UNIVERSITY OF CINCINNATI Date: October 31, 2006

I, Ryan Lake_________________________________________________, hereby submit this work as part of the requirements for the degree of: Master of Science in: Mechanical Engineering It is entitled: Integration of a Small Engine Dynamometer into an Eddy Current Controlled Chassis Dynamometer

This work and its defense approved by:

Chair: Dr. Randall Allemang___________ Dr. David Thompson_____________ Dr. Jay Kim

_

_______________________________ _______________________________

Integration of a Small Engine Dynamometer into an Eddy Current Controlled Chassis Dynamometer

A thesis submitted to the Graduate School of the University of Cincinnati

In partial fulfillment of the requirements for the degree of

MASTER OF SCIENCE

In the Department of Mechanical, Industrial, and Nuclear Engineering of the College of Engineering

2006 By Ryan Douglas Lake B.S., University of Cincinnati, 2004

Committee Chair: Dr. Randall Allemang Committee: Dr. David Thompson Dr. Jay Kim

Abstract The task of tuning an engine from scratch can be very time consuming and difficult if the right equipment is not utilized. Several different types of dynamometers with feedback control systems exist that enable a tuner to simplify the process. However, most of these systems are designed for specific applications and engines. Typically, the proper equipment is determined based on the budget and requirements of the tuner. The most common engines for Formula SAE (FSAE) cars are usually motorcycle engines or something similar. Unlike the usual car engines, which have separate transmissions, these engines and transmissions are built together. A complete custom engine dynamometer stand and corresponding connection between the transmission output shaft and the dynamometer is necessary. Different types of dynamometers were researched to determine their pros and cons. The nine inch Land and Sea water brake absorber and dynamometer stand utilized by the University of Cincinnati’s FSAE team since 1998 was researched to determine its performance characteristics. The Mustang Chassis Dynamometer and corresponding eddy current absorber purchased in 2003 were researched as well. The eddy current absorber is capable of maintaining low RPM speeds compared to the water brake. This key feature could be taken advantage of if a connection system is developed to utilize the eddy current absorber in the chassis dynamometer as the absorber for the engine dynamometer. Various designs were investigated and evaluated. The details of these designs and the pros and cons of each setup are discussed. The final design was tested and utilized for tuning the 2006 FSAE engine saving a significant amount of time and effort. During this testing period, small problems with the system arose and were corrected as they surfaced. The system is still in a state of testing, and recommendations are presented that will enable a final setup to be permanently installed.

i

ii

Acknowledgments I would like to extend a great deal of thanks to three groups of individuals: my committee members, my colleagues in SDRL and Bearcat Motorsports, and my family, without your support I would not have been able to complete my research and this thesis. First, I would like to express my sincere appreciation to my advisor, Dr. Randy Allemang. I can not thank you enough for your continued assistance and input into the development, review, and completion of this thesis. In addition to the support for my thesis, the opportunity to work with you as a teacher assistant for the Auto Design I, II, and III courses during the 2005 and 2006 academic years has been exceptional. The position not only provided me with the funds to live during the time, but enabled me to further develop my leadership, team work, and vehicle design skills. The time I spent with the FSAE program is invaluable to my personal and professional development. To my other committee members, Dr. David Thompson, and Dr. Jay Kim, I truly appreciate your assistance and input enabling me to complete this thesis. To my colleagues in SDRL and Bearcat Motorsports, your academic support and assistance in reviewing, and suggesting changes to improve the quality of my thesis are greatly appreciated. I wish all of you the best of success in the remainder of your academic and professional careers. Thanks to the Bearcat Motorsports organization for the use of the engine and chassis dynamometers and to the 2005 and 2006 team members for their assistance in tuning and operating the dynamometers, especially to Dave Moster, Jeff Kenney, and Greg Curlin for their time spent helping develop and test my thesis. Last and most important, to my family, especially my wife, Stephanie, I could not have completed this thesis and masters degree without your endless love and support.

iii

Table of Contents TABLE OF CONTENTS .......................................................................................................iv LIST OF FIGURES.................................................................................................................v LIST OF TABLES..................................................................................................................vi LIST OF ACRONYMS.........................................................................................................vii CHAPTER 1: BACKGROUND.............................................................................................1 1.1 PURPOSE OF DYNAMOMETERS ........................................................................................1 1.2 HISTORY OF DYNAMOMETERS ........................................................................................3 1.3 WATER BRAKE DYNAMOMETERS ...................................................................................6 1.4 ELECTRICAL DYNAMOMETERS .....................................................................................10 CHAPTER 2: UNIVERSITY OF CINCINNATI’S FSAE DYNAMOMETERS............15 2.1 NINE INCH LAND AND SEA WATER BRAKE ABSORBER ................................................15 2.2 MD-95 MUSTANG CHASSIS DYNAMOMETER ...............................................................20 CHAPTER 3: COUPLING THE ENGINE TO THE CHASSIS DYNAMOMETER ....24 3.1 DIRECT DRIVE DESIGN .................................................................................................26 3.2 GEARBOX DESIGN .........................................................................................................30 3.3 CHAIN AND SPROCKETS DESIGN ...................................................................................32 CHAPTER 4: OPERATING THE CHAIN AND SPROCKETS DESIGN .....................37 4.1 INITIAL BREAK-IN .........................................................................................................37 4.2 TUNING ..........................................................................................................................38 4.3 OTHER ISSUES ...............................................................................................................42 4.4 THE CORRECT EDDY CURRENT ABSORBER .................................................................47 CHAPTER 5: CONCLUSIONS AND FUTURE RECOMMENDATIONS ....................50 5.1 CONCLUSIONS................................................................................................................50 5.2 FUTURE RECOMMENDATIONS .......................................................................................54 REFERENCES ......................................................................................................................57 APPENDIX A: DYNAMOMETER COMPARISON .......................................................59 APPENDIX B: MD-95 CHASSIS DYNAMOMETER SPECIFICATION.....................61 APPENDIX C: TELMA CC 80 RETARDER SPECIFICATIONS ................................62 APPENDIX D: F4I REDUCTION/SPEED TABLES .......................................................69

iv

List of Figures FIGURE 1: ROPE BRAKE (A) PRONY BRAKE (B).................................................................... 3 FIGURE 2: COMPUTER AND REAL-TIME CONTROL SYSTEM DYNAMOMETER ...................... 5 FIGURE 3: WATER BRAKE DYNAMOMETER CROSS-SECTION ............................................... 6 FIGURE 4: TYPICAL PERFORMANCE CURVES FOR A WATER BRAKE DYNAMOMETER .......... 9 FIGURE 5: PERFORMANCE CURVES: DC OR AC DYNAMOMETERS...................................... 11 FIGURE 6: EDDY CURRENT DYNAMOMETER CROSS SECTION AND END VIEW ................... 13 FIGURE 7: PERFORMANCE CURVES FOR EDDY CURRENT DYNAMOMETER ......................... 14 FIGURE 8: OUTER CASING, INTERNAL ROTOR, BEARING, AND SEAL ................................. 15 FIGURE 9: PICTURE OF THE WATER BRAKE BEARINGS, LEVER ARM, AND LOAD CELL ..... 17 FIGURE 10: NINE INCH WATER BRAKE PERFORMANCE CURVES VS. MEASURED CURVES . 18 FIGURE 11: 2005 FSAE CAR ON MD-95 CHASSIS DYNAMOMETER ................................... 22 FIGURE 12: ACTUAL 2005 TORQUE CURVE VS. DESIRED TORQUE CURVE ......................... 24 FIGURE 13: TOP AND REAR VIEW OF DIRECT DRIVE POTENTIAL SETUP ............................ 27 FIGURE 14: ISO VIEW OF DIRECT DRIVE POTENTIAL SETUP ............................................. 28 FIGURE 15: TOP, FRONT, AND ISO VIEW OF GEARBOX POTENTIAL SETUP ........................ 31 FIGURE 16: TOP, FRONT, AND ISO VIEW OF CHAIN AND SPROCKETS SETUP ..................... 32 FIGURE 17: SPLINED ADAPTOR AND SPROCKET ................................................................. 33 FIGURE 18: CHAIN ROUTING .............................................................................................. 35 FIGURE 19: CHAIN TENSIONING BOLTS ............................................................................. 35 FIGURE 20: CHAIN ANCHOR ............................................................................................... 37 FIGURE 21: PERFORMANCE ELECTRONICS FUEL TABLE, 0 TO 12,000 RPM....................... 42 FIGURE 22: ARROW ON ROTOR .......................................................................................... 43 FIGURE 23: LEVER ARM ALTERATION ............................................................................... 46 FIGURE 24: K-40 PERFORMANCE CURVES VS. MEASURED CURVES................................... 48 FIGURE 25: CC 80 PERFORMANCE CURVES VS. MEASURED CURVES ................................. 49

v

List of Tables TABLE 1: WATER BRAKE PERFORMANCE CURVE SEGMENT DESCRIPTIONS ........................ 9 TABLE 2: DC OR AC PERFORMANCE CURVE SEGMENT DESCRIPTIONS ............................. 11 TABLE 3: EDDY CURRENT PERFORMANCE CURVE SEGMENT DESCRIPTIONS ..................... 14 TABLE 4: METHODS FOR DETERMINING EQUIVALENT INERTIA IN POUNDS ....................... 21 TABLE 5: MAJOR PROS AND CONS OF EACH DESIGN .......................................................... 36 TABLE 6: ROLL TO ENGINE RPM CONVERSION VALUES ................................................... 39 TABLE 7: LEVER ARM MODIFICATION ............................................................................... 46

vi

List of Acronyms AC – Alternating Current AFR – Air to Fuel Ratio DC – Direct Current ECU – Engine Control Unit FSAE – Formula Society of Engineers MPH – Miles per Hour RPM – Revolutions per Minute TP – Throttle Position VAC – Volts Alternating Current

vii

Chapter 1: Background 1.1 Purpose of Dynamometers

The main purpose of a dynamometer is to absorb energy. This energy is delivered in a rotational form and is transferred from the energy source using a driveshaft or other type of mechanical system. To absorb energy, the rotation must have resistance. If instrumentation is utilized that measures the rotational speed (tachometer) and the resistance force (load cell) at a distance from the center of rotation, many parameters related to the energy source can be calculated. Revolutions per minute (RPM), torque, and horsepower are three significant parameters. The RPM can be determined directly from the tachometer and any relevant gear reductions. The torque can be determined by the load cell value and the distance from the center of rotation. The raw horsepower is linearly related to RPM and torque. To get comparative values of horsepower, corrections for temperature and humidity must be taken into account. Engine, chassis, and shock dynamometers are the most common types used today. An engine dynamometer is used to tune or test an internal combustion engine. A chassis dynamometer can be used in the same fashion; however it is capable of also testing driveline performance, mileage accumulation, and many other characteristics. This is the most popular use and will be the background for this thesis. The average person will never need to use or own a dynamometer. However, to any person or organization that is concerned with making modifications to their vehicle or engine in order to produce more power, a dynamometer provides a simple way to determine how the modifications improve performance. A simple five to ten second pull on a dynamometer before and

1

after the modifications are made, will be able to answer that question. This information is crucial to the avid weekend racer and to professional race teams world wide. Engine dynamometers are not only useful for determining the benefit of upgrading engine components, but also for generating an entire timing and fuel map for the vehicle’s Engine Control Unit (ECU). Most engine dynamometers are capable of holding a motor at a certain speed, utilizing a feedback control process, while modifications are made to the ignition and fuel map to either produce a desired air-fuel-ratio (AFR) or to achieve the maximum amount of torque and horsepower. Original equipment manufacturers, race teams, and after market manufactures specifically will conduct this type of tuning and testing on their engine systems on an engine dynamometer. As the name implies, an engine dynamometer is used to tune just the engine, no transmission or other driveline components are used, meaning the parasitic losses are at a minimum. Once an ignition and fuel map are generated and the motor is installed into the automobile, a chassis dynamometer can be used to determine many other parameters: efficiency of the drive train system, estimated fuel economy, emissions, acceleration, and reliability to name a few. Dynamometers are very important tools that are used for many different types of applications related to the internal combustion engine. Without the use of dynamometers, the advances in engine performance would not be as simple to test and/or verify the overall effect that an added component contributes.

