Synchronous Charge Trapping Final Report

Transcription

1 Synchronous Charge Trapping Final Report To: Karen Den Braven Prepared by: Team Short Circuit Alex Fuhrman Andrew Hooper Ty Lord Cole Bode Team HEV shop phone: (208) Date: May 9,

2 Short Circuit University of Idaho College of Engineering Moscow, ID May 9, 2010 NIATT University of Idaho 115 Engineering Physics Building Moscow, ID Attention: Dr. Karen Den Braven Director of NIATT, U of I CSC Team Subject: Two Stoke Exhaust Gas Trapping Interim Report We are submitting a final report for the synchronous charge trapping (SCT) engine designed for the University of Idaho s Rotax direct-injected two stoke engine. The final report attached contains our final synchronous charge trapping engine concept designed to reduce exhaust emissions and decrease brake specific fuel consumption. This report presents the final concept design and the process used to decide upon this design. The final concept has been stress analyzed in all areas where the previous synchronous charge trapping design failed and others to prevent all perceived modes of failure. Additional information on the project can be found on our team website: If you have any questions, concerns or suggestions please contact us. Thank you for your advice and input on our design and the opportunity it has presented. Respectfully, Short Circuit 2

3 Table of Contents Table of Contents... 3 Figures... 4 Tables... 4 Executive Summary Introduction Background Objective Concepts Considered Redesign Peter Britanyak s System Rotary Disk System Parallel Rotary Valve System Concept Selection Peter Britanyak s System Rotary Disk System Parallel Rotary Valve System Selected Design Stock Cylinder Modifications Valves Valve Shaft Pulley System Electronic Controls Product Manufacturing Stock Cylinder Modifications Valves and Counterbalances Pulley System Product Evaluation Specifications Plan for Performance Testing DFMEA Economic Analysis

4 8.0 Recommendations Appendixes Appendix A Stress Analysis of Valve Systems Appendix B Drawing Package Appendix C Purchased Parts Appendix D DFMEA Appendix E Controller Architecture and Code Figures Figure 1: Cutaway view of Peter Britanyak s design... 9 Figure 2: Cutaway view of Rotax 670 rotary valve engine Figure 3: Cutaway of Parallel Rotary Valve system compression (left) and exhaust (right) Figure 4: Allowable valve shaft torque with Peter s design with different material valves Figure 5: Current Valve Design Figure 6: Cutaway view of Parallel Valve system Figure 7: Exhaust side view of Parallel Valve system Figure 8: Magneto side view of belt drive Figure 11: Flowchart of Variable Exhaust Trapping Figure 12: Valve Position Management Figure 13: Valve Position Order of Operations Figure 14: Cylinder with Machined Pocket Figure 15: Modified Cylinder with Welded Insert Figure 16: Modified Cylinder with Removable Insert Figure 17: Valve Face Figure 18: Valve and Counterbalances on Shaft Tables Table 1: The five modes used for snowmobile testing for EPA and NPS... 7 Table 2: Specifications for SCT design... 7 Table 3: Decision Matrix used for Valve system decision Table 4: Parallel Rotary Design Estimated Cost

5 Executive Summary The Short Circuit team designed and manufactured a system to limit the amount of fresh air charge that is allowed to escape through the exhaust of a two stoke engine. This system was designed for the National Institute for Advanced Transportation Technology and will be used on the University of Idaho clean snowmobile engine. The team developed the prototype design and manufactured the system. The path taken to determine and manufacture the design will be discussed in this paper. The team had two concepts to choose from; an in-pipe exhaust valve that would increase back pressure, or a synchronous charge trapping (SCT) design that would change the exhaust port timing with crank position. The team chose the synchronous charge trapping system to accomplish the goals of the project. The system designed by Short Circuit is a revision of a system that was designed by a master s student from the University of Idaho in 2009 and The key difference in the new system is the rotational motion of the valve rather than a reciprocating motion. The valves rotate on a shaft running parallel to the crankshaft. This shaft rotates counter-crank wise, and is driven by a double sided cogged belt. The tensioning system is also used as a tuning device. This allows the timing of the valve to be advanced or retarded. A hybrid stepper motor controls advance based on throttle position and engine speed. Many components of the system were manufactured to assemble the prototype engine. These components were made at the University of Idaho mechanical engineering machine shop. The parts that are in direct exhaust flow were made of stainless steel, while other components were made of aluminum. The positive attributes of this design are that it is simple, with only a few additional moving parts, and durable. The stresses that are seen on critical components are very low. These qualities will appeal to manufacturers of two stroke engines. 5

