A4- O'CALLAGHAN, WETZEL, FRANCO

Introduction

Engineering 103 was comprised of exploring and experiencing the design and modeling process for building bridges.  This process was followed by a sequence of designing, building, and testing through the use of a variety of tools such as the computer software West Point Bridge Designer 2012, model materials such as K’nex tools, and innovative in-class testing of the finished versions of the various bridge designs.  The objective was to understand how the bridge would, in theory, behave when the real-world bridge project was completed, as well as the effectiveness of each group's bridge and the groups' abilities to work in teams efficiently.

Each group entered their design into the class competition.  The different models were tested in comparison to each other which resulted in determining a model that satisfied the best cost to weight ratio.  As the experimental process followed an in-depth analysis, many steps were executed by each team.  The teams went through several phases which included planning, documenting, computer modeling, static analysis and finally, physical modeling.  Each failure, as well as each success, was observed and analyzed in order to gain more knowledge of the module through blog documentation.

The Design Process

Throughout the course, the students learned what happens behind the scenes when a bridge is being designed.  When the class was first introduced to the topic of bridge designing, group seven had the obvious goals of designing and building the most efficient bridge possible.  This was determined by who had the lowest cost to weight ratio, or the bridge that could hold the most weight relative to its cost.  The instructor had many ideas as to how to help the students understand how to complete this goal.   

Before the class could start designing their own bridges with K’nex pieces, they were instructed to learn about bridge designs through their own research and to use a bridge designing application, West Point Bridge Designer 2012.  This program let the students design their own truss bridges from a lateral point of view.  The students had the option of using different types of materials as well as different lengths and thicknesses of these materials.  After the design was complete, the program then took the design and turned it into a 3D animation, demonstrating a truck driving over the bridge.  If the truck made it all the way across the bridge, the bridge was successful.  However, if the truck failed to clear the bridge, it meant that the bridge design failed and at least one member of the bridge had too much compression or tension acting on or against it.  The program also showed how much tension and compression was on each member and calculated the cost of the bridge based on the member size, length, slenderness, cross section, and material type.  West Point Bridge Designer 2012 really helped the students to learn which designs were most efficient and which designs dispersed the weight of the truck most evenly.  After playing and experimenting with WPBD, the group decided that a deck truss would work best.  This is when the truss sits atop the bridge.  This was picked because it was most common out of the group’s designs and the group felt it would be easiest to build out of K’nex pieces. 

The instructor also required the students to complete an exercise where they had to figure out the tension and compression on each member of their bridge.  This gave the group an estimate of how their bridge would fail and what changes they should make, if any.  Group seven decided that they should add more support pieces to their bridge and try to add some middle cross section pieces, rather than just connecting the two sides of the bridge together on the top and bottom.  At this point the bridge was composed of nine cubes lying side by side.  In the three middle cubes, there was an X shape along the two vertically standing edges of the box with a gusset plate at the middle of the X.  A K’nex piece was used to connect the X’s front to back to keep the bridge from twisting. This can be seen in Figure 1.

Figure 1. Bridge before Modification

This was the one problem that was noticed with the bridge design – the bridge would twist in opposite directions when weight was placed on top of it.  This was taken into careful consideration when the group redesigned their bridge.  Up to this point, this was the final design chosen for the three foot long bridge, but could only hold a maximum of thirteen pounds.  Only after testing the bridge which resulted in this outcome did the design change once more.  The bridge as designed was not very efficient and the group hoped to increase the amount of weight their bridge was able to support.   After the bridge was modified one more time, the group predicted that the bridge would be able to support seventeen pounds. 

The Final Design

Figure 2. Drawing of Final Bridge Design

After designing, testing, and analyzing the three-foot bridge in numerous ways, the design was ready to be finalized. The final design for the bridge consisted of nine squares, making up the length of the bridge.  The plan for the final design bridge can be seen above in Figure 2. Each square had an X through it. In other words, each square was composed of four right triangles, their right angles at the center of the square. To give the bridge width, enabling for vehicles to travel across, 3.375” chords were extended to and from the corners of each square, respectively, between the two side trusses of the bridge.

