ASME HPV 2011 Report_California State University Northridge
Content
2011 ASME Human Powered Vehicle Challenge California State University, Northridge Design Report
Vehicle Number: Pending
Table of Contents Section
Page
Abstract………………………………………………………………………………………….3 Abstract ………………………………………………………………………………………….3 Introduction…………………………… Introduction ………………………………………………………… ……………………………………………………… …………………………..4 ..4 Design Description………………………………………………………………………………5 Description………………………………………………………………………………5 Analysis………………………………………………………………………………………….12 Analysis………………………………………………………………………………………….12 Testing…………………………………………………………………………………………...22 Testing …………………………………………………………………………………………...22 Safety…………………………………………………………………………………………….27 Safety …………………………………………………………………………………………….27
Cost Analysis……………………………………………………………………………………28 Analysis……………………………………………………………………………………28 References...…………………………………………………………………………………..…30 References... …………………………………………………………………………………..…30 Attachment #1 — Form Form 6 and 3-View 3-View Drawings……………………………..…………...…...31
Page 2 of 33
Abstract
A Human Powered Vehicle (HPV) is an efficient, highly engineered vehicle that runs on human muscle power. It can have everyday applications — from from commuting to work, to carrying goods to market. The American Society Societ y of Mechanical Engineers International Human Powered Vehicle Challenge (HPVC) provides an opportunity for students to demonstrate the application of sound engineering design principles in the development of sustainable and practical transportation alternatives. The mission statement of the California State University, Northridge (CSUN) Human Powered Vehicle team is as follows: To design and manufacture a Human Powered Vehicle that is practical, high performing, pe rforming, and furthers the art of HPV design. The design goals were safety, speed, weight reduction, comfort, and ease of use. The CSUN Human Powered Vehicle team has designed a tadpole recumbent vehicle. The vehicle configuration is a fully faired recumbent recumb ent tricycle, with two wheels in the front that will provide steering and one single wheel in the rear that will provide forward motion. The vehicle has been designed to be lightweight and very stable. The primary material that was used to fabricate the body of the HPV was carbon fiber. Compared to traditional materials such as, Aluminum 6061 and 302 Stainless Steel, carbon fiber is lighter, more rigid and can have hav e a yield strength that is comparable to steel when fabricated correctly. Due to its high strength strength to weight ratio, it was an ideal choice of material material to manufacture a vehicle that will be powered a human. fiber,acts which is stiff in both tensile and compressive directions, due to itsbyfiber weaveCarbon orientation, as the reinforcement to a structure when coated with epoxy epo xy resin. Aerodynamic considerations were critical in designing the vehicle. To reduce the aerodynamic drag force on the HPV, the team designed a full fairing with an integrated visor and disc wheel covers. Two different configurations will be used for the sprint event and the endurance event. In the sprint s print event, a low drag coefficient is of paramount interest; therefore, the visor will be used in the vehicle v ehicle configuration – completely completely enclosing the rider and bicycle. bic ycle. During the endurance event, the vehicle and rider are required to perform on a course with other riders and sloping terrain terrain for an extended period of time. time. Accordingly, heat removal, weight, and visibility become prime considerations. The visor will be removed for this event allowing forced convective heat transfer, visibility all around the rider, and a lowered system mass. Multiple software packages were used to design and analyze the vehicle. The SolidWorks CAD and Simulation package was the primary design and analysis tool used. It was used to create a solid model of the entire vehicle assembly and perform Finite Element Analysis (FEA) and Computational Fluid Dynamics (CFD) (CFD) analysis. In addition to the the SolidWorks Simulation package, NEiWorks was used to analyze the structural belly-pan, which was complemented with physical testing. Various physical tests were performed to ensure the structural strength and safety of the vehicle and to verify FEA. The tests that were performed were top and side load load safety tests of the roll bar, tensile testing of various layers of carbon fiber, abrasive testing and an adhesive strength test to decide on the best adhesive to bond cured carbon to cured carbon.
Page 3 of 33
Introduction
A Human Powered Vehicle (HPV) is an efficient, highly engineered vehicle that runs on human muscle power. It can have everyday applications — from from commuting to work, to carrying goods to market. Furthermore, human powered transportation is often the only type available in underdeveloped or inaccessible parts of the world. Well designed vehicles can be a valuable form of sustainable transportation. By increasing mechanical advantage, human h uman powered vehicles afford the rider the benefits of increased range and shorter travel times. The American Society of Mechanical Engineers En gineers International Human Powered Vehicle Challenge (HPVC) provides an opportunity for students to demonstrate the application of sound engineering design principles in the development of sustainable and practical transportation alternatives. The 2010-2011 California State University, Northridge (CSUN) Human Powered Vehicle team has designed a tadpole recumbent vehicle. The vehicle configuration is a fully faired recumbent tricycle, with two wheels in the front that will provide steering and one single wheel in the rear that will provide forward motion. The vehicle has been d designed esigned to be lightweight as well as very stable and can be seen in Fi Figures gures 1 and 2. Multiple software packages as well as physical testing were used to design and analyze the vehicle. The SolidWorks CAD and Simulation package was the primary design and analysis tool used. It was used to create a solid model of the entire vehicle assembly and perform Finite Element Analysis Anal ysis (FEA) and Computational Fluid Dynamics (CFD) analysis. In addition to the SolidWorks SolidWorks Simulation package, NEiWorks was used to analyze the structural belly-pan, which was complemented with physical testing.
Figure 1: Vehicle CAD Model with Fairing
Figure 2: Vehicle CAD Model without Fairing
Page 4 of 33
Design Description The mission statement of the 2010-2011 CSUN HPV H PV team is as follows: To design and manufacture a Human Powered Vehicle that is practical, high performing, and furthers the art of HPV design. The design goals were safety, speed, weight reduction, comfort, and ease of use. These criteria were used when deciding between recumbent and upright vehicle designs. Table I displays the decision matrix created to evaluate the design concepts. The design goals were assigned weights and the design alternatives were graded on a scale from 0 to 10, 10 being
excellent and 0 being failure. The recumbent 3 wheel design came up with the highest total score, 7.3 out of a possible p ossible 10. The 3 wheel tadpole design allows for greater stability for the rider. This design scores high marks for safety, comfort, and ease of use. This makes the vehicle a viable choice for personal transportation.