2

1.2 History of Dynamometers

There are many different types of dynamometers that are used around the world today including hydraulic, electrical, frictional, air brake, and hydrostatic. Different variations of some of the dynamometers are available. Hybrids have also been developed; one example is an electrical and hydraulic dynamometer connected together in series. One of the earliest type of dynamometer to be developed, called a rope brake, relied solely on friction between a rope and a drum. The drum was attached to the power source and the frictional force was regulated by adding or subtracting weights. See Figure 1a below. [1] This dynamometer was invented and used in the earlier part of the 19th century. “Its successor, the Prony brake, also relied on mechanical friction and like the rope brake required cooling by water introduced into the hollow brake drum and removed by a scoop.”[1] Figure 1b shows a typical Prony brake. [2]

a

b

Figure 1: Rope Brake (a) Prony Brake (b) The first water brake dynamometer was developed by William Froude in 1877. Within four years of the invention, “Heenan and Froude was established and produced the first commercial dynamometers.” In 1952 the first rolling test rigs, an early form of chassis dynamometers, were produced. In 1976, “Consine Dynamics [supplied] the first

3

direct digital controlled, DC Chassis Dynamometer with full road load simulation. In 1983, Consine Dynamics joined Froude Engineering Ltd. to form Froude Consine. This merger expanded the company product range to include ‘state of the art’ vehicle test chassis dynamometers with digital controllers.” [3] In 1985, a United States Patent was published that described the invention of an eddy current absorption unit that was to be used to measure torque from an engine. [4] The Dynojet Research company invented the “first single roller, inertia, chassis Dynamometer for motorcycles in 1989” [5] Modern advances in technology, including the use of computers and feedback control systems, have significantly impacted the capabilities of dynamometers. The computers provide a method of interaction between the operator and the dynamometer. Parameters that are desired, typically horsepower, torque, speed, AFR, plus many more are easily displayed, logged, and saved to a file for easy accessibility. Operational parameters can also be inputted into the computer and in return will interact with the control system that regulates the dynamometer. For example, if there is a particular speed that the operator needs to maintain while a certain engine parameter, such as fuel and/or timing is changed, that speed can be entered into the computer and would become an operating parameter for the control system of the dynamometer. The software that communicates with the dynamometer’s control system varies from each manufacturer. However, most have the same general concept and once trained on one, other company’s software should be easy to learn. Control systems that are used for operating the dynamometers have been developed over the past 20 years and are available in both open and closed loop form. “In openloop mode, the dynamometer control is set to a percentage of available dynamometer

4

output or load. In this mode, the resulting load is independent of throttle position (TP), RPM, or vehicle speed.” [5] When operating in closed loop mode, which is more commonly used, a constant speed can be maintained during changes in TP. When this happens, the load that is applied to the dynamometer is changing with the change in TP. The closed loop control system monitors the load and the speed. It counteracts the change in TP by increasing or decreasing the amount of load necessary to maintain the desired speed. The computer and control systems can be used on most types of dynamometers. However, the effect of the computer and control system depends on the type of dynamometer that is used. The diagram below shows a chassis dynamometer that utilizes a computer and real-time control system during operation. [6]

Figure 2: Computer and Real-Time Control System Dynamometer

5

1.3 Water Brake Dynamometers

The hydraulic dynamometer, more commonly known as the water brake, uses water, a water tight casing, and a rotor that spins inside of the casing. The casing and rotor represent a pump; however, since the goal of a dynamometer is to absorb energy, they are designed to be inefficient water pumps. The energy absorbed is converted into heat, therefore heating the water inside the casing. The water brake is available in various different configurations, three of which are described in more detail below. A typical cross section of a water brake is shown below in Figure 3. [7]

Figure 3: Water Brake Dynamometer Cross-Section The first and most popular water brake dynamometer is commonly referred to as a variable fill machine. As the name implies, the load is controlled by changing the amount of water that is inside of the casing. The change is typically controlled with a valve on the inlet and a separate valve on the outlet side. Needle valves are used instead of ball valves due to their ability to change the flow rate in minute increments. Loading is slowly changed by opening or closing the inlet valve, and quickly changed by opening or closing the outlet valve. This allows small changes in torque resistance by only

6

changing the inlet valve. This process is fairly simple and can be done manually by a person or controlled by a computer to hold back the desired amount of torque. A second type of water brake dynamometer is called the constant fill machine, or “the classical Froude or sluice plate design”. [1] With this machine, the load is not changed by varying the amount of water, but by inserting pairs of thin plates between the rotor and casing. This reduces the clearance between the rotor and casing therefore increasing the amount of torque that can be absorbed. The opposite will occur if you remove a pair of the sluice plates. Each setup is not capable of controlling a large variation in torque, and to change the amount of torque that it can handle, the unit must be disassembled, and then reassembled with more plates added, or removed. This is a tedious, manual process that could take a significant amount of time. A third type of water brake dynamometer is called a disc dynamometer. The loading on this machine is controlled by a combination of plates and the amount of water inside of the casing, similar to the variable fill machine described above. The small clearance between the plates results in intensive shearing of the water which will resist the applied torque, and by changing the amount of water with the needle valves; more or less torque can be absorbed. A small variation in this machine is to have perforated discs instead of solid discs. This will enable the machine to absorb more torque. Each setup is not capable of controlling a large variation in torque; however it is more adjustable than the constant fill machine due to the variable amount of water in the casing. Similar to the constant fill machine, disassembly and reassembly with inserted plates are required to change the range of torque that the machine can absorb. Again, this is tedious and time consuming.

7

Water brake dynamometers are considered a low cost machine when compared to other types of dynamometers that are capable of absorbing the same amount of torque. Water brake dynamometers are available in many sizes. Water brake absorbers are available for a shifter cart, which is a high performance go cart that can produce about 30 horsepower, as well as a diesel engine that can produce 2,000 horsepower or more. Due to this factor, the cost of a water brake dynamometer is going to vary significantly based on the horsepower requirements. The water brake and a few required components for a shifter cart cost nearly $4,000 [8] and a diesel engine water brake cost nearly $36,000. [9] Other instrumentation, which is usually required for accurate and informed tuning, will add additional costs to these prices. As with any type of machine, the water brake dynamometer has its limits. Depending on the particular unit that is purchased, it has a maximum operating speed, and a maximum amount of power absorption that it can withstand. These two factors also limit the amount of torque that it can resist. Horsepower is linearly related to speed and torque as shown in the following equation:

HP =

Torque * RPM 5252

Equation 1: Equation relating horsepower, torque (ft-lbf), and RPM Based on this equation and testing conducted on individual units, a performance curve can be developed so that the operating capability of the absorber is identified. An example of a water brake’s performance curves can be seen below in Figure 4 with a corresponding description of each line segment in the following table. [1] From this figure, it is noticeable that the water brake dynamometer is inefficient at low speeds. It is

8

difficult for this type of machine to absorb torque and horsepower at roughly 40% of its maximum speed. This presents problems for tuning engines in the lower RPM ranges.

Figure 4: Typical Performance Curves for a Water Brake Dynamometer Line Segment a b c d e

Description Dynamometer full. Torque increases with square of speed. Performance limited by maximum permitted shaft torque. Performance limited by maximum permitted power. Maximum permitted speed. Minimum torque corresponding to minimum permitted water flow.

Table 1: Water Brake Performance Curve Segment Descriptions Another inefficiency that is important to keep in mind is the slow response of changing the load. Compared to electrically controlled dynamometers, water brakes require 10-100 times longer to respond to a 90% change in load. See Appendix A for the difference between each electrical dynamometer. [10] This is really only a problem when you do not have a smooth torque curve. For example, when tuning from scratch, or if an intake is poorly designed such that a large dip or peak occurs, the load cannot be changed quickly enough to keep the engine from running away from you or nearly stalling out. This problem can result in longer time spent tuning areas of the ECU map where the torque increases or decreases rapidly.

9

1.4 Electrical Dynamometers

As the classification implies, an electrical dynamometer’s load is controlled by electricity. Computers and control systems for these dynamometers are necessary to accurately control the electricity supplied to the machine. Electrical dynamometers can be one of three types: Alternating Current (AC), Direct Current (DC), or Eddy Current. Each electrical dynamometer has its advantages and disadvantages which will be discussed below. See Appendix A for a thorough comparison. [10] A DC dynamometer is comprised of a DC motor generator that is mounted to a substructure along the axis of rotation. The motor’s outer casing is free to pivot as a means to transmit resistance torque to a load cell attached to a lever arm. “The rotational speed of a DC motor is proportional to the voltage applied to it, and the torque is proportional to the current.” [11] Voltage is linerly related to current and resistance per Ohm’s Law. V = I *R

Equation 2: Ohm's Law According to Ohm’s Law, in order to change the torque, i.e. a change in current, the voltage must be changed. To change the voltage “an electronically-controlled switching device made of thyristors [or] transistors” [11] is used. The dynamometer’s control system, including the switching device allows for very quick changes in load. The DC dynamometer is capable of handling a 90% change in load 50 times faster than the water brake. A DC dynamometer also has the ability to act as a starter for an engine dynamometer. This applies to all engines that mount their starters to their transmission’s bellhousing instead of to the engines themselves. Another benefit of a DC dynamometer is its ability to produce torque down to zero RPM, also known as stall torque. An 10

example of a DC dynamometer’s performance curves can be see below in Figure 5 with a corresponding description of each line segment in the following table. [1] Compared to the water brake dynamometer, the DC dynamometer is capable of producing torque much lower in RPM, providing a broader operational envelope for engine tuning. However, the DC dynamometer does have its drawbacks as well. Including a high cost per horsepower capacity, a high rotational inertia, they are relatively large and do not operate well in higher RPM ranges compared to a water brake with the same power absorbtion capacity.