 

Figure 3. Final Bridge Design

Having learned that the weak joints were those composed of a chord clipped into the gusset plate and lying perpendicular to the gusset plate, all but eight of these joints were replaced. The old joints allowed sliding of the chords which were responsible for preventing twisting of the bridge, causing the bridge to twist under light loads. The new, stronger joints were composed of two interlocked three-hundred-sixty degree grooved gusset plates with chords snapped securely into their sockets for better support against twisting of the bridge. By analyzing the way the bridge twisted as the weight of its load was increased, the best set of joints to leave unchanged was determined. Because the price of each new joint was quadruple the price of each old joint, eight old joints were carefully chosen to remain intact, saving $24,000 from being added to the cost of the bridge.  The final design is pictured above in Figure 3.

Overall, the bridge was priced at $409,500 and was comprised of two-hundred-sixty-five parts, as calculated on the spreadsheet in Figure 4. During the competition, the bridge collapsed under the weight of 14.7 pounds after undergoing too much tension, as seen in the images below, Figures 5 and 6.

 

Figure 4. Bill of Truss Materials

Figures 5 and 6. Breaking Point

Final Results

The bridge was designed as a deck truss made up of equal size cubes in a line.  When weight was placed in the bucket, the bridge had a tendency to wiggle back and forth.  In other words, the top level of the bridge would move back and forth while the bottom stayed still.  This resulted in more and more twisting as the bucket of sand got heavier.  After the bridge reached its maximum load weight, 14.7 pounds, the bridge twisted so much that a member from the top level snapped out of the gusset plate that it was connected to.

The group tried to prevent this twisting from occurring.  After adding in the X shapes all along the bridge, the front and back pieces of the bridge were able to be connected in three layers top to bottom instead of just two.  The group thought this would support more weight because the bridge wouldn’t be able to twist as far if more pieces were holding it together.  However, this really did not help at all since the bridge repeatedly broke in the same place and only held half a pound more.  This was insignificant compared to the cost of all the members we had to add to create the X shape design.

Conclusion

          As engineers of all fields, the class gained valuable experience by undergoing the bridge design process. One did not need to be a major in civil engineering to learn that there is much testing, analyzing, and teamwork required for the final product of an engineering design process to be successful. Although the bridge designed by group seven was not successful compared to other bridges in the competition, it was successful in putting the group through a small-scale engineering project that required them to understand all of the major—and some of the minor—steps in any engineering design process. The strategies, plans, elevations, calculations, computer testing, physical modeling, and physical testing are all aspects of engineering projects that the group members will grow to be very familiar with, even as mechanical and electrical engineers. 

Future Work

Based on the final product of this project, one of the changes that could be applied in the future would be to change the way the K'nex grooved gusset plates are joined to each other.  During the design process, the group learned that grooved gusset plates tend to slip apart under tension.  Therefore, whatever pieces join the chords of the bridge together need to be very strong in order to successfully sustain the amount of weight and force distribution on the bridge.  Additionally more care should be given to the variety, magnitude, and location of the possible angles of the structures that form the bridge.  There might be new undiscovered advantages if there were more angles available for use in the bridge structure, but K’nex pieces limited which joint angles could be used. 

Term Review

Last week in class, our group tested our 3' bridge design. Before doing so, we'd submitted a survey about the bridge to our professor who put the info provided by the survey into an Excel sheet. Using the Excel sheet, the entire class was able to compare the cost and strength of each bridge, side-by-side. As a class, we decided to test the bridges in order of highest to lowest predicted cost-to-strength ratio. Which group went first? None other than ours. Even worse than the 17 pounds we'd predicted our bridge would hold, it only supported 14.7 pounds in the final test. To make things worse, this meant our cost-to-strength ratio was even  greater than we'd estimated.

Now that the final competition is over, I can look back at all the steps we'd taken throughout the course and honestly say that I learned something worthwhile about every single one of the topics identified in the goals for the course. Working with a group required teamwork. Planning, documenting, computer modeling and physical modeling were all part of the design process which our group followed throughout the course. We used some computer software I'd never used before, and we used some computer software that I've already been familiar with for years. Lastly, we used static and forensic analytical methods to learn more about the weaknesses of our bridge designs.

