Applying Combustion Chamber Surface Temperature to Combustion Control of Motorcycle Engines

Transcription

1 Technical Paper Applying Combustion Chamber Surface Temperature to Combustion Control of Motorcycle Engines Applying Combustion Chamber Surface Temperature to Combustion Control of Motorcycle Engines Satoshi ICHIHASHI *1 Key Words: Motorcycle engine, Knocking, Combustion chamber surface temperature (TCCD), Combustion control 1. Introduction Following the expansion in use of fuel injection systems to replace carburetors, engine control systems in motorcycles have become increasingly important. Electronic fuel injection with oxygen sensor feedback control and a 3-way catalytic converter provide for adaptive fuel control, and thus contribute to certain and stable reduction in emissions. In comparison with the technological progress on a day-by-day basis for injection control, ignition control has seen almost no outstanding advancement in past years. With the situation where motorcycles are used as an essential mode of transportation in developing countries, there are increased demands for strengthening of emissions and fuel consumption regulations such as EURO/BS6 which is based on UN-GTR No.2 (WMTC). Although emission regulations for fourwheeled vehicles are generally one step further forward than those for motorcycles, it is difficult to transfer some improvement technologies to motorcycles for reasons of characteristics and cost differences. For example, use of the high-frequency vibration detection type knocking control system, which is commonly used in four-wheeled vehicles, is limited to several sport-touring motorcycles, and is not used for a wider range of motorcycles because of difficulties in layout and unacceptable extra cost. This gives a strong motivation for development concentrating on motorcycles, especially in the field of combustion control. This research aims to develop a new concept of combustion control for common motorcycles. The new combustion control focuses on the effect of ignition timing control with measurement of the surface temperature of the engine combustion chamber. This is because, although ignition timing adjustment is widely known to have a major effect on the state of combustion, it has not been used positively so far because positive ignition control increases has risk of knocking which is directly connected to fatal damage in the engine. In this respect, it is important to develop parameters which can measure the risk of knocking and integrity of combustion. To measure the integrity of combustion, we propose the use of differential temperature of the combustion chamber (TCCD) for gauging the combustion. When the engine representative temperature (TW) is subtracted from the combustion chamber surface temperature (TCC), the temperature difference between combustion chamber surface and the engine coolant can calculated, and we named this difference TCCD. In this research, we found that the TCCD Received 18 November 216, Content reprinted from SAE Technical Paper , 216 SAE International. Further distribution of this material is not *1 System Development Department, R&D Operations --

2 Keihin Technical Review Vol.6 (217) parameters have behavior characteristics similar to Table 1 Specification of test engine that of torque and injection fuel quantity, moreover the TCCD rises sharply when knocking combustion starts. TCCD is a parameter which can be used to measure the quantity of heat discharged through the combustion chamber wall. It is understood that the engine runs at the optimal conditions under an appropriate cooling system working range. Discharge heat quantity is proportional to the total generated heat quantity. Total generated heat from combustion is proportional to engine torque. Based on these preconditions, the TCCD, torque and injection fuel has a proportional relationship in principles. Regarding TCCD behavior when engine knocking occurs, temperature boundary layer in the combustion chamber is eliminated by a pressure wave and the TCCD rises sharply. Our challenge in this research is to make simple and practical combustion control for motorcycles by using the TCCD characteristics (1)-(4). Displaced volume Stroke Bore Compression ratio Valve train layout Cooling type Maximum Output Maximum torque Emission Test Engine 1 196cc 4.mm 68.mm 8.:1 OHV 2valves Forced Air cooled 4.1kw@36rpm 12.4Nm@2rpm N/A 3. Measurement Method Test Engine 2 69cc 84.mm 2mm 12.6:1 DOHC 4valves Water cooled 49kw@7rpm 68Nm@6rpm EURO3 In this research we use thermocouple type temperature sensor and a prototype thermistor temperature sensor which is under consideration for production of a system. The thermocouple specification shown in Table 2. The thermocouple sensor has small sensing tip making it suitable for attachment to the combustion chamber surface. The thermocouple sensor was used to find optimum thermistor sensor position which has best better response for combustion difference. The specifications of the prototype thermistor 2. Test Engine The specification of the test engines are shown sensor specification are shown in Table 3 and its exterior shell shape is shown in Figure 1. In the air cooled engine pre-testing, 8 points of thermocouple sensor mounted on combustion Technical Papers in Table 1. We use two different engine for this research. Test engine 1 is forced air cooled general purpose engine which use for pre-tests, and Test engine 2 is water cooled single cylinder engine which used for high performance sport motorcycle. The air cooled general purpose engine has typical low cost specification in market. Water cooled engine has bigger bore and highest compression ratio in the market, and thus it has a high risk of the knocking combustion. The water cooled engine was used to investigate the relationship between the combustion chamber temperature and the extent of knocking. chamber to measure temperature behavior and ranges. Thermocouple sensor use 1.mm diameter sheath type. This is mount on combustion chamber by swaging. Thermocouple position in air cooled engine pre-testing on Figure 2, and measuring points listed on Table 4. In the water cooled engine test (Figure 3). We applied thermocouples, No.2 to No.6, and an indication pressure sensor, No.7 to understand effect of the combustion as related to surface temperature. Each position of the sensor shown in Figure 3, and measuring position are listed on Table. -11-

