Investigation on the transient behavior of a high compression two-wheeler single cylinder engine close to idling.
The introduction of stricter emission legislation and the demand of increased power for small two-wheelers lead to an increase of technical requirements. Especially the introduction of liquid-cooling over air-cooling allows the introduction of higher compression ratios, which improves power output as well as thermodynamic efficiencies and thereby fuel consumption.
But an increase in compression ratio also introduces further challenges during transient behavior especially close to idling. In order to keep the two-wheeler specific responsiveness of the vehicle, the overall rotational inertia of the engine must be kept low. But the combination of low inertia and high compression ratio can lead to a stalling of the engine if the throttle is opened and closed very quickly in idle operation. The fast opening and closing of the throttle is called a throttle blip.
This paper describes the development of a procedure to apply reproducible blipping events to a vehicle in order to derive a deeper physical understanding of the stalling events. Goal is to identify influencing factors that determine whether a blip will lead to a stalling of the engine or not. These factors contain e.g. the timing of the blip within the working cycle or calibration parameters like average idle speed and ignition timing.
The investigations are carried out on a motorcycle with a water-cooled 400cc single cylinder engine from the performance segment with high compression ratio and an engine management system with port fuel injection. Several thousand blipping events are automatically applied and evaluated with the focus on the stalling probability.
CITATION: Heikes, H. and Jost, F., "Investigation on the Transient Behavior of a High Compression Two-Wheeler Single Cylinder Engine Close to Idling," SAE Int. J. Engines 10(1):2017
The robustness of internal combustion against stalling at sudden throttle opening is an important measure for dimensioning of the engine (e.g. overall inertia), development of the engine management system (e.g. accurate fuel metering and ignition timing) and calibration. In order to perform an objective optimization a reliable methodology for applying throttle blips (also known as tip-ins) and evaluating the system response is mandatory. This paper will focus on the following aspects:
1. development of a representative throttle blip which can be applied automated and reproducible
2. classification of the mechanisms which lead to a stall and introduction of stall modes
3. introduction of the stall probability (ratio of all stalling throttle blips compared to all applied throttle blips in a certain operating point) as an objective measure for comparison
4. identification and investigation of influencing factors on the stall probability
5. definition of a procedure for an optimized calibration of idle operation
Motorcycle engines are usually equipped with one throttle valve per cylinder to create small intake manifold volumes and therefore ensure good transient behavior. This transient responsiveness of the induced intake air is much higher than in typical automotive applications. According to the sensitivity towards stalling during fast transients single cylinder engines can be considered the worst case as combustion only occurs once per 720 [degrees]ca to accelerate the crank train. The kinetic energy stored within the crank train (e.g. crank, con rod, piston and fly wheel) is the only source to overcome the necessary work of compression in the next working cycle. Stalling can take place if the crank speed and therefore the kinetic energy stored are too small to ensure a complete compression process in the next working cycle. In this case the piston will revert its moving direction before top dead center and the crank is turning backwards until the engine comes to a complete stop.
Several design parameters are influencing the necessary work of compression and therefore the blip stall behavior. Aside of the overall inertia, the compression ratio is one of these factors. Therefore engines with high power density and high performance are in focus. Another parameter is the displacement of the engine and the present cylinder charge.
Derived from the physical understanding of the blip stall phenomenon a representative vehicle from the single cylinder segment is chosen. Table 1 contains the relevant technical data. All investigations have been conducted in idling operation as this can be considered the worst case situation for fast throttle transients because of the low engine speed and therefore the low kinetic energy of the crank train. Furthermore the gradient of positive or negative throttle blips can be considered as independent from engine speed. But the influence of a fast changing throttle on the cylinder charge is more critical at low engine speeds, as full load air charge is already reached at throttle angles significantly below the maximum throttle opening.
Figure 1 depicts the experimental setup of the demonstrator vehicle. It consist of two main building blocks. The upper part is the data acquisition system, where the analog signals for ignition current, injection voltage, crank voltage from the inductive sensor and the voltage of the throttle position sensor are measured with a time based resolution of 0.01ms. These signals are a basis for the automated evaluation of the blipping events. Especially the crank signal has a high relevance, as it is used to transform the measured time basis into a crank angle basis for all the other measured signals. The lower part gives more details on the blip automation system. This system is based on an embedded micro controller which also reads in the crank and the ignition signal. The controller creates a trigger signal for the throttle position actuator which contains an electrical motor. This motor is connected via a lever to a bowden cable which replaces the original bowden cable on the mechanical throttle body of the motorcycle. The crank signal is utilized to create the blipping event at a defined crank position in the working cycle. The ignition signal is used as a reference to indicate the phase of the crank signal and therefore distinguish between firing and gas exchange top dead center.