I think the only thing that was not beneficial to me was the fact that this Engineering 103 lab was disconnected from the Engineering 103 lecture section of the overall course. If the bridge lab could be formatted to follow the general pattern of the assignments for all the other labs, I think I would feel like the Engineering 103 lecture section of the course was more important to me, since it is something I am required to attend. Basically, I just felt that attending the lecture session wasn't very important because I was told to ignore what they told everyone about assignments being due and ways to earn extra credit. On the other hand, I feel that having individual blog posts benefitted me alot. I think this is something that every lab should have. Although teamwork is important--especially as engineers--we are not paid to do the work for our classes, which means some students simply do not care as much as other students. Some students know that other students will do the work for them so that their own grade is not negatively affected by the actions of a lazy student. I've experienced this in Engineering 101 and 102, and the students who did not do all of their work did not have anything held against them. In the real world, they could potential be fired from their job, but in class, they won't even receive an F because someone else will end up doing the group assignment this student was responsible for. Instead, individual assignments reflect the work of each student, even if the assignments show similar results because of two or three students being in a group together.

- Melissa Wetzel

Bridge Process

During class last week, my group and I tested a three-foot bridge we'd designed over the course of the previous week. This bridge held only 14.2 pounds and its point of failure was caused by weak joints for the cross beams. Under 14.2 pounds, the bridge twisted and its members popped apart. To fix this, we replaced the weak connections with stronger ones. The original connections consisted of a Knex chord laying in the gusset plate joint. These connections were weak because the chords could slide freely through the plates, although being held against them. The new connections were created with two grooved 360-degree gusset plates joining the chords' ends like two puzzle pieces interlocking. This type of connection does not allow for movement and helps minimize the bridge's horizontal displacement. Next week we will test out three-foot bridge and compare our results to the rest of the groups' results to see which group has the best overall bridge.


Now that my group and I have almost completed the bridge design process for the term, I am realizing how much I have learned, specifically about bridge design. In class yesterday, Dr. Mitchell put three rough bridge analyses on the board. All were the same width and comprised of three triangles in the same order, however each was a different height. According to the analyses, a bridge whose height is the taller, or whose triangles are the least obtuse, will have the least amount of compression and tension on its members under a given load. Another important key to designing bridges that I've learned is that joints tend to be the weak spots. Before taking this class, I always thought that chords were the "weakest links" in bridges, but it is truly the "links" which are weaker. In testing the Knex, I have read information about which joints can handle the greatest stress force before failing, but we have never analyzed the force it would take to snap a chord in half. This is because of weak joints.

A3 - WETZEL

Truss Overview:

Free Body Diagram & Calculations:

Replication of Analysis in Bridge Designer:

In order to make the results of the hand analysis correspond to the online Bridge Designer program, some simple scaling was used. While using Bridge Designer, I mentally set each grid square to be 3"x3". Because my bridge constraints set the bridge length to be 36", that meant 18 squares on the grid would correspond to the length of my hand-drawn bridge. The height, needing to be 10", I scaled at about 3.3 squares on the Bridge Designer grid. This enabled me to use Bridge Designer to see if my own calculations were right about this bridge by creating angles similar to the ones I used. As long as the triangles composing my bridges corresponded, the tension and compression forces would also correspond. I was happy to see that the results calculated by Bridge Designer were very closely related to the results I'd calculated myself.


Knex Bridge Designer Analysis:

Just as I've done for the hand and computer analyses of the small bridge, I've now designed the Knex bridge using the Bridge Designer program. In order to scale the draft's calculations to the actual bridge, I will need to measure the actual bridge and set a value for each square on the grid in the program. After we test the bridge and see the weight it fails under, I will be able to scale the real load with the one I've used in the designer to determine the stress on the Knex bridge when it is holding a specific load.

Given the testing information about Knex joints, I might use this analysis as a guideline for the Knex bridge. I can say that, theoretically--and based on the computations of the Bridge Designer--the bridge will be able to hold x amount of weight when there are y joints. But if I optimize the strength of the joints by having more chords meet in a single joint, the bridge should be able to hold more than the program suggests. The testing information about Knex shows that a joint is stronger when more member meet at it. At the same time, the Bridge Designer's unique formula "members + 3 = 2*nodes" leads users to do the same thing: have multiple members meet at a single node. By following this pattern, my group and I should be able to increase the strength of our bridge enough for the cost to strength ratio to benefit from the modifications.