Table I. Overall Design Concept Decision Matrix Design Goals Safety
Speed
Weight W eight
Comfort
Ease of Use TOTAL
Weighting Factors
Design Alternatives
0.25
0.30
0.20
0.15
0.10
1
Upright
6
7
9
4
9
6.9
Recumbent (2 Wheel)
7
8
6
8
6
7.15
Recumbent (3 Wheel)
10
5
5
10
8
7.3
The primary material that was used to fabricate the body of the HPV was carbon c arbon fiber. Compared to traditional materials such as, Aluminum 6061 and 302 Stainless Steel, carbon fiber is lighter, more rigid and can have hav e a yield strength that is comparable to steel when fabricated correctly. Due to its high strength strength to weight ratio, it was an ideal choice of materi material al to manufacture a vehicle that will be powered by a human. Carbon fiber, which is stiff in both tensile and compressive directions, due to its fiber weave orientation, acts as the reinforcement to a structure when coated with epoxy resin. resin. Some of the general properties for each material considered are listed in Table II.
Table II. Material Property Comparison 6061 Aluminum
Density (lb/in3)
Tensile Strength (psi)
Yield Strength (psi)
Brinell Hardness
0.0975
45000
40000
95
Page 5 of 33
302 Stainless Steel
Density (lb/in3)
Tensile Strength (psi)
Yield Strength (psi)
Brinell Hardness
0.29
90000
40000
150
Density (lb/in3) 0.0636
Plain Weave Carbon Fiber Tensile Failure Stress Compress Failure Stress (psi) (psi)
511983
222000
Tensile Modulus (psi) 33358679
There were three main structural components of the vehicle that were fabricated: the belly pan, the frame, and the roll bar. Both the frame and roll bar were constructed by machining out a mold from pink extruded foam as shown below in Figure 3. The molds’ purpose was to create the geometry needed for each e ach part and remained within the structure of the carbon fiber.
Figure 3: Pink Foam Mold of Frame
Figure 4: Belly Pan Vacuum Bag Process
The fibers in the carbon fiber weave are oriented a specific wa way y to allow for maximum strength to occur in the direction of the fibers. Structural strength of the carbon became maximized by following this pattern of orientation. Curing techniques techniqu es for the carbon were the most intricate part of the fabricating process. Both the roll bar and frame were cured using a vacuum bagging technique. Figure 4 shows the belly pan vacuum bag process. A structural belly pan was fabricated to provide a flat mounting platform to attach components. The strength and ease of manufacturing made this design ideal. The belly pan was constructed by sandwiching an aramid honeycomb core between 8 layers of carbon fiber composites. This was one more layer than FEA recommended in order to maximize the structures overall strength. The honeycomb core increased the compressive strength of the belly pan while adding only minimal weight. The dynamic components of the vehicle were broken down into two categories: standard bicycle parts, and original manufactured components. The parts that were produced for this project were the steering system, front wheel mounts, rear dropout brackets, and drive train train path. The decision matrix that assisted in determining the chosen designs for each category is shown in Table III. Great emphasis was placed on manufacturability man ufacturability and how well parts would perform in competition settings. The design goals were assigned weights and the design alternatives were graded on a scale from 0 to 10, 10 being excellent and 0 being failure. Page 6 of 33
Table III. Manufactured Parts Decision Matrix Design Goals
Weight Design Alternatives
0.25
Ease of Manufacture
Aesthetics
Cost
Weighting Factors 0.30 0.10 0.10
Performance 0.25
TOTAL 1
Steering Wheel with Cable
6
2
9
6
6
5.1
Wheel with Shaft
3
3
9
3
8
4.85
Steering Levers
6
9
6
9
9
7.95
Cross-Pivot Levers
3
9
6
9
6
6.45
Wheel Mounting Dual Headset
6
9
8
3
6
6.8
Hub and Spindle
9
6
9
9
6
7.35
Chain Path
Through Frame Axle
3
9
9
9
6
6.75
Off-set axle
6
6
9
6
6
6.3
One Chain
6
9
3
9
3
6.15
When it came to sizing all of the manufactured parts, there were several parameters considered. All of the parts had to have a minimum factor of safety of 2. Also, many of the parts had to be designed to be compatible with other standard bic bicycle ycle parts. To simplify the assembly process, all of the bolts that assemble the manufactured parts are all the same. This allows team members to use the same tools on all of the components of the vehicle. The primary material that was used was 6061-T6 Aluminum. 6061- T6 Aluminum was chosen for its strength, low w weight, eight, price, and availability. While the axles were able to be made on a lathe, many of the custom parts had to be made on mill or CNC machine. For some of the more simple parts, p arts, a Bridgeport 2 axis Accurate Control mill was used. The complex parts that required more detail were cut on a Haas VF2, 3 axis CNC mill. The CNC parts were programmed using u sing Mastercam which converts SolidWorks part files into machine code. Most, if not all of these more complicated parts which would normally take hours to machine could be produced in less than an hour. Overall, the total machining time is extremely low considering the amount of machined parts being produced. The steering system consists of steering levers attached to a hub and spindle setup. The levers are able to rotate opposite each other and attach underneath the rider’s seat on an axle that is mounted through the frame. The levers will then attach to a push push-pull -pull mechanism that serves as the steering linkage. The movement of the push-pull mechanism is channeled to the spindle through a bell-crank located at the centerlevers of theinto vehicle. The purpose of the is to The convert the linear motion of the steering rotational movement of bell-crank the front wheels. Page 7 of 33
handlebars are made with 7/8‖ inch OD aluminum tubing; each side is bent to two 45 degree angles, the first near the linkage system which connects to the steering rods located o on n the underside of the HPV. The tubes are bent twice at 45 degrees instead of a single 90 degree bend so that the tubes will not collapse; this will allow the aluminum steering arms to be much stronger than if the tubes were to be bent with a single 90 degree bend. The steering handlebar assembly is pictured in Figure 5.
Figure 5: Steering Handlebar Assembly The front wheel hub and spindle design has a built in 10 of caster and 4 of camber in order to give the vehicle excellent handling. Figure 6 shows the front wheel mounts. The edge of the frame is used as a brace to prevent any displacement of the front wheels from the hub. The spindle is a C-shaped design that fits around the hub similar to most vehicles. The camber and caster work together to change the angle of the wheel when the vehicle is turned. These wheel mounts were designed and built ―in-house,‖ providing the ability for a custom design not available in stores. The rear dropout brackets shown in Figure 7 are another unique feature of the vehicle. Most bicycles have the rear dropouts permanently permanentl y fixed to the frame. This vehicle utilizes a two piece rear dropout design. An aluminum plug with two tapped holes on both sides of the frame remains permanently fixed. The actual dropout is a separate piece that is custom bu built ilt and bolts onto the permanent plugs.
