Variations of Exhaust Gas Temperature and Combustion Stability due to Changes in Spark and Exhaust Valve Timings

The effects of spark timing and exhaust valve timing change on exhaust gas temperature during cold start of an SI engine are studied through engine bench tests. It is observed that the exhaust gas temperature increases when the spark timing and valve timing are retarded individually or simultaneously, due to late combustion or slow flame speed. However, using COVimep it is also investigated that the combustion stability during cold start deteriorated under retarded exhaust valve timing condition. To increase exhaust gas temperature for fast warmup of catalysts while maintaining combustion stability, the retarded spark timing will be useful for the cold start period. Furthermore an optimal condition for spark and valve timing should be found and applied for the increase of exhaust gas temperature in the cold start period. INTRODUCTION Three-way catalyst(TWC), the successful application for the emission after-treatment of SI engines, is very effective to lower the emission levels from vehicles. The conversion rates of CO, HC and NOx of a TWC are very high, between 80 and 90%, after it becomes fully heated up to a normal operating temperature. But they also have inherent problems related to catalytic chemical reaction. Since the catalyst stays at lower temperatures during cold start period of a vehicle, harmful species such as CO and HC pass through the TWC without catalytic reaction and the level of exhaust emissions becomes very high in this period. Therefore, the key technologies to meet the stringent emission regulations such as LEV, ULEV and SULEV of CARB, and to save the air quality in urban areas are closely related to reduce the time required to reach light-off temperature of a catalysts in the cold start period(1,2). Previous studies showed change of spark ignition timing significantly affects exhaust gas temperature in cold start period(3,4). When spark ignition timing is retarded, the start of combustion is delayed, resulting in a lower maximum cylinder pressure. On the other hand, flame stays up to a later stage of the expansion stroke and the exhaust gas temperature is higher than the normal spark ignition timing cases. Although energy loss is considerable with retarded spark timing, rapid warmup of catalyst in a cold engine start situation can be achieved due to an increase in the exhaust gas temperature. Recent development in engine control unit (ECU) and variable valve timing (VVT) technology is also very helpful to minimize the warmup time of catalysts in cold start. A VVT system can change the intake or exhaust valve timings to optimize the gas exchange processes, and the engine operating parameters such as engine speed, load and coolant temperature change accordingly (5). Changes in the intake and exhaust valve timings affect flame speed, temperature and residual gas fraction in the cylinder, and these changes control the combustion processes directly. Therefore, a proper change of valve timing can raise the exhaust gas temperature for rapid warmup of the catalysts in the cold start period. However such changes of spark and valve timings can affect the combustion stability that leads to the idle quality and emission compositions. It would be meaningless If exhaust temperature were increased with the sacrifice of stability that results in the increase of HC emission. Therefore, a proper change should mean the rapid warmup of a catalyst while maintaining combustion stability. The main objectives of this study are to optimize the spark ignition timing and exhaust valve timing and to increase the exhaust gas temperature during the cold start period for rapid warmup of catalysts. At first, the effects of exhaust valve timing and spark ignition timing on cold start operation are investigated through engine bench tests. Exhaust valve timing is changed using a variable timing camshaft and spark ignition timing is changed by an external ECU. The changes in combustion characteristics and exhaust gas temperature are measured and analyzed. In addition, the variations of combustion stability with the change of these timings are also investigated. EXPERIMENTS SETUP EXPERIMENTAL SETUP – A 2-liter, naturally aspirated, four-cylinder SI engine is used as a test engine and its specifications are described in Table 1. Fig 1 shows a schematic diagram of the experimental setup. Pressure of cylinder #1 is measured using a spark-plug type Kistler 6052B pressure transducer. Measuring timings are synchronized with the crank angle encoder which generates pulse by one degree of crank angle change. Therefore 720 pressure data are acquired in one cycle of cylinder #1. Measured pressure signals are converted to voltage signals by a charge-to-voltage amplifier, and acquired and analyzed by a data acquisition system. Pressure in