PowerPoint

BikeEeeeeeeeeee.pdf

Home >Engineering homework help >Mechanical Engineering homework help >PowerPoint

A Soft Ride at BikeE A Case Study for The Mechanical Design Process

Introduction During the 1990s, BikeE was one of the top manufacturers of recumbent bicycles in the world. A recumbent bicycle is one where the rider is seated or lying down. In 1995 BikeE introduced the AT model with an active rear suspension. Later that year the AT was named one of Bicycling Magazine’s Best New Products of the year. From 1996 - 2002, BikeE sold over 15,000 ATs. Many are still on the road today. This case study was used as an example in the 3rd edition of The Mechanical Design Process1. What makes it a unique case study is that it takes a product from need, through concept development to detail design. It uses many best practices detailed in the book. The Problem: Design a rear suspension for a recumbent bicycle The Method: The engineers used the best practices in The Mechanical Design Process in great detail. Advantages/disadvantages: This case study shows many best practices in detail.

1 Note that the author of the Mechanical Design Process was also the lead engineer for the development of the BikeE AT.

Figure 1: The BikeE AT

Background The BikeE Corporation manufactured bicycles that allowed the rider to pedal in a seated position. This position is more comfortable than that with a traditional bicycle, with little pressure on the neck, wrists, and arms. On these bikes, it is easy to see the scenery or traffic as the rider's head is upright. Additionally, since the frontal area is small, they can be very fast. These types of bicycles are easy to maneuver and fun to ride. The BikeE CT was the first model produced. It was introduced in 1992. The CT (Fig. 2) is characterized by a cantilevered rear stay (the part between the aluminum extrusion body and rear wheel). This stay is cantilevered to provide a little flexibility much like a diving board, a simplistic suspension system. It was important in the design of this bike to have as much flexibility as possible in the rear stay as 75 percent of the rider's weight is over the rear wheel. Additionally, some limited shock absorption is offered by the flexibility of the wheel and the foam cushion on which the rider is sitting. In 1995 BikeE undertook a project to design an actual suspension system for the rear wheel. The resulting product was the AT. This case study describes the evolution of the AT.

The Design Plan BikeE was a small company developing its first bicycle rear suspension system. The design team consisted of a design engineer, a product manager, a technician, manufacturing manager, machinist, materials specialist and an industrial designer. The core of this team was the design engineer, the product manager and the manufacturing manager. The core team drafted a list of tasks, as shown here. Note that these are fairly generic.

1. Generate engineering specifications. 2. Design two concepts. 3. Develop prototypes PI. 4. Test PI prototypes. 5. Select one concept. 6. Develop P2 prototypes. 7. Field test P2 prototypes. 8. Generate product documentation 9. Produce production plan.

Figure 2: The BikeE CT

These, they put together in a Gantt Chart, Fig. 3 to show the relative timing for the tasks.

QFD Development To ensure that they understood the problem, the team developed a Quality Function Deployment (QFD) chart (Fig. 4). For the BikeE suspension system, the main customers were bicycle riders; and were generally both the purchasers and the users. In interviewing current riders of BikeE products, the team realized that there were two types of riders for the new product. First was the rider who rides solely on streets. The BikeE was initially designed as a road or commuting bicycle. However, there were some riders who wanted to go off-roading, so the second group of rider-customers were those who wanted to ride on rough roads or trails. Additional customers considered were bicycle shop sales people and mechanics (often the same people). These people had to be enthusiastic to sell the product, answer questions about it, and repair it. Within the company, manufacturing, assembly and shipping personnel were also considered as customers but not shown in the example QFD. To gather customer information for the BikeE suspension system, the team developed a survey and distributed it to current BikeE owners. Below is a sample of the questions included in the survey:

Q1. How many miles do you ride your BikeE each week? (Circle the best choice.) 1. <5 miles 2. 5-10 miles 3. 10-30 miles 4. >30 miles

Figure 3: Gantt Chart for BikeE AT Tasks

Figure 4: QFD for BikeE AT

Q2. What surfaces do you often ride on? (Circle all that apply.)