6 1.0 Introduction 1.1 Background The synchronous charge trapping system is a University of Idaho senior design project sponsored by (NIATT) the National Institute of Advanced Transportation Technology. NIATT is a federally funded university transportation center dedicated to developing solutions to current transportation problems within the northwest and the United States. The main benefactor of the project is the U of I Clean Snowmobile Challenge (UICSC) team. The UICSC team is a NIATT funded student design project that competes in the SAE clean snowmobile challenge. The clean snowmobile challenge was created in 2000 due to increased pressure from conservationists and a push from industry to do something about the high toxic emissions and noise levels produced by snowmobiles. This is also because snowmobiles commonly ride in environmentally sensitive areas, such as Yellow Stone National Park. The clean snowmobile challenge is supported by the snowmobile community, conservationists groups, Society of Automotive Engineers (SAE), the Environmental Protection Agency (EPA), National Park Service (NPS), the Department of Energy (DoE), and others. 1.2 Objective The Short Circuit team came into the project planning to create a synchronous charge trapping system to be adapted to the UICSC team s direct injection two stroke engine while minimizing cost and complexity. The goal of the synchronous charge trapping system is to decrease hydro carbons, carbon monoxide emissions and decrease brake specific fuel consumption. The UICSC team needs to meet an Emissions score (E-Score) of 170 to meet NPS standards for exhaust emissions. The E-Score is calculated from an EPA five mode test where hydrocarbons, carbon monoxide and nitrogen are measured at five mode points shown in Table 1. The data is placed into the five mode test equation to generate an E- Score. An E-Score of 120 meets the EPA s standards for snowmobiles for 2012 and passes emissions at competition. The short circuit team hopes to achieve an emission score of greater than 170 without the use of catalyst along with other specific goals created by the UICSC team. 6

7 Table 1: The five modes used for snowmobile testing for EPA and NPS Mode Point Speed [% of Rated] Torque [% of Rated] Weighting [%] Idle 0 5 Along with the primary goal of obtaining an E-Score of 170 without a catalyst we developed other specifications for the SCT design to be considered successful and a candidate for use in the UICSC team s competition snowmobile, see Table 2 for details. Table 2: Specifications for SCT design General Requirements Specific Requirements Acceptable performance Durability Survive extreme temperature and stress cycles 50hr maintenance interval Performance Stock torque Decrease BSFC by 30% at cruise Emissions Reduce emissions E score > 170 Design Fit in Original Equipment Manufacturer (OEM) chassis OEM fit and finish 7

8 2.0 Concepts Considered 2.1 Redesign Peter Britanyak s System Peter Britanyak, a former master student at the University of Idaho, designed the initial synchronous charge trapping system, shown in Figure 1, for his master s thesis. The design he created used a mechanically actuated reciprocating valve in the exhaust port in order to trap the intake charge of a direct-injected 600 Ski-doo engine. The stock 600 carbureted cylinders were modified by machining out the stock exhaust port and welding in a new aluminum flange. The aluminum flange allowed an aluminum insert to be bolted into the modified cylinder which carried the valve, valve shaft, and high temperature graphalloy bushings. Peter s design is completely mechanical and uses an eccentrically positioned linkage attached to the crankshaft of the engine. This eccentric linkage rotates a common shaft that moves two separate linkages moving the valves in a reciprocating motion. With this system Peter was able to achieve as much as a 30% increase in torque over stock torque while matching stock brake specific fuel consumption (BSFC), and a 10% decrease in BSFC by matching stock torque up to 3500 revolutions per minute (RPM). The main issue with Peter s system is that it is incapable of running higher than 3500 RPM. Components of the valve train became overstressed and failed at RPMs over 3500 RPM. The valve shaft sheared when engine RPMs were greater than 3500 RPM due to high accelerations of the valve during operation. These high accelerations were caused due to the reciprocating motion of the valve changing direction twice during every engine stroke. While testing, Peter was able to show his design had greater efficiency than the stock direct-injected 600 engine, but it was unable to survive at RPMs exceeding Our first concept considered was to take Peter s design and redesign the valves, shafts and possibly linkages to allow the system to survive up to 8000 RPM reliably. The main component that would need to be redesigned would be the valve. In order to get the system to survive at 8000 RPM reliably, the valve would need to be lightened significantly. The valve, originally built from stainless steel, would need to be constructed of a material such as titanium sheet metal in order to reduce the moment of inertia. 8