Figure 6: Front Wheel Mounts
Figure 7: Rear Dropout
For the drive chain path, a two-chain system was the best choice to meet the design criteria. The two chains are connected with a transfer hub that mounts to an axle attached to the frame (Figure 8). This design takes advantage of the carbon fiber frame by eliminating the need for a separate structure that would be bulky bulk y and heavy. It also allows the ch chain ain to travel very close to the frame. The transfer hub consists con sists of a cassette mount designed for a regular rear bicycle wheel. of transfer mounting the fullthe 13front gear pedals cassette, there is a 15 tooth,The and15two 20 gear tooth gears that act asInstead a power between and the rear wheel. tooth Page 8 of 33
is intended for use in the sprint event. The 20-20 gear combination will p provide rovide a lower gear ratio for use during the endurance event. A 10 speed cassette is used at the rear wheel to provide a range of overall gear ratios for the most efficient power transmission from the rider to the rear wheel.
Figure 8: Transfer Hub The parts that were purchased were all parts p arts common to bicycles that are readily available from stores and would be impractical to manufacture. It was determined that the purchase of these parts would make the project more time efficient and economical. Parts that were designated for purchase include: the rear wheel, 10 speed cassette, tires, tubes, transfer hub, and the front wheels. The front and rear wheels were custom made and laced by hand. Air resistance can can approach 90% of the total retarding retarding force on a bicycle. Consequently, aerodynamic considerations were critical in designing the vehicle, particularly for the sprint event. The power required to overcome aerodynamic drag increases with velocity cubed, limiting a rider’s rider’s top speed. Three principal ways of decreasing decreasing wind resistance applied to the design of the aerodynamic include: decreasing the frontal area, streamlining bicycle components, and smoothing the surfaces of the fairing, roll bar, and rider. To reduce the aerodynamic drag force on the HPV, the team designed a full fairing with an integrated visor and disc wheel covers. Design alternatives were evaluated evaluated by means of a design matrix given below in Table IV. The design goals were assigned weights and the design alternatives were graded on a scale from 0 to 10, 10 being excellent and 0 being failure. Table IV. Aerodynamics Design Concept Decision Matrix Design Goals Frontal Area
Low Drag Coefficient
Heat Dissipation
Rigidity/Weight
Cost
Ease of Manufacture
Weighting Factors
TOTAL
Design Alternatives
0.35
0.20
0.20
0.15
0.05
Exposed rider, roll bar, & enclosed rear wheel
10
6
10
10
10
6
4
10
6
10
10
10
8
Enclosed front wheels/Exposed rider & roll bar Enclosed rider & Roll bar/Exposed disc wheels, event specific
0.05
1
10
8.7
8
8
6.2
6
6
9
Page 9 of 33
With a score of 9 out of a possible 10, the best choice was det determined ermined to be an event specific design. The sprint configuration shown below in Figure 9 (left) completely encloses the rider and major vehicle components. This configuration will lead to an optimal drag coefficient. coefficient. The endurance configuration shown in Figure 9 (right) will not incorporate the canopy, leading to a lowered system weight, increased ventilation and visibility.
Figure 9: Sprint configuration (left) and Endurance configuration without canopy (right)
For correctly streamlined fairing geometries, pressure drag is minimized. In order to minimize boundary layer separation over the surface, a long profile of 103 inches was modeled using the NACA 0015 profile generated in Excel and imported as guide curves within the SolidWorks environment (Figure 10). Using curvature combs, the computed airfoil data was modified to meet rider geometry and vehicle dimensions. dime nsions. Particularly, the widest point moved back towards the rear of the vehicle. Figure 10: Surface geometries governing frontal area reduction
To facilitate the fabrication of the fairing skin, visor and stiffeners, a male plug was fabricated out of a 2ft x 4 ft x 8 ft solid block of 1 lb density expanded polystyrene pol ystyrene foam (EPS). An initial rough cut, shown in Figure 11, formed by using a hot wire 3-axis CNC machine, was followed by hand carving to produce the streamlined shape of the fairing. Final symmetry was checked using cross sectional area templates. Areas of concern, in particular the pedal box, box , shoulder, and head clearance were checked for conformance. Surface filling followed by surface sealing stages were then applied to the foam sculpture. A joint compound was applied to the entire surface to fill in all gaps and surface voids then sanded as shown in Figure 12. Two coats of a water based primer (StyroPrime) followed by several coats (3-4) of a liquid plastic (StyroSpray) supplied by Industrial Polymers were applied to the foam surface. This barrier allowed for the application of a polyprimer p olyprimer (PLC). The entire surface was then sanded, cleaned, and sprayed with an epoxy resin. A mold release was tthen hen applied (5 layers) in preparation preparation for a wet layup. Page 10 of 33
Figure 11: Rough cut foam block shown with ¼ inch routed carving templates (left), hand carving to final shape (right)
Figure 12: Filling in all gaps and surface voids (left), surface sealing stage (right)
The wet layup consisted primarily of an inner carbon fiber layer and an outer ou ter hybrid carbon/Kevlar layer. Once the initial shell had cured, additional hybrid fiber rei reinforcement nforcement layers were bonded into the tub and canopy components at specified locations, near the rider, interfacing with the road during rollover. A framing rib structure layup consisting of tw two o layers of carbon fiber and Lantor Soric XF foam was fabricated. The layup components were chosen to increase stiffness and reduce overall weight, due to t o lower resin retention, when compared with equivalent all fiber lay-ups. Following alignment of the ttub ub and canopy skins, the ribbing structure was bonded to the inner surfaces of each. A female fiberglass fiberglass tool was fabricated to drape form lexan for a canopy visor. The tool was build with a shoulder to allow the lexan to be formed with a corresponding shoulder to ensure a flush mate between between the visor and canopy skin when bonded. Aluminum sheet metal brackets were fabricated and bonded to the inner surface of the tub at specified locations to mount the tub to the frame. Page 11 of 33
Analysis
The rollover/side protection component of the design (Roll bar) consists of a single, continuous carbon fiber feature which sweeps around the rider’s seat and h ead (Figure 13). The roll bar is rigidly and directly attached to the main body of the vehicle (frame-belly pan) on the mounting faces. The integrated design of the belly pan, fframe rame and roll bar has pr provided ovided a single solid and strong base structure for mounting the rest re st of the components such as wheels, crank, seat, etc. Since the strength of the roll bar is of paramount p aramount importance in any roll over scenario, extensive finite element analyses (FEA) have been conducted to ensure the integrity in tegrity of the structure in these situations. Using NEiWorks NEiWorks composite capabilities and through an iterative design optimization process, different carbon fiber layups have been examined and optimized to achieve minimum weight, minimum material waste and maximum strength based on ASME requirements. First, in a series of static analyses the roll bar has been subjected to 800 lb top (12 degrees from vertical) and 500 lb side loads with the objective of finding the optimal layup as stated above. With the roll bar held fixed on the mounting faces, the experimentally obtained material properties, the Hill composite failure criterion criterion and the allowable bond shear stress of 1000 psi (determined by by numerous testing samples) were used to define the boundary bo undary conditions and computational constraints of the analyses. anal yses.