1. Smooth pavement 2. Rough pavement 3. Gravel 4. Packed dirt paths 5. Forest trails

Q3. If your BikeE had a rear suspension system, what types of surfaces would you ride on? (Circle all that apply.)

1. Smooth pavement 2. Rough pavement 3. Gravel 4. Packed dirt paths 5. Forest trails

Q4. If your BikeE had a rear suspension system, how often would you expect to adjust or maintain it? Consider the maintenance similar to checking and adding air to your tires. (Circle the most frequent acceptable time period.)

1. Never 2. Once every 3 months 3. Once a month 4. Once a week 5. Every ride

Q5. What is most important to you? (Rank 1 through 5.) ___ Smoothing road surface bumps (e.g., rough surfaces, manhole covers) ___ Absorbing pothole shocks ___ Low maintenance ___ Cool looks ___ Maintenance ease

Q6. What is your weight? ____ Q7. Describe your experience using your BikeE to go to work or school.

The design team used the results of the survey plus interviews with bike shop personnel to develop the list of requirements shown in Table 1 and in the QFD. To find the relative importance of the requirements they presented the list to potential street riders and sales-repair people and told them that they had 100 pennies to distribute amongst the requirements and to

Table 1: Preliminary List of Customers' Requirements

put the most pennies on those they thought most important. The average results for each group are shown in the first columns of Fig. 4. To find the true market opportunities the team compared the customers’ requirements to three existing benchmarks. The first was the current cantilever product shown in Fig. 2 (the BikeE CT). The second was a traditional mountain bike system and the third was a recumbent competitor's system. Although there are many mountain bike configurations to choose from, only one was used here. To determine how well the competitors met the requirements, the design team used questionnaires to evaluate them. The average results from street riders are shown in the far right columns of Fig. 4. Important points to note are that:

1. The BikeE CT gives a poor ride on streets, but even the competition is not very highly thought of by the street riders.

2. The BikeE CT, with its semi-rigid rear stay, does little to eliminate bumps but can handle different rider weights and heights.

3 Neither the mountain bike nor the competitive recumbent do a good job of not pogoing (bouncing up and down with pedal pressure), or adjusting for rider height or weight.

4. None are easy to adjust.

These were all important factors considered in the design, especially those that were considered very important to the customers. The team then developed engineering specifications for the BikeE suspension as shown in the column headings in Fig.4. Some important points about these are:

1. In order to measure the "energy transmitted on a standard road," a standard road and a method of measuring the energy content of the road and that transmitted needed to be devised.

2. Some specifications are subjective. For example, it may be possible to actually measure "pogoing" and "looks" but, it may be easier to use a test panel and note percentage of subjects who notice pogoing or like the looks.

3. The "# of tools to adjust" is listed twice. The first time, for weight and height, will have a target of 0. This could have been listed as two different measures, but was combined, as the target is the same. The second, for adjusting the ride hardness, will have a target of 1.

Conceptual Design The design team took a careful, functional approach to the design of the swingarm. During the design of the semi-rigid rear stay for the CT model they had missed some functions and wanted to take more care there. For the suspension system, the "most important" function can be worded in a couple of different ways. The team used "transfer forces between wheel, chain, and frame and absorb peak loads between wheel and frame," which is really two overall functions - transfer and absorb. These helped them define the boundaries of the system: the wheel, the chain, and the frame of the bike, as shown in Fig. 5; and that the primary type of flow is energy.