3 Applying Combustion Chamber Surface Temperature to Combustion Control of Motorcycle Engines Table 2 Specifications of thermocouple sensor Table Measuring points for water cooled engine Type Core wire Material Core wire/sheath diameter Max. operation temperature Response time Sheath tube material Table 3 K-type with sheath tube Chromel-Alumel.2mm/1.mm 6degC 12msec. (in boiled water) SUS34 Specifications of prototype thermistor sensor No Place Thermistor type sensor Intake side squish surface (reference for sensor) Intake squish center Exhaust side squish Far side squish from spark plug Secondary plug side squish surface Indication pressure sensor Sensor element Thermal constant τ Dissipation constant Operation temperature range Insulation resistance Shell material NTC thermistor type sec (in still air).7mw/degc -~3degC Min. Mohm (VDC.) A661 Table 4 Measuring points for air cooled engine No Place Intake valve side Intake side near spark plug Center of combustion chamber Exhaust side near spark plug Exhaust valve side Exhaust valve outer side Between IN./EX. Valves Intake valve outer side 9 4 M P-.8 Fig. 3 Measuring points on water cooled engine We apply prototype thermistor sensor in position No.1, because, in this testing, position No.3 had best temperature response, but it is not possible to machine a M thread & hole in position No.3. This is due to cooling water channel shape in cylinder head. The blue mark shown as No.7 is position of the indication pressure sensor. Fig. 1 Prototype thermistor sensor 4. Pre-test Result on Air Cooled Engine Fig. 2 Measuring points on air cooled engine In the pre-test of the air cooled engine, we endeavored to understand the outline of the combustion chamber surface temperature. All measurements were taken under no-load conditions with extra engine cooling using a blower for exhaust systems. We obtained following new knowledge concerning combustion chamber surface temperature. 1. The temperature range is degc to 2degC at normal operation (Figure 4). 2. The temperature difference in the combustion -12-

4 Keihin Technical Review Vol.6 (217) chamber is approximately 4 to degc at point to point (Figure ). 3. In the air cooled engine a rich mixture of fuel gives a slightly lower temperature (Figure ). 4. In the air cooled engine high rpm gives higher temperature, but each point tendency does not make differences (Figure 6).. The surface thermocouple temperature response is approximately 3ms (Figure 7). Full Scale: degc rpm Fig. 7 Time First firing Approx.3ms Temperature response No.1 Temp. Cylinder temperature Engine rpm Response time of surface temperature at engine start Ne2rpm CO.% Ne2rpm CO 7.%. Test Result on Water Cooled Engine TCC (degc) Intake Valve Spark plug Exhaust Valve Intake Valve Spark plug Exhaust Valve Fig. 4 Ne2rpm no-load Temperature distribution CO (%) Fig. Surface temperature difference by CO % TCC (degc) at CO % The first step for water cooled engine test was for determine the thermistor type sensor position for controls. Figure 8 shows each position temperature measurement examples for each position with the ignition timing difference. This test engine has knocking sensitive area at rpm TH4~6deg. The intake side point (Nos. 1, 2 & 3) have better response in the knock levels compared to exhaust side (No. 4, & 6). Our assumption is that the fact that temperature increases when knocking occurs is due to combustion chamber geometry, because with the larger bore engine, knocking starts at the intake valve side due to lower temperature as compared to Temperature (degc) rpm TH4 18 level Knock Ignition timing (BTDC degca) (Nm)/Knock level (bar) Technical Papers Fig. 6 Temperature difference by engine rpm Fig. 8 Surface temperature (vs Ignition timing) -13-