Mountin Mou ntin faces faces
Figure 13: Integrated components of the vehicle including the roll bar, frame, and belly pan
Figure 14 presents the optimized carbon fiber layup la yup obtained with the static analyses. This optimized layup (left in Figure 14) is an 8 layers carbon fiber sandwich consisting of uni and plain weave fibers with a horizontal wrap direction (right in Figure 14) following the shown stack up layup.
Wrap direction
Figure 14: The optimized carbo carbon n fiber layup of the roll bar (left) along with the wrap direction (right)
Page 12 of 33
The displacement and composite status of the optimized o ptimized model for the two load scenarios are shown in Figures 15 and 16. The maximum displacements obtained with the optimal layup were 0.96 inch and 0.87 inch for the side and top loads, respectively. Furthermore, the composite failure indexes reported in both cases indicated the healthy status of the structure. The static analyses have then been extended to dynamic (impact) analyses for further examinations in order to account for a more realistic roll over scenario.
Figure 15: 800 lb top load displacement
Figure 16: 500 lb side load displacement
Since it is expected that in the case of a roll over the roll bar would initially come in contact with the ground on its sides, a side collision has been considered for the impact analysis. To simulate this side impact in the software (Figure 18) the roll bar has been carefully shot with initial velocity of 25 miles/hr in a straight line trajectory from a distance of 1 ft to hit a plane which is simulating the ground. Upon examining Figure 17 and Figure 18 it can be seen that the obtained optimal layup has brought about not only a healthy structure (failure indices less than one) but also an acceptable (ASME rule) side deflection in case of the real impact. The magnitude of this deflection at the roll bar’s shoulder (denoted by node 628) was 0.93 inch.
Figure 17: The Roll bar side impact displacement results of node 628
Figure 18: The roll bar side impact composite max failure
Page 13 of 33
The frame-belly pan is an assembly of two components which are manufactured separately but are bound together to form the base skeleton of the vehicle. The frame itself is holding the crank housing on the front while supporting the back wh wheels eels directly on its rear handles. The belly pan on the other hand serves as a mounting structure for the front wheels while acting as a stress reliever for the frame. The geometry (Figure 19) has been designed as narrow as possible to decrease the surface area and reduce excessive flexing in X and Y directions. Following a roll bar-like approach, an efficient carbon fiber layup was developed through a system of static and dynamic d ynamic analyses while taking into account the stress-generated Back Wheel Supports failures that may occur in the critical segments of the two components. Following this procedure the analyses were resolved into crank housing, structure flexing and wheel supports.
Seat Location Crank Housing
The crank housing, which involves the frame only (Figure 20), has been manufactured by inserting the crank insert (an Aluminum tube) into the pre-cut opening in front of the frame which has then been wrapped and positioned with multiple layers of carbon fiber. This analysis was concerned with determining the sufficient amount carbon fiber layers required to secure the crank in place and avoid any failure. Figure 19: Frame-belly pan assembly with w ith critical segments
The failures may occur due to the coupled loads that are exerted during the pedaling. Following the EN 14766 (European Mountain Bicycles Committee) recommendations, the crank housing should withstand a 400 lb couple load (Figure 20). The frame boundaries were fixed sufficiently far from the crank housing to provide a valid solution in the region of interest. After several iterations, the analyses have determined that 8 layers of plain carbon fiber with a layup schedule of (0/+45/0/0/0/+45/0/0) would be capable of capturing stress disturbances in the crank housing. The simulation results of the segmental body of the frame Figure 20: The crank housing (Figure 20) showing total displacement and composite failure indexes are presented in Figures 21 and 22. They were determined by implementing the optimized layup stated above. From Figure 21, it is seen that the total displacement reported is 0.005 inch illustrating the strength and rigidity of the crank housing. As for the composite failure index result, it also indicates a healthy status for the crank housing.
Page 14 of 33
Figure 21: The displacement results of the crank housing
Figure 22: The composite failure index result of the crank
Once the optimal layup for the crank housing was determined, the same layup was applied to the entire frame to quantify quan tify the amount of flexing present in the X and Y directions. To avoid boundary condition related inaccuracies, the belly pan was rigidly connected to the frame, constituting a single continues model. Consequently Consequentl y the analysis has also aimed to define de fine an optimal layup for the belly pan to help the frame in increasing the overall vehicle rigidity. The structural flexing can occur either due to the static loading of the rider’s weight or the d dynamic ynamic loading of a rough surface contact (such as a bump), along with the cyclic pedaling loads at the cranks. To simulate the effect of these three loads simultaneously while reducing the t he computational resources, a procedure presented in Figures 23a-23e has been followed to efficiently replace the complexity of a dynamic analysis an alysis with a conservative static analysis. To initiate this approximation, a separate static analysis (including only the rider’s weight) was performed. Next using the resulting displacements, an equivalent stiffness (Keq) has been calculated for the frame and belly bell y pan combined. At this point the frame-belly pan assembly has been mathematically modeled by a single mass at the center of gravity attached to the equivalent spring (Figure 23a). Using this simplified model, an arbitrary load (representing the surface bump) has been approximated with a conservative static load (P0) as shown in Figure 23b. The value of this conservative static load was determined based on a scenario where the vehicle experiences a 3 inch high bump. It can be seen from Figure 23c that the spring will be displaced by less than 3 inches at the top of the bump. Assuming the worst case scenario of the full 3 inches displacement of the spring, P 0 was determined (using the Keq and displacement) to be 1.5xWrider where Wrider is the average rider’s weight. Now that a single known kno wn static equivalent load is determined to approximate the dynamic loads of the road the frequency spectra chart (Figure 23d) was used to take into account the impulse (short period) nature of this load load.. It can be seen from the chart (shown in red) that this conversion can be done by multiplying the P0 by the magnification factor of two. Collectively it could be said that: P0dyn = 3x Wrider = 700 lb. Applying the obtained force at the CG and accompanying that with the 400 lb pedaling force (as discussed in the crank housing analysis) anal ysis) a full loading scenario of the structure was achieved. In addition the structure was fixed at the wheel locations locat ions (front and back) as shown in Figure 23e.