Definition of Blip Event
As a first step around two hundred blipping events have been applied manually with the goal to create very short blip durations. Within this two hundred events the engine stalled four times. Out of this four stalling events the throttle position profile with the shortest duration is defined as the reference target blip as shown in Figure 2. The throttle is opened to around 80 % of the maximum throttle opening within 45 ms and closed again immediately within 50 ms. The derived target blip has an overall duration of 100 ms. Converted towards the crank angle basis this transfers to around 1000 [degrees]ca at idling speed. In the second step this target throttle blip has been reproduced via the automation system. As visible the manual and the automated throttle blips are matching well with a slightly longer duration of the automated blip. The reproducibility of the automated blips is adequate. To create an even worse blipping event, coming from the 100 ms target blip another shorter reference blip is defined. The target for the blip duration is 50 ms and it is not utilizing the full throttle opening but is limited to around 45 %. By this measure the overall blip duration of the automated blips can be reduced to around 70 ms which is slightly higher than the target of 50 ms but still sufficient. Dependent on the instantaneous engine speed during the blip this translates to about 720 [degrees]ca which is the length of one working cycle. As 45 % throttle opening already leads to the maximum cylinder charge at idle speed, this allows for a transient jump between idle air charge, full load air charge and back to idle air charge within three consecutive working cycles. This shorter and therefore more critical 70 ms blip will be used throughout all the following investigations.
Apart of the blip duration, the positioning of the blip within the working cycle is a main influencing factor. Therefore the blip start position is defined as shown in Figure 3. This value describes the distance between blip start and the firing top dead center of the working cycle in which the blip occurs. The throttle position profile is depicted over the crank position for three different blip start positions. Dependent on this position the influence of the throttle blip during the intake event is different.
Figure 4 illustrates the high engine speed dynamic which is induced by a sudden throttle blip. The blip start position in this case is exactly at firing top dead center. The lower part contains the throttle position as well as the high-resolution signal of the inductive crank sensor over time. This signal is used to calculate the instantaneous engine speed from every tooth on the trigger wheel which is given in the upper plot as well as a crank angle basis for all measured signals. The Average engine speed in idling is around 1700 rpm. During the compression stroke of the working cycle in which the blip takes place the high cylinder charge leads to a very strong decrease of the crank speed to values around 350 rpm afterwards the combustion accelerates the engine again to about 2800 rpm. In this case there is no stalling and the engine returns to idle speed because the throttle is already closed when the next intake stroke begins.
Observed Stall Mechanisms
In order to detail out the mechanisms leading to a stalling of the engine first the definition of the numbering of the working cycles is discussed in Figure 5. The numbering always starts with the working cycle 1 (WC1) which is the one before the working cycle with the actual blip event. The working cycle with the blip is numbered as WC2. It is followed by the working cycle WC3 right after the blip start and in certain cases by on additional work cycle WC4. The figure displays a non-stalling throttle blip with a blip start position of 400 [degrees]ca aTDC.
In general several different stall mechanisms can arise from a throttle blip . Based on the definition in the figure above the blip induced stall can occur in three different working cycles. Reason for the stall can either be the fast throttle opening or the fast closing event of the throttle blip. Derived from this systematic depicted in Figure 6 four different stall modes can be distinguished. The detailed processes which lead to the stalling are explained on the basis of representative example plots.
Figure 7 contains an example for the stall mode A. As a reference for the pressure and the engine speed behavior of non-stalling blip events these signals are added as dashed grey lines in the following plots. The engine is running in idling with closed throttle and therefore high intake pressure vacuum at intake valve closing (1) in the working cycle before the blip (WC1). The blip start position is at 250 [degrees]ca aTDC which is before the intake stroke of WC2. During the following intake phase no significant pressure drop can be measured in the intake (2).The ambient pressure level at intake valve closing indicates that the maximum cylinder charge is trapped inside of the combustion chamber. Because of the low average crank speed in the working cycle with the blip, the kinetic energy is not high enough to deliver the needed work to compress the cylinder charge. Therefore the piston does not reach TDC. It stops and starts to accelerate backwards (3) driven by the pressure from the compression phase. As soon as the intake valves are opening again the spring forces lead to a deceleration and the engine stops moving (4). For this behavior no measures can be taken within the ECU aside an increase of the average idling speed. This is the typical stall behavior if no combustion takes place in the working cycle containing the blip.
