2005-01-0025 - On the Use of a Honda 600cc 4-Cylinder Engine for Formula SAE Competition

SAVE OFFLINE

2005-01-0025 SAE TECHNICAL PAPER SERIES

On the Use of a Honda 600cc 4-Cylinder Engine for Formula SAE Competition Mario Farrugia, Mike Rossey and Brian P. Sangeorzan Oakland University

Reprinted From: Electronic Engine Controls 2005 (SP-1975)

2005 SAE World Congress Detroit, Michigan April 11-14, 2005 400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A. Tel: (724) 776-4841 Fax: (724) 776-5760 Web: www.sae.org

The Engineering Meetings Board has approved this paper for publication. It has successfully completed SAE’s peer review process under the supervision of the session organizer. This process requires a minimum of three (3) reviews by industry experts. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of SAE. For permission and licensing requests contact: SAE Permissions 400 Commonwealth Drive Warrendale, PA 15096-0001-USA Email: [email protected] Tel: 724-772-4028 Fax: 724-772-4891

For multiple print copies contact: SAE Customer Service Tel: 877-606-7323 (inside USA and Canada) Tel: 724-776-4970 (outside USA) Fax: 724-776-1615 Email: [email protected] ISSN 0148-7191 Copyright © 2005 SAE International Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE. The author is solely responsible for the content of the paper. A process is available by which discussions will be printed with the paper if it is published in SAE Transactions. Persons wishing to submit papers to be considered for presentation or publication by SAE should send the manuscript or a 300 word abstract to Secretary, Engineering Meetings Board, SAE. Printed in USA

2005-01-0025

On the Use of a Honda 600cc 4-Cylinder Engine for Formula SAE Competition Mario Farrugia, Mike Rossey and Brian P. Sangeorzan Oakland University Copyright © 2005 SAE International

ABSTRACT The Formula SAE® rules require the use of a 20mm intake restrictor. The presence of the restrictor necessitates the design or retuning of fuel and spark strategies that, in turn require the use of a programmable engine control unit (ECU). This paper describes a process used to establish the fuel and spark strategies for a standard production motorcycle engine operating with a restricted air intake. Honda 600cc engines were controlled by three different ECUs: a Haltech, DTA and an “in-house” ECU. Simple calculations of injection duration are suggested to provide a baseline fuel map from which the engine could be started, and then fuel maps are tuned by experiment. Similar baseline numbers for ignition timing are given. Experimental dynamometer testing was performed to determine spark/fuel hooks, MBT timing and whether to inject onto open or closed intake valves. 1-D engine simulations were developed using both Wave® and GTPower® to provide a simulation environment for the FSAE engine development. The 1-D models were validated with steady-state flow bench data and dynamometer tests.

INTRODUCTION The Formula SAE® (FSAE) rules dictate a 610cc maximum engine displacement and a 20mm intake restrictor. Motorcycle engines are a practical choice for this student competition as the popular 600cc engine is available on many motorcycles and these engines usually have a high rpm capability that provides the possibility of high power. The presence of the restrictor necessitates the design of a fuelling strategy that usually employs fuel injection and a programmable engine control unit (ECU). Over the last few years Oakland University’s Formula SAE team has chosen to use the Honda 600cc engine. The choice of an engine controller has been based on cost and availability and has evolved from those commercially available from Haltech [1] and DTA [2], to the recent use of an Oakland-designed ECU [3]. The development of the Oakland ECU is detailed by Farrugia [4]. The Haltech system required the use of Hall Effect