9 Figure 1: Cutaway view of Peter Britanyak s design 2.2 Rotary Disk System Our second design option was designing a synchronous charge trapping system using a rotary disk in the exhaust port, similar to the rotary intake valves on a 670 Rotax engine like the one shown in Figure 2. The rotary disk design is similar to Peter s design in the way it traps the intake charge in the cylinder using a mechanical system, but differing in the type of valve used to trap the charge. Peter s design is a normally open system, meaning that the exhaust port is always open except when the valve is down to trap the charge. The rotary disk system uses a normally closed system, meaning that the exhaust port is always closed except when the opening in the disk aligns with the exhaust port allowing the exhaust charge to escape. This system would require a gear driven system from the crankshaft of the engine driving the sheet metal disk in a circular motion 360 degrees. The disk would be located as close to the cylinder wall as possible and rotate perpendicular to the crankshaft s rotation. 9

10 Figure 2: Cutaway view of Rotax 670 rotary valve engine 2.3 Parallel Rotary Valve System Our third design option was designing a synchronous charge trapping system using a parallel rotary valve system, shown in Figure 3. The parallel rotary valve design is also similar to Peter s design in the way it traps the intake charge in the cylinder using a mechanical system, but differing in the motion of valve used to trap the charge. The parallel rotary valve system uses a valve similar in design to Peter s valve but rotating 360 degrees parallel to the crankshaft. The two rotating valves are located 180 degrees from each other on a common valve shaft. The valve system incorporates a counter balance for each valve to ensure the valve is balanced on the shaft. Instead of using high temperature bushings inside the aluminum insert this design uses two high temperature bearings on the outside of each engine cylinder. The valves are rotated using a belt and pulley system that is ran off the engines crankshaft rotation. The 360 degree rotation of the valves eliminates the stresses in the shaft caused by the reciprocating motion of Peter s initial design. 10

11 Figure 3: Cutaway of Parallel Rotary Valve system compression (left) and exhaust (right) 3.0 Concept Selection For the benefit of NIATT and the UICSC team we chose to work on a cost effective design. As a team it was important to us that our selected valve design would balance total cost, machining difficulty, efficiency, and variability. Knowing these parameters we listed the pros and cons for each design. 3.1 Peter Britanyak s System Pros: The majority of the components are already designed and built. This would save time in the machine shop and save money since we wouldn t have to machine new parts or purchase a new set of cylinders. Peter s design is also proven to work with a 30% increase in torque while matching stock BSFC, and a 10% decrease in BSFC when matching stock torque. Cons: The biggest problem with Peter s system is the high accelerations of the valves causing the valves and their components to be overstressed to the point of failure. From our calculations we found the max allowable torque induced on the valve shaft without damage would be 80ft*lbs. Peter s valve shaft saw this max value at 4000 RPM and would see 450ft*lbs of torque at 8000 RPM. Peter s system is also non-variable and doesn t have the potential to be variable. See Figure 4 for stresses in the valve shaft with Peter s original design along with the valve being constructed of lighter material. For a more detailed analysis of why Peter s design failed see Appendix A. 11

