Archive for December, 2010
Final mechanical system
The next generation of green thinking needs to be about something other than being green. Sustainability is important and is certainly something that needs to be taken into the design process, but it needs to be viewed different. This new movement of sustainability needs to be about making an emotional connection with a product to move forward with environmentalism in mind. The project here is meant to take that idea of excitement and emotional connection to sustainability, and put it into a vehicle. Motorsports are the ideal medium for showcasing excitement with a focus on sustainability and there is no better embodiment of emotional connection in motorsports than the motorcycle. To that end, a 1996 Suzuki GS500e motorcycle will be converted from a 2 cylinder, 4 stroke combustion motor to a mechanical hybrid, by means of an internal source of compressed air. What follows is the culmination of a two semester mechanical engineering design project at San Francisco State University resulting in the successful completion of the mechanical systems for this concept motorcycle power plant.
This air gas hybrid will be built on the process ideas put forth by the Scuderi split cycle air gas hybrid motor, however, applied to a readily available, normal gasoline motor. The goal is to show that increased performance and efficiency can be achieved, without significant impact, by creating a hybrid engine system from what is already available. This project will be split into two main foci, converting one of the two cylinders in the motor to act as a compressed air source and developing a controllable system for storing and applying the stored air charge.
Before an in-depth discussion regarding the development of a new system can occur, one must first evaluate current technology and the basics of the internal combustion engine. An internal combustion engine is a complex gas power cycle where stored energy, in the form of chemical fuel, is transformed into kinetic energy through controlled combustion and allowed expansion. This expansion, via a crankshaft, becomes a useable force. However, combustion of a fuel is a non-cyclic process. The working fluid, an air-fuel mixture, undergoes a permanent chemical change and after use is discarded to be replaced with a new intake charge. For the purpose of thermodynamic evaluation, this gas power cycle must be represented as an air standard cycle. For calculation purposes, a mass of air operates in the complete thermodynamic cycle and heat is added and rejected via external reservoirs. This process is to be considered reversible. For the purpose of analysis, certain assumptions must be made. The working fluid is considered to behave as an idea gas with constant specific heats. The combustion process is replaced with heat addition; exhaust with heat rejection.
The following terms are to be defined for ease of discussion:
TDC; Top Dead Center: Position of the piston at top of stroke
BDC; Bottom Dead Center: Position of the piston at bottom of stroke
Stroke: Distance between TDC and BDC
Bore: Diameter of the piston
Compression ratio: ratio of maximum volume to minimum volume VBDC/VTDC
Engine displacement = (# of cylinders) x (stroke length) x (bore area)
MEP: mean effective pressure: A const. theoretical pressure that if acts on piston produces work same as that during an actual cycle
Wnet = MEP x Piston area x Stroke = MEP x displacement volume
Figure B1: The four stroke process
Figure B1 shows the basic four stroke internal combustion engine. This process consists of four distinct strokes, hence the name. During the intake stroke, the piston moves down in the cylinder and the intake valve opens. This draws a fresh fuel/air mixture into the cylinder. As the piston moves back up, the intake valve closes, and the air is compressed. This is the compression stroke. At the beginning of the power stroke, the now highly compressed fuel/air mixture is introduced to spark via an ignition source. This causes combustion, adding heat to the process. A rapid expansion of gasses applies pressure to the piston, forcing it down, converting the potential energy stored in the fuel to kinetic energy. After the power stroke, the exhaust valve opens and the piston returns to TDC, exhausting the combusted fuel/air mixture, finishing the exhaust stroke.
As earlier mentioned, for the purpose of analysis, this four stroke process is converted to an idealized air standard cycle. In this case, the Otto cycle is used, represented in Figure B2.