5 Applying Combustion Chamber Surface Temperature to Combustion Control of Motorcycle Engines the exhaust side. Figure 9 shows a test engine which was accidentally damaged during the measurement. It can be noted that the engine damage only appears on the intake valve side, a fact that supports our assumption. From these results, it was determined to locate the prototype thermistor at point No.1. The following equation shows the relationship between combustion chamber temperature (TCC) measured by the thermistor sensor, the coolant temperature (TW) and difference between them (TCCD) which was used as a parameter to measure combustion. TCCD = TCC TW (1) Figure shows detailed measurement results of TCCD, engine torque and knock levels at rpm TH48 (most sensitive point for knocking). By the ignition advance, both torque and TCCD will increase proportionally until minimum ignition advance with best torque (MBT), after which torque is saturated and knocking starts due to over advanced ignition, at which stage the TCCD rises sharply in accordance knocking level. The results in Figure 11 show the TCCD with peak indication pressure at various engine running conditions. The TCCD has liner proportional characteristics with the peak indication pressure. These characteristics suggest important new knowledge, because if a specific TCCD corresponds to a specific peak indication pressure (Pmax), one can specify the indication mean effective pressure Intake Exhaust (IMEP) in particular engine running conditions. Figure 12 shows the relationship between Pmax, (Nm)/TCCD (degc) TCCD (degc) IMEP (bar) Fig. 4 Engine torque TCCD Knock level rpm TH48 Ignition timing (BTDC degca) Relationship between torque, TCCD and knocking level NE 19 - rpm TH 4. to WOT R² = Knock level (bar) 3 ALL 4 ALL 4 ALL 2 29 ALL 24 ALL 19 ALL PMAX (bar) Fig deg 9.deg 1deg 2deg 2deg 3deg 3deg 4deg 48deg 6deg WOT TCCD and peak indication pressure rpm MBT Pmax (bar) Fig. 9 Test engine which damaged by knocking Fig. 12 Relation of Pmax and IMEP with MBT -14-

6 Keihin Technical Review Vol.6 (217) IMEP and MBT. Figure 13 shows TCCD of different Air fuel ratio. Air fuel ratio has some influence on TCCD, but it does not affect control to a major extent. TCCD (degc) Fig. 13 rpm. 11. (Nm) 12. Relationship between torque and TCCD at different A/F ratios 12 determined by the TCCD. In specific engine running conditions, there is no significant rate of change for each discharge heat value. Conventional fuel injection systems are designed on this principal. For example, the fuel injection system determines ignition timing and injection fuel quantity according to engine running conditions as determined by a specified sensor in order to reproduce specific engine output. This suggests that the balance of the discharge heat presents no significant change and that each discharge heat value has its proportional relation (Formulation 3). Engine Control Unit 6. Principles This research aims to develop a new concept of combustion control for common motorcycle. On this basis, the system should be minimized from current conventional fuel injection system. Figure 14 shows schematic of this system. The prototype system only adds TCC sensor to the conventional system. TCCD control principles based on combustion engine basics with the following preconditions. 1. The heat generated by the engine and the cooling are balanced within the normal operation condition of the engine. 2. The engine runs on normal combustion without misfiring. 3. The ratio of distribution of the heat discharged distributions is reproducible in specific engine running conditions. Figure 1 shows image of TCCD. This is the principle of TCCD control. If engine runs specific rpm and torque, combustion condition can be Air temp sensor Injector TCC Sensor Fig. 14 Combustion Gas Temp. based TCCD Target Temperature TCCD difference for ignition timing correction Fig. 1 Qtq h1 Combustion Chamber Qtotal Spark Plug Qw Schematic of system Qw λ x TCC sensor Cylinder Head Qex Coolant temp sensor (TW) Crank sensor TCC Water Jacket h2 High Temperature TW sensor Image of TCCD temperature TW Technical Papers -1-