Figure 5: Performance curves: DC or AC Dynamometers Line Segment Description Constant torque corresponding to maximum current and excitation. a Performance limited by maximum permitted power. b Maximum permitted speed. c Table 2: DC or AC Performance Curve Segment Descriptions An AC dynamometer is very similar to the DC dynamometer with exception of the type of motor used and the method of controlling the load. An AC motor is used in place of the DC motor. “These asynchronous machines consist essentially of an induction motor, the speed of which is controlled by varying the supply frequency. The power supply comprises a rectifier, an intermediate DC circuit and an inverter to produce the variable frequency supply.” [1] The change from a DC motor to an AC motor is beneficial by reducing the inertia of the system. The AC dynamometer has higher RPM

11

capabilities and reacts to the 90% change in load twice as fast as the DC dynamometer, therefore 100 times faster than a water brake. Other than these differences, the pros and cons of the DC and AC dynamometers and performance curves are similar. An Eddy Current dynamometer’s load is not controlled by an electrical motor. However, it is still classified as an electrical dynamometer because it still utilizes electricity to control the operation. There are two types of eddy current machines. The first is cooled by ambient air, while the other is cooled with a closed-loop water system. As expected the air-cooled system is much simpler to install and to maintain, as well as less expensive due to the ‘extra’ equipment necessary to operate the water-cooled system. The air-cooled system is open to the atmosphere while the water-cooled system is encased similar to a water brake dynamometer. The load on both types is controlled in the same manner, which “makes use of the principle of electro-magnetic induction”. [1] A magnetic field that is parallel, but offset, to the axis of rotation is generated by a coil wrapped around a magnetic pole. These coils are stationary while one or two rotors, made of ferrous material, are rotated in close proximity of the coils. The rotor is attached permanently to the shaft that is connected to the power source. Figure 6 below shows a cross section and end view of an eddy current dynamometer where 40 and 50 are the rotors, 60 is the coils of wire, and 200 is electromagnets running parallel to rotation. [12]

12

Figure 6: Eddy Current Dynamometer Cross Section and End View When the coils are powered and the rotor is turning, “circulating eddy currents and the dissipation of power in the form of electrical resistive losses” [1] occur. The circulating eddy currents generate energy, i.e. heat, and as a result heats up the rotors and surrounding air. The load is controlled by changing the current that is passed through the coils. Like the DC and AC systems, the current is regulated with a control system that enables a response time to a 90% change in load that is 10 times faster than the water brake; however it is twice as slow as the DC machine, and 10 times slower than the AC machine. The eddy current machine is also capable of reaching a higher RPM and has a lower inertia compared to the other electrical dynamometers making it a more viable option for smaller displacement and horsepower applications that utilize a higher RPM range. It is also more affordable and smaller than the DC and AC dynamometers; however larger and more expensive than a water brake. Eddy current machines, like the water brakes, are not capable of acting as a starter for an engine; therefore an external starting source is necessary. The eddy current dynamometer is not capable of generating stall torque like the DC and AC machines, but it is significantly better than the water brake. A typical eddy current’s performance curves are shown in Figure 7 with a corresponding description of each line segment in the following table. [1] 13

Figure 7: Performance curves for Eddy Current Dynamometer Line Segment a b c d e

Description Low speed torque corresponding to maximum permitted excitation. Performance limited by maximum permitted shaft torque. Performance limited by maximum permitted power. Maximum permitted speed. Minimum torque corresponding to residual magnetization and friction.

Table 3: Eddy Current Performance Curve Segment Descriptions Electrical dynamometers are much more expensive, when compared to a similarly horsepower rated water brake. The eddy current is the cheapest of the three available, while the AC and DC are similar in price. For an absorber that can withstand 200 horsepower, a water brake system costs nearly $4,700 [13], while the eddy current costs $12,950 [14], and the AC costs $49,950 [15]. A water brake system capable of absorbing 1,600 horsepower costs $14,450 [16], while an AC dynamometer capable of absorbing 1,250 horsepower costs $539,500. [15] The prices are for kits that include various necessary components and software that enable the dynamometer to operate.

14

Chapter 2: University of Cincinnati’s FSAE Dynamometers

The University of Cincinnati’s FSAE team owns two different dynamometers. The first is a nine inch water brake absorber manufactured by Land and Sea that was purchased in 1998 to be adapted to an engine dynamometer that was built by the FSAE team. The second is an MD-95 Mustang Chassis Dynamometer that was purchased in 2003 to enable the FSAE team to utilize eddy current control to finalize the custom tuned maps that are generated on the water brake engine dynamometer.

2.1 Nine Inch Land and Sea Water Brake Absorber

The nine inch Land and Sea dynamometer is designed to be a fast responding toroidal flow water brake absorber. It is a variable fill type machine that utilizes public water supply that is regulated through manual load control valve purchased from Land and Sea. Both inlet and outlet valves are needle valves, which permit small increments in load change. The rotor inside of the casing is constructed of plastic and is “divided into pockets by radial vanes set at an angle to the axis of the rotor.” [1] The disassembled water brake is shown with one of the two outer casings in the following figure.

Figure 8: Outer Casing, Internal Rotor, Bearing, and Seal

15

The water brake is instrumented with a water temperature gauge incased in the outer housing of the dynamometer. This sensor provides the tuner with a reference point as to how hard the dynamometer is working. From experience, anything under 150ºF is considered a safe operating range; however most tuning can be completed below 125ºF. When tuning in the higher RPM range, 12,000 to 15,000, it is hard to apply enough load without obtaining a temperature between 125ºF and 150ºF. At and above 150ºF, the grease in the bearings of the absorber begin to liquefy and increasing the risk of damage. Sealed ball bearings are used in this system combined with radial axle seals. The inner seal on the bearing is removed and zerk fittings on the casing allow the bearing to be greased. The grease also acts as a seal for the bearings, so when elevated temperatures are maintained and grease is lost, it needs to be replaced or bearing and seal failure are inevitable. Attention to this temperature value is critical to the operation of the dynamometer. If the temperature continues to rise and reaches 212ºF or higher, there is no longer water in the absorber, it has turned to vapor. Vapor cannot resist torque, so the engine’s RPM will increase rapidly, which can damage the engine and the absorber. The nine inch absorber was purchased to enable the FSAE team to tune Yamaha FZR 600 motorcycle engines. It was later modified to work with the Honda CBR 600 F4 and F4i motorcycle engines. The transmissions on these motorcycle engines are built together with the engine as one unit. The water brake’s shaft was not connected to the crank like a normal engine dynamometer, but to the output of the transmission where the front sprocket is located. A front sprocket had an adapter plate welded to it that would be used to connect to a driveshaft.

16

The shaft that the rotor inside of the absorber rotates about is supported with two flange-mounted Dodge bearings. The end of that shaft has an adapter flange welded on that allows a driveshaft with two u-joints to be connected between it and the adaptor plate on the front sprocket. The speed is measured with an encoder wheel and Hall Effect sensor that is placed between the two bearings. The torque is measured with a load cell that is attached to a two foot lever arm which is connected to the outer casing of the water brake. The setup can be seen below in Figure 9. A data acquisition system is used to record these parameters, as well as exhaust gas, coolant, and water brake temperatures.

Figure 9: Picture of the Water Brake Bearings, Lever Arm, and Load Cell The engine dynamometer is built to tune motorcycle engines that have the engine and transmission integrated together. It is built specifically to tune a 1989-1995 Yamaha FZR 600cc four cylinder motor. The stock motor is capable of producing 76 horsepower at 10,000 RPM and 45 ft-lbs of torque at the crank at 8,250 RPM. [17] In this engine, 6th gear is the highest gear and when engaged a reduction of speed by 1.9, and multiplication of torque by 1.9. The rev limiter is set at 14,000 RPM, so with the reduction, the maximum RPM of the engine dynamometer during tuning would be about 7,300 RPM. 17

The peak horsepower is shifted to 5,263 RPM. The peak torque would be at 4,342 RPM and with the multiplication would rise from 45 to 85.4 ft-lbs. Like the Yamaha FZR 600cc motor, the Honda CBR 600cc F4i motor is a four cylinder engine that has an integrated transmission. The dimensions and component layouts are similar between the two, making the Honda CBR motor a viable replacement for the out-of-date Yamaha FZR motors. The stock Honda CBR motor is capable of producing 95.5 horsepower at 12,500 RPM and 42.6 ft-lbs of torque at the crank at 10,250 RPM. [18] Again 6th gear is the highest gear and when engaged it has a reduction of speed by 2.14, and multiplication of torque by 2.14. The rev limiter is set at 14,500 RPM, so with the reduction, the maximum RPM of the engine dynamometer during tuning would be about 6,775 RPM. The peak horsepower is shifted to 5,841 RPM. The peak torque would be at 4,789 RPM and with the multiplication would rise from 42.6 to 91.1 ft-lbs. These values, for both the FZR and CBR engines, are within the absorptive range of the performance curves for the Land & Sea Water Brake shown in Figure 10.

Figure 10: Nine Inch Water Brake Performance Curves vs. Measured Curves

18

Land & Sea recommends that the dynamometer should not be used for an application that is near the upper or lower limits of the particular absorber. The operating range that is utilized for tuning the FZR and F4i engines is in the lower limit of the nine inch water brake. However, the absorber works well from about 4,000 engine RPM up to redline, with the exception of areas in the torque curve where there is a sharp increase/decrease. The reason for the problem is the inability to change the load fast enough to compensate for the change in torque. The torque and horsepower curves that are generated by the FSAE tuned engine in 6th gear are shown in the Figure 10 as well. They nearly cross the capacity envelope of the nine inch absorber at nearly 2,000 RPM, which is approximately 4,000 engine RPM. This confirms the lack of control issues that are prevalent when trying to tune below 4,000 engine RPM. In September 2005 the FSAE team upgraded the water brake dynamometer by purchasing the auto-load control equipment available from Land & Sea. This additional equipment includes the auto-load control valve, the electrical components that monitor torque and speed, and a hand-held control interface that is used to input the desired speed and/or sweep rate. It was not installed during the school year for the 2006 FSAE car, but should be installed and tested in the fall of 2006 to determine if it functions well for this application.