the intake plenum chamber is measured by a Kistler 4045B, an absolute pressure sensor. A programmable ECU changes spark ignition timing in order to set the timing values at the test conditions. Other signals such as engine speed and exhaust gas temperature are stored in the data acquisition PC. Table 1. Specification of test engine Items Specifications Type 4 cylinder, spark-ignition, Inline, DOHC Bore 82 mm Stroke 93.5 mm Compression ratio 10.3 Idle speed 800±100 rpm Spark timing BTDC 10° ±5° Intake valve timing BTDC 8°/ABDC 40° Exhaust valve timing BBDC 50°/ATDC 10° Valve overlap 18° Fig. 1. Schematic diagram of experimental setup In the test engine, a variable timing camshaft that can change the phase of cam events is installed for changing valve timing. Fig. 2 shows the variable timing camshaft and modified sprocket for the experiments. Fig. 3 shows the variable timing camshaft installed in the cylinder head. The cam sprocket and chain pulley can be disassembled from camshaft while the engine is on the test bench. Exhaust cam phase can be changed by simply turning the camshaft when the sprocket and pulley are disconnected. As shown in Fig. 2(b), there are 16 keyholes on the pulley and 15 holes on the camshaft mount. Consequently, the minimum change of cam phase is 3° crank angle(CA). (a) Valve train (b) Cam sprocket and pulley of variable timing camshaft Fig. 2. Variable timing camshaft and sprocket Fig. 3 Variable timing camshaft mounted on the cylinder head COMBUSTION ANALYZER – The measured pressure data is used for the calculation of rate of heat release (ROHR) to observe the change of combustion characteristics with the change of these conditions. The data is also applied to measure coefficient of imep variation (COVimep), an index to evaluate combustion stability. For these reasons, a PC-based combustion analyzer was programmed using LabVIEW by National Instruments. Fig. 3 shows the main screen and diagram of the combustion analyzer. In order to analyze the A/D converted voltage-based pressure data, the engine specifications and measurement conditions should be supplied. The analyzer calculates the pressure from voltage data and volume from synchronized crank angles in each cycle. So the cycle-by-cycle P-V diagram in addition to P-θ diagram is gathered. These P-θ curves are numerically differentiated to calculate the ROHR. Numerical cyclic integration is also provided to P-V curves, to find imep of each cycle. These imep values are statistically evaluated for the COVimep with this equation(6); (%) 100 × = imep imep imep X COV σ Fig. 4. Main screen of combustion analyzer TEST CONDITIONS Since the goal of this study is to investigate the effects of spark timing and exhaust valve timing on the exhaust gas temperature, especially in a cold start period, the test engine is soaked at 20°C before each test. The exhaust valve timing is changed to BBDC 50° ±12° CA. Similarly, in order to investigate the effects of spark timings, spark ignition timing is changed to BTDC 10° ±5° CA. In each case, the same amount of fuel is supplied, through the control of fuel injection pulse width using an external ECU(Motec M8). Because the engine is started under the cold start condition(20°C), the stoichiometric feedback control of fuel supply is not applied. The test conditions for the baseline case that has original spark and exhaust valve timings are determined through a preliminary test, and a proper fuel injection duty map for starting and stable operation of the engine was established using the external ECU. The same fuel injection duty map is applied to other test cases. Exhaust gas temperature is measured from beginning to 200 seconds after engine start, and cylinder pressure is measured at 300 seconds after engine start. In the preliminary tests, it was observed that the varied significantly with the extreme change of valve timing. Therefore COVimep was measured with the exhaust valve open at BBDC 50° ±6° CA, to see the proper effects of valve timing changes. RESULTS AND DISCUSSION Effects of exhaust valve timing – In order to investigate the feasibility of valve timing change for raising exhaust gas temperature, the effects of exhaust valve timing change on cold engine performance are experimentally studied. Fig. 5 shows exhaust gas temperature variations with the change of exhaust valve timing. As shown in this figure, exhaust gas temperature increases when exhaust valve timing is retarded 12°CA from the baseline case. On the contrary, when exhaust valve timing is advanced 12°CA, a small decrease in exhaust gas temperature was observed compared with the baseline case until 100 seconds after the engine starts. However, exhaust gas temperature with the advanced exhaust valve timing slightly increases after 200 seconds. It is considered that the advanced blowdown process causes an increase in exhaust gas temperature. In spite of such