Figure 5: The boundary of the suspension design problem. To help understand the function of the system, the design team drew a simple free body diagram (Fig. 6) based on Fig. 5. The arrows in Fig. 6 represent the forces due to the chain tension, the wheel pushing up, and the frame loading on the suspension system to balance out the other two forces. This problem is essentially a two-dimensional problem as side loads are small on a bicycle. Note that during the design of the BikeE CT, the one with no suspension, the force due to chain tension was not considered properly. It was later learned that the highest fatigue stress in the rear stay (i.e., the part that connects the frame to the wheel in Fig. 2) was due to chain forces and not the vertical forces due the rider's weight and resulting vertical wheel force.

Figure 6 The forces on the suspension system

Based on this understanding, the team decomposed the main functions into sub- functions, as shown in Fig. 7. The team decomposed the force from the wheel into as many functions as they could think of. Note that they focused on interfaces, how the energy gets in and out of the systems, and what has to happen internally to the energy. They divided the transfer of energy into concern for large bumps and small bumps based on their experience that it is difficult to get a smooth ride over little bumps and still have a system that can take the large hits. They also divided the energy storage from the energy dissipation, the spring from the damper.

Figure 7: Functional decomposition for the suspension system While developing the sub-functions they also remembered that the system to be designed would probably have to carry a fender and the brakes. They noted these as secondary functions. The team used the functional model as a basis for a morphology (Fig. 8). Here, the ideas generated for each main function are shown.

Figure 8: Morphology for the suspension system

Concept Evaluation The BikeE team used a number of different methods to reduce the total number of concepts generated with the morphology. As they reviewed the results of the morphology they eliminated many of the potential alternatives because they couldn't pass a simple go/no-go screening. Thus, they combined their generation and evaluation efforts. From the morphology members of the design team developed conceptual sketches in their design notebooks (Fig 9). Each concept sketch was followed by notes abstracted from notebooks. These were broken down into two main sub-problems: the bicycle's structure and the energy storage/dissipation method. The sketches labeled with an "S" are structural and those labeled with an "E" are energy management. During the exercise of developing these concepts the team found that they learned much about the project:

• Three of the four structural concepts used pivots. The development of pivots requires a new sub-project complete with requirements, function consideration, and concept generation.

• The energy management with springs and dampers can be implemented with an off-the-shelf unit or made from basic components.

• The four structural ideas can be combined with the three energy management ideas in up to twelve different ways. They needed to reduce these to two ideas before prototyping.

Figure 9: Concept sketches

They did take care not to eliminate too much, too soon, however. After using their judgment, they had four structures and three energy storage/dissipation methods. They thought that all 12 combinations of these were feasible-either conditional or worth considering. In considering the twelve remaining concepts, the team compared each to customer's requirements in Fig. 4 asked each the six technology feasibility questions. None of the team had ever designed and built a suspension system before, so the technology assessment of each of the concepts was based on what they had read about. They concluded:

1. Probably any of the 12 alternatives could be made to meet most of the performance requirements itemized in the QFD, within what they knew at this point in the project.

2. The truss structure could not pass the technology-readiness test at all. No one on the team was convinced that they knew much about the critical parameters, sensitivities, and failure modes of a flexible member on a bicycle.

Thus, this screening reduced the number of concepts to use the air shock, coil/oil shock, or elastomer on either the pivot on the crank, the pivot on the body structures, or the linkage system. The go/no-go screening left three structures and three energy storage/dissipation systems to consider. In this case study we will focus only on the team's effort to decide on an energy storage dissipation system. Keep in mind that none of the team had ever designed or built a suspension system before. The criteria for evaluation were developed from the customers' requirements in Fig. 4. The resulting list of eleven items is shown in Table 2. Some of the customers' requirements were left off or combined to keep the list to the most important for consideration. Since the list was new, the team reevaluated the importance of the 11 criteria by allocating 100 points among them. The results shown in Table 2 are compromise values for the entire team.