19

2.2 MD-95 Mustang Chassis Dynamometer

The MD-95 Mustang Chassis Dynamometer is designed to be used for low horsepower and low speed testing and tuning. It utilizes an air-cooled eddy current brake that is capable of absorbing 150 horsepower at speeds up to 80 MPH continuous. It is also capable of reaching 100 MPH on an intermediate basis, and can measure 200 horsepower. The additional 50 horsepower can be attributed to the inertial acceleration of the dynamometers rollers. The speed ratings are limited by the bearings that support the rollers. The complete MD-95 specifications sheet can be viewed in Appendix B. A chassis dynamometer is used to measure the horsepower and torque generated at the wheels of a vehicle. This dynamometer uses cradle type rollers, instead of a single large roller, that are precision machined and dynamically balanced. There are four rolls total; one for the front and rear side of each drive tire. The rolls connected to the eddy current absorber are grooved to provide positive traction between the rolls and the tires. They are connected together with couplers between each roller and between the roller and absorber shaft. The other rolls are smooth and are not connected together. The dynamometer is rated at 600 pounds of base mechanical inertia. The eddy current absorber requires 230 VAC single phase to generate power to the coils. The speed is measured with an encoder wheel and Hall Effect type sensor that is connected to the shaft of one of the grooved rolls. The torque is measured with a S-type load cell on a lever arm connected to the body of the absorber. The base mechanical inertia is a phrase that chassis dynamometer manufacturers use to classify the size of the dynamometer and “is a function of the mass and inertia of all

20

the rotating elements of the dynamometer system.” [19] The manufacturers describe the ‘inertia’ of the dynamometers in an equivalent weight, in pounds, to eliminate confusion in the sales and marketing. The equivalent weight enables a customer to purchase a dynamometer based on the lightest vehicle that will be tested. The unit of pounds is much easier for these customers to understand, versus the actual rotational inertia of the system in lbf-ft-sec2. A dynamometer unit is designed to have a certain base mechanical inertia. Once built, the unit will be tested to determine the actual base mechanical inertia to be used in the software to calculate the horsepower associated with accelerating the inertia. Since the rotational inertia of the system can be calculated based on the design, the equivalent weight can be determined in two similar methods. These methods are shown in Table 4 below. First Method Relating Equation

Second Method

I=(W/g)*R2

Roll Radius (in)

RR

4.2875

Tire Radius (in)

RT

10

αR=T/I

((4.2875 / 12) * F) / 2.234

System Inertia (lbf-ft-sec2)

I

2.234

Roller Radius (ft)

R

Roll Angular 0.3573 Acceleration (rev/sec2)

Radius Constant

C=R2/g

Tire Angular 0.0040 Acceleration αT=(RR/RT)*αR (rev/sec2)

0.42875 * 0.1599 * F

Equivalent Weight (lbs)

W=I/C

Tire Linear 563.044 Acceleration (ft/sec2)

aT=RT*αT

10 * 0.06857 / 12 * F

F=maT

F=(W / g) * 0.057 * F

Force Balance g = 32.174 ft/sec2

Simplified Equation Equivalent Weight (lbs)

(F * g) / (0.057 * F) = W W=g/0.057

563.044

Table 4: Methods for Determining Equivalent Inertia in Pounds

21

According to Mustang’s recommendations, the weight of the lightest car to be tested is to be greater than the base mechanical inertia of the dynamometer. The MD-95 Chassis Dynamometer is useful to the FSAE team because the car built by the team is light weight, about 650 pounds with a driver, and produces nearly 90 horsepower. Since the dynamometer is built for full-size vehicles and has cradle rollers, the spacing between the rollers is too large to accommodate the car’s 20 inch diameter tires. The gap is reduced from 17.25 inches down to 11.625 inches by shifting the smooth rollers closer to the grooved rollers. The FSAE car is also narrow compared to a full-size vehicle. The unused portions of the rollers are covered to protect personnel when the dynamometer is operating. The chassis dynamometer is not placed down in a pit so a ramp is used to get the car off and on the dynamometer, as well as hold the front wheels elevated to the same height as the rear wheels during testing. The car is held in place with 2.5 inch wide heavy duty straps. There are two used in the rear and one in the center holding backwards, and one or two in the front holding down. The picture in Figure 11 below illustrates the setup.

Figure 11: 2005 FSAE car on MD-95 Chassis Dynamometer 22

The dynamometer is controlled using the Mustang supplied control panel and corresponding software. The MDSP 7000 Series Dynamometer Controller Software is utilized and has many testing and tuning options. An entire manual exists on Mustang’s website, and there are a few available around with the dynamometer, that is dedicated to describing all of the software’s capabilities. The testing that is generally conducted on the FSAE car utilizes the Constant Speed Test and the Power Curve. As the name Constant Speed Test implies, the car is held at a constant speed while the TP is changed

to allow for final tuning of the engine. The Power Curve test is used to generate horsepower and torque curves for the vehicle and can be controlled in two different methods. The first is titled the fixed-sweep-time mode and the second is titled the vehiclesimulation-loading mode. The fixed-sweep-time mode is conducted by entering in a

starting test speed, an ending test speed, and the duration of the test in seconds. The controller will allow the cars speed to increase one MPH at a time between the starting and the ending speed. The vehicle-simulation-loading mode is conducted by using the same values as entered in the fixed-sweep-time mode in addition to the vehicle’s weight, the power at 50 (MPH), and the simulated inertia. This test will randomly change the resistance on the dynamometer to vary the speed between the starting and ending speed. Both tests are typically operated under full, not partial, throttle. This provides the tuner with the maximum amount of horsepower and torque that the car can generate. Throttle changes should not be made during testing because it will provide the tuner with corrupt information about the actual performance of the car.

23

Chapter 3: Coupling the Engine to the Chassis Dynamometer

The troubles experienced with the water brake engine dynamometer during tuning for the 2005 FSAE car exposed a weakness of the team. This weakness was the long amount of time that it takes to tune the engine. The intake and exhaust system that was built for the car was designed to place the peak torque at 8,000 RPM. The design did not take into consideration torque elsewhere because wave and Helmholtz equations were used to determine the intake runner length. The resulting torque that was generated from the built intake did place the peak torque near 8,000 RPM; however, in the RPM range between 3,700 and 6,700 the torque curve is uneven, as shown in the following torque curve in Figure 12.

Figure 12: Actual 2005 Torque Curve vs. Desired Torque Curve At 3,700 RPM the engine produces around 32 ft-lbs of torque. At 5,100 RPM it is down to nearly 30.5 ft-lbs and then back up to 32 ft-lbs at 5,700 RPM. After that it drops

24

off sharply to 26 ft-lbs at 6,700 RPM and then increases quickly up to 36 ft-lbs at 8,200 RPM. The dip that occurs between 5,700 and 8,200 RPM made tuning the engine very difficult with the water brake dynamometer. The steep increase and decrease in torque would cause the engine either to run away from the operator or nearly stall out and die. After numerous hours were spent trying to tune the 5,700 to 8,200 RPM range of the map, it was decided that the map was as complete as possible with the water brake. The map would be finessed on the chassis dynamometer once the car was operational. When driven for the first time, it was evident that there was a large dip in the torque curve. When taking off from a stop and accelerating the car would seem to bog down until the RPM would reach about 7,500 and then the car would accelerate very fast due to the drastic increase in torque. The car was then placed on the chassis dynamometer hoping that it would be able to maintain the desired speed to keep the engine from running away from the tuners. The engines RPM was held at 6,500 in 3rd gear allowing the tuners to try to work with the fuel and ignition map to fill in sections of the dip. During this testing, the driver’s back is only about four inches away from the header and the radiator. Complaints were made regarding the heat after four to six minutes of testing. Once the car had cooled down, tuning continued on the map where the torque curve was steep. The chassis dynamometer was capable of holding the car’s speed where desired, allowing the tuners to quickly make final adjustments to the map in the locations that the water brake was not capable of maintaining. Small improvements were made in these portions of the map; however in the end, the map still had the large dip which could not be fixed with tuning.

25

With the ability to quickly tune the sections of the map on the chassis dynamometer where it was difficult with the water brake, the desire arose to utilize the chassis dynamometer as an engine dynamometer. The problem and origination of this thesis is how to design and integrate the output of the transmission on the engine to the eddy current absorber. Even though the project would result in the integration of the two dynamometers, the need to quickly operate each system independently was still required. In order to preserve this ability of independent operation, serious modifications to both the engine and chassis dynamometers were not practical. Restricted modifications were those that either permanently disabled one of the systems or required a significant amount of time to reassemble to its original configuration. This included repositioning the components’ locations on the engine dynamometer and the removal of one or more of the chassis dynamometer’s rollers.

3.1 Direct Drive Design

Solid models of both the chassis and engine dynamometer were generated in Solid Edge®. With the above limitations in mind, the investigation began by maneuvering the solid models of each system in an assembly to determine possible configurations that would result in parallel alignment of the transmission shaft and the eddy current absorber. The first proposed design yielded axial alignment of the transmission shaft and the roller at the opposite end of eddy current absorber on the chassis dynamometer. The bottom of the chassis dynamometer frame was supported by 3.5 inch square blocks. This places the

26

center line of the chassis dynamometer’s rollers and eddy current absorber about 10.75 inches above the ground. As built, the engine dynamometer places the center line of the transmission shaft at nearly 18.25 inches above the ground. The sub-frame that holds the engine on the engine dynamometer frame is held in place with four bolts. It is easily removed once the four bolts holding the driveshaft to the transmission is detached. When removed and placed directly on the ground, the center line of the transmission shaft would be at 10.75 inch, which would line up perfectly with the chassis dynamometer’s rollers. An adaptor flange would need to be manufactured to enable the use of the existing driveshaft utilized for the water brake system. A hole would need to be cut in the end of the chassis dynamometer frame to accommodate the driveshaft. A Solid Edge® assembly was generated to show how this potential design would be setup. Pictures from the assembly are shown in Figure 13 to help visualize this setup.

Figure 13: Top and Rear View of Direct Drive Potential Setup A bracket to prevent relative motion between the chassis dynamometer frame and the engine sub-frame would need to be built and allow linear adjustment in both vertical and front-to-rear directions and rotational adjustment to enable parallel alignment. Fine

27

adjustment would be necessary to prevent misalignment, therefore vibration, of the output shaft and the end of the roller. It would be difficult to create a simple bracket that would permit this type of adjustment. An additional picture, shown in Figure 14, shows a close-up of the design. The engine sub-frame sits below the chassis dynamometer frame, which complicates the installation of a simple bracket for adjusting the rotational alignment. A significant amount of design work would be necessary to develop this bracket, in addition to manufacturing and installation.

Figure 14: ISO View of Direct Drive Potential Setup However, in order to accomplish this setup other considerations had to be taken into account. Since the engine sub-frame would be removed from the engine dynamometer frame, the existing wiring and data acquisition harnesses would not work, as well as the radiator. Adding length to both wiring harnesses and routing new coolant lines would be simple and relatively inexpensive. The extra wiring could be wound up and the old

28

coolant lines reinstalled when returned to the original configuration. A more expensive option, though not as logical, would be to use a separate radiator and separate harnesses for the proposed design, leaving the engine dynamometer untouched except for the removal of the driveshaft and the four bolts holding down the sub-frame. Another concern is the high rotational speeds that would be prevalent with the direct drive of the eddy current brake. The chassis dynamometer’s maximum continuous speed is 80 MPH. With the 8.575 inch diameter roller, this correlates to a maximum of 3,135 RPM. This maximum RPM limits the continuous operational RPM range in each gear. The entire engine RPM range can be tuned in 1st gear. In 2nd gear the engine can only reach about 11,500 RPM, and in 3rd only about 9,000 RPM. A spreadsheet that compares the crank’s RPM to the transmission’s output shaft RPM in each gear is located in Appendix D. Though this design seems practical, sustained high speeds and the necessity to remove the engine sub-frame, which also required additional wiring and coolant system modifications, prompted the need for a different design. The new design would simplify the process of switching back and forth between each engine dynamometer setup and reduce the high rotational speeds.