If the air charge is ignited during the compression phase and a combustion occurs, the behavior looks slightly different. This stall mode B is depicted in Figure 8. The base situation is the same as in the situation mentioned before. The engine is running in idling with closed throttle (1). The blip start is at 170 [degrees]ca aTDC which is before the intake stroke. The ambient pressure level at intake valve closing indicates that the maximum cylinder charge is trapped inside of the combustion chamber (2). Because of the low average crank speed in the working cycle where the blip occurs, the kinetic energy is not high enough to deliver the needed work to compress the cylinder charge and overcome the pressure build up from the starting combustion before top dead center. Therefore the piston does not reach TDC. It stops and starts to revert backwards (3) driven by the pressure rise from combustion until the intake valves open again. As soon as the intake valves open the overpressure inside of the combustion chamber is released into the intake manifold, leading to a pressure overshoot because of the completely closed throttle (4). As the pressure from combustion is significantly higher than the pressure rise from compression only, the maximum backwards rotation speed is higher than in Figure 7. After the intake valves are opened the engine starts to decelerate again as it starts moving towards TDC again against the gas forces in the intake manifold. After that the engine rotation is fluctuating back and forth with high gradients until the engine stops rotating. In order to avoid this behavior an ignition timing close too or after TDC seems to be preferable. This measure inside of the ECU will be investigated in detail at a later point.
Figure 6. Systematic of blip stall mechanisms Stall relevant Stall Number of working throttle phase WC cycles between in blip event WC with Blip Start and WC with stall opening WC2 [DELTA] WC = 0 WC3 [DELTA] WC = 1 closing WC4 [DELTA] WC = 2 Stall relevant Air-Fuel ratio throttle phase in blip event opening [lambda] = [[lambda].sub.target] WC2: [lambda] [not equal to] [[lambda].sub.target] (Misfire) closing WC2: [lambda] = [[lambda].sub.target] WC3: [lambda] [not equal to] [[lambda].sub.target] (Misfire) Stall relevant Stall Reasons for stall throttle phase mode in blip event opening A kinetic energy of crank train is not sufficient for necessary work of compression B additional in-cylinder pressure rise from combustion before TDC C no torque produced in WC2 [right arrow] engine speed drop [right arrow] if cylinder charge in WC3 still high because throttle still open, stall via mode A closing D high torque produced in WC2 [right arrow] engine speed rise [right arrow] throttle closes between injection and intake stroke [right arrow] cylinder charge In WC3 low and rich [right arrow] no torque produced in WC3 [right arrow] early inflammation of cylinder charge during intake phase of WC4
The third and fourth stall modes C and D described in the systematic in Figure 6 are highly dependent on the air-fuel ratio of certain working cycles. The rapid throttle opening also has influence on the mixture formation and the charging process . This work however will focus only on the influence of the qualitative behavior of the air-fuel ratio in the individual working cycles. The stall mode C can occur if, during the working cycle where the blip starts (WC2), the ECU is not injecting the correct fuel mass to match the rapidly increasing cylinder charge. This would lead to a significantly lean air-fuel ratio and therefore a delayed combustion or even a misfire. The low amount of torque produced from such a combustion will lead to a reduction of the engine speed in the following working cycle WC3. If the throttle is still open in this phase the engine can stall in this working cycle following stall mode A or B. This behavior has never been observed on the used demonstrator as the ECU is always reacting with proper post injections to ensure the desired air-fuel ratio. Although it has not been observed here, this stall mode is part of the systematic, as it is usually referred to as one of the key root causes for engine stalling at throttle transients close to idling.