sensors on the crankshaft or camshafts, with accompanying magnets. The DTA system uses the stock crank wheel but required that two teeth be ground off (missing) from the Honda’s stock 12 tooth configuration. The Oakland ECU was designed to use all 12 teeth on the crank wheel and allows greater flexibility in number of teeth as well as the sensors used for crank, cam, temperature, pressure, etc. All of these controllers use a look-up table strategy for controlling fuel and spark. The look-up tables are typically based on either throttle position (TPS) or intake manifold pressure (MAP) as the input Load parameter to the controller. Since the fuel quantities for the restricted engine will be significantly different from the unrestricted engine, and since the production engine values are not readily available, the look-up tables must be generated from an arbitrary starting point. A simple process for generating fuel tables will be described herein. The requirement of a single intake restrictor also necessitates a redesign of the stock intake manifold. Manifold pressure in a restricted engine is quite unlike that in the similar unrestricted engine. The choice of intake plenum, and runner lengths and diameters affect engine performance, however physical building and testing of manifolds is both expensive and time consuming. Modern simulation tools may help to reduce the cost and time required to design the restricted intake system, provided the simulation models can be sufficiently validated for the designer to have confidence extrapolating with the tools. The Oakland University Formula team developed 1-D engine flow models in both WAVE [5] and GT-Power [6] to facilitate engine design choices (in addition to just the intake manifold). These models were found useful in understanding experimental phenomena that were encountered. This paper describes a process used to tune a restricted engine for the FSAE competition. The engine and programmable ECU were already determined, and the restricted intake system was already installed on the engine. A simple, systematic process of establishing and building the spark and fuel tables and test the engine is described. Some preliminary results from a 1D engine flow simulation are also described and compared with experimental results. It was felt that a

published description of this process may be of some help to other student formula teams.

Next, if the stoichiometric air-to-fuel ratio is 14.6, then the mass of fuel required per cylinder per cycle would be,

ENGINE MAPPING Once a programmable ECU was available, the first task was to configure the hardware and software to operate the engine. The first priority was to establish the baseline fuel flow and spark timing tables with which to start and run the engine. ENGINE DETAILS

ma 1.78 × 10 −4 kg = 14.5 AFR −5 = 1.23 × 10 kg

Mass of fuel = m f =

For gasoline of Specific Gravity of 0.75 [8]

To get started, some basic engine parameters must be known. For the Honda F4i engine used in this study, some of the fundamental engine parameters are summarized in Table 1 below:

1.23 × 10 −5 kg Volume of fuel = V f = kg 0.735 l −5 = 1.64 × 10 l

Engine Type Bore Stroke Engine Displacement Compression Ratio IVO at 1mm lift IVC at 1mm lift EVO at 1mm lift EVC at 1mm lift

The Honda stock injectors were flow tested with a Cosworth IC-5460 Engine Control System. The flow rate was measured by pulsing the injectors for 8ms, while counting the number of injection events, and measuring the total volume of fuel collected in a graduated cylinder. See Table 2 for sample fuel injector calibration measurements. A fuel flow bench feature was also implemented in the in-house ECU for this kind of test. The average volume for the injectors was 0.0280 ml per 8 ms pulse.

Intake Valve lift Exhaust Valve Lift Firing Order Idle speed

F4I 67.0 mm 42.5 mm 599 cc 12:1 22º BTDC 43º ABDC 38º BBDC 7º ATDC 8.4 mm (author’s measurement) 7.2 mm (author’s measurement) 1-2-4-3 1300rpm

Table 1: Honda CBR600 F4i Parameters [7] FUEL TABLE Before starting the engine, some initial calculations were performed to establish a preliminary fuel look-up table. The approach taken was to calculate how much fuel was necessary for stoichiometric combustion in each cylinder, assuming that each cylinder was filled with air at atmospheric pressure (100% volumetric efficiency). The fuel quantity for idle conditions was then calculated for an expected volumetric efficiency (based on ambient pressure). For one cylinder of 150cc filled with air (only) at 100kPa and 293K, using the ideal gas law we have

PV 100 × 10 3 Pa ⋅ 150 × 10 −6 m 3 Mass of air = ma = = J RT 287 ⋅ 293K kg ⋅ K = 1.78 × 10 − 4 kg