29

3.2 Gearbox Design

After further study, a second design was developed that counteracted the issues that arose from the first design. Instead of directly connecting the transmission shaft and the roller, a gear box would be utilized to reduce the rotational speed. Commercially available units were deemed too expensive and in most cases too large for the power capacity required. The FSAE team had spare parts from old engines, so the gears that were used in the transmission for 2nd gear, which is a 2.063 reduction, were used to develop a small, simple gearbox that would be manufactured. In this design, the connection between the output of the transmission and the end of the roller involved a few steps. Instead of removing the engine sub-frame and the hassle of the wiring and cooling components, the removal of the water brake from its shaft would be necessary. This can be done with a bolt and wrench in about five minutes. The water brake would be replaced with a coupler that would connect to the input shaft of the gearbox. The gearbox would be secured to the water brake sub-frame at a location that would place the output of the gearbox at a 12.75 inch height. This is the same height as the rollers with the chassis dynamometer sitting on top of with 3.5×5.5 inch blocks, with the 5.5 inch dimension in the vertical direction. The existing casters would need to be replaced with shorter ones, with shims to provide small height adjustment. A similar bracket, as described in the previous design, would be needed to restrain the engine dynamometer frame and still allow adjustment. Again, similar problems exist with this bracket that was present in the Direct Drive design, though a less elaborate design would work because the frames are closer together. Adaptor flanges would need to be manufactured for the

30

output shaft of the gearbox and the end of the roller. Another driveshaft with two Ujoints would need to be built to connect the system together. A Solid Edge® assembly was generated to show how this potential design would be setup. Pictures from the assembly are shown below in Figure 15 to help visualize this setup. Notice that this configuration utilizes the main engine dynamometer frame and therefore the existing radiator and all electrical components.

Figure 15: Top, Front, and ISO View of Gearbox Potential Setup Compared to the first design, instead of only being able to tune throughout the entire RPM range in 1st gear, now it is possible in every gear except for 6th. Also, other than the simple removal of the water brake from the shaft, no modifications to the engine dynamometer would be necessary. The Gear Box design met the desired goals, however in the process, the complexity increased significantly and the time commitment required in manufacturing all of the one-off components increased. Again, a design that was simpler was desired.

31

3.3 Chain and Sprockets Design

A third idea, which was intended to reduce the amount of manufacturing time, was to replace the gearbox with a chain and sprocket. This third idea was short lived because it prompted the final design that was built. The third design is the simplest and most feasible design that was developed. As stated, a chain and two sprockets are utilized in a similar method as on a motorcycle. A stock front sprocket would be installed on the output shaft of the transmission. A rear sprocket would be connected to the one of the grooved rollers on the chassis dynamometer. The Solid Edge® assembly pictures in Figure 16 shows the setup.

Figure 16: Top, Front, and ISO View of Chain and Sprockets Setup

32

To connect the rear sprocket to the roller a couple of ideas were investigated. The first was to bolt the sprocket to the end of the roller. This placed the roller and chain really close to the bearing support for the roller. A second design involved an adaptor that bolted to the end of the roller, but spaced the sprocket away from the bearing support. The third idea, which was utilized, used the uniform grooves on the roller as a spline. A ¾ inch thick steel adaptor is used that has a water jetted spline pattern on the inside diameter to hold the sprocket in position. The sprocket is held to the adaptor with six 5/16 inch bolts. The roll to sprocket adaptor is shown in Figure 17.

Figure 17: Splined Adaptor and Sprocket The adaptor is split into two halves. There are two reasons for splitting the adaptor. The first, and most important, is to prevent the sprocket from moving. A bolt on each side is used to hold the halves together once installed over the grooved rollers. When tightened down, the sprocket does not move. The second is to allow for alignment adjustment. When the two bolts are loose, the adaptor can be shifted along the grooved roller axially to facilitate the planar alignment of the front and rear sprocket. Also, when loose, the sprocket can be checked for trueness. A dial indicator is used to verify that the sprocket is turning true in the axial direction, meaning no wobbling, and can be adjusted with a soft tipped hammer. The radial trueness is adjusted by loosening the six bolts that 33

hold the sprocket to the adaptor. The holes in the sprocket are oversized to 3/8” to allow for this adjustment. The trueness is adjusted with shims and is verified with a human eye and a reference point. Once adjusted, the bolts are retightened and the trueness is checked again in both directions. A long straight edge is used to verify that the rear sprocket and front sprocket are in the same plane. The closer the sprockets are to being in the same plane results in a longer life of the chain and sprockets. However, like the axial alignment required in the other designs, this does not have to be perfect and is capable of withstanding minor misalignment. To make this design work, no modifications to the engine dynamometer are necessary. However, the upper and lower scatter shields that surround the driveshaft must be removed. The driveshaft, which is held in place with four bolts on each end, must also be removed. Next, the engine needs to be slid as far backwards in the adjustment grooves as possible. Finally, the rubber mounts between the engine subframe and the main dynamometer frame need to be removed and hard mounted. All of which can be done in about twenty-five to thirty minutes. The adaptor that is used to connect the transmission shaft to the driveshaft is replaced with a stock front sprocket. An opening in the back of the chassis dynamometer frame is used to run the lower section of the chain through. The cutout was located incorrectly when it was made, but was fixed by elevating the chassis dynamometer frame up an additional two inches. It is now supported with 3.5×5.5 inch blocks, with the 5.5 inch dimension in the vertical direction. The top of the chain is high enough that it clears the top side of the chassis dynamometer frame shown in Figure 18. Instead of using similar brackets that were discussed in the other two designs, two ½ inch bolts are used. See Figure 19 below. Not only does this

34

connect the two dynamometers together, but also permits linear adjustment needed to tension the chain. Small angular adjustment is also possible which enables parallel alignment of the rollers and output shaft of the transmission.

Figure 18: Chain Routing

Figure 19: Chain Tensioning Bolts

A 55 tooth sprocket made for a 525 pitch chain purchased from Sprocket Specialists is utilized. This is the smallest rear sprocket that would fit around the roller and splined adaptor without any interference issues. The front sprocket is from the stock Honda CBR F4i engine and has 16 teeth. This sprocket combination results in a 3.438 reduction. This large reduction significantly reduces the high rotational speeds and enables the freedom to tune the engine in any of the six gears. In comparison to the direct drive design, instead of obtaining the 3,135 RPM at about 9,000 engine RPM in 3rd gear, now the speed is only 912 RPM. However, in comparison to the direct drive and the gearbox designs, even with proper maintenance, the chain will stretch and the teeth on the sprockets will wear out. In the other two designs, the driveshafts have u-joints which are lubricated with grease and would last significantly longer than the chain and sprockets.

35

The goal for this project was to determine a method of connecting the University of Cincinnati’s engine dynamometer to the chassis dynamometer in a design that was quick and simple. The final design that was used is just that. To provide an overview of the benefits and drawbacks of each design that was researched, the following table was generated. Direct Drive Design Pros Cons Small amount of manufacturing

High rotational speeds

Cheap

75 minute change over time

Simple

Misalignment issues

Gearbox Design Pros Cons Slower A lot of rotational manufacturing speeds 60 minute Misalignment change over issues time

Chain and Sprockets Design Pros Cons Very little manufacturing

Chain breakage

30 minute change over time

Expensive

Complicated

Simple

Chain and Sprockets Wear Out

Expensive

Misalignment is not a big issue Much slower rotational speeds

Table 5: Major Pros and Cons of each Design

36

Chapter 4: Operating the Chain and Sprockets Design

4.1 Initial Break-In

The first time that the Chain and Sprockets design was run; four issues immediately arose that needed attention. The first involved the rubber mounts from which the engine sub-frame is mounted. After noticing the engine and sub-frame moving while loaded with the eddy current, the rubber mounts were removed to create a solid mount to prevent undesired movement. Once this problem was solved, another one developed that was similar in fashion. Since the rubber mounts were removed and no elastic translation was available, the energy was then used to lift the front of the engine dynamometer off of the ground. The logical fix for this issue was to anchor the front side of the engine dynamometer to the ground. A solid anchor was considered, however due to the necessity to adjust the slack in the drive chain, a flexible or adjustable anchor was required. Turn buckles, which permit adjustment, were researched and deemed inconvenient due to available lengths with appropriate load ratings. Two links of chain were used instead. One end was anchored to the ground while the other was bolted in tension from the bottom up allowing for adjustment. One of the chain anchors is shown in Figure 20.

37

Figure 20: Chain anchor

Another problem arose when the drive chain repeatedly required tightening. Upon further investigation, slotted holes were found on the engine sub-frame that enabled adjustment for the alignment of the water brake. The bolts were loosened and the engine was slid as far back as it could go in the slots, then tightened back down. The fourth problem was the tight and loose spots in the chain. By process of elimination, the rear sprocket was deemed responsible. The sprocket was visually out-ofround when the roller was spinning. To fix the problem, the bolt holes in the sprocket were oversized, and shims were inserted between the roller and the sprocket. The reason behind the problem is one of two, or a combination, issues. The bolt holes in the sprocket are not concentric with itself, or the bolt holes on the adaptor are not concentric with the roller. Neither was investigated further since the problem was significantly reduced. There are still minor tight and loose spots in the chain, but they do not affect the performance of the dynamometer. Once these problems were repaired, no other issues developed until later in the tuning process.

4.2 Tuning

The goals when retuning the 2005 intake and exhaust systems were to get familiarized with the new setup and to obtain a better understanding of the Mustang Dynamometer software. Before any accurate measurements were taken, the software’s parameters had to be corrected for the new setup. The first parameter that had to be changed was the roll diameter.

To determine the torque applied to the roller, the software uses the diameter of the roller at which the tires apply the force. With the Chain and Sprockets design, the rear sprocket is larger than the roller. Thus the diameter that the force is applied is increased from the 8.575 inch roller diameter to the 10.948 inch pitch diameter of the 55 tooth 525 rear sprocket. Also, since no tires are used and the gear ratios and final reduction ratio are all known, the engine RPM can be accurately calculated from the measured eddy current RPM using the Roll to Engine RPM Conversion. The correct value, shown in Table 6 for each gear, must be entered based on which gear is engaged during tuning. 1st 17.748

2nd 12.919

3rd 10.317

4th 8.901

5th 7.972

6th 7.353

Table 6: Roll to Engine RPM Conversion Values The final parameter that needs to be changed is the equivalent weight, or inertia. With the Chain and Sprockets setup, a weight of 400 pounds is used. Any significant value higher than this, such as 450 or more, will result in the dynamometer surging at the locations of the map where the torque increases sharply. The same happens if the weight is decreased drastically. This occurs because the design only utilizes two of the four rolls; therefore the inertia of the chassis dynamometer is smaller than the 600 pounds of base mechanical inertia. Without the two idle rollers, the Solid Edge® model of the system estimates that the equivalent inertia should be 390 pounds. This is near the 400 pounds that is physically used during tuning. Other nearby values, ± 25 pounds, can be tested to see if surging is reduced further, but 400 pounds appears to work well. Once these values are changed, the software is setup and ready for tuning to begin. The engine is typically started in neutral and allowed time to reach normal operating temperature. Once warm, the engine is shutoff and placed into gear. Nearly all of the