12 Figure 4: Allowable valve shaft torque with Peter s design with different material valves 3.2 Rotary Disk System Pros: A rotary valve would not see the high accelerations that are seen on Peter s design due to the valve never changing its direction of motion. Cons: The cost and complexity play very big parts in this system. Our goal for this system was to adapt an existing 670cc Rotax rotary disk system to the UICSC 600 direct-injected system. This would require the purchase of an entire 670cc Rotax engine. The rotary disk system on this engine is set up for the intake of the engine, requiring our team to modify it making the rotary disk on the exhaust side. Doing this would be a complex process, with no viable way to make the valve system variable. 3.3 Parallel Rotary Valve System Pros: The valves in the parallel rotary system would see very low accelerations due to their circular motion. This would allow the engine to run at the required 8000 RPM without having reliability issues. The valves in this design would not restrict exhaust flow in the exhaust port. The machining process for this system would be similar to Peter s system, saving both modeling and machining time as we would be able to follow Peter s process. This system also has the potential to be variable unlike the other two designs. Cons: This system would have a high cost associated with it due to the fact that new cylinders would need to be purchased and then machined in the shop. We would also have to spend time modeling the parallel system to fit the cylinders. 12

13 From our evaluation of the pros, cons, and estimated cost of each design we were able to compare the four different specifications using a decision matrix to quantify each design, shown in Table 3, with a score of 4 being ideal and 0 being non-ideal: Table 3: Decision Matrix used for Valve system decision Criteria Peter s Parallel Disk Cost Complexity Efficiency Variability Total Score Selected Design Through careful evaluation team Short Circuit decided to design, build, and test the parallel rotary valve system. This is the only design, out of the three systems, that had the potential to meet our specifications; most importantly the specifications for reliability and variability. This system consists of four main components; these include the valve, valve shaft, pulley system, and electronic controls. 4.1 Stock Cylinder Modifications The modifications required to make the stock Ski-doo 600 cylinders work with the rotary valve system are very similar to the modifications Peter made to his stock cylinders. The cylinders will be machined to accept a fixed insert that slips into the cylinder on the exhaust side. The cylinders will also be modified to accept two high temperature bearings that are pressed directly into the cylinder on each side. The fixed insert is welded permanently inside the modified cylinder. The welded insert seals the coolant inside the water jacket of the modified cylinder while also allowing for enough room to house a removable insert. The removable insert slips into the welded insert and is fixed in place using fasteners 13

14 through the flanges of both the removable and welded inserts. The removable insert houses the rotating valve and shaft along with keeping a smooth transition for exhaust flow from cylinder to exhaust pipe. 4.2 Valves The valves on the parallel rotary system (Figure 6) are very similar to Peter s designed valves. The valve face reflects the same geometry of the cylinder in order to maximize a tight fit in the exhaust port, see Figure 5. The valve was also designed with a counter-weight opposite the face for the purpose of having a better balanced shaft. This improves the reliability by minimizing radial forces on the valve shaft. The two legs of the valve allow the shaft to slide along the shaft and be fixed onto the shaft with high-temp roll pins. Rotation Figure 5: Current Valve Design Figure 6: Cutaway view of Parallel Valve system 4.2 Valve Shaft A single steel valve shaft was used in this system. High temperature bearings were pressed into each side of both cylinders water jacket which allows the shaft to be inserted through the cylinders and then attached to the valves, and removed for maintenance purposes. On the magneto side of the engine the shaft is connected to the pulley system. See Figure 7 for the fully assembled prototype. 14

15 Figure 7: Exhaust side view of Parallel Valve system 4.3 Pulley System The 600cc Ski-doo direct-injected engine s crank shaft rotates toward the exhaust ports (clockwise looking from the magneto side). For this parallel rotary system to work the valves rotate in the opposite direction of the crank. To do this a pulley system was implemented that spins the valve shaft in the direction shown below in Figure 8. This pulley system maintains a 1:1 ratio while also allowing for variability. This is done by the implementation of a pulley bracket allowing the two smaller pulleys to rotate about the valve shaft. The belt driving pulley is attached to the crank of the engine with an aluminum hub that is bolted to the magneto of the engine. The magneto side bracket, called the mag bracket, holds the idler pulleys, shaft driving pulley, valve shaft and tensioner plate. The mag bracket is mounted to the crank case of the engine. The mag bracket also uses a sealed ball bearing to allow the valve shaft to rotate freely. The shaft driving pulley is attached directly to the valve shaft using set screws. The two idler pulleys which are used to adjust the timing of the valve with relation to the crank are attached to the tensioner plate which is allowed to rotate independently about the valve shaft on a brass bushing. The tensioner plate is attached to the stepper motor s lead screw and the stepper motor is attached to the mag bracket. This allows the tensioner plate to be rotated clockwise or counter clockwise allowing the valve timing to be retarted or advanced. 15

