Alex Crease, Joe Kochevar, Eric Schneider, Paul Titchener
ENGR 2320 Final Project Report Alex Crease, Joe Kochevar, Eric Schneider, Paul Titchener
2
Table of Contents Executive Summary ....................................................................................................................................... 3 Design: .......................................................................................................................................................... 4 Overview (Paul): ........................................................................................................................................ 4 Specific Member Design (Alex): ................................................................................................................ 4 Construction Design (Joe): ........................................................................................................................ 5 Analysis: ........................................................................................................................................................ 6 Analysis(Joe): ............................................................................................................................................. 6 Forces in each member from Solidworks (Eric) ......................................................................................... 7 Compression Member Buckling (Alex): ..................................................................................................... 8 Instron Testing (Paul): ............................................................................................................................... 8 Failure Analysis (Eric) ............................................................................................................................... 10 Conclusion ................................................................................................................................................... 12
Table of Figures Figure 1:Solidworks model of the bridge. The arch is constructed out of T beams, the X sections connecting the arcs are constructed out of half popsicle sticks glued together and the radial tension members are constructed from string. ......................................................................................................... 4 Figure 2:T Beam Construction....................................................................................................................... 5 Figure 3: T Beam Construction ...................................................................................................................... 5 Figure 4: This figure shows the compressive load curve of a T beam the full length of the T beams used on the bridge. The beam supports 450N before buckling ............................................................................ 6 Figure 5: Solidworks weldments model of the bridge. This simulation was run with a 1000N force at the center of the bridge. ..................................................................................................................................... 7 Figure 6: Loads and factors of safety for bridge members. .......................................................................... 8 Figure 7:Our final joints had the string wrapped around the wood, not glued straight to it ....................... 8 Figure 8:Example setups in the Instron mechanical tester. These pictures show testing of T beams in compression. ................................................................................................................................................. 9 Figure 9: Loads supported by various popsicle stick members. ................................................................. 10 Figure 10: One of the right-hand strings was covered in glue, which we believe is what pulled the hook in this image so far off-center. Alternatively, we may have placed the board off-center at the beginning. Either way, the off center loading ended up pulling the compression members in towards the center of the bridge. ................................................................................................................................................... 11 Figure 11: Here you can see the rightmost joint failing inwards. Note the massive compression and buckling occurring in the bracing X member. Under the intended loading these members were designed to only experience tension loads, and therefore were not built take the loads that they experienced in practice when the load was off-center. ...................................................................................................... 11 ENGR 2320 Final Project Report Alex Crease, Joe Kochevar, Eric Schneider, Paul Titchener
3
Executive Summary This popsicle bridge design was very similar to the previous spaghetti bridge design. It makes approximates an arch using strengthened compression members (in this case T-beams) and spreads the load out into a number of tensile members that connect evenly along the arch. The predicted failure point for the bridge was in the central tension members where they joined the arch. This was determined through Instron testing each member of the bridge and compar- ing that strength to predicted loads in each member. Our predicted failure load was 815N, from the force required to break the strings. In testing, the actual failure load was 490N. The actual failure mode was largely due to con- struction errors. A tension member was accidentally coated in glue during the construction pro- cess, which meant that it stretched differently from the other members. This caused a load im- balance, loading one side of the bridge far beyond expectations and causing failure. This imbal- ance also created a torque around the axis of the bridge (along the deck), which caused a tor- sional failure.
ENGR 2320 Final Project Report Alex Crease, Joe Kochevar, Eric Schneider, Paul Titchener
4
Design Overview (Paul): The bridge was built in an approximated arch out of T beams. This design was chosen with strength to weight ratio in mind, given that an arch design distributes equal loads to each sec- tion of the bridge. The outer hoop was designed to be in pure compression, and inner, radial members were designed to be in pure tension. Due to the fact that popsicle sticks will buckle if subjected to compression loads, these members were made out of T beams. Popsicle sticks were glued as in FIGURE BELOW and then used to approximate an arch. This bridge was constructed out of a limited quantity of materials. Therefore, it was decided to make our radial tension members out of the provided string rather than popsicle sticks. While the string has a significantly lower elastic modulus than the popsicle sticks, and therefore stretches, the ultimate tensile strength is quite strong. It was tested to hold more than 100 N, however broke in the grips of the Instron rather than breaking the string itself.