39

tuning is conducted in 3rd gear. As stated in Chapter 2 under the Mustang Dynamometer section, the Constant Speed Test is used. This test is started and conducted in the following manner. A starting speed is entered. In 3rd gear this correlates to about twelve MPH. Prior to an additional tuner releasing the clutch, the computer operator will press the Start Test button and provide a corresponding increase in throttle. The clutch is then released and the tuning process is underway. The Constant Speed Test allows the tuner to step up or down by one MPH increments. This correlates to a 312 RPM change at the crank in 3rd gear. Since the fuel and ignition tables are incremented by 800 RPM (12,000 maximum RPM setup), an change in two, or three MPH is necessary to advance/retard to the next column of cells. The large final reduction that the Chain and Sprockets design utilizes results in really slow rotational speed. Experience has shown that the eddy current dynamometer reacts to changes in load better when it is rotating faster. Thus, to enable tuners the ability to easily tune the lower RPM range of the engine, 6th gear is used from about 1,800 to 4,000 RPM. In 6th gear, a one MPH increment correlates to a 227 RPM change at the crank. Again, since the fuel and ignition tables are incremented by 800 RPM an change in three or four MPH is necessary to advance/retard to the next column of cells in 6th gear. While tuning in 3rd gear, the sharp incline in torque caused the eddy current dynamometer to surge. When the weight parameter was decreased to 400 pounds, the surging lessened, but was still prevalent. Knowing that the dynamometer reacts better at faster speeds prompted tuning this area of the map in 6th gear as well. This results in a 40% increase in speed, as well as a 29% decrease in torque that the eddy current must absorb. The eddy current speed never surpasses 1,700 RPM in 6th gear, which is nearly

40

half of the 3,135 RPM top speed. Instead of producing 484 ft-lbs peak torque, the engine now produces 345 ft-lbs peak torque. The higher speed and lower torque significantly reduced the surging that was occurring in 3rd gear and further simplified tuning in the difficult portion of the map. As mentioned previously, eddy current dynamometers get hot when they are used continuously. To assist in maintaining cool conditions, fans are used along with some homemade ductwork directed at the eddy current absorber. To try to prevent overheating, the tuning is limited to about eight minutes in the 2,000 to 7,000 RPM range and lower loading conditions. It is also limited to about four minutes at when exceeding 7,000 RPM and higher loading conditions. This is a general guideline for tuning with the eddy current absorber, however first hand exposure will provide a better understanding. Experience has shown that the dynamometer does get to a point where it is too hot and will not be able to maintain the desired speed. Obviously, when this occurs it is time to stop tuning and allow the eddy current rotors to cool down. When tuning is to be stopped it is recommended that the speed be decreased while still running the Constant Speed Test. Reduce the throttle to about 20% and bring down the speed to about twelve MPH

before the clutch is disengaged. This enables the dynamometer to stop faster and as a result prevents the chain and transmission from coasting to a stop from high speeds. The cooling time will vary, depending on how hot the eddy current brake got. In the case where it gets too hot to hold the desired speed, it should sit for at least 45 minutes with fans blowing on the absorber. In other instances, 30-35 minutes is sufficient.

41

4.3 Other Issues

As tuning continued, other issues developed including a serious vibration that would occur above 11,000 engine RPM in 3rd gear, which corresponds to about 34 MPH. However, the vibration is not prevalent in 6th gear at 34 MPH. It appears that a resonance occurs between the engine, engine dynamometer, and chassis dynamometer near 11,000 engine RPM. The main cause of the vibration is unknown. However, the chassis dynamometer is not anchored to the concrete, and this is suspected to be the root cause of the problem. Once a final layout is determined for the Center Hill facility, the chassis dynamometer can be anchored down to determine its affect on the vibration. Unfortunately, the vibration prevented tuning the engine for the 2006 FSAE car above 11,200 RPM using the Chain and Sprockets design. The ECU’s fuel and ignition maps ended with two columns of cells at 11,200 and 12,000 RPM. An example of the fuel table used in the Performance Electronics software is shown in Figure 21 below.

Figure 21: Performance Electronics Fuel Table, 0 to 12,000 RPM The peak torque had been achieved at 8,500 RPM, and the effect of the restrictor was linearly reducing the torque at about 1.2 ft-lbs per 500 RPM. To enable the completion of the table, i.e. the 11,200 and 12,000 RPM columns, an interpolation was performed to

42

generate the values. Once completed, a quick sweep of the 11,200 RPM column was conducted to verify that the AFR was acceptable. Changes that were made resulted in corresponding interpolation changes to the 12,000 RPM column. The plan was to later confirm these values and to continue tuning out to 14,000 RPM on the water brake dynamometer. This was deemed unnecessary and a waste of time due to data that showed that while on the course, drivers typically shifted before getting to 10,750 RPM. Another problem was discovered after the eddy current absorber got red hot after an extended tuning period. There were arrows cast into the eddy current rotors indicating the intended rotational direction, shown in Figure 22. The Chain and Sprockets design rotated the rollers/rotors in the opposite direction. As installed in the Center Hill facility, the only way that a car can be tested on the chassis dynamometer also resulted in the backwards rotation of the rollers. The Chain and Sprockets design was based on the assumption that the rollers would rotate in the same direction that a FSAE car on the chassis dynamometer would make them rotate. However, this assumption was incorrect due to the unintentional installation error. The Figure 22: Arrow on Rotor MD-95 spec sheet was compared to the MD-100 Mustang Dynamometer spec sheet. The MD-95 spec sheet, shown in Appendix B, said that the dynamometer is uni-directional and the front rolls are not coupled to rear rolls. The MD-100 spec sheet said that the dynamometer is belted for bi-directional capability.

43

Mustang Dynamometer was contacted to verify that no damage was caused to the eddy current absorber by rotating it in the opposite direction. Technical support verified that the only difference that makes the MD-95 uni-directional and the MD-100 bi-directional is the belt that connects the front and rear rolls. Another small problem that occurred twice was the master link clip breaking. Luckily the chain never separated sparing catastrophic damage to the engine. The cause of this is unknown; however the problem could possibly lie in the chain hitting the diamond plate roll cover and chassis dynamometer frame. This typically occurs at times when running really rich and other times throughout the tuning process. If this is found to be the cause a piece, of rubber screwed to the roll cover and dynamometer frame could help. Always check to make sure the master link clip is still attached after each run. An additional problem that occurred involved exceeding the maximum output of the load cell. This had been occurring since the tuning had begun, but was not evident to the tuners. The testing and tuning of the 2005 intake and exhaust setup, as well as the tuning for the rebuilt 2004 intake and exhaust setup were all conducted during this time. The torque curves for these setups had an interesting flat spot between 8,000 and 11,200 RPM that did not exceed 36 ft-lbs. Once tuning began for the 2006 engine and the same flat spot occurred, it was predicted that the restrictor was prohibiting anymore power from being produced. At one point in time, the surging in 3rd gear required the use of 6th gear to tune the 8,000 RPM column. The maximum engine torque of 36 ft-lbs in 3rd gear all of a sudden was 45 ft-lbs in 6th gear. The torque curves from each gear agreed except where the torque superseded 36 ft-lbs. This prompted research into what was causing this inconsistency. The trace viewer enables the tuner to choose which parameters to display

44

in a graph format. The force measured by the load cell did not surpass 355 lbs. The lever arm was manually loaded, which confirmed this maximum output. The measured length of the lever arm is one foot. This results in a maximum torque measurement of 355 ftlbs. With the Chain and Sprockets design, and an approximated 45 ft-lbs of torque at the crank, the eddy current brake would need to be able to hold back 464 ft-lbs in 3rd gear. In 6th gear, the eddy current brake would need to be able to hold back 331 ft-lbs. Since this is lower than the 355 ft-lbs rated capacity, the dynamometer was able to provide the tuners with the correct engine torque in 6th gear and not in 3rd gear. To fix the problem, a couple of options were available. The first was to change the ratio of the sprockets used in the Chain and Sprockets design. The ratio would need to decrease, and since the rear sprocket was as small as possible without causing interference, then the front sprocket had to be larger. The largest front sprocket available for the F4i engine has 17 teeth. This would make the reduction ratio change from 3.438 to 3.235. Though helpful (464 ft-lbs would be reduced to 436 ft-lbs), it would not be enough to still tune in 3rd gear. A second option was to tune everything in 6th gear. This was possible, but not desirable due to the vibration that was prevalent and the higher rotational speeds. The third and final option was to extend the lever arm on the eddy current absorber. Since the absorber was apparently capable of holding back the torque generated by the engine and existing reduction, this was deemed the viable solution. The original lever arm was twelve inches long. The only way to accommodate a longer lever arm, in a short time frame was to extend it to 21.25 inches. This length enabled the load cell to be directly mounted to the main frame of the chassis dynamometer with minimal work, and therefore increased the ft-lb measurement capability of the load cell as shown

45

in Table 7 below. Figure 23 shows this modification, in addition to the two posts that were welded to the new lever arm to allow calibration weights to be easily installed. They were welded at 15 inches from the center of the absorber, and resulted in a more accurate calibration than the previous method.

Original Modified

Lever Arm Length (ft) 1 1.77

Maximum Measurable Torque (ft-lbs) 355 628.6

Table 7: Lever Arm Modification Figure 23: Lever Arm Alteration Since the engine and the corresponding reduction were producing enough torque to overload the load cell, a concern developed that the absorber was being overloaded as well. The following question was asked: Why would a load cell that is only be capable of measuring 355 pounds be attached to a twelve inch lever arm, if the absorber can hold back more than that? At the point in time when this occurred, the performance curves for the eddy current absorber were not acquired. (See next section for further details) Therefore, the only information available was the maximum 150 horsepower that Mustang specified on the MD-95 spec sheet. The engine was not capable of producing more than 95 horsepower, which is about 63% of the rated horsepower absorption. Taking this into consideration, as well as the fact that the dynamometer had been holding back the applied load for a couple of hours of testing, proved to be enough information to continue tuning without damaging the eddy current.

46

4.4 The Correct Eddy Current Absorber

The eddy current absorber that is utilized in the MD-95 Mustang Chassis Dynamometer is built by Telma, Model CC 80. It has a total of sixteen coils, eight per rotor which are powered by 24 volts, and has a maximum braking torque of 589 ft-lbs. It has a maximum rotational speed of 4,500 RPM. These and other technical specifications can be found in Appendix C. These specifications explicitly provide the performance curves for the CC 80. When compared to the specifications from Mustang for the MD-95, they are quite different. The MD-95 specs out 150 horsepower absorption, while the CC 80 specs out over 450 horsepower. The MD-95 spec sheet also states that it is capable of measuring 200 horsepower. The chassis dynamometer is capable of measuring more than it is capable of absorbing due to the inertia of the rollers that are accelerated during testing. Despite that, the difference between the maximum horsepower absorption, and therefore torque absorption, is substantial. According to Mustang, the MD-95 was equipped with a K-40 Klam absorber. However, prior to providing this information, they admitted that in the past poor records were kept. When the performance curves for the K-40 were received, shown below in Figure 24, there was an obvious discrepancy between the horsepower and torque that was actually absorbed by the eddy current during tuning.

47

Figure 24: K-40 Performance Curves vs. Measured Curves Further research found that the dimensions and the number of coils for the K-40, compared to the actual eddy current absorber, were drastically different as well. Again, Mustang was contacted. This time the technical representative said that the eddy current absorber was probably a K-70, not a K-40. This answer was presented without any additional investigation. The dimensions and performance curves for the K-70 were much more consistent with the actually measurements. However, the ‘guess’ provoked a more detailed search of the eddy current absorber to try to find an identification tag. A tag was found under the lever arm mounting plate. The information was completely different than what was previously researched. Instead of a Klam absorber, it said that it was manufactured by Telma. The information on this tag and the chassis dynamometer tag can be found in Appendix B. Telma was contacted directly to obtain the performance curves for their CC 80 model. The performance curves and the measured horsepower and torque curves from 3rd and 6th gear can be seen below in Figure 25.