16 Figure 8: Magneto side view of belt drive 4.4 Electronic Controls To ensure the parallel rotary system will work in all five engine mode points the valve system needed to be variable. Electronic controls would change the time at which the valve is in the exhaust port. At lower RPMs where gas trapping is most effective the valve traps the intake charge as designed. However, at high RPMs where the tune pipe is harmonically tuned, the valve needs to be free of the exhaust port or it would impede the effects of the tune pipe. Changing the valve timing is achieved by rotating the outer, smaller pulleys about the valve shaft. Doing this consequently rotates the belt around the pulley system, changing the position of the belt on each pulley as the two pulleys rotate. Rotating the two smaller pulleys about the valve shaft is accomplished by using a Hayden Kerk eight series external hybrid stepper motor linear actuator as shown in Figure 9. The stepper motor is composed of a rotating lead-screw which causes a threaded nut attached to the arm of two smaller pulleys to move thus achieving rotation in both directions. This rotation and angular position is determined through throttle position and RPM by a Cypress Semiconductor CY8CKIT-001 PSOC 5 microcontroller as shown in Figure

17 Figure 9: Hybrid stepper motor linear actuator Figure 10: PSoC Microcontroller The throttle position and RPM signals are extracted from the type 262 E-Tec Computer on board the snowmobile. The throttle position signal is taken from the throttle position sensor (TPS) located at pin J1-A18. RPM is taken from the tachometer at pin J1-A16. Throttle position outputs a 0-5VDC signal while RPM outputs a square wave signal 0-14V with 6 pulses per revolution and an internal pull up resistor to critical 12V. The PSOC 5 microcontroller uses two on board analog to digital converters to acquire both throttle position and RPM. Additional electrical components were added to the board s circuitry area to implement a voltage divider that pulls down the RPM signal to the 5V required by the system PSOC system. The software required stepping the motor and the pulley arm to a certain angular position and required C code that take the two TPS and RPM signals and combine them in an algorithm. From this algorithm, and through dyno testing on the engine stand, a program was written to the microcontroller to vary the valve system at all RPM ranges and throttle positions. Basic software models for signal processing and motor control can be seen below in Figures 11, 12 and

18 Basic Flowchart for Variable Exhaust Trapping INPUT: Throttle Position Controller OUTPUT: Valve Position INPUT: RPM (rotations per minute) Figure 9: Flowchart of Variable Exhaust Trapping Valve Position Management Task Manager (MAIN) TPS RPM TPS Position Position RPM System Setup Read Throttle Position Read RPM Valve Position Valve/Motor Control Figure 10: Valve Position Management 18

19 Valve Position Order of Operations Startup/Reset System Initialization Read Throttle Position Read RPM Infinite Loop Valve/Motor State Logic Valve/Motor Control Figure 11: Valve Position Order of Operations 5.0 Product Manufacturing 5.1 Stock Cylinder Modifications The stock Ski-doo 600 cylinders were modified to house two separate inserts that hold the valve and valve shaft inside. One insert is welded into the modified stock cylinder and the other is removable to allow for the valve to be removed easily from the cylinder assembly. The welded insert was used to make the exhaust port large enough to house the valve and valve shaft assembly. The cylinders were first modified by fill welding inside the cylinder water passage to eliminate any thin spots that might occur after machining. A pocket was then machined into the exhaust side of the cylinders to house the welded insert. The cylinders were then fill welded a second time to eliminate any pockets left after the machining was complete. The cylinder was then machined a second time to achieve final dimensions, shown in Figure 14. The welded insert, which was machined using the HASS CNC mill, was then placed inside of the pocket in the cylinder with gasket material between the face of the welded insert and the back wall of the cylinder pocket. The welded insert was then welded into place inside the modified cylinder, see Figure 15. The now modified cylinder with the welded insert was machined again to final 19