= 0.0164ml

Injector

# 1 1 2 2 3 3 4 4

run run run run run run run run

1 2 1 2 1 2 1 2

Fuel Press [psi] 50 50 50 50 50 50 50 50

Volume [ml]

Pulse Count

77 78 78.5 78 79 79 77 78

10876 10997 11318 11162 11470 11508 10928 11033

Flow [ml / 8 ms] 0.0283 0.0284 0.0278 0.0280 0.0275 0.0275 0.0282 0.0283

Table 2: Fuel Injector Experimental Data Experiments on other injectors showed that the fuel flow rate is approximately linear with injector open time, that is, the actual time that the injector needle is open. It was determined that the time to open the Honda injectors was 0.2 to 0.5ms. This is the time required to activate the solenoid and open the injector, before any fuel is released. The actual injection open time would be (8 – 0.5) ms, but the small difference was not important here as the purpose is to just establish a baseline from which to begin dynamometer testing. Assuming then a linear relationship, the pulse time required for stoichiometric combustion can be calculated as:

8 x = 0.0280 0.0164

So, for this case, the injection duration, x, would be about 4.7 ms. This calculation presumed a cylinder filled with air at 100kPa, which relates to WOT, 100% volumetric efficiency. At idle most engines would run close to 40kPa, which considering the ideal gas law would imply that there would be close to 40% of the mass of air at wide open throttle. Therefore we would need 40% of the 4.7ms, that is 1.9ms at idle. For the first trials of the engine, we did not have an idea of how the volumetric efficiency changes with rpm. Therefore, our initial fuel table was only a function of load. That is, our fuel injection duration was 4.7ms at WOT for all speeds, and 1.9ms at zero throttle for all speeds. The intermediate throttle positions were linearly interpolated between these end values. The initial fuel table is shown in Figure 1, which is in the form of a wedge. It is not dependent on speed, simply 1.9ms at zero throttle and 4.7ms at WOT.

Degrees of Advance

Engine Speed rpm

Load

Figure 2: Initial Ignition Table (Screen Shot from DTAwin [2]) IDLE SPEED CONTROL The engine was started with these initial fuel and ignition tables. Experiments on the dynamometer were performed to find the relevant fuel quantities required at the various operating conditions. The Oakland Formula SAE engine does not have an idle speed control motor, which is a widespread method of idle speed control. Idle speed was controlled on our setup by means of spark timing.

Injection Duration msec

Engine Speed rpm

Load

Figure 1: Initial fuel table (Screen Shot from DTAwin [2]) IGNITION TABLE The Honda Service Manual [7] states that the spark advance is thirteen degrees before TDC at idle. Thirty degrees advance at high rpm is quite normal for engines; hence the initial table was set to have 13o advance at idle (1300rpm) and 30o advance at 6000 rpm. It is also quite common for racing engines not to have any load offset to timing i.e. no vacuum advance. Hence the initial ignition table was setup to be only a function of speed. Refer to Figure 2.

Around idle speed conditions, rpm in the range of 1500rpm and very small TPS, the ignition timing table was adjusted to achieve speed control. At 1500rpm, our choice of idle speed, the ignition timing was set to the value at which the engine runs well, say 15o BTDC. At higher rpm, say 2000, a purposely low value of ignition timing was used, say 10o BTDC while at lower rpm, 1000rpm, a higher ignition timing value was specified, 20o BTDC. This ignition strategy slowed down the engine if it tried to idle too fast, but aided the engine if it tried to idle too low. This system worked very well and was capable of properly maintaining engine to idle from cold start to fully warmed-up conditions. During warm-up, engine controllers typically employ coolant temperature compensation that enriches the fuel strategy because of a fuel deposition on walls and denser air charge (due to less heating of the air in the manifold and intake port). The strategy described above presumes that the coolant temperature compensation is active, and does not replace the need for coolant temperature compensation.