Page 15 of 33
Figure 23b Figure 23a
Figure 23e Figure 23c
Figure 23d
Figures 23a through 23e show the process of defining the dynamic loading of the belly pan p an and frame
The displacement and composite status results obtained from these analyses showed that the same layup schedule introduced in the crank housing could be extended to the entire body of the frame. Furthermore it was that the optimal layup for the in includes cludes a 0.75 inch Nomex honey comb core determined which is sandwiched between 7 layers of belly plainpan weave carbon fiber. Figures 24 and 25 summarize the belly pan and frame final layup schedule along with their zero angle (wrapping) directions.
Figure 24: The layup sche schedule dule (left) and the wrap dire direction ction (right) of the belly pan
Figure 25: The layup schedule (left) and the wrap direction (right) of the frame
Page 16 of 33
The simulation predictions of the deflection of the frame-belly pan assembly in Y (up and down) and X (side to side) directions performed with the optimized layups are presented in Figure 26 and Figure 27. From the displacement results in Y and X (0.09 inch and 0.10 inch, respectively) it is seen that the profile flexing of the structure is nearly negligible. To further investigate the results obtained for the frame optimized layup, a separate simulation was employed on the end segment of the frame whose rear supports are holding the back wheel. Unlike the previous analysis (see Figure 23e) where the rear supports were held fixed, this time they were loaded directly by b y two vertical forces
Figure 26: The displacement displacement results of the structure in Y direction
(same on each whose magnitudes were determined by support) taking into account the reaction forces produced by P0dyn (see Figure 23e). It has been determined that approximately 2/3 of this load (400 lb) is transferred to the rear supports (Figure 28). By changing the boundary conditions to include the roll bar in the analysis, and fixing the structure far from the rear supports, each component was assigned with its previously obtained optimal layup. The first round of analysis indicated a small de-lamination at the area near the end of the right support based on the reported composite failure indices (Figure 29). Figure 27: The displacement results of the structure in X direction
Figure 28: Shows the rear support FEA model along with 200 lb loads on each support
Figure 29: Shows the small de-lamination at the right handle
Page 17 of 33
To solve the problem the original frame layup la yup was locally improved via adding two extra layers of carbon fiber. The subsequent analysis an alysis presented in Figure 30 confirmed the effect of the imposed correction.
Figure 30: Shows the healthy status of the handles after addition of extra layers
Page 18 of 33
Using SolidWorks flow simulation to perform flow analysis on the solid model of the fairing, the design team optimized the fairing geometry. The flow simulations have aimed at predicting the fairing (and eventually the full model) drag coefficient, defined as:
Where F is the fairing drag force, A is the fairing frontal area; ρ is the air density and V is the vehicle velocity.
To set up these simulations, an incoming uniform air flow of 40 mile/hr, temperature of 59F, and turbulence intensity of 1% have been specified at the boundaries of the computational domain. This computational domain have further been equipped with robust mesh refinement and extended end tail (Figure 31) to increase the accuracy of the results. Through an iterative design optimization process the fairing geometries have been improved at each step with the objective of achieving a streamlined (low pressure drag) shape. Figure 32 and 33 present the 6th and 8th (final) iterations along with their flow patterns and pressure fields. The rise- drop- rise of the pressure fields at the nose-shoulder -tail of the models are in full accordance with the theoretical expectations of a streamlined shape. However the steady stream of the flow over and past the th e final design has formed a more moderate nose-tail nos e-tail pressure gradient resulting in a lower drag coefficient of 0.06. This pressure gradient is shown in detail with a pressure profile graph extending from the tip to the back b ack of the fairing (Figure 34). It is seen that the nose-tail pressure drop governing the fluid flow for the final fairing becomes nearly recovered (0.014 psi), reducing the pressure drag to a great extend. ex tend. In fact an approximate 57% improvement in drag coefficient from iteration 6 to the final iteration 8 was realized as tabulated below in Table V.
Figure 31: The computational extended region and mesh refinement
Figure 32: 6th iteration along with its flow trajectory and pressure field
Figure 33: 8th iteration (final) along with its flow trajectory and pressure field
Figure 34: The pressure gradient profile of final fairing
Page 19 of 33
Table V. Aerodynamic Drag Data Configuration
Drag Coefficient -Cd
Drag Force (lb)
2
Frontal Area-AF (in )
Analysis of Fairing ONLY
Iteration 6Sprint Configuration Iteration 7Sprint Configuration Iteration 8Sprint Configuration
0.094
2.24
841.35
0.064
1.72
945.69
0.06
1.44
906.73
Analysis of Full Model
Iteration 8Sprint Configuration (Full Model) Iteration 8Endurance Configuration (Full Model)
0.073
1.44
1035.9
0.209
5.64
948.36
Finally to achieve the true magnitude ma gnitude of the drag coefficient for the final fairing, a full model simulation was considered by including all the exposed components of the v vehicle. ehicle. Results for the full model analysis are tabulated in Table V. From sprint configuration in Figure 35 (left), it can be seen that the frontal area has increased to 103 1035.9 5.9 in2 and the drag coefficient has also increased to 0.073. These values are still smaller when compared to the other shapes considered. Figure 35 (right) shows the endurance configuration which has a drag coefficient of 0.209 and a frontal area of 948.36 in 2.