48

Figure 25: CC 80 Performance Curves vs. Measured Curves

The graph in Figure 25 depicts an important comparison. In 3rd gear, the measured torque never surpasses 75% of the available torque. In 6th gear, it never surpasses 53%. Since 3rd gear uses a higher percentage of the available torque, the dynamometer will become hotter faster when compared to tuning in 6th gear. Another important factor that is displayed in this graph is the comparison of the curves generated by 3rd gear and 6th gear. The 6th gear curve is stretched along the x-axis and compressed along the y-axis compared to the 3rd gear curve. This is caused by the increased reduction by tuning in 6th gear versus 3rd gear. It is easier for the software to control the speed and corresponding load when the rate of change in torque is decreased. This is evident during tuning due to the decrease in the dynamometer surging in the RPM range where the slope is the largest.

49

Chapter 5: Conclusions and Future Recommendations

5.1 Conclusions

When it comes to motorsports, everyone wants to know how well the engine in the vehicle performs. The all important horsepower and torque are what the average, everyday person wants to know when it comes to the vehicle’s performance. They want to be able to see how a bolt-on part/kit adds to the performance of the engine. In more cases today, they want to be able to see how changing the fuel and ignition table on their stand alone ECU affects their horsepower and torque. The larger the number, the faster they can get from point A to point B. Companies that manufacture, sell, and market high performance parts for the car enthusiast all compete for the most gain in horsepower and torque for the money. Race teams are particular interested in obtaining the absolute best horsepower and torque curves possible. With the combination of a well-performing engine, and a wellperforming suspension, a racecar should dominate in its race. In order to produce these optimal curves, engine tuning is a necessity. As presented, there are many choices for the type of dynamometer that can be used for this process. Each one has its benefits and drawbacks, including limitations on speed, the ability to quickly change the required load, and the initial expense. Typically, the more expensive, the better the dynamometer is overall. Thus, when it comes to choosing a dynamometer, the price is always a large part of the consideration.

50

Another very important aspect is the control system that regulates the speed and load of the dynamometer. A feedback control system, though expensive, significantly reduces the amount of time and effort required to tune an engine. Since time equals money, the payback period should be reduced with the purchase of the control system. The University of Cincinnati’s FSAE team is interested in saving as much time as possible. Time does not necessarily equal money for this organization. However, the less amount of time spent tuning the engine for the FSAE car results in more practice time and chances for under-designed components to break prior to attending the collegiac competition. Using the manual controlled water brake dynamometer to tune the ECU’s fuel and ignition map has historically prevented the car from finishing at an early enough date. The purchase of the Mustang Chassis Dynamometer controlled with an eddy current absorber and the corresponding MDSP 7000 Series Dynamometer Controller Software provided the team with the benefit of a better dynamometer. However, the

downside to tuning on the chassis dynamometer is the requirement for the car to be operational. Engine tuning needs to be ongoing while the car is built, not started once the car is completed. In addition to this, the time spent on the chassis dynamometer is risky for the driver. The 1,200-1,500ºF header is within inches of the drivers back and is only isolated from the cockpit by a piece of reflective insulation adhered to a sheet of aluminum. Also, if the engine were to self-destruct during tuning, the driver is directly inline with the rotational parts that tend to fail and could be injured. The front tires of the car are elevated and supported with a ramp. If the straps were to break, or something similar, the driver could be injured. These reasons are more than enough justification to restrict this method of complete map tuning from the FSAE program.

51

Since the eddy current dynamometer’s load and speed control capabilities are very desirable, and it is unrealistic to utilize it continuously in the chassis dynamometer form, an engine dynamometer setup would be advantageous. This would eliminate the risks that a driver would be exposed to, as well as eliminate the unnecessary wear and tear on the cars driveline and tires. To obtain an engine dynamometer setup in combination with the chassis dynamometer, various designs were investigated. The Chain and Sprockets design showed the most promise and simplicity when integrated into the chassis dynamometer. The design only moderately disables both the engine and chassis dynamometer. When connected together with this design, the chassis dynamometer can be ready to run in as little as two minutes, by simply removing the chain. The engine dynamometer, utilizing the water brake instead of the eddy current Chain and Sprockets design, can be ready to run in about thirty minutes. First, this requires the removal of the chain and front sprocket and then the reinstallation of the lower driveshaft scatter shield and front sprocket adaptor. Next, the engine must be moved forward to its water brake operating position. Once forward, the rubber mounts can be reinserted between the engine subframe and the main dynamometer frame and bolted down. Finally, the driveshaft can be reconnected and the upper scatter shield attached prior to hooking up the water supply. Table 5 provides an overview of the pros and cons of each design that was researched. Along with the short changeover time discussed above, the Chain and Sprockets design has important advantages, including misalignment and reduced rotational speeds when compared to the other two designs. The chain can handle minor misalignment, and the sprockets provide the speed reduction. The simplistic bolted

52

connection is also beneficial. No brackets are necessary to enable angular alignment which is required for the Direct Drive and Gear Box designs. A small amount of manufacturing is required. The splined adaptor and sprocket are the only items that require any amount of manufacturing. Even though minor problems occurred with the initial proposed design, small additions or modifications quickly corrected the issues. Once continuous tuning began and other problems developed, more changes were necessary to enable accurate tuning. With the exception of the vibration problem and the cause of the master link clips breaking, each problem was eliminated. The vibration at the higher RPM ranges did not prevent the FSAE team from developing a well tuned ECU map. Nearly all of the engine’s power is below the RPM range in which the vibration is prevalent. In the area where vibration was a problem, interpolation and in-car testing/confirmation were sufficient in completing the fuel map. Dealing with the technical representatives from Mustang can be complicated. They have a tendency to be vague and do not appear to know what eddy current is installed in the MD-95 Chassis Dynamometer that the FSAE team owns. The K-40, manufactured by Klam, was their educated guess. If the chassis dynamometer had a K-40 installed in it, then it would have never been able to hold back the torque that was generated. When the specs did not match for the K-40, the K-70 was the next assumption. This too was found to be incorrect when an identification tag was found on the bottom of the eddy current absorber and identified it as a CC 80 manufactured by Telma. The dimensions, number of coils, and other specs for this absorber matched the absorber in the chassis dynamometer. The performance curves also encased the measured horsepower and

53

torque curves. The absorber is appropriately sized for this application. The upper or lower limits of its capabilities are exceeded during the tuning process. The eddy current enables the tuner to hold the engine at a slow speed. The water brake is inefficient at lower speeds. With the nine inch absorber and in 6th gear, the output speed of the transmission is slow compared to the limits of the absorber. This situation makes it very difficult to control the load. The eddy current, though close to its limits, at slow speeds significantly improves the ability to tune the bottom end of the map.

5.2 Future Recommendations

The addition of the Chain and Sprockets design, which utilizes the eddy current dynamometer, to the University of Cincinnati’s FSAE program significantly, reduced the amount of time required to tune the engine for the team’s cars. However, for the 2006 car’s engine, no tuning was attempted over 11,200 RPM. The chassis and engine dynamometer would vibrate excessively, and no attempt was made to supersede this speed. If the chassis dynamometer were to be anchored to the ground the vibration might be reduced. It was not anchored after it was moved to accommodate the engine dynamometer’s location for the Chain and Sprockets design. The Center Hill facility is to be rearranged and a new location for the chassis dynamometer is still undetermined. If the permanent location has not been finalized by the tuning period for the 2007 FSAE car, the chassis dynamometer should be anchored down to the concrete regardless.

54

Since vibration is a problem with the Chain and Sprockets setup at 11,000 RPM and higher, the water brake should be used at these speeds. At the higher speeds the water brake is advantageous because the torque changes slowly. The automatic load control that was purchased for the nine inch water brake needs to be installed and tested. If the auto load control is integrated into the water brake system and functions as advertised, it will be extremely beneficial. It should reduce the level of expertise and time that is required to operate the manual control valve to maintain the engine’s RPM. Rubber padding needs to be installed to keep the chain from hitting the diamond plate covers and the chassis dynamometer frame. It is believed that this will keep the chain master link clip from breaking. Always have a few extra around in the event that the master link clip does break. After each tuning run, check to make sure that the clip is still present on the master link. If it continues to be a problem, the rivet type, versus the clip type, master link may be necessary. This will complicate reconnecting the chain and sprockets because the chain becomes continuous once riveted together. The engine dynamometer frame will need to be moved closer to the chassis dynamometer using the chain tensioning bolts. This will allow the chain to be installed around the front sprocket, which can then be slid onto the transmission output shaft. Another recommendation is to install some type of temperature probe near one of the rotors on the eddy current absorber. This will allow tuners to observe the temperature that the rotors are emitting, and determine what temperature is safe. For example, if the eddy current brake cannot hold a constant speed at 1000ºF, then they can monitor the temperature and when it reaches 850ºF or so, tuning can be stopped to allow it to cool down.

55

To simplify the process of returning the engine to its proper location in the engine dynamometer sub-frame, some form of adjustable hard mount needs to be added. This was going to be done when returned to the water brake setup, but it has not been used in that fashion since tuning began with the Chain and Sprockets design. The re-alignment of the driveshaft when setting up the water brake dynamometer is described in the following steps. Locate the marks/indentations where the bolts have been tightened down in the past. Slide the engine forward in the slots until the bolts line up with the marks. Once close, connect the driveshaft between the transmission output shaft and the water brake absorber. Place a dial indicator on the driveshaft between the two u-joints, as close to the engine as possible. Continue adjusting the engines position until the dial indicator reads true. Determine some method, probably a bolt and jam nut, which will provide a positive stop for moving the engine forward anymore. Before anything is welded, disconnect the ECU. Finally, a new clutch system needs to be installed. In order to operate the clutch in its existing location, operators climb over the chassis dynamometer. This is a tripping and falling hazard, and in an emergency is difficult to access.

56

References [1]

Plint, Michael and Martyr, Anthony (1999) Engine Testing Theory and Practice. 2nd edition. Woburn, Butterworth-Heinemann.