Figure B2: The Otto cycle
Process 1-2 is an isentropic compression. Work (Win) is applied via the crankshaft, compressing the air. If one applies the First Law of Thermodynamics;
U2-U1 = Q – Win
Q = 0 (since, reversible adiabatic compression)
Win = U2-U1
Process 2-3 is a constant volume heat addition;
U3-U2 = +Qin – W
W = 0 (since, it is a constant volume process)
Qin = U3-U2
Process 3-4 is an isentropic expansion;
U4-U3 = Q – Wout
Q = 0 (rev. adiabatic expansion)
Wout = U4-U3
Finally, process 4-1 is a constant volume heat removal;
U1-U4 = – Qout + W
W = 0 (no piston work)
Qout = U4-U1
The thermal efficiency is given by:
Here y=1.4 at ambient temperature
Effectively, the efficiency of the internal combustion process can be increased through a rise in compression ratio.
The mass air flow through a 4 stroke motor is governed by the following equation:
where D is the displacement, N is the rotational velocity, and 120 is a factor of 60 to convert from revolutions per minute into second and 2 for the amount of crankshaft rotations per intake stroke. A motor that were to intake air every stroke would simply have 60 in the denominator.
Often times forced induction is used in internal combustion motors to increase performance by artificially increasing compression ratios. Superchargers; either crank driven or exhaust driven turbochargers; feed the motor positive air pressure, drastically increasing volumetric efficiency. Unfortunately, superchargers are limited in that they add a parasitic load to the motor, requiring substantial amounts of energy to be used in their operation. Also, their size and weight makes them difficult to use in applications where packaging and total weight are important concerns, such as on a motorcycle.
Certain valve considerations need to be addressed due to their relevance to this project. Modern 4 stroke motors often regulate their air intake and exhaust through the use of poppet valves. Figure B3 shows one of these valves.
Figure B3: Poppet valve
This valve can be considered an active valve, as it requires an external actuator for control. Camshafts are used to open these valves by depressing them. They offer high levels of control, however, due to force being required operate them, can rob up to 10% of the produced power of the engine.
Common in two stroke motors are reed valves. These reed valves, as seen in figure B4, are passive valves, self-actuating.
Figure B4: Reed valve
This one way valve is actuated via air pressure differential, allowing flow in the direction of the arrows, only when pressure is greater before the valve then after. Reed valves often see use in 2 stroke motors where intake and exhaust occur with every stroke of the motor. When extra control is not needed, they make for a far more simple option than the entire valve train that’s required with the use of poppet valves.
For high performance motorcycle applications, where packaging needs prevent the use of external forced induction systems, but an increase in performance is desired, an internal supercharger is proposed. This system would convert one of two cylinders from a normal four stroke Otto cycle to an air pump. For future application, this air pump would feed pressurized air to the adjoining cylinder. To operate, valving changes must be made to the compressor cylinder, allowing intake to occur every rotation of the crankshaft. Based on earlier stated equations, mass flow rate through this cylinder will double. As displacement of the motor for compression is now reduced by ½, mass flow rate returns to the original value, however, thermal efficiency increases. As such, overall performance can increase, without a significant increase in weight or size by keeping everything internal to the motor, utilizing already present components.
Three prominent possibilities were evaluated for conversion of the valve train for this application. The motor used for conversion, as seen in figure E1, is a 487cc parallel twin from a 1996 Suzuki GS500E motorcycle.
Figure E1: 487cc parallel twin engine
First, custom camshafts were considered. Figure E2 shows the stock camshafts.
Figure E2: Stock GS500E camshafts
The proposed design was to create new camshafts where the lobes were modified in such a way to allow opening of the intake and exhaust valves twice as often. This option would be the simplest and most efficient, as cam timing could be perfected for the application, however, due to the incredibly high strength of steel needed for camshafts, fabrication requires specialized equipment and high costs. Quotes received from various camshaft fabricators were nearly $2000 for the pair.
Second, a replacement head was considered. A replacement head, on a split cylinder (v twin or opposed twin motor configuration) would be a simple solution, allowing for increases in efficiency, compression parameters, and simple fabrication. However, due to the availability presented, a parallel twin configuration was chosen. A new head design would require a far more complicated design to ensure normal operation of the unmodified cylinder, making costs much higher and fabrication vastly more difficult.
The third, and selected, solution was to remove the stock valves all together and replace with reed valves mounted externally to the motor. While not offering the same level of performance and ease of packaging as the previous two options, externally mounting reed valves is a lower cost alternative utilizing passive valves that come at a much lower cost. Off the shelf valves are utilized, residing in machined reed cages attached to the intake and exhaust ports.