Figure 35: Flow trajectory and pressure field for sprint configuration (left) and endurance configuration (right)
Page 20 of 33
To substantiate the effectiveness of the fairing design, a rough evaluation of the expected performance at constant velocity, no winds, level ground, a conservative rolling resistance coefficient of 0.008, overall drag coefficient coe fficient of 0.073, negligible transmission loss, and typical system mass of 88 kg (based on expected vehicle mass of 27 kg and rider mass of 61 k kg) g) was input to the Power equation given below.
frontal area, (m²) : projected frontal : aerodynamic drag coefficient : rolling resistance, resistance, typically .003 < < .006 at 10 m/s G: grade, percent divided by 100 (zero for this case) M: total system system mass (kg) (W) P: cyclist power, (W) t: time, (s) V: velocity, (m/s) : wind velocity, head or tail winds (zero (zero for this case), (m/s)
: acceleration of the vehicle, (zero for this case)
W: weight of the system, cyclist and machine, (N) : air density, (kg/m³) at standard conditions
Based on the idealized model, it can be noted that for a stipulated top speed goal of 40 mph, the power input required is 376 3 76 Watts using the fully enclosed sprint fairing configuration. Under the same parameter values, using the endurance fairing configuration, not including pit stops, the team can expect an average velocity of 25 mph corresponding to 62 miles of distance covered for the 2.5 hour duration of the endurance event based on the measured average rider power input of 206 watts. To maximize performance during the endurance event, research has shown that the rider must release three units of heat for every unit un it of power input to the pedals. With an expected power input of 200 watts, 600 watts of heat must be removed to avoid rider discomfort and significant reduction in efficiency. Due to the rider ’s head being exposed to air flow, it was calculated that 100 watts may be removed by convective heat transfer. The calculated radiative heat loss through the fairing was 100 watts. To facilitate heat loss due to sweat evaporation evapo ration and natural convection over the remainder of the rider ’s body, a NACA duct was fabricated and installed in front of the visor.
Refer to page 28 for a detailed det ailed cost analysis of the vehicle.
Page 21 of 33
Testing
Various physical tests were performed to ensure the structural strength and safety of the vehicle and to verify FEA. The tests that were performed were top and side load load safety tests of the roll bar, tensile testing of various layers of carbon fiber, abrasive testing and an adhesive strength test to decide on the best adhesive to bond cured carbon to cured carbon. The roll bar test was performed to ensure that tha t the impact of a possible crash would not endanger the safety of the rider. rider. The structural strength of the roll bar, specified specified by ASME, should have a top load requirement that can with withstand a 600 lb load at an angle of 12 degrees with respect to the vertical axis and an d a deflection of less than 2.0 inches. The test setup, shown in Figure 36, mounted the roll bar to a metal fr frame ame with heavy duty straps. The top of the roll bar was pulled at an angle downward of 12 degrees from the vertical by a hois hoist. t. The hoist was attached to a scale at the other end, which digitally read read the tensile force in pounds. At a force of 670 lb with a minimal and non-measurable deflection, the top load test passed the test requirement. When compared to FEA, the deflection was less than the predicted val value ue of 0.87 inches at an 800 lb load.
Figure 36: Top load roll bar test setup
The side load test, shown in Figure 37 and Figure 38, was setup by applying heavy duty straps and clamps to the bottom of the roll bar to restrain it from from movement. A second set of straps were tied to to the side of the roll roll bar, at shoulder height, and aattached ttached to a hoist. The other end of the hoist was attached to a scale which read the tensile force as the roll bar was pulled horizontally to the side. side. The ASME rules specify that the the roll bar should withstand withstand a minimum force of 300 lb to the side of the roll bar at shoulder height with a deflection less than 1.5 inches. The side load test passed its specification requirement with a load of 306 lb and a deflection of 1.1 inches. However, this deflection was slightly higher than the FEA prediction of 0.96 inches.
Page 22 of 33
Figure 37: Side Roll Bar Setup
Figure 38: Side Roll Bar Result
Tensile tests were performed on various samples of carbon fiber, each with a specific number of layers. This testing was conducted to verify the material strength properties of the carbon fiber used in FEA. Each sample was layered at a zero and forty-five degree fiber orientation mimic theconsisted same process performed when fabricating eknown ach component of the vehicle. Thetotest setup of creating carbon samples with a each neck length and clamped into a tensile machine fixture as shown in Figure 39 and Figure 40. The results and data for each sample are outlined in Table VI below.
Figure 39: Tensile Test Setup
Figure 40: Tensile Samples after Test
Page 23 of 33
Table VI. Tensile Testing Final Results Tensile Testing # of Carbon Layers
Gauge Length (in)
Peak Load (lb)
Width (in)
Thickness (in)
Area 2 (in )
Stress (psi)
1
1.5
257.11
1
0.012
0.012
21425.83
3
2.75
676.04
1
0.035
0.035
19315.43
4 6 8
3 5.75 6
1604.84 2499.61 3499.45
1.25 1.25 1.25
0.048 0.053 0.070
0.060 0.066 0.088
26747.33 37729.96 39993.71
It was shown that as the number of layers of carbon increase, the peak stress prior to failure increases as expected. At 8 layers, as shown in Table VI, there is sufficient strength to handle an ultimate yield stress of 39,993.71 psi. The stress values found matched closel closely y with the material properties used for the carbon fiber for FEA. Abrasion testing was performed to experimentally determine the best possible fairing material to resist resist abrasion in case of a vehicle roll over. Two procedures were performed in order to experimentally determine the best possible material. In the first procedure, a special fixture, shown in Figure 41, was built and four test samples were attached to t o the apparatus shown in Figure 42.
Figure 41: Abrasion Fixture
Figure 42: Abrasion Test Setup
The test samples were subjected to abrasion by b y dragging the test sample for 1 100 00 ft from rest against asphalt, in order to determine the change chan ge in mass after abrasion. The material with the lowest change in mass after this abrasive process proce ss was considered best. The second procedure pro cedure was called a skid test, consisting of replicating an initial velocity of 20 mph, with the fixture shown in Figure 41. The samples were released with this initial velocity and the distance they the y traveled before coming to a complete complete stop was measured. The material with the lowest stopping distance was deemed as being best b est because the sliding distance after a rollover would be minimized. Table VII shows the results from from the abrasion test. It can be seen that the carbon/Kevlar samples had the lowest change in mass of 0.0519 g, 0.0634 g, and 0.0486 g, respectively, before and after the experiment.