DYNAMOMETER TESTING

ENGINE GEOMETRICAL DATA

The engine was coupled to a Clayton water brake dynamometer for testing. Sixth gear was used during these tests to match the maximum engine rpm to maximum allowable dynamometer rpm. The ECU allowed two control strategies, either TPS or MAP. The look-up tables were in the form of a Load parameter (either TPS or MAP) versus the engine rpm. TPS was used as the load parameter in this situation since this is a direct input in the dynamometer setup, i.e. the Load location within the look-up tables was set by adjusting the TPS manually. Engine speed was then set by manipulating the dynamometer loading. Optimal ignition timing and fuel injection duration were determined at all available speed discretizations in the table at WOT, and several more at part throttle. Ignition and Fuel hooks were performed at the various test conditions. As expected, the MBT timing was higher at higher rpm. The MBT timing was high also where the volumetric efficiency was poor. Volumetric efficiency was calculated from measurements of the mass air flow using a laminar flow element.

Engine details were obtained from the engine manual and by measurement. General information such as bore, stroke, valve lifts, valve opening and closing angles were available in the manual. Connecting rod length, valve, port and runner dimensions were physically measured. The F4i heads and cams available were on operational engines and hence the valve lift with crank angle could not directly be measured as the cams are directly acting on the valves and it was not possible to probe them with a dial indicator. The cam lobe was measured instead, as this was reachable with a dial indicator. Valve lift profile was mathematically determined from the measured cam lobe. The valve lift profile was determined by mathematically imposing a flat follower on the measured lobe. The calculated valve lift matched very well with the angles specified in the manual; total error was 3 degrees. FLOWBENCH DATA

110

3000 rpm 4000 rpm

100

5000 rpm

T o rq u e , lb -ft

6000 rpm 90

6500 rpm 8000 rpm 9000 rpm

80

70

60

50 15

20

25

30

35

40

45

50

Spark Advance, deg BTDC

Figure 3: Spark advance hooks at WOT Additional tests were conducted to determine the best timing for start of fuel injection. Best performance was measured with fuel injected onto open valves, versus closed valves. It was found that injection onto open valves gave 6% more torque at the point of worst volumetric efficiency (6500rpm). This was a worthwhile improvement given the fact that it did not involve any extra hardware.

1D ENGINE SIMULATION Wave and GT-Power models were built for the FSAE engine to help understand engine characteristics and provide tuning tools.

55

An available cylinder head from an older F2 was tested on a flowbench. The flowbench tests showed that the stock intake hardly restricts the head. The restricted intake manifolds used on Oakland’s FSAE 2002 and 2003 cars were also flowed. The restricted intakes showed much lower flowrate. It was also found that there was not much variability between 2002 and 2003 intakes, and not much variability between the runners of the 2003 intake. Hence there was no need to trim the individual cylinders injector timings. The coefficients of discharge were then calculated from the measured flows so that they were available for engine modeling. Figure 4 shows the flow rates through the inlet and exhaust ports for a fixed differential pressure of 28” of water. Coefficients of discharge for 1mm valve lift discretizations were calculated from the measured flow data.

Port Flow Rates versus Valve Lift at 28" H2O differential pressure 140.0 120.0 100.0

F lo w , S C F M

Effect of Spark Advance on Torque for various engine speeds

80.0

Intake port Cylinder 1

60.0

Intake port Cylinder 2

40.0

Exhaust port Cylinder 1

20.0

Exhaust port Cylinder 2

0.0 0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

Valve lift, mm

Figure 4: Port flow rates against valve lift Steady state flow bench models were built to simulate the flow bench tests. The pressure difference between