As reed valves have been decided upon for this application, selection and sizing was first faced. While common on small displacement, low performance motors, reed valves for larger motors are less common. Available options were evaluated based on their physical dimension and the horsepower and displacement of the motors they were removed from. Polaris runs a particular design of reed valve on many of their larger motors that are of appropriate availability and sizing. The specific valves used were from 600cc Polaris personal watercraft. Figure D1 shows the reed valve relative to the size of the motor.
Figure D1: Reed valve mocked up on exhaust port
Housings were designed for these valves using the Rhino3D surface modeling software. Figure D2 shows the original valve housing design in a Rhino3D screenshot.
Figure D2: Rhino 3D screenshot of original valve design.
This original design called for a tapered profile in an attempt to minimize turbulence in the housing. Unfortunately, limitations of the milling machine required some design changes to allow for clearance and capabilities of the cutting tooling.
The final model was converted into orthogonal views and brought into the Cut2D 2.5 axis CAM software. Figure D3 shows a screenshot of the valve housing being processed through CAM software.
Figure D3: Valve housing model in Cut2D
Toolpaths for the CNC mill were created in Cut2D, then processed and saved in G-code for output to the mill controller.
The Tormach 1100 CNC mill was used for the fabrication of the valve housings.
Figure D4: Tormach 1100 CNC mill
4” x 4” x 3” billets of 6061 aluminum were used for the raw material. Billets were faced and then machined from the top and bottom in separate stages.
Figure D5: Facing the billets on the mill
Figure D6: Billet after facing
Figure D7: Machining top profile
Figure D8: Housing after top machining operation
Figure D9: Finished valve housing with nested valve
Figure D10: Finished valve housings mounted to cylinder head
Shows a brief video of the valve fabrication process from billet to housing.
In addition to the design for housings of the reed valves, the stock poppet valves needed to be removed from the cylinder head. Simple removal would not suffice, as the valve journals that normally house the stock poppet valves would be left open, making compression impossible.
Figure D11: Motorcycle head cutaway
Figure D11 shows a cutaway of an overhead cam motorcycle motor. The poppet valves are clearly visible, along with the path for ventilation that would occur with only their removal. A seal was created, using a bolt of proper diameter, o rings for both sides of the valve journal, appropriately sized washers, and a nylon lock nut. After removal of the stock valves, this replacement valve rod was put in place, the motor reassembled and prepared for testing.
Figure D12: Cylinder head with valves removed; valve journals visible
Figure D13: Bolt and o ring seal for valve journal
Figure D14: Journal seal in place
Figure D15: Motor reassembled, ready for testing
To verify proof of concept, one must revisit the earlier mentioned equations for mass air flow rate relations.
Once again, for a motor with an intake cycle on every rotation of the crankshaft, the denominator of 120 changes to a value of 60. If one were to compare the mass flow rate of the two cylinders, this equation states that the mass flow rate of the compressor cylinder should be twice that of the untouched cylinder, given than density, displacement, and rpm must be constant. To verify that the motor does in fact intake air twice into the compressor cylinder for every intake stroke on the untouched cylinder, the motor must be ran. A 12V source was applied to the starter motor on the engine, turning it over at a constant speed. Through observation, the engine can be seen to actuate the reed valves on the compressor cylinder at twice the rate of the untouched cylinder.
This is how the motor behaves under that power source, showing the timed movement of the valves.
With the compressor parts fabricated and installed in the bike, attention can be drawn on to utilizing this compressed air source. The air must be routed from the compressor outlet to a storage/buffeting tank to the engine intake and considerations must be made for the behavior of the compressed air supply. When considering components for this system, thought must be given to the needs of the motor, packaging constraints, strength, heat and weight concerns. For the piping, we want something as light as possible, that doesn’t absorb heat as it passes near the motor, and can be shaped to fit the packaging limitations. For this purpose, composite intake charge pipes were fabricated from carbon fiber to insulate the charge air, minimize weight and for ease of shaping into the complex curves needed. The air storage tank must also fit the packaging and be of proper volume. Here, an air to air intercooler was used to also benefit heat concerns. With an evaluation of the behavior of the compressed air supply during a variety of driving conditions performed, a blow off valve was added to better regulate air flow during off throttle conditions.