Page 24 of 33
Table VII. Abrasion Test Results--Dragged for 100 ft from rest at 20 mph Test #
Material/Configuration
masso (g)
massf (g)
Δm Δm (g)
l (in)
w (in)
Area 2 (in )
1
Plain weave Carbon Fiber 0o
40.9651
40.8523
0.1128
2.5
2.5
6.25
2
Plain weave Carbon Fiber 0o
41.6812
41.5924
0.0888
2.5
2.5
6.25
o
3 4
Plain weave Carbon Fiber 45 Plain weave Carbon Fiber 45o
39.4177 41.2423
39.3147 41.1422
0.103 0.1001
2.5 2.5
2.5 2.5
6.25 6.25
5
Kevlar plain weave #549
39.986
39.9057
0.0803
2.5
2.5
6.25
6
Kevlar plain weave #549
40.4099
40.3357
0.0742
2.5
2.5
6.25
7
Kevlar plain weave #2469 (smaller fibers)
40.3965
40.326
0.0705
2.5
2.5
6.25
8
Kevlar plain weave #2469 (smaller fibers)
41.6584
41.544
0.1144
2.5
2.5
6.25
9
Kevlar Twill weave
40.1187
40.0434
0.0753
2.5
2.5
6.25
10
Kevlar Twill weave
40.5889
40.5183
0.0706
2.5
2.5
6.25
11
Carbon/Kevlar hybrid (Kevlar direction of travel)
40.5306
40.4469
0.0837 0.0837
2.5
2.5
6.25
12
Carbon/Kevlar hybrid (Kevlar direction of travel)
42.0082
41.9596
0.0486 0.0486
2.5
2.5
6.25
13
Carbon/Kevlar hybrid (Carbon in direction of travel)
40.3251
40.2732
0.0519
2.5
2.5
6.25
14
Carbon/Kevlar hybrid (Carbon in direction of travel)
42.3793
42.3159
0.0634
2.5
2.5
6.25
Table VIII shows the results from from the skid test. It was shown tthat hat the plain weave carbon orientated at 45 degrees stopped with the shortest distance of 15.25 ft. However, the carbon/Kevlar samples also performed well on this test. Based on the results of both tests, the carbon/Kevlar was chosen for the fairing. Table VIII. Skid Test Results--Initial velocity from 20 mph to a complete stop Material/Configuration Material/Configura tion
Stopping Distance (ft)
Plain weave weave Carbon Fiber 0o
18.917
Plain weave weave Carbon Fiber 45o
15.250
Kevlar plain weave #549 Kevlar plain weave #2469 (smaller fibers)
17.250 16.208
Kevlar Twill weave
17.875
Carbon/Kevlar hybrid (Kevlar in direction of travel)
16.167
Carbon/Kevlar hybrid (Carbon in direction of travel)
17.667
Adhesive testing was performed to experimentally determine the best adhesive to use for bonding cured carbon to cured carbon. This was performed during fabrication when bonding the roll bar to the frame frame and the frame to the belly pan. The ideal material needed to be able to withstand maximum force without without experiencing shear failure. A total of nine test samples, shown in Figure 43, Figure 44, and Figure 45, of carbon fiber were created and bonded with three different types of adhesives: epoxy resin, Hysol, and 3M DP-420. Page 25 of 33
Figure 43: Epoxy Sample
Figure 44: Hysol Sample
Figure 45: DP-420 Sample
A tensile testing machine was used to accurately determine d etermine the peak load prior to shear failure. The ultimate stress before failure for each sample was determined by b y using the peak load and dividing it by the measured bond cure area. A total of three samples were used to find the average peak load per adhesive. The average peak stress for DP-420, epoxy resin, and Hysol were calculated to be 1852 psi, 1376 psi, and 2583 psi, respectively. These results are tabulated in Table IX. Ultimately, Hysol was chosen due to its high stress tolerance.
Table IX. Adhesive Test Results 2
Sample
Run
Length (in)
Width (in)
Area (in )
Peak Load (lb)
Peak Stress (psi)
DP-420
1 2 3
0.562 0.573 0.530
0.598 0.563 0.649
0.336 0.323 0.344
436 652.3 770
1297 2022 2238
Average
619.333
1852
Std Dev
169.275
493
DP-420 DP-421
Epoxy
1
0.507
0.592
0.300
544
1812
Epoxy
2 3
0.534 0.528
0.576 0.580
0.308 0.306
310 463.78
1008 1307
Average
439.26
1376
Std Dev
118.911
407
Epoxy
Hysol
1
0.708
0.594
0.421
1006
2392
Hysol
2
0.665
0.579
0.385
1108.475
2879
Hysol
3
0.678
0.582
0.395
978.006
2478
Average
1057.188
2583
Std Dev
72.531
260
Page 26 of 33
Safety
A recumbent vehicle design provides many man y safety benefits to the rider. Greater safety is possible because of the near impossibility of taking a ―header‖ over the front wheel or of catching a foot or pedal on o n the ground when cornering. There is far greater comfort in an almost complete absence of pain or trauma in the rider’s hands and wrists, or back and neck, or crotch. This design also allows better visibility forward and to the side for the rider compared with that For purposes, for a diamond-framed with dropped canopy was designed tobicycle give the rider at leasthandlebars. 180 deg degrees rees ofsafety visibility duringthe thevehicle’s sprint event. During the endurance event, the canopy will be removed, thus giving the rider an even greater range of visibility.
As the case with any vehicle, certain precautions need to be taken tak en to ensure that all of the riders of the recumbent vehicle and other othe r competitors are safe. First and foremost is having a seatbelt since the vehicle requires the rider to lie in a reclining position. The design incorporates a simple lap belt that will wrap around the frame and secure the ride com comfortably fortably within the seat. Brakes are required on any vehicle to prevent any unwanted contact with surrounding vehicles or obstructions. The vehicle incorporates two different brake systems. The front brakes are cable driven and are to be used by the rider while riding. The rear brake will be used as a parking brake that will assist the rider while getting getting in and out of the vehicle. Recumbent vehicles can be difficult to get in and out of and th this is will make driver changes safer, quicker, and more efficient. All nuts and bolts will be checked prior to each event in order to maintain their security. Safety wire and Loctite are two of the most common fastener security techniques and will be utilized to ensure that no mechanical parts become loose. Since many of the vehicle parts will be custom made out of aluminum, it is important that there are no sharp ed edges ges or loose chips. Each machined part has been cleaned c leaned up using files and deburring tools. S Similarly, imilarly, all open end tubes are capped using either plastic plugs or Delrin plugs. Certain areas with a high p potential otential for pinching, such as the chain and gears, must be evaluated for safety. Most of the pinch points are behind or below the rider so they are of minimal concern. The key area is the chain that runs down between the rider’s legs. A chain guard is incorporated to prevent any leg contact.