Figure 1:Solidworks model of the bridge. The arch is constructed out of T beams, the X sections connect- ing the arcs are constructed out of half popsicle sticks glued together and the radial tension members are constructed from string. Specific Member Design (Alex): There were two types of members that we used in our design, the T Beam members and the flat tension members. These members were designed to take on compression and tension loads, respectively. The overall bridge design was similar to our spaghetti bridge design in that t was an approximation of arc of 1060 mm in diameter. In order to conserve popsicle sticks, we used a member length of 260 mm, or 2 popsicle stick lengths, so we used 6 members in the ENGR 2320 Final Project Report Alex Crease, Joe Kochevar, Eric Schneider, Paul Titchener
5
arch for our approximation, meaning 12 T beams in total, 6 per side. The tension members were used for cross beams to allow the bridge to withstand the applied load without twisting. These were made by clamping the flat faces of beams to one another. Because the popsicle sticks act extremely well under tension, they were made out of half popsicle sticks. Strings were used to support the main load as they can take extremely high tension loads. This meant that our bridge would deflect a lot under a load because the string stretches a lot under tension. With the string as part of our bridge design, we were able to reduce the number of popsicle sticks in our bridge to 80, making it extremely light with a high strength to weight ratio. Construction Design (Joe): One of the first decisions we made was to construct our bridge using individually-manufactured T-members. As such, we needed a way to quickly and effectively manufacture T-member seg- ments of repeatable sizes and build integrities. To accomplish this, we laser cut cardboard molds, allowing us to position, glue, and clamp up to a dozen individual t-members simultane- ously.
Figure 2: T-Beam construction Experimental Instron results are included later in the report, and one of the reasons we feel comfortable predicting the performance of our bridge with those experimental results is because of this highly repeatable construction setup.
Figure 3: T-Beam gluing in the molds ENGR 2320 Final Project Report Alex Crease, Joe Kochevar, Eric Schneider, Paul Titchener
6
Analysis: Our analysis consisted of designing and testing repeatable members and determining the amount of force that those members need to support. This is covered in more detail in the fol- lowing sections. Experimental Analysis Basis (Joe): By using a repeatable construction design, we were able to test a number of our segments on the Instron, and extrapolate our data to reliably analyze the rest of the bridge. Our tests showed that each T-member could endure approximately 450 N of compressive load, which, when con- sidering the construction geometry, gives us a minimum factor of safety of 1.73 in the compres- sion members when supporting a load of 200 lbs.
Figure 4: This figure shows the compressive load curve of a T beam the full length of the T beams used on the bridge. The beam supports 450N before buckling ENGR 2320 Final Project Report Alex Crease, Joe Kochevar, Eric Schneider, Paul Titchener
7
Forces in each member from Solidworks (Eric) We used the Solidworks model of the bridge shown in Figure 1 and a weldment assembly to estimate the force in each member of the bridge. We used this essentially to see how the forces broke down in the bridge geometrically.
Figure 5: Solidworks weldments model of the bridge. This simulation was run with a 1000N force at the center of the bridge. The weldment simulation gave the following results, which were very useful for comparing to our experimental results. Type of member % of total load Instron determined strength rating Compression members (T-beams) %26 of total load 260 N at 200 lb load, pieces rated for 450 N in compression Tension members (string) %13.5 of total load 135 N at 200 lb load, testing for strength rating not successful ENGR 2320 Final Project Report Alex Crease, Joe Kochevar, Eric Schneider, Paul Titchener
8
Cross bracing %2% of total load 20 N at 200 lb load, pieces rated for 2000 N in tension End to end tension member %20 of total load 200 N at 200 lb load, rated for 2000 N in tension Figure 6: Loads and factors of safety for bridge members. We used Solidworks to estimate the forces and ignored the calculated stresses because of the difficulty in modeling stress in wood, which has wildly different behaviors parallel and perpen- dicular to the grain. The distribution of forces matched our hand calculations well, and so we used these predicted forces along with our Instron results to guess which parts of the bridge would fail. When string was glued directly to wood, for example, the joint held about 100 N, which was less than the force that Solidworks predicted on the string members, so we attached the strings in a stronger manner, looping the string over the top of the T-beams.