7 Applying Combustion Chamber Surface Temperature to Combustion Control of Motorcycle Engines Qtotal = Qtq + Qw + Qex (2) Qtq Qw (3) K = Qw/Qtq (4) Qtotal: Total heat quantity Qtq: Heat quantity which translated to torque Qw: discharge heat through the wall Qex: discharge heat through the exhaust gas K: Ratio of Qtq and Qw which in specific engine condition Qw = K * Qtq = U * A * (TCCD) () U: Overall heat transfer coefficient from combustion gas to coolant A: Sum of cooling area Engine specified parameters x, λ and engine running condition specified parameter h2, can be viewed by constant of Ku as a whole in Equation 6. Overall heat transfer coefficient U determined by equation 7. The result is U only determined by h1. x 1 Ku = + λ h2 1 U = = 1 x h1 λ h2 h1 1 + Ku * h1 (6) (7) h1: Heat transfer coefficient through combustion gas to wall h2: Heat transfer coefficient through coolant x: Thickness of combustion chamber wall λ: Heat conductivity of combustion chamber material Ku: constant which determined external of combustion chamber Heat transfer coefficient h1 comes from Eichelberg equation (Equation 8). In this equation Cm is related rpm, both p and Tg are related combustion. This means h1 determined by Engine rpm and combustion. h1 = 7.6Cm 1/3 P 1/2 Tg 1/2 [W/(m 2 K)] (8) Cm: Piston speed (m/sec.) p: Cylinder pressure (Mpa) Tg: Combustion gas temperature (K) Therefore, equation transform to equation 9, equation 9 shows TCCD determined by torque and combustion conditions. This equation is important, because if there is a difference in combustion conditions, such difference can be determined from the difference in TCCD value. Qtq TCCD = = Qtq U * * A 1 + Ku * h1 A * h1 (9) Figure 16 shows relationship between TCCD, torque and knocking level depending on ignition advance. In a conventional fuel injection system, the ignition timing is a fixed parameter which is determined by the fuel RON value. In Figure 16 BTDC19degCA is the equivalent of 9 RON fuel which is normally available in Europe. However if the engine were to use by regular fuel in Japan at 9 RON, the 19degCA ignition advance would be too much and thus there being risk of TCCD (degc) (Nm) RON RON RON9 RON9 RON RON Fig. 16 rpm TH48 Production IG-Timing Ignition timing (BTDC degca) 2 1 Konck level (bar) TCCD behavior for 9 RON and RON -16-

8 Keihin Technical Review Vol.6 (217) engine damage due to knocking. Figure 17 shows different x-axis to the data of the Figure 16 graph. If ignition timing were controlled by TCCD, the engine would be able to handle both 9 and RON fuel with one TCCD set point (39degC). Figure 18 shows the comparison of torque for different compression ratios. This engine has a nominal production compression ratio of ε12.6. For this test, a difference in compression ratio of +/-. was implemented by the adjustment of cylinder gasket and cylinder head gasket thickness. The +/- (Nm) rpm TH TCCD set point RON9 RON9 RON RON 2 1 Knock level (bar). compression ratio difference produced a torque variation of 3.6% at nominal production ignition timing (BTDC 19degCA). Figure 19 shows the case where ignition timing is controlled at TCCD 39degC. As shown, we were able to limit variation in torque to 1.6%FS. TCCD control can reduce torque variation by approximately % and improve torque level by approximately.nm. (Nm) 6 Knock level Knock level variation 1.6%FS TCCD (degc) Knock level 1 Knock level (bar) Fig. 17 TCCD (degc) and knock level behavior in RON9 and RON Fig. 19 Compression ratio comparison (x axis -TCCD) 7. Control Algorithm for Prototype Technical Papers (Nm) 6 Kock level Kock level variation 3.6%FS Fig. 18 Ignition Timing (BTDC degca) Kock level Compression ratio comparison (x axisignition timing) 1 Knock Level (bar) Figure 2 shows an outline of target TCCD temperature calculation. The target TCCD map was acquired at stable running conditions beforehand. There is also a correction value table depending on the engine temperature conditions. This correction value added to or subtracted from the target TCCD depends on the coolant temperature. The temperature transition delay simulates the movement of coolant with conventional filtering methods. The system calculates the difference between TCCD targets with actual TCCD and applies ignition advance or retard according the TCCD difference. -17-

9 Applying Combustion Chamber Surface Temperature to Combustion Control of Motorcycle Engines Figure 21 shows logging trace examples for each parameter on WMTC phase3. a retarded ignition of -degca with conventional ignition control to prevent hard knocking. With conventional ignition control, the retarded Target TCCD MAP X-Engine rpm Y- Z- Target TCCD Fig. 2 Target correction value (degc) Correction table +/- + Reverse or cooling side temp. Target TCCD (degc) Target TCCD calculation Temperature behavior reproduced by Filtering Target temp. ( based: Immediate) Time Filtered Target TCCD ignition map gives slow acceleration according to the retarded amount. With TCCD control, and the target TCCD set +degc as in Figure 22 above, the TCCD target gives a controlled advance ignition. The result in acceleration time improved by approximately.sec with TCCD control, and also variation of Display range TCCD: -4degC/NE: -rpm Target TCCD TCCD the acceleration time was reduced whether it was retarded in accordance with the ignition map or not. However there was no audible knocking occurring NE Throttle during the tests including those for 9 RON fuel. Figure 24 shows the WMTC fuel efficiency comparison between TCCD and conventional ignition Fig. 21 TCCD control logging trace (WMTC phase3) Display range: Temp. -4degC/ -Nm TH -deg/ig correction +/-2degCA Ignition correction 8. Implementation to Vehicle TCCD Figure 22 show logging trace of acceleration test at 6th gear. Test conditions are -1kph With TCCD W/O TCCD Throttle TH48deg roll on with RON fuel. This trace is a comparison of the system with and without TCCD control. The dotted line shows data without TCCD Fig. 22 Top gear acceleration control and the solid line shows data with TCCD RON STD 4 control at the target of +degc. To keep higher target TCCD, ignition correction starts from advance side (about 9degCA) and it TCCD control RON STD -3deg RON STD -deg RON9 STD -deg 2 reduce to small advance (2degCA around). Although RON STD 9 TCCD keep advance ignition, there is no audible knocking happen during this test. Figure 23 shows result of an acceleration test with 9 RON & RON fuels. The acceleration time with conventional ignition control depends on the amount of ignition Conventinal RON STD -3deg 1172 RON STD -deg RON9 STD -deg Time (msec) retard (-3degCA or -degca). 9 RON fuel required Fig. 23-1kph acceleration test result -18-