Page 27 of 33
Cost Analysis
This section provides a summary of cost analysis for this project. In order to t o properly illustrate the overall cost of a single quantity production of a human powered vehicle, each and every component of the vehicle must be properly tabulated to have a detailed cost analysis. Mass production costs for this vehicle will also be included in this section since this method method is more cost efficient. The tables below provide the division of labor cost, material cost, and ov overhead erhead equipment cost. The cost for the utilities u tilities and production facility was not included in this analysis since this varies with location.
Table X. Labor Cost Estimate Quantity
Hourly Rate
Hours per Week
Design Engineer
3
$28.00
40
$3,360.00
$13,440.00
Project Supervisor
1
$36.00
40
$1,440.00
$5,760.00
Carbon Fiber Team
3
$15.00
40
$1,800.00
$7,200.00
Machinist
1
$20.00
40
$800.00
$3,200.00
Welder Assembly Team
1 3
$18.00 $15.00
40 40
$720.00 $1,800.00
$2,880.00 $7,200.00
$9,920.00
$39,680.00
Labor
Cost per Week
Total Cost
Cost per Month
Table XI. Material Cost Estimate Individual Vehicle Cost
Materials
Quantity
Unit Cost
Total
10 Vehicles Cost Quantity
Total
Carbon Fiber (1/32'')
14
$65.00
$910.00
140
$9,100.00
G-10
1
$85.00
$85.00
10
$850.00
Fairing Mold Foam Aluminum
1 3
$350.00 $700.00
$350.00 $2,100.00
1 30
$350.00 $21,000.00
Composite Supplies
1
$120.00
$120.00
10
$1,200.00
Epoxy Resin
1
$95.00
$95.00
10
$950.00
Adhesive
1
$80.00
$80.00
10
$800.00
Nomex Honeycomb
1
$650.00
$650.00
10
$6,500.00
Subsystem Components
1
$1,200.00
$1,200.00
10
$12,000.00
Total Cost
$5,590.00
$52,750.00
Page 28 of 33
Table XII. Overhead Equipment Cost Manufacturing Equipment
Quantity
Cost
Total
TL-2 CNC Lathe
1
$27,000.00
$27,000.00
Vacuum Bagging Pump
1
$470.00
$470.00
Computer/Software
3
$1,000.00
$3,000.00
Welding Machine
1
$3,200.00
$3,200.00
CNC Milling Machine
1
$22,000.00
$22,000.00
Tools
1
$1,500.00
$1,500.00
Total Cost
$57,170.00
Table XIII. Cost Analysis Summary Cost of Single Vehicle
6 Year Production-10 Vehicles Per Month
Materials
$5,590.00
$4,024,800.00
Labor
$9,920.00
$7,142,400.00
$57,170.00 $72,680.00
$57,170.00 $11,224,370.00
$72,680.00
$15,589.40
Overhead Equipment Total Cost Total Cost Per Vehicle
As Table XIII shows, the cost per vehicle is greatly g reatly reduced as the number of vehicles being produced increases. This is a direct result of the high cost of overhead equipment. For a production run of 6 years, assuming 10 vehicles a month are produced, the total cost per vehicle is $15,589.40. However, if only one vehicle is produced, the cost is $72,680.00. It should be noted that the vehicle presented at competition does not cost nearly this much. All overhead equipment was provided and the students were not paid for their labor. Therefore, the only costs incurred were the material costs.
Page 29 of 33
References
"ASME - Human Powered Vehicle Challenge (HPVC)." ASME - Home. Web. <http://www.asme.org/events/competitions/human-powered-vehicle-challenge-(hpvc)>. ―Cycles—Published Standards.‖ European Standards.‖ European Committee for Standardization. 2005. Web.
<http://www.cen.eu/cen/Sectors/TechnicalCommitteesWorkshops/CENTechnicalCommit tees/Pages/Standards.aspx?param=6314&title=Cycles>. Kyle, M. W. "Aerodynamics of Human-Powered Vehicles." Proceedings of the Institution of Mechanical Engineer. Part A, Journal of Power and Energy (2004): 141-54. Print.
Ryan, R. "Fluid Dynamics." Dec. 2011. Lecture. Salary.com - Salary Information, Job Search, Education Opportunities and Career Advice.
Web. <http://www.salary.com/>. Shields, D. "Sculptor/Propmaker Teaching Session." Personal interview. 31 Dec. 2011. Wilson, David Gordon, Jim Papadopoulos, and Frank Rowland. Whitt. Bicycling Science. Cambridge, MA: MIT, 2004. Print
Page 30 of 33
2011 Human Powered Vehicle Challenge West Sponsored by ASME and Montana State University Form 6: Vehicle Description Description Due April 11, 2011 (Dimensions in inches, pounds) Competition Location:
Montana State University
School name:
California State University, Northridge
Vehicle name: Vehicle number :
Bicycle Engineering @ Northridge (BE@N) Unknown
Vehicle type
Unrestricted
Speed__X_____
Vehicle configuration Upright
Semi-recumbent
Prone
Other (specify)
Frame material
Carbon fiber composite
Fairing material(s)
Carbon fiber with Kevlar fabric sandwiched
Number of wheels
3
X
Vehicle Dimensions Length
103.08 in
Width
Height
53.28 in
Wheelbase 37.56 in
Weight Distribution
Front
60%
Wheel Size
Front
20 in
Frontal area
Sprint Configuration: 1035.9 in and Endurance Configuration: 948.36 in
Steering Braking
40%
Total
100%
Rear
27.5 in
2
Front
_X
Front
Estimated Cd
Rear
37.56 in
2
Rear Rear
Both
__X
0.073
Vehicle history (e.g., has it competed before? where? when?) This vehicle is a clean sheet design and has not competed in any event.
Page 31 of 33
Attachment 1 — F Figure igure 1: 3-View Drawing Without Fairing (dimensions shown in feet)
Page 32 of 33
Attachment 1 — Figure Figure 2: 3-View Drawing With Fairing (dimensions shown in feet)
Page 33 of 33
