 1/4"
LN 7.8 Compression 1.14E02 1/4" x 1/4"
MN 8.46 Tension 7.66E03 1/8" x 1/8"
MO 3.9 Tension 3.54E03 1/4" x 1/4"
NO 8.46 Compression 1.24E02 1/4" x 1/4"
Table 3a. Length, crosssectional width, and number of pieces to be laminated for each member
Member Label Wood CrossSection Size (in)
Length (in) Amount
AB 1/8 5.7 4
AC 1/8 5.2 4
BC 1/8 5.7 1
BD 1/8 5.2 4
CD 1/8 5.7 4
CE 1/8 5.2 4
DE 1/8 5.7 1
DF 1/8 5.2 4
EF 1/8 5.7 4
EG 1/8 5.2 4
FG 1/8 5.7 2
FH 3/16 5.2 4
GH 1/8 5.7 4
GI 1/8 5.2 4
HI 1/8 5.7 4
HJ 3/16 5.2 4
IJ 1/8 5.7 2
IK 1/8 5.2 4
JK 1/8 5.7 4
JL 1/8 5.2 4
KL 1/8 5.7 1
KM 1/8 5.2 4
LM 1/8 5.7 4
LN 1/8 5.2 4
MN 1/8 5.7 1
MO 1/8 5.2 4
NO 1/8 5.7 4
AA' 1/8 4.2 1
BB' 1/8 4.2 1
CC' 1/8 4.2 1
DD' 1/8 4.2 1
EE' 1/8 4.2 1
FF' 3/16 4.2 1
GG' 3/16 4.2 1
HH' 3/16 4.2 3
II' 3/16 4.2 1
JJ' 3/16 4.2 1
KK' 1/8 4.2 1
LL' 1/8 4.2 1
MM' 1/8 4.2 1
NN' 1/8 4.2 1
OO' 1/8 4.2 1 *Table 5a. simplifies this data to show totals for each cross sectional size
Table 4a. Balsa wood properties based on flexural test
Trial Length (cm)
Width (cm)
Moment of Inertia (cm^4)
Load (N)
Length (cm)
Deflection (cm)
Elastic Modulus (GPa)
1 0.318 0.318 0.001 0.196 25.000 1.800 6.704
2 0.318 0.318 0.001 0.196 25.000 2.200 5.485
3 0.159 0.159 0.000 0.098 10.000 1.000 6.178
4 0.159 0.159 0.000 0.098 10.000 1.500 4.119
Table 5a. Cross sectional area of Members
Cross Sectional Size Number of Members Pieces Laminated
1/4" x 1/4" 38 4 x (1/8" balsa wood)
1/8" x 1/8" 8 4 x (1/16" balsa wood)
3/8" x 3/8" 4 4 x (3/16" balsa wood)
1/4" x 1/8" 4 2 x (1/8" balsa wood)