10 Keihin Technical Review Vol.6 (217) control. In this test, the engine variation is simulated by ignition timing difference based on the production ignition setting (STD) and ignition retarded by -3degCA and -degca respectively from the STD ignition map. In conventional ignition control fuel efficiency becomes worse at retarded ignition. With TCCD control, fuel efficiency is improved approximately pressure peak reduction, and thus decrease TCCD. Vehicle test using different octane level fuels in various acceleration situations have been implemented, resulting in that, the proposed combustion control concept/algorithm are the simple and practical ways to common motorcycle fuel injection systems. The new techniques are expected to be realized on mass-production motorcycles in the near future. 3%. FE (L/km) WMTC fuel efficiency. STD IGMAP STD IGMAP - retard TCCD with -3 retard STD IGMAP -3 retard TCCD with STD IG TCCD with - retard References (1) Hiroyasu, H., Easy understand Internal combustion engines Nissin publish Co th Edition. (In Japanese) ISBN /3- (2) Kaneko, Y., Fundamentals of Internal 3.3 Combustion Engine Sankaido Co Fig. 24 WMTC weighted total TCCD vs Conventional ignition control comparison (WMTC FE) First Edition. (In Japanese) ISBN C33 (3) Yamaha motor Co. Motorcycle editorial board. Motorcycle. Fundamental knowledge for 9. Summary In this paper, new concept of TCCD combustion beginners (In Japanese) Sankaido Co rd Edition ISBN (4) Ichihashi, S., Control apparatus for internal combustion engine International patent No.PCT/ Technical Papers controls, aiming to obtain high performance, and better fuel efficiency for common motorcycle, have been provided and substantiated. Strong correlation between TCCD and in-cylinder pressure peak was found in our pre-test, which has been proven in principle latter as described because combustion chamber temperature is dependent on the quantity of heat generation and heat conductivity. Therefore, existing thermistor sensor is proposed for knock detection instead of complicated and expensive knock sensor indication system. Furthermore, ignition timing control is developed for avoiding retard condition which cause in-cylinder JP21/84796 Acknowledgements The author would like to express his gratitude to KTM A.G. R&D EMS team who supported test engine and vehicle. All testing data acquisition in this research are from the work package of motorcycle knock control system project in Keihin Corporation, the author s colleagues, Tomomi YONEMARU and Takashi TAKANOHASHI, contributed a lot to testing in the early stage of this project. -19-

11 Applying Combustion Chamber Surface Temperature to Combustion Control of Motorcycle Engines Definitions/Abbreviations TCC TCCD Temperature of combustion chamber [degc] Temperature difference between TCC with TW [degc] TW Engine coolant temperature of water cooled engine [degc] CO Carbon monoxide [%] OHV BTDC Overhead valve type valve train layout Before top dead center crank angle [degca] DOHC Double overhead camshaft valve train layout RON NE MBT Research octane number of the fuel Number of Engine revolution [rpm] Minimum ignition advance with best torque [degca] STD Production ignition setting [degca] Author S. ICHIHASHI Will motorcycle become a dinosaur, is my obsession in recent days. However, I still believe motorcycle is the last and only thing to feel real truth, especially in a computer age that virtual reality technology is filling up our world. Riding a motorcycle is sometimes accompanying with cold, wet, hot, and even risk. Such real experience (sometimes a terrible experience) help us to understand the hard situation in real-world. To feel happy in sunny day, let s ride motorcycle! (which gives strong motivation to motorcycle development...) (ICHIHASHI) -2-