Table 2: Decision Matrix for energy management system

The three alternatives considered for energy management were the Cane Creek air shock, the coil/oil system, and the elastomer. Evaluation was done with the Coil/oil as datum. The other two options were rated using the -, S, + scale. The air shock had the best scores, implying it was the best concept. Note that in order to evaluate the elastomer, a test bike was built (Fig. 10). This bike not only allowed testing the elastomer, it also allowed the first experiments with the pivot on the body. Even though it was early in the project to build a bicycle, it was necessary because knowledge about the elastomeric suspension was just too low to evaluate without it.

Product Design Most of the effort to design this system was accomplished by a single engineer, Bob, a member of the team. Only a small part of Bob's work will be presented, as the entire history of the design process is too long. This discussion is organized, namely, from constraints to configuration to connections to components. Integrated into this flow is concern for

Figure 10: Elastomer test bike

materials and manufacturing processes, as Bob considered these concurrently. The first step Bob took was to understand the constraints on the system. Functional constraints were developed above, but here Bob was initially concerned with the spatial constraints. As shown in the layout drawing he made (Fig. 11), the other components that constrain the suspension system are the frame, the wheel, and the chain. The wheel is shown in two different positions, fully extended and fully compressed. Bob had to represent the chain as an envelope determined by what gear the bike was in and the compression of the suspension system. Although not shown, the suspension system could not interfere with the seat, the ground, and the rider; and the suspension system had to support a fender and a kickstand. All the items that are shown in italics provide spatial constraints on the suspension system. In fact, as with most designers, Bob initially focused on the constraints shown in the layout drawing and then checked the evolving product to ensure it cleared what it had to and that the accessories could be mounted.

Figure 11: Layout drawing Bob initially noted that there were only two new components, the swingarm (the member that connects the frame, the air shock and the wheel) and the air shock itself. However, by the time this product made it to production there were over 15 components on the BOM (Bill of Materials). In other words, to make this a producible system, each major component needed many pieces. A key decision Bob had to make while laying out the product was where to position the swingarm and air shock relative to each other and on the frame. He had to try many configurations (one is shown in Fig. 12), and for each he analyzed the forces, stresses, the loads on the air shock and the amount of potential pogoing due to the chain not being aligned with the swingarm. Additionally, he had to trade off the performance of the system versus its ability to fit within the envelope defined by the other components. This was especially challenging when he considered the chain, as

the swingarm needs to clear it regardless of what gear the bike is in, how far the suspension is deflected and the chain deflection when going over a bump. Figure 12 shows the result of Bob's effort with the swingarm roughed in and the positions of it and the air shock figured out. The swingarm used by Bob in this drawing is similar to that used on the experimental bike (Fig. 10). The layout drawing in Fig. 12 helped Bob ensure to himself that the configuration met all the constraints. He rechecked this frequently throughout the rest of his effort.

Figure 12: One layout of the swingarm Bob's next challenge was to design the connections or interfaces between components. This proved to take the greatest part of the effort, which is common. Essentially there are four connections, each will be discussed and the unique problems that Bob faced with each will be emphasized.

• Connection between wheel and swingarm. This was the easiest joint to design as all common bicycle wheels have the same axle diameter and connect to bicycles with a component called a "dropout" (if the nuts that hold the wheel on are loosened, the wheel drops out of the bicycle frame). The geometry and shape of a dropout was well refined and Bob had no reason to change it.

• Connection between the swingarm and the air shock. This too was well refined in that the air shock manufacturer supplied bushings to allow the shock to pivot relative to the swingarm (one degree of freedom). The drawing in Fig. 13 shows the bushings that allowed the shock to pivot. The only design issue was how to get the load from the swingarm to the air shock, a problem addressed shortly.

• Connection between the frame and the air shock. This also makes use of the same one-degree-of-freedom interface on the air shock, as shown in Fig. 13.

The big challenge here was in ensuring that the forces transmitted through the shock could be distributed in the aluminum wall of the frame, a topic considered later.