20 dimensions to eliminate any warping that occurred during the welding process. The removable insert was machined using the HAAS CNC mill and placed inside of the welded insert. A bolt pattern around the outside flanges of the welded insert and removable insert allow them to seat together, eliminating any exhaust gas leakage, shown in Figure 16. After both inserts were complete and inside the modified cylinders, the HAAS CNC mill was used to drill the valve shaft hole through the cylinders and make the seats on each side of the cylinders for the high temperature roller bearings. Figure 12: Cylinder with Machined Pocket Figure 13: Modified Cylinder with Welded Insert Figure 14: Modified Cylinder with Removable Insert 20

21 5.2 Valves and Counterbalances The valves were the most complicated to manufacture due to complex geometry of the face of the valve and the material used in order to survive the high exhaust temperatures. The valve which was machined from stainless steel was machined using the HAAS CNC mill. They clamped on to the mill using aluminum blocks that would be later sacrificed during the machining process. The backside of the valve face and legs were machined first. The valve was then flipped over and was held onto the mill by the legs of the valve and the face of the valve was then machined, see Figure 17. Then the using the HAAS CNC mill the counter balances were machined from stainless steel, see Figure 18. The next step in manufacturing the valves and counterbalances was drilling the valve shaft through the legs of the valve, and then the counter balances. The final step was drilling the holes for the roll pins which hold the valves and counterbalances onto the valve shaft. This final process was completed using a SHARP manual mill; a block was placed between the legs of the valve in order to eliminate any leg deformation during the drilling process. Figure 15: Valve Face Figure 16: Valve and Counterbalances on Shaft 5.3 Pulley System The pulley system was much easier to manufacture than the other components in our design, such as the inserts and valve. Some of the parts such as the belting driving pulley and the shaft driving pulley were left unmodified from the vendor. The mag bracket was machined out of aluminum using the HAAS CNC mill. The aluminum belt driving pulley hub, idler pulley shafts and the brass bushing were machined using the HASS CNC lathe. The tensioner plate was machined from aluminum using the SHARP manual mill. In order to use sealed needle bearings the idler pulleys were also modified using the SHARP 21

22 manual mill and the needle bearings were pressed into the center of the pulleys. The needle bearings spin on fixed shafts that were pressed into the tensioner plate to allow the idler pulleys to float on the shafts for proper belt alignment. 6.0 Product Evaluation 6.1 Specifications Many of our original product specifications dealt with how our design performed after being tested on the University of Idaho s Borghi Saveri dynamometer. Unfortunately, after spending most of our time manufacturing our design we were unable to get the engine on the dyno for testing. We have developed a testing plan that will be performed this summer by the University of Idaho clean snowmobile team. This testing plan will be covered in the next section of this paper. We were able to design our system to theoretically meet our durability specifications. Whether our design actually meets the durability specifications will have to be discovered in the future during testing. We were also able to meet the space requirements specifications by designing and manufacturing the parallel rotary valve system to fit inside the OEM chassis with OEM quality. 6.2 Plan for Performance Testing The parallel rotary valve engine will be tested against the stock Ski-doo 600 direct-injected engine with the stock tuned pipe. The parallel rotary valve engine will use the stock head, Peter s pipe and stock muffler. The parallel rotary valve engine will not use the stock tuned pipe. The initial tuning of our engine will begin by tuning the engine at low speed and low load and then work up through the RPM band and load. There will be a routine check of all critical components between each tuning session, see the DFMEA section of this paper for more detail. The torque will be tested against the stock engine at each of the five mode points by matching the stock BSFC. Then the BSFC will be tested against the stock engine at each of the five mode points by matching the stock torque. Emissions will be tested, after the engine has been tuned through all of the RPM and load range. The parallel rotary valve engine emissions will be tested against the stock Ski-doo 600 direct-injected engine, using the EPA five mode tests. In the fall, after the performance and emissions of our design has been tested against the stock engine, the engine will be placed into the competition chassis. Then the sound emissions will be tested against the stock engine. After all the testing is complete the UICSC team will have to analyze the data and decide which engine will be used in next year s competition. 22