the air source and air sink in the 1D codes was 7kPa, which approximated the 28” of water during the flow bench tests. The mass flow rate determined from these models was compared to the experimental values. Models were modified until agreement was good. Similar steady-state flowbench models were also developed for the intake system, including the restrictor and the exhaust ports. The steady-state modeling of the intake system was very important as the restricted intake fundamentally affected the overall engine model. Since the restrictor determines the maximum amount of air that the engine can breath, its model proved to be very important for the overall engine performance. The experimental and simulated volumetric efficiency are shown in the Figure 5. The model does predict the low volumetric efficiency point quite well. It is expected that the model for the restrictor needs some refinement to get better agreement in the higher rpm range. STUDY ON HIGH MAP VALUES AT IDLE Running the Honda engine in both stock and restricted intake configurations revealed very high MAP sensor readings at idle conditions. The MAP sensor value was in the 70 to 80 kPa range at idle which is very high when compared to expected values for typical engines. A possible explanation for the high MAP value could be the presence of back flow of cylinder gases through the intake valves. This phenomenon was analyzed using the engine simulation, which also predicted high MAP at idle. Furthermore the residual gases were predicted to be around 70% at idle while only 3% at WOT. The 1-D software showed intake runner gas velocity to be directed out of the cylinder for long periods of time, which is explainable due to the high valve overlap periods that this engine has.

experimental

DYNAMOMETER DAQ As part of the dynamometer apparatus, LabView [9] software was used to monitor the engine and dynamometer. Various temperatures were read into the DAQ and displayed on the computer monitor. Dynamometer speed and torque and an extra oxygen sensor were also read into the DAQ system. Measured data was archived in text files for later analysis. Speed control of the dynamometer was also implemented in LabView. The water-brake dynamometer control is accomplished with separate fill and drain valves, and hence a standard PID control loop was reconfigured within LabView to split the output into separate open and close signals. The system worked quite well, however it is believed that the system could be improved by reconfiguring the dynamometer to drain continuously and then use the controller to fill the dynamometer using a control loop. The disadvantage to the latter system is that a very stable pressure water supply would then be required. The DAQ system was also setup to “overhear” data that the ECU sends over the RS232 connection to the ECU GUI running on the PC. In this manner data that is readily available in the ECU system such as spark timing and fuel injection duration were made available to the LabView system, and hence the stored data would have these parameters as well.

CONCLUSION The use of a Honda engine for the FSAE competition at Oakland University has provided a training ground for the participating students. The use of ECU’s to control the engine was mastered starting from basic principles. Engine dynamometer monitoring and control were also performed. Engine simulation studies were performed using two commercial products and provided backup to experimental findings. All this thanks to the 20mm restrictor!

ACKNOWLEDGMENTS

1D model

We would like to thank Dan Agnew of Managed Programs Rochester Michigan for providing technical guidance on the use of 1D software and for providing access to his flowbench facility. We would also like to thank Ricardo and Gamma Technologies for providing their 1D engine simulation packages. Thanks also goes to National Instruments for donating the DAQ system for the dynamometer.

Engine speed, rpm

Figure 5: Experimental and modeled volumetric efficiency

REFERENCES 1. Haltech Engine Management Systems, Sydney Australia, www.haltech.com 2. DTA Competition Engine Management Systems, Salford England, www.dtafast.co.uk

3. Oakland University’s Engine Management site, www.oakland.edu/~mfarrugi/Formula SAE and www.reataengineering.com 4. Farrugia M., Farrugia M., and Sangeorzan B. P., 2005, “ECU development for the Formula SAE Engine,” 2005-01-0027” 5. WAVE, Ricardo, www.ricardo.com 6. GT-Power, Gamma Technologies, www.gtisoft.com 7. Honda Service Manual 2001-2003 CBR600F4i, Honda Motor Co. Ltd 8. Heywood J. B., 1988, “Internal Combustion Engines”, McGraw Hill, Table D4 9. LabView, National Instruments, www.ni.com

ACRONYMS 1D DAQ ECU GUI MAP MBT PID TPS WOT

One Dimensional Data AcQuisition Electronic Control Unit Graphical User Interface Manifold Absolute Pressure Minimum timing for Best Torque Proportional Integral and Derivative (control) Throttle Position Sensor Wide Open Throttle