Figure 13: Bushings in shock

• Connection between the swingarm and the frame. This was the most time consuming interface designed during the project. It was essentially a one degree- of- freedom pivot but it had to connect the two blades of the swingarm to the thin aluminum wall of the frame. There are many ways to make a single-degree-of- freedom joint. Those that were considered included mounting the pivots to the side wall of the frame, the bottom of the frame and in a structure outside the frame. Additionally, the actual pivot bearings considered were a ball bearing (making use of an off-the-shelf assembly used for mounting the cranks and pedals), a bronze bushing, a plastic bushing, and a flexible joint (a solid piece of material that flexes within its elastic limit to provide a pivot that works over a small angle). The actual placement and configuration of the connection is dependent on the pivot types.

Finally, Bob developed the components. There are too many to detail all of them here and so only a few are discussed. The Swingarm Body Bob based his design on the cantilever structure from the earlier bike that worked fairly well and looked good. Additionally, the prototype, shown in Fig. 10, using a simple straight member seemed to work well, but Bob realized this needed extensive analysis and testing. Thus, even though most bicycles are based on a truss and trusses are strong structural shapes, Bob focused on a cantilever swingarm.

For structural members in bending, the strongest shape is an I-beam. For the swingarm, this is not practical. The rectangular section used in the prototype (Fig. 10) and earlier models approximates an I-beam in that much of its material is far from the neutral axis, but the industrial designer on the team felt that this looked crude and not sleek enough for this new product. She wanted an oval shape. Shapes that were considered are shown in Fig. 14, note that the clearance of the chain forced special attention to the interference between components. Other sub issues included:

• Verification of the strength of the swingarm. • Developing a manufacturing method to make oval tubes out of round ones or

finding a vendor who could supply them at a competitive price. A decision had to made whether to make the tubes or buy them.

• Developing a manufacturing method to blend oval shape with dropouts, the fitting that mounts to the wheel.

Figure 14: Shapes considered for the swingarm . The load in the swingarm needed to be transferred to the air shock. Bob faced the issue of designing a simple component in a compact space, that cleared the air shock and chain, transferred the loads between the swingarm and the shock, and was easy to manufacture. Two ideas were explored (Fig. 15): the "spider" and the "tube." Each of these introduced its own unique challenges in connecting the swingarm to the air shock.

Figure 15: Connections for the air shock to the swingarm. The “spider” on top and the “tube” on the bottom

Bob analyzed the strength and, working with manufacturing, the ease of manufacturing these configurations. The important points he addressed were:

• The spider configuration is essentially two opposing trusses, thus most of the structure is in tension. The shape of the truss or spider legs is not straight, as they have to clear the chain. Bob made the depth of the legs large to distribute the stress over a broad section of the swingarm to keep the stresses down.

• The tube configuration makes the stresses much more complex than the spider. The force flow is as shown in Fig. 16. The air shock puts the finger it is attached to in bending. This makes the weld 'at point A very critical because the stress is highest at this point. The fingers put the tube in both torsion and bending. Thus, Bob had to size the tube to take these loads and also design the welds to the swingarms to be of equal strength. Figure 16: Force flow in the air shock connection

Bob analyzed and tested each of the components developed here to find the modes of failure. In every case he wanted be sure that any failure would result in a safe situation. For example, even though the swingarm was designed to take the high forces generated from heavy people riding off curbs and other similar expected uses, he considered what would fail if someone rode the bike off a cliff (which someone later did). Integral to Bob's thinking were the materials to use for the components. Because BikeE and other bicycle manufacturers traditionally used 4130 steel for frames, Bob chose to use this material. However, the member of the design team that represented manufacturing wanted Bob to consider brazing rather than welding the parts together to ease the manufacturing time required. Bob spent significant time studying how the heat used in each of these processes would affect the strength in high-stress areas.