Analysis Process

Last week in class, I worked with the other members in my group to analyze our Knex bridge design using online bridge designer software created by Johns Hopkins University. It was difficult for us to put our exact bridge design into the software because of a special restraint the program called for: "Members + 3 equal twice the nodes." Basically the problem was that our bridge is symmetrical about a single node--not about a single member. This caused our bridge to have an even number of members. When you add 3 to any even number, you come up with an odd number which, when divided by, is a decimal. But for our bridge, we cannot have a fraction of a node, so we needed to change our bridge design in order to use the online design program. The fact that this program wouldn't accept a bridge whose exact center point was a node rather than a member, led me to question our bridge's strength. This week in class, I will discuss with my group mates ways we can modify our current bridge design so that it follows the "members + 3 = 2*nodes" rule.

As far as analysis goes, I do think this method of analysis is efficient for a real bridge if additional factors are considered. On a real bridge, the weight of each bridge component needs to be taken into consideration when determining the tension and compression on chords. The steal beams, metal bolts, thick plates, and even the paint, all add to the total dead weight of a bridge which is a factor significant in determining the forces on each bridge member. If this dead weight is not accounted for, the bridge would fail before reaching its calculated load capacity. After thinking about this, I decided that I'd like to analyze the twisting of truss bridges. One bridge design that my group mates and I tested ended up twisting and snapping. I think it is important to know how strong and, as a result, heavy, a cross beam needs to be so that it is not adding to the bridge's dead weight, nor decreasing the bridge's weight capacity.

Analysis Desires

Last week, my group mates and I tested our 2' bridge in a competition against the rest of the groups in our class. Our bridge would have cost over $305,000 to build full-scale, but it only supported 17.4 pounds of sand during the test. This was the first time we tested one of our bridges using the apparatus, and we were able to get a good view of how our bridge failed. This week we will be testing our 3' bridge which was designed completely differently. We decided to cut costs by using different gusset plates and longer chords. We believe our bridge will be able to support more weight than our last bridge. And even though this bridge is a foot longer than the last, it's nearly $50,000 less expensive.

In designing the bridge using Knex, something that is challenging is having to run several tests and analyze each one to determine where the weaker points are. In West Point Bridge Designer, a table of information was presented after each simulated test. The info included statistics regarding each individual chord's tension and compression forces, and slenderness. I would like to be able to determine the tension and compression forces across the top and bottom of the Knex bridges we've been designing, but the problem is that WPBD bridges normally fail because of weak chords giving out. The pattern I've noticed with the Knex bridges is that the joints, or "Gusset plates," have been failing before the chords. I am not sure how we could calculate the tension and compression forces.

Knex Process

Last week in class, our group discussed the bridge designs we'd submitted on paper and chose a strategy for building our Knex bridge. After modeling one of the paper drafts which used a number of 3.375" chords, we chose to shorten the average length of the chords and increase the number of joints. Doing so raised the cost of our bridge but also increased its strength greatly. We are going to test the bridge in class this week and pay close attention to the way the bridge collapses. I would like to figure out how to disperse the load weight across the bridge so that the bridge doesn't have a weak point in the future.

There are important differences between designing a bridge using Knex and designing a twenty-foot "real" steal bridge. The most obvious, I believe, is the length to weight ratio of members. Knex are very light plastic pieces which seem to be very strong for their weight. Without having yet compared a Knex bridge's strength-to-weight ratio to that of a steel bridge, I am imagining that the steel bridge would have more dead weight in proportion to its strength. On this note, a steel bridge with the same design as a Knex bridge may not be able to support as much load as a Knex bridge due to the different strength-to-weight ratios. The steel bridge may give out sooner because of this, or it may hold together longer due to the use of better Gusset plates.

While experimenting with the Knex bridge designs, I have learned that the Knex grooved Gusset plates tend to be the weak points in the designs. The smooth plastic pieces easily slide apart under tension, causing the bridge to fail. On a "real" bridge, there would be stronger Gusset plates due to a greater amount of friction between materials and the usage of bolts for additional support. Both of these elements would help strengthen the steel bridge, but neither of these options are available to strengthen the Knex bridges, making the steel bridge potential stronger.