23 6.3 DFMEA For the design failure mode effect analysis we came up with several key components that could fail and cause a complete system failure. The type of failure that could cause complete catastrophe is burning down or over heating the engine during testing. If the engine does burn down, and destroys the cylinder walls inside our cylinder, a new set of cylinders will have to be purchased and new welded inserts will have to be machined and then the inserts will have to be welded inside the modified cylinder. This will cause a huge cost to the UICSC and will cause the testing to be set back by at least three weeks. To prevent the engine from burning down and destroying the cylinder walls we recommend that the oil lines to the engine be inspected before any dyno testing is done, and the engine be tuned while using an air fuel sensor along with keeping the exhaust gas temperatures in the safe range. The second type of catastrophic failure that could occur is the failure of the valve inside the exhaust port. If the valve was to come apart and launch pieces of valve or valve shaft into the cylinder while the engine was running the cylinders, along with the crank of the engine, could be destroyed. In order to help ensure that this type of failure does not occur we recommend that the valves and valve shaft be inspected after first idling for 10 minutes on the dyno, then after running at 3000rpm on the dyno, and finally after running 10 minutes at wide open throttle and full load. For a full analysis of our DFMEA see Appendix D. 7.0 Economic Analysis We decided to do an economic analysis of our design by outlining the costs of our design from a start to finish application. The table below shows what our design would cost a person to take a stock Ski-doo 600 direct injected engine and modify it to use the parallel rotary design. Table 4 below outlines the costs of materials, components, machine work and welding. 23

24 Table 4: Parallel Rotary Design Estimated Cost Item Cost of Item Number of Items Total Stock $ $ Ball Bearings $ $ Valve Shaft $ $20.62 Belt $ $15.02 Belt Driving Pulley $ $8.75 Shaft Driving Pulley $ $21.99 Idler Pulleys $ $24.24 Needle Bearings $ $21.22 Retaining Rings $ $0.15 Machine Work $60/hr 100 hr $6, Welding Work $30/hr 5hr $ Total $6, Recommendations Our recommendation is that the University of Idaho clean snowmobile team continues to progress with our current design. The synchronous charge trapping has been proven in concept by Peter Britanyak s first generation design. Our design uses his design s best features, without having any highly stressed component, allowing it to be reliable. We also recommend that our design should be tested by a graduate student on the UICSC team so that the engine s performance can be thoroughly tested and documented. We recommend that the UICSC uses our DFMEA and our testing plan to test the engine s performance. The engine should be tuned for the optimal combination of power, durability, and exhaust emissions. After efficiency and exhaust emissions are tested, without any major component failures, the durability of the design will be proven. At the conclusion of testing the overall success of the engine can be determined. If the engine is successful at meeting our goals it should be mounted in the University of Idaho clean snowmobile team s chassis and used in competition. 24

25 Appendixes Appendix A Stress Analysis of Valve Systems 25

26 26

27 27

28 28

29 29

30 30

31 31

32 Appendix B Drawing Package 32

33 33

34 34

35 35

36 36

37 37

38 38

39 39

40 40

41 41

42 42

43 43

44 44

45 45

46 46

47 47

48 48

49 49

50 50

51 51

52 52

53 53

54 54

55 55

56 56

57 57

58 58

59 59

60 60

61 61

62 Appendix C Purchased Parts 62

63 63

64 64

65 65

66 RPN DETECT OCCUR SEV Appendix D DFMEA ITEM AND FUNCTION POTENTIAL FAILURE MODE(S) POTENTIAL EFFECT(S) OF FAILURE POTENTIAL CAUSE(S) OF FAILURE CURRENT DESIGN CONTROLS RECOMMENDED ACTIONS Engine Burn Down Burn Down Burn Down Damage to cylinder 4 Oil line falling off 2 None 2 16 Damage to piston 4 Overheating 2 Damage to valve system 4 Prolonged run at Mode1 2 EGTs, and Head Temp sensors EGTs, and Head Temp sensors Regular Inspection Dyno Experience, Good Lambda, Keep EGTs in check Dyno Experience, Good Lambda, Keep EGTs in check SEV OCCUR DETECT 1=No Harm Done 3=Moderate Damage To System 5=Total Destruction Of Valve System Or Engine 1=Remote 3=Every Week 5=Every Test Period 1=Certain 3=Moderate Delectability 5=Undetectable 66