The stall mode D however has been detected and is depicted in Figure 9. The steps (1) and (2) are comparable to the behavior in stall mode B (Figure 8) with a slightly later blip start position. But in this case the kinetic energy of the crank is sufficient to provide the work of compression (3). The working cycle during the blip is finished and the working cycle after the blip (WC3) starts with a full load combustion leading to a high rise in engine speed. The fuel injection during WC3 is rather long as the opened throttle still indicates a full load air charge. Now the throttle is closed before the intake stroke of WC3 takes place. This leads to a small, close to idle air charge in WC3 as can be seen from the high intake vacuum at (4). The air-fuel ratio is for this WC is on the rich side, as the fuel has already been injected based on the high throttle opening. The rich mixture leads to a misfire in WC3 resulting in no engine speed rise during power stroke (5). The residual gas which is kept inside of the combustion chamber now contains a rich mixture out of air and fuel which has been heated up during the compression stroke. As visible in (6) the fresh air charge is inflaming during intake stroke in the following working cycle WC4. This inflammation starts before the intake valves are closed so the influence on the engine speed is limited but there is already a clear pressure rise inside of the intake manifold which exceeds the ambient pressure by far (the flat spot at around 1150 mbar is caused by a limitation of the pressure sensor). As soon as the intake valves are closed the high pressure inside of the combustion chamber and the starting compression are decelerating the engine towards stand still. The root cause for the inflammation during the intake phase is not completely clear but it is assumed that this is a result of the mixture of the stoichiometric fresh air charge with the rich residual air charge from the preceding misfiring working cycle. There seems to be a relevant level of pre-reactions in the rich mixture. These local chemical reactions can have an accelerating impact on the kinetics of auto-ignition . A comparable behavior of back firing cycles preceded by extremely slow-burning cycles is also described in .
Investigating all blipping events which have been measured for this work show that more than 98% of the investigated blips are stalling according to stall mode A and B. Table 2 contains the total count of all applied blips of 20800. Out of this number only 757 stalls are observed. This is not the regular frequency which can be observed in series production as several parameters have been varied to test the sensitivity of the stalling behavior towards them (details to be provided in the following sections). It has to be noticed that the 747 stalls following mode A and mode B have a strong dependency. Both depend on the needed work to compress the cylinder charge. Mode B in addition has to overcome the pressure rise from combustion. It is not possible to eliminate all of them by a retardation of the ignition. From this investigation it is not possible to state how many of the 721 stalls can be omitted and how many would transfer into stalls following stall mode A.
Investigation of blip/stall behavior
The applied blips can lead either to a regular combustion and an increase of the engine speed or to a stall of the engine. As an objective measure for the stability against stalling after a throttle blip the stall probability is introduced as follows:
stall probability = [number of stalling blips / number of applied blips] [??] 100 [%] (1)
As described the position of the blipping event within the working cycle seems to have a high influence on the stall behavior. In order to investigate this influence the blip start position is varied within steps of 10 [degrees]ca from firing top dead center to 720 [degrees]ca after top dead center. For every position 100 blipping events are applied. As the positioning of the blipping events is based on the tooth signal of the trigger wheel it is not possible to place blipping events within the tooth gap. This leads to two areas of about 20 [degrees]ca length where no significant number of blips are investigated. These areas are marked in Figure 10 close to 360 and 720 [degrees]ca. This figure shows the total number of applied blips at a certain blip start position as well as the number of stalled blips. It is visible that only blips with a starting position before 450 [degrees]ca are leading to a significant number of stalls. Relevant for a blip to lead to a stall is the trapped charge inside of the combustion chamber. The comparable flat behavior of the number of stalling blips in Figure 10 between 0 and 450 [degrees]ca shows that it is almost irrelevant for the mentioned stall mechanism when the blip starts as long as it is before or in the beginning of the intake phase. The diagram includes all the stall mechanisms.
Why this is the case can be seen in Figure 11. This depicts the intake manifold pressure at intake valve closing over the blip start position. As soon as the blipping event starts within the intake phase the pressure at intake valve closing starts to drop. This is the case as the reduction of the intake vacuum because of the opened throttle will not become completely effective before the intake event is finished. The reduced pressure indicates a reduced trapped air charge and therefore a reduced needed work of compression. For later start positions the pressure at intake valve closing is not influenced anymore for this working cycle.
Aside of the blip start position, the engine speed is an important influencing factor. In this case not the average engine speed over several working cycles should be evaluated because of its overlying fluctuations. These fluctuations especially in idling and for single cylinder engines are induced by the cyclic variations of the combustion process. A fast combustion with early combustion