WPBD vs. Knex

Last week, our class learned about the apparatus that is going to be used to test our bridges. I took note that the Knex bridge design we test needs to have a flat surface on the top of it in order for the apparatus to sit evenly across the bridge. Mr. Jay Bhatt came in to speak to the class and showed us some very useful resources that I will use to learn more about bridges as my group and I continue to improve our designs. Towards the end of the class period, my group mates and I discussed ways to use the Knex to design a bridge that would be strong when compared to its cost. We all decided that the best thing for the group would be for each of us to design Knex bridges separately and compare our design results. This week we will talk about the ideas we had behind our bridge designs and, out of the three of our design, which one we think will be the best.

As we transition from using West Point Bridge Designer on the computer to actually building our designs using Knex, there are going to be a number of differences. Primarily, I must point out that WPBD had a set number of connections for the road surface of the bridge. Not being able to move these connections or change the number of them affected the designs of the bridges. Additionally, WPBD allowed chords of any length to be created, their materials to be changed, and their diameters to be altered. Using Knex, we will be limited in the sizes which will affect the strength of the bridge, the shape of the bridge, and the size of the bridge. Only certain triangular patters will be able to be created. Because the longer pieces are more slender, it will be very important to leave them out of the design. The more slender a member, the weaker it may become as weight is added to the bridge.

- Melissa Wetzel

A2 - WETZEL

The goal of my design was to have a low cost and high strength. Never having designed or tested model bridges before beginning this class, my best idea was to use shorter chords as members of the bridge. Although joints each cost at least $1,000, I decided that their use in between short chords would help make the bridge stronger and tighter once weight was added. By tighter, I mean that I am imagining as weight is added to the bridge's load, the chords will push into the joints and the displacement of the joints will be towards the center of the bridge. As the members move towards the center like so, the stronger, shorter chords will lean into the joints, tightening the joint sections.

Elevation and Plan for the Knex bridge design.

I made sure that the web of the truss was composed entirely of triangles--most of which, in my case, happen to be right triangles. I am hoping that these hold up as well as equilateral triangles may have. My prediction is that if the right angle of one triangle does not have support backing it, it will give out before other parts of the bridge. This is something that I will pay close attention to so I know what the best angles are to use as the groups approach the drafting of their final designs. Something I purposely did not do was make the plan of the bridge out of triangles. I figured that the weight-bearing members all ran vertically, so I did not want to add dead weight to the bridge or raise its cost by webbing its ceiling.

The Truss Bill of Materials

I did my best to keep my design simple so that there would be no useless chords or joints in the price, but I am not yet sure that each chord that I've left will make the bridge strong enough. I was both surprised and glad to see that the total cost of my first Knex bridge design was below $200,000--a price I'd been trying to get my bridge down to in West Point Bridge Design.

West Point Bridge Design

Last week each student in the class submitted their own bridge design to our professor, James Mitchell, online via a survey. In class we compared the results of each bridge side-by-side. My bridge, coming in at close to $590,000, was one of the most expensive. The least expensive bridge had a cost of just below $250,000. I clearly wasn't going to be winning this first competition.

For the rest of that class period, each group was to analyze the bridges each member of the group had created. My group members and I decided that we would benefit the most by attempting to build a bridge entirely from scratch on West Point Bridge Design, keeping one goal in mind: have a cost below $250,000. After about 50 minutes had passed, the least expensive bridge we could design would cost about $440,000. I was happy with this, however, because in about a day I'd learned how to reduce the cost of a bridge by over $100,000.

My group and I decided we would play around on WPBD before our next class together, but our group's bridge had already been submitted and there was no way we could do anything to change our design's official status until the next class. I am very much looking forward to seeing the results of the competition between groups in class tomorrow. I'm sure our bridge is not the least expensive, but maybe it is the safest or the most durable. On another note, the group is not having any trouble working together. So far we haven't run into any issues that we cannot fix by simply looking in a book or asking a question to someone with a good understanding of bridges.

Going back to the WPBD program, I have noticed that some aspects of it are not very realistic, however, the program helps the used understand many important areas of civil engineering. As we discussed in class, when you test a bridge, the program exaggerates the displacement of the bridge's road surface so that the user can easily see how significant the changes are that they make to their bridge.