67 DETECT RPN OCCUR SEV ITEM AND FUNCTION POTENTIAL FAILURE MODE(S) POTENTIAL EFFECT(S) OF FAILURE POTENTIAL CAUSE(S) OF FAILURE CURRENT DESIGN CONTROLS RECOMMENDED ACTIONS Valve Valve Release Valve Release Shear Pins Fail Shear Pins Fail Melt Valves Melt Valves Fling metal into cylinder 5 Damage to inserts and shaft. 4 Fling metal into cylinder 5 Damage to inserts and shaft. 3 Damage to inserts and shaft 5 Destruction of valves 5 Over Stress and or deformation 2 None 2 20 Over Stress and or deformation 2 None 2 16 High Shear Forces on Pins 4 None 2 40 High Shear Forces on Pins 4 None 2 24 Valve material having low melting temp None 2 10 Valve material having low melting temp None 2 10 Inspect shear pins and safety wire after 10 mins idle then 10 min at 3000 rpm and 10 min at mode 4 Inspect shear pins and safety wire after 10 mins idle then 10 min at 3000 rpm and 10 min at mode 4 Inspect shear pins and safety wire after 10 mins idle then 10 min at 3000 rpm and 10 min at mode 4 Inspect shear pins and safety wire after 10 mins idle then 10 min at 3000 rpm and 10 min at mode 4 Inspect valve after every 30 min of running for deformation initially at elevated temps Inspect valve after every 30 min of running for deformation initially at elevated temps SEV OCCUR DETECT 1=No Harm Done 3=Moderate Damage To System 5=Total Destruction Of Valve System Or Engine 1=Remote 3=Every Week 5=Every Test Period 1=Certain 3=Moderate Delectability 5=Undetectable 67

68 RPN DETECT OCCUR SEV ITEM AND FUNCTION POTENTIAL FAILURE MODE(S) POTENTIAL EFFECT(S) OF FAILURE POTENTIAL CAUSE(S) OF FAILURE CURRENT DESIGN CONTROLS RECOMMENDED ACTIONS Belt Breaks/Melt Falls off Halts testing until it can be replaced 2 High Heat, High Tension 1 None 2 4 Halts Testing until fixed 1 Too loose 3 None 2 6 Inspect during valve inspections for melting and fraying Inspect tension during valve inspections Skips Cog Changes Valve timing and position 2 Loose Belt, High Belt Load 2 None 2 8 Inspect tension during valve inspections SEV OCCUR DETECT 1=No Harm Done 3=Moderate Damage To System 5=Total Destruction Of Valve System Or Engine 1=Remote 3=Every Week 5=Every Test Period 1=Certain 3=Moderate Delectability 5=Undetectable 68

69 RPN DETECT OCCUR SEV ITEM AND FUNCTION POTENTIA L FAILURE MODE(S) POTENTIAL EFFECT(S) OF FAILURE POTENTIAL CAUSE(S) OF FAILURE CURRENT DESIGN CONTROLS RECOMMENDE D ACTIONS Controller Overheat s Board Gets Shorted Board looses stepper motor position Could Destroy electrical components 2 Destroy Component s, Sink lots of current burning up itself and EMM 4 Puts valve in wrong angular position and would make engine run bad 1 High engine compartmen t temperature s 1 Electromachanic al Vent flaps which open when compartment is at a specified temperature 5 8 Conductive object touches it. Wires rub through insulation. 1 None 5 Bad software loops 2 None Preventative steps to ensure compartment temp is not out of range Enclose in non-dielectric box for protection. Place wires in safe areas. Write good code and calibrate when testing. SEV OCCUR DETECT 1=No Harm Done 3=Moderate Damage To System 5=Total Destruction Of Valve System Or Engine 1=Remote 3=Every Week 5=Every Test Period 1=Certain 3=Moderate Delectability 5=Undetectable 69

70 Appendix E Controller Architecture and Code 70

71 71

72 72

73 ADC Code 73

74 74

75 ADC Measured Values Voltage (DC) HEX Value 0 0x000C x018D x030E x04CD x065F x087F x09AD x0B x0D x0EAA x0FEA 75