The intake charge pipes for the motorcycle were fabricated from carbon fiber. This composite material allows for a pipe of significantly lower weight than a steel or aluminum pipe while maintaining the same strength characteristics. Carbon fiber also offers far better insulation properties, so the intake air does not absorb any heat as the pipes travel near the head of the motor, as required by packaging.
To fabricate these pipes, the lost foam technique was utilized. To do this, the pipes were laid out in extruded polystyrene foam(EPS), in this case a foam wreath.
Figure P1: EPS piping
Figure P1 shows the EPS tubing cut to shape and mocked up in location on the bike. After a final shape is established, the pipes are wrapped in 4 layers of coaxial woven carbon fiber.
Figure P2: EPS pipe wrapped in carbon fiber
Figure P2 shows the pipe with the carbon in place. As each layer is added, a high temperature epoxy resin is applied, fully saturating the carbon fiber. Eventually, this resin will cure into a solid form, reinforced by the carbon fiber fabric. To ensure maximum strength, the pipe with the resin saturated carbon fiber was wrapped in a heat shrink tape. This squeezes out excess resin and any air bubbles caught between the layers of carbon fiber. It also ensures a solid, even pipe is made so that it is free of air leaks.
Figure P3: Charge pipe unwrapped
After the resin has fully cured, the heat shrink tape is removed from the pipe, leaving a final high gloss surface finish and strong, solid pipe as seen in figure P3. At this point, however, the pipe is still full of foam. To remove this foam, acetone or gasoline is poured into the pipe. Polystyrene foam was chosen originally for this application as it is soluble in gasoline and acetone. The foam dissolves away, leaving a light weight, strong, hollow carbon fiber charge pipe in complex shapes that would be difficult to otherwise mold and fabricate.
The air tank on the bike serves multiple purposes. Beyond just acting as a storage tank for air, it buffets the air pulses coming out of the motor. Because the compressor pulses on an off, the air stream exiting is more sinusoidal. Also, the intake on the motor only opens for part of every other stroke, so the inlet creates pressure waves as well. To prevent this, typically a motor will use an intake plenum to buffet these intake pulses. Packaging constraints on the motor prevent the use of an intake plenum here, however, with proper sizing and placement, the air tank can satisfy these needs. Further, the use of an intercooler offers an opportunity to remove heat from the intake stream, offering higher thermodynamic efficiencies and as a result, more power and improved fuel efficiency.
Figure IC1: Intercooler placement
Figure IC1 shows the intercooler, a standard part from a Mazda Millenia sedan, in place on the frame under the seat. The positioning is a bit of a trade off. It places the air tank in close proximity to the motor intake, so it is usable as an intake plenum, however, air flow across the heat exchanger is reduced. Eventually, ducting will be added to deal with this air flow concern. This particular intercooler was used due to its small size. To act as an intake plenum, an ideal internal volume of 3 times the cylinder displacement is needed. The Millenia intercooler is a bit larger than this value, but not significantly. It is also the smallest non custom intercooler available at a reasonable cost.
The compressor system on this motorcycle creates pressure any time the motor is running. Because of this, there will be times when the throttle plate is closed and the motor is not consuming air, where this constant air build up would eventually cause problems. To avoid this situation, a blow off valve was installed on the charge pipe between the intercooler and the intake. This valve operates based off a vacuum signal from the motor intake. When the throttle plate closes, vacuum is applied to a diaphragm on the top of the blow off valve. This causes the valve to open and vent out from the pipe.
Figure BV1: Blow off valve in place
Figure BV1 shows the final placement of the blow off valve. This placement is the only place inside the frame perimeter the valve could fit. Also, putting the valve near the intake allows for a constant velocity flow during gear changes. At this time, the throttle plate is shut with the motor still spinning at high rpm. Without the blow off valve, air flow velocity would come to a stop. With the blow off valve, the only change in air velocity flow occurs in the short section of pipe between the blow off valve and the intake.