increase, the exhaust gas temperature is still higher when the valve timing is retarded. Consequently, it was concluded that retarded exhaust valve timing is beneficial for increasing exhaust gas temperature while ensuring stable engine operation under the conditions tested. Exhaust valve open timing ( CA, reference @ 130 ATDC) -12 0 12 T em pe ra tu re ( o C ) 0 200 250 300 350 400 50 Sec 100 Sec 150 Sec 200 Sec Fig. 5. Exhaust gas temperature curves with the change of exhaust valve timing Fig. 6 shows rate of heat release (ROHR) curves with the change of exhaust valve timing. The pressure curves to calculate ROHR are obtained by averaging 200 consecutive cycles under each test condition. The ROHR curves for ±12°CA cases are obtained from the cylinder pressure data. As shown in Fig. 6, the ROHR curves reach their peak values at around 50° CA, and decrease later on. This means that combustion rate and flame propagation speed are highest at around 50° CA and they are rapidly decreasing after that point. Note when the exhaust valve timing is advanced, the peak value is higher but combustion ends earlier than the other cases. On the other hand, when the exhaust valve timing is retarded, the peak value at 50° CA is relatively lower but heat release from fuel continues for a longer time. These phenomena can be explained by flame speed. When flame speed is faster, burning rate at the earlier stages is higher, resulting in a higher value of ROHR at around 50° CA. However, a rapid decrease in ROHR occurs because considerable amount of fuel is already burned. The opposite is true when the flame speed is lower. So the retarded exhaust valve timing is beneficial for higher exhaust gas temperature because heat release continues to a later stage of the expansion stroke. Crank Angle (CA) -150 -100 -50 0 50 100 150 R O H R (J / o C A ) -10 -8 -6 -4 -2 0 2 4 6 8 10 12 12CA Retard 12CA Advance Fig. 6. Rate of heat release curves with the change of exhaust valve timing When using the variable timing camshaft, it is possible to change the exhaust valve open and close timings but there is no way to change the cam profile. It means when the valve open timing is retarded or advanced, the valve close timing should be changed accordingly. Only the exhaust valve timing is changed in this experiment, and the intake/exhaust valve overlap period must be altered because there is no change of intake valve timing. During idle operation of an SI engine, pressure in the intake manifold is much lower than that in the exhaust manifold. This causes a backward flow of exhaust gas during valve overlap period and increases the amount of residual gas in the next cycle (7). When the exhaust valve timing is advanced, the exhaust valves are closed earlier, and the overlap becomes shorter. On the contrary, when the exhaust valve timing is retarded, the overlap becomes longer, and therefore, the amount of residual gas increases. The higher the residual gas fraction, the lower the flame speed and engine stability, especially in idle conditions. However, flame lasts longer due to a slow burn process, and the exhaust gas temperature increases. Effects of spark timing Fig. 7 shows variations in exhaust gas temperature with the change of spark timing. As shown in this figure, exhaust gas temperature increases when spark timing is retarded to BTDC 5°CA from the baseline case. On the contrary, when spark timing is advanced to BTDC 15°CA, the exhaust gas temperature rapidly decreases compared with the baseline case. Fig. 8 shows the ROHR with the change of spark timing. As shown in this figure, it is obvious that the crank angle at which maximum heat release occurs moves to the right side, resulting in an increase in the exhaust gas temperature. Spark timing ( CA, BTDC) 5 10 15 T em pe ra tu re ( o C ) 0 200 250 300 350 400 50 Sec 100 Sec 150 Sec 200 Sec Fig. 7. Exhaust gas temperature curves with the change of spark timing Crank Angle (CA) -150 -100 -50 0 50 100 150 R O H R (J / o C A) -15 -10 -5 0 5 10 15 BTDC 5CA BTDC 10CA BTDC 15CA Fig. 8. Rate of heat release curves with the change of spark timing Based on the above results, it is obvious that there are two important factors to hold the flame longer time in the combustion chamber, in order to obtain a higher exhaust gas temperature. First, a retarded exhaust valve timing increases the valve overlap period and residual gas fraction under idle and cold start conditions. The flame propagation becomes slower and late burn occurs to raise exhaust gas temperature. Second, when spark ignition timing is sufficiently retarded, start of combustion is delayed and the flame lasts longer before the exhaust valves open. This late burn or partial burn phenomena are favorable for increasing the pressure and temperature of the exhaust gas. Exhaust valve open timing ( CA, reference @ 130 ATDC) -12 0 12 Sp ar k tim in g ( o C A , B T D C )