Product Evaluation Three of the Engineering Specifications on the QFD (Fig. 4) were for vertical accelerations during different riding conditions: maximum acceleration on a standard street, maximum acceleration on a 2.5 cm standard pothole, and maximum acceleration on a 5 cm standard pothole. Translating these specifications to a P-diagram, the street surface or pothole is the input signal, the maximum acceleration is the quality measure and the targets are, as given in Fig. 17.

Figure 17: P-diagram for BikeE suspension

Also shown in the P-diagram are the control and noise parameters. The design team had control over the dimensions of the swingarm (e.g., its length, the location of the air shock on it), some of the internal settings in the air shock and the recommended air pressure for the shock. What they did not have control over was:

• The actual air pressure in the shock • The weight of the rider • The temperature • The dirt buildup on the shock • The age of the shock

These parameters are all noises. The designers knew that the customer would consider the system a quality product if it met the engineering specifications and was insensitive to these noise factors. They realized that the first two “noises” can be somewhat managed through the geometry of the shock and the swingarm, but they also had to put limits on rider weight and suggested pressures in the owners’ manual. The manufacturer of the chosen shock had done a good job making the unit temperature, dirt and age insensitive. BikeE also planned to sell shock rebuild kits as an aftermarket item2. The engineers at BikeE had some simulation capability, but this was only sufficient to ensure that the performance was in the range of the targets. They felt that the best results could be found with physical hardware. Thus they built a test bike and instrumented it for measuring acceleration. They also set up a test track with 2.5- and 5- cm potholes. Tests were performed with riders of differing weights and with pressures different than those recommended. They also experimented with dirt on the shock and with heating and chilling it. Their goal was to find the best 'configuration of the parameters they could control and be insensitive to the noises. Cost The cost to make and sell a BikeE AT is broken down in Table 3. As can be seen, purchased parts are over half of the direct cost. To make all the custom parts and assemble those with the purchased parts took about 9 hours. BikeE’s margin, the amount they made on each bicycle sold to a dealer, was 29% or $171.

2 A Crane Creek shock on the author’s bike lasted 15 years with heavy use before it needed a rebuild Copyright David G. Ullman 2015 19

Table 3: Cost breakdown for a BikeE AT.

Design for Assembly BikeE engineers knew that one way to improve the margin was to make custom components easier to assemble. One of the most complex assemblies on the early Bikes was the seat frame, Fig. 18. This frame had nine different components requiring 20 separate operations to put together. These included positioning and welding operations, which took 30 minutes. They knew that there was room for improvement, thus they undertook a Design for Assembly exercise as part of the AT development. The new seat, Figure 19 has only four components requiring 8 operations and about eight minutes to assemble. The team used a template to compare the old and new seats. The form for the new seat is shown in Fig 20. The total score is 86/100 where the old seat was 70/100. The main difference between the two was in the Overall Assembly measures. For the old seat,

Figure 18: Old BikeE seat frame

the part count was poor and there was no base part for fixturing. Further, part mating was improved.

Figure 20: DFA Template for the new seat

Figure 19 The new BikeE AT seat frame

Conclusions The BikeE team developed the AT model using many best design practices. This resulted in an award winning and successful product.

Links • US Patent 5,509,678; Recumbent Bicycle • DFA Template link; http://highered.mcgraw-

hill.com/sites/0072975741/student_view0/templates.html

Author This case study was written by David G. Ullman, Emeritus Professor of Mechanical Design from Oregon State University and author of The Mechanical Engineering Process, 5th edition, McGraw Hill. He has been a designer of transportation and medical systems and holds five patents. More details on Professor Ullman can be found at www.davidullman.com.

http://highered.mcgraw-hill.com/sites/0072975741/student_view0/templates.html

http://www.mhhe.com/ullman4e

http://www.davidullman.com/

A Case Study for The Mechanical Design Process
Introduction
Background
The Design Plan
QFD Development
Conceptual Design
Concept Evaluation
Product Design
Product Evaluation
Design for Assembly
Conclusions
Links
Author