[2]

The University of Texas at Austin, Department of Mechanical Engineering. www.me.utexas.edu. ME 224L: Home-Made Prony Brake. http://www.me.utexas.edu/~lotario/me244L/labs/pmdc/pronybrake.html

[3]

Froude Hofmann. www.froudehofmann.com. About Us: History. http://www.froudehofmann.com/about_1.htm

[4]

Free Patents Online. www.freepatentsonline.com. Eddy Current Patent. http://www.freepatentsonline.com/4509374.html

[5]

Dynojet Research. www.dynojet.com. About Us: History. http://www.dynojet.com/about_us/index.php

[6]

National Instruments. www.zone.ni.com. Achieve Flexibility in Your Automotive Dynamometer Applications. http://zone.ni.com/devzone/conceptd.nsf/webmain/D1D95AF0DB0AC49E86256 CCA00514200

[7]

Wikipedia. www.wikipedia.org. Dynamometer. http://en.wikipedia.org/wiki/Dynamometer

[8]

Land and Sea. www.land-and-sea.com. Kart Engine Dynamometer – Pricing. http://www.land-and-sea.com/kart-dyno/kart-dyno-price.htm

[9]

Land and Sea. www.land-and-sea.com. Diesel Dynamometer – Pricing. http://www.land-and-sea.com/diesel-dynamometer/diesel-dynamometer-price.htm

57

[10]

Land and Sea. www.land-and-sea.com. Dynamometer Comparison. http://www.land-and-sea.com/dynamometer/dynamometer-comparison.htm

[11]

Wikipedia. www.wikipedia.org. Electric Motor. http://en.wikipedia.org/wiki/Electric_motor

[12]

Free Patents Online. www.freepatentsonline.com. Eddy Current Patent. http://www.freepatentsonline.com/4937483.html

[13]

Land and Sea. www.land-and-sea.com. PWC Dynamometers - Pricing. http://www.land-and-sea.com/pwc-dyno/pwc-dyno-price.htm

[14]

Land and Sea. www.land-and-sea.com. Eddy Current Brakes - Pricing. http://www.land-and-sea.com/eddy-current-dynamometer/eddy-currentdynamometer-price.htm

[15]

Land and Sea. www.land-and-sea.com. A/C Drives and Brakes - Pricing. http://www.land-and-sea.com/ac_dynamometer/ac-dynamometer-price.htm

[16]

Land and Sea. www.land-and-sea.com. Auto Engine Dynamometers - Pricing. http://www.land-and-sea.com/dyno/dyno-price.htm

[17]

Heath Agdog. http://heath.agdog.com. FZR – FZR Specifications. http://heath.agdog.com/fzr/specifications/

[18]

Sport Rider. http://www.sportrider.com. Dynamometer Charts – Honda 2001 Cbr600F4i. http://www.sportrider.com/bikes/street_bike_dyno_charts

[19]

Mustang Dynamometers. http://www.mustangdyne.com. Contact Us – FAQ. http://www.mustangdyne.com/faq.htm

58

Appendix A: Dynamometer Comparison [10]

Feature

Water Eddy Inertia Brake Current Rollers Absorbers Absorbers

Friction Brakes

Hydraulic AC Motor- DC MotorPumps Generators Generators

High Hp (steady state) capacity?

Excellent

n/a*

Good

Poor (cooling required

Fair (cooler required

Fair

Fair

High RPM capability?

Excellent

n/a

Fair

Good

Poor

Fair

Poor

Low RPM torque (steady state) capacity?

Fair

n/a*

Very Good

Excellent

Excellent

Excellent

Excellent

Stall torque (0 RPM range) test capability?

Poor

Fair (startup only)

Poor

Good

Good

Stability of RPM load control?

Good

n/a*

Very Good

Poor

Fair

Excellent

Very Good

n/a*

Very Good (less than 0.05 seconds)

Poor (less than 1.0 seconds)

Fair (less than 0.75 seconds)

Excellent (less than 0.005 seconds)

Very Good (less than 0.01 seconds)

Typical Good response time (less than 0.5 to 90% load seconds) change?

Excellent Excellent (with encoder) (with encoder)

Testing simulation under computer control?

Good

n/a*

Excellent

Fair

Not typically available

Excellent

Very Good

Motoring capability?

requires separate motor

requires separate motor

requires separate motor

requires separate motor

requires separate motor

Excellent

Excellent

Suitability to long duration testing?

Excellent

n/a*

Fair

Fair (water cooled)

Good (with cooler)

Excellent

Good

Ease of engine starting?

requires separate starter

requires separate starter

requires separate starter

requires separate starter

requires separate starter

Excellent

Excellent

Hp capacity vs. weight?

Excellent

Poor

Fair

Fair

Fair

Poor

Poor

Hp capacity vs. size?

Excellent

Excellent

Fair

Good

Good

Fair

Poor

Portability of absorption unit

Excellent

Determined by weight

Fair

Fair

Good

Poor

Poor

59

RPM, Torque, & Hp Accuracy?

Determined Determined Determined by Data by Data by Data Acquisition Acquisition Acquisition

Determined by Data Acquisition

Determined by Data Acquisition

Determined by Data Acquisition

Determined by Data Acquisition

Data Repeatability?

Determined Determined Determined by Data by Data by Data Acquisition Acquisition Acquisition

Determined by Data Acquisition

Determined by Data Acquisition

Determined by Data Acquisition

Determined by Data Acquisition

Resistance to hysteresis (breakaway friction)

Good

Excellent

Good

Good

Poor

Good

Fair

Immunity to untrunioned parasitic drag losses

Excellent (if mounted directly)

Poor

Good

Good

Fair

Excellent (with in-line transducer)

Excellent (with in-line transducer)

Affordability per Hp (steady state) capacity?

Excellent

n/a*

Good

Fair

Fair

Poor

Poor

Installation affordability?

Excellent (if water is available)

Fair (if pit or lift required)

Very Good (115v AC required)

Good (but cooler required)

Good (but cooler required)

Good (electrician required)

Good (electrician required)

Affordability of maintenance?

Excellent

Excellent

Excellent

Fair

Good

Excellent

Good

60

Appendix B: MD-95 Chassis Dynamometer Specification

Horsepower Maximum Speed Loading Inertia Axle Weight Controls Rolls

Frame Lift Air requirement Power Req. Notes

200-hp maximum measurement capability 150-hp maximum absorption 100 mph intermittent 80 mph continuous Air cooled eddy current power absorber 600 lbs base mechanical inertia 6,000 lbs (2,727 kg) maximum Pentium based PC control system, MD-7000 Control Platform Precision machined and dynamically balanced Front roll set may be grooved horizontally 8.5" diameter balanced rolls 41" face length 26" inner track width 108" outer track width 17.25" roll spacing Heavy-duty structural steel frame Between roll lift with integrated lock 80-100 PSI dry, regulated, oil free 115 VAC single phase, 60 Hz, 15 amps (computer) 230 VAC single phase, 60 HZ, 30 amps (dyno controls) Vehicle restraint kit required Dynamometer is uni-directional Front rolls are not coupled to rear rolls No additional mechanical inertia

Eddy Current Absorber Identification Plate Mfg Telma Mfg in Made In France Model # Type CC 80 Voltage V 24 NOC105996 Code Nº 10 2 104

Mustang Dynamometer Identification Plate Mfg Model # Md-95 Mfg Serial # 18320 Mfg Date Feb-03 Mystery # 6369

NOTE: It is believed that the horsepower ratings provided in the MD-95 spec sheet are incorrect. The actual horsepower capacity can be seen in the performance curves in section 4.4.

Appendix C: Telma CC 80 Retarder Specifications

62

63

64

65

66

67

68

Appendix D: F4i Reduction/Speed Tables Main Out Reduction

Engine Speed (rpm) 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000 7500 8000 8500 9000 9500 10000 10500 11000 11500 12000 12500 13000 13500 14000 14500

Crank 45 82 1.822

1st 12 34 2.833

2nd 16 33 2.063

3rd 17 28 1.647

4th 19 27 1.421

5th 22 28 1.273

6th 23 27 1.174

Final 16 55 3.438

Output Transmission Speed (rpm) 1st 96.84 193.69 290.53 387.37 484.22 581.06 677.91 774.75 871.59 968.44 1065.28 1162.12 1258.97 1355.81 1452.65 1549.50 1646.34 1743.19 1840.03 1936.87 2033.72 2130.56 2227.40 2324.25 2421.09 2517.93 2614.78 2711.62 2808.46

2nd 133.04 266.08 399.11 532.15 665.19 798.23 931.26 1064.30 1197.34 1330.38 1463.41 1596.45 1729.49 1862.53 1995.57 2128.60 2261.64 2394.68 2527.72 2660.75 2793.79 2926.83 3059.87 3192.90 3325.94 3458.98 3592.02 3725.06 3858.09

3rd 166.59 333.19 499.78 666.38 832.97 999.56 1166.16 1332.75 1499.35 1665.94 1832.53 1999.13 2165.72 2332.32 2498.91 2665.51 2832.10 2998.69 3165.29 3331.88 3498.48 3665.07 3831.66 3998.26 4164.85 4331.45 4498.04 4664.63 4831.23

69

4th 193.09 386.18 579.27 772.36 965.45 1158.54 1351.63 1544.72 1737.80 1930.89 2123.98 2317.07 2510.16 2703.25 2896.34 3089.43 3282.52 3475.61 3668.70 3861.79 4054.88 4247.97 4441.06 4634.15 4827.24 5020.33 5213.41 5406.50 5599.59

5th 215.59 431.18 646.78 862.37 1077.96 1293.55 1509.15 1724.74 1940.33 2155.92 2371.52 2587.11 2802.70 3018.29 3233.89 3449.48 3665.07 3880.66 4096.25 4311.85 4527.44 4743.03 4958.62 5174.22 5389.81 5605.40 5820.99 6036.59 6252.18

6th 233.74 467.48 701.22 934.96 1168.70 1402.44 1636.18 1869.92 2103.66 2337.40 2571.14 2804.88 3038.62 3272.36 3506.10 3739.84 3973.58 4207.32 4441.06 4674.80 4908.54 5142.28 5376.02 5609.76 5843.50 6077.24 6310.98 6544.72 6778.46

Engine Speed (rpm) 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000 7500 8000 8500 9000 9500 10000 10500 11000 11500 12000 12500 13000 13500 14000 14500

Corresponding Wheel/Roll Speed (rpm) 1st 28.17 56.35 84.52 112.69 140.86 169.04 197.21 225.38 253.55 281.73 309.90 338.07 366.24 394.42 422.59 450.76 478.94 507.11 535.28 563.45 591.63 619.80 647.97 676.14 704.32 732.49 760.66 788.84 817.01

2nd 38.70 77.40 116.11 154.81 193.51 232.21 270.91 309.61 348.32 387.02 425.72 464.42 503.12 541.83 580.53 619.23 657.93 696.63 735.34 774.04 812.74 851.44 890.14 928.84 967.55 1006.25 1044.95 1083.65 1122.35

3rd 48.46 96.93 145.39 193.85 242.32 290.78 339.25 387.71 436.17 484.64 533.10 581.56 630.03 678.49 726.96 775.42 823.88 872.35 920.81 969.27 1017.74 1066.20 1114.67 1163.13 1211.59 1260.06 1308.52 1356.98 1405.45

70

4th 56.17 112.34 168.51 224.69 280.86 337.03 393.20 449.37 505.54 561.71 617.89 674.06 730.23 786.40 842.57 898.74 954.92 1011.09 1067.26 1123.43 1179.60 1235.77 1291.94 1348.12 1404.29 1460.46 1516.63 1572.80 1628.97

5th 62.72 125.44 188.15 250.87 313.59 376.31 439.02 501.74 564.46 627.18 689.90 752.61 815.33 878.05 940.77 1003.48 1066.20 1128.92 1191.64 1254.36 1317.07 1379.79 1442.51 1505.23 1567.94 1630.66 1693.38 1756.10 1818.82

6th 68.00 135.99 203.99 271.99 339.99 407.98 475.98 543.98 611.97 679.97 747.97 815.96 883.96 951.96 1019.96 1087.95 1155.95 1223.95 1291.94 1359.94 1427.94 1495.93 1563.93 1631.93 1699.93 1767.92 1835.92 1903.92 1971.91