Figure 7: Our final joints had the string wrapped around the wood, not glued straight to it Compression Member Buckling (Alex): In order to determine the load that each compression member could take without buckling, as- suming the T beams were constructed as designed, they would have a radius of gyration of 4.89 mm, giving us a maximum compression length of 531.8 mm without buckling with pin joints. Given that our compression members are fixed and not pin joints, the members would buckle at a length of 265 mm. As our members were designed to be 260 mm long, they would not fail un- der compression given our loading scenario. Instron Testing (Paul): Once we knew the forces that each member needed to take in order to support a 1000N force at the loading point (slightly more than 200 lbs) we tested each member of the bridge to ensure that it would not break under the forces applied to it. We tested several setups: ENGR 2320 Final Project Report Alex Crease, Joe Kochevar, Eric Schneider, Paul Titchener
9
1. T Beam compression members. We tested individual T beam members at the same length as we intended to use them in the final structure. This allowed us to know precisely how they would buckle and com- press as they were loaded. 2. Popsicle stick tension members One of our plans for the bridge design was to construct tension members through gluing multiple popsicle sticks together. This test determined the strength of a face to face glue joint between two popsicle sticks in compression or tension. 3. String glued to wood This test showed the strength of a joint where a piece of string was simply glued directly to the face of a piece of wood. 4. String Our second design for the radial tension members, and the one that we decided to use, was to use piece of string rather than popsicle sticks. This test was meant to show us the maximum load that a piece of string could carry. However, this test was inconclusive because the string either slipped of the grips of the Instron or broke within the grip.
Figure 8:Example setups in the Instron mechanical tester. These pictures show testing of T beams in compression.
ENGR 2320 Final Project Report Alex Crease, Joe Kochevar, Eric Schneider, Paul Titchener
10
Results of experimental testing: Member Tested Force (Direction) Failure Mode Popsicle Stick 163N (Compression) Buckling Popsicle Stick 2000N (Tension) Splitting Along Grain Single Length T Beam 550N (Compression) T member (see figure be- low) splitting along the grain of the web Full Length T Beam 450N (Compression) Buckling 2 Popsicle Sticks with Wood Glue joint 1400N (Tension) Glue joint shearing 2 Popsicle Sticks with Gorilla Glue Joint <500N. Broke during clamping Glue joint shearing String glued to a popsicle stick (no knot) 110N (Tension) Glue joint shearing Figure 9: Loads supported by various popsicle stick members.
Failure Analysis (Eric) Our bridge failed at 490 N, in a twisting motion. One of the T-beam to T-beam joints was pulled inwards, putting the bracing X in compression until the joint failed. We believe that as load was applied a glue-covered string on one side of our bridge stretched differently from the other strings, pulling the hook off center and causing imbalanced forces from one side to the other. ENGR 2320 Final Project Report Alex Crease, Joe Kochevar, Eric Schneider, Paul Titchener
11
Figure 10: One of the right-hand strings was covered in glue, which we believe is what pulled the hook in this image so far off-center. Alternatively, we may have placed the board off-center at the beginning. Either way, the off center loading ended up pulling the compression members in towards the center of the bridge.
Figure 11: Here you can see the rightmost joint failing inwards. Note the massive compression and buck- ling occurring in the bracing X member. Under the intended loading these members were designed to only experience tension loads, and therefore were not built take the loads that they experienced in practice when the load was off-center. In the time-lapse in Figure 11 you can see the T-beam to T-beam joint push its way inwards un- til the joint gives. If we had further strengthened the bracing members for compression the joint would have held for longer. We didnt expect the bracing members to be in compression, which is why we didnt strengthen for that load, but its become clear to us that bracing members like this need to be much stronger than the predicted force because they are the members that ab- sorb the force created by slop in any of the other systems. ENGR 2320 Final Project Report Alex Crease, Joe Kochevar, Eric Schneider, Paul Titchener
12
Conclusion Our bridge was the lightest in the class, at 300 grams, and had the second highest strength to weight ratio, 166. Our bridge failed at the point it did largely because of a construction failure, where a tension member got coated in gorilla glue and failed to stretch at the same rate as the other members, causing a force imbalance. We expect that if that string hadnt been coated the bridge could have held much more, possibly up to the 200 pound limit. We believe this mostly because of the testing we did on individual elements - we knew how much force the main com- pressive and tensile elements could hold before they would fail and where they would fail. How- ever, we assumed that the loading would be more centered than it was and the behavior of the bridge changed drastically and unfortunately when slop in our system caused non-ideal loading.