Without even realizing it, users of WPBD are learning how civil engineers use computers as tools to help them perfect their designs. This program calculates the stress on each member of the bridge and displays it in a chart within a fraction of a second after running the test mode for a bridge design. This saves time for the civil engineer and allows them to make more changes to their designs without having to go through the process of manually calculating all the forces acting on each member of each bridge design.

Personally, my favorite part of WPBD is that it calculates the cost of my bridge for me. Every time I change the size, material, or structure of a member, the cost automatically updates. This makes it easy to avoid heading in the wrong direction with a design and ends up saving alot of time. The program provides standard options for bridge materials and dimensions which helps users learn which materials are the best for designing bridges.

Overall, the program is unrealistic in the sense that it does not account for certain forces which act on bridges. Wind is something that can significantly affect a bridge in reality, but the program does not provide for such a force. All of the things that program does do accurately definitely help the user speed up the process of designing a bridge.

Research Questions

During class in week 3, Mr. Jay Bhatt, Drexel University's research librarian, will be visiting to help students with strategies for research. In week 2, Mr. Bhatt spoke to the entire Engineering 103 student body about using the library's website as a main research tool. He showed us where to find different databases to search all sorts of subjects and he helped us to understand which resources to use to find certain types of information.

After completing my first bridge design for Engineering 103 using the West Point Bridge Designer computer program, I have begun to notice patterns that fail and patterns that are successful for bridge truss designs. I have specifically noticed that on alot of designs I came up with, the top of my truss was weaker than the bottom. I understand that this is because of the way weight is dispersed across the bridge, but I would like to learn more about strong triangular designs commonly used in trusses and what it is about these designs thats makes them so popular.

I am hoping that Mr. Bhatt will be able to help me find resources to answer the following questions:

1. Does the design of a truss above/below the road surface of a bridge affect the bridge's potential strength? If so, does placing a truss above the roach surface of the bridge make the bridge stronger or weaker?
2. Is it better for a truss to be made of small triangles of equal size or large triangles of varying sizes?
3. Is there a certain way to go about designing a truss? Are there any base-designs or patterns to follow other than using all triangles?

A1-WETZEL

The goals I had in mind for my bridge design were:

• to build a two-lane truss bridge without the use of pillars
• for each shape made by the truss to be a triangle--the strongest shape
• for the truss design to include several small triangles stacked upon each other

My bridge design from the "Drawing Board" view.

A truck crossing the bridge in "Test" mode.

A table displaying the "Load Test Results" regarding each member of my bridge design.

While designing my bridge, I decided that instead of using small triangles, I might benefit from using larger triangles as a way of dispersing weight across the structure. I also learned that a member is too slender to be useful if its slenderness rating is above 300. Taking advantage of this information, I decided I could make my bridge stronger by keeping the slenderness rating below 200 for each member.

Currently, my bridge design cost is $585,603.91. After learning more about bridge design and developing strategies for keeping costs low and strengths high, I think that I could potentially lower the cost of this bridge to around $380,000--a difference of over $200,000.

From designing this bridge, I have learned that the weaker area of a bridge can be the top of the truss. For bridges whose trusses are above them, I'd always believed the bottom of the truss to be weaker. By experimenting with different bridge designs, I've notices that during failed tests, the top of my trusses were almost always the weaker sections. This is something I will try to understand better as I research bridges and learn more about successful bridge designs.

Teamwork

Posted by Melissa Wetzel

As a group, Ana, Rebecca and I have read and discussed Professor Mitchell's website about teamwork and concluded that it will be to our advantage for us to be open and accepting of each other as teammates. The importance of clear communication between teammates is more than obvious, and as a group we understand that communicating our ideas will not only make it easier to excel in our work, but will enable us to expand on each other's ideas and collaborate on our work. While collaborating with each other, each teammate will be able to “bring to the table” their best qualities and skills in each of the areas we will study. At the same time, we will be able to learn from each other and view challenges from different perspectives which also has the potential of benefiting our team.

Naturally our team is not going to be perfect. When people exchange ideas and try to conceptualize with each other, conflict is bound to arise. As a group we have decided to respectfully accept each other's ideas and remain patient when one of us does not understand something. We are a team so we can only succeed together—not individually. It is in each of our best interests to help each other along the way, but to each equally contribute to the group's work. We will plan who is to complete which tasks and we will follow up with each other so that we are all on the same page as to how the team is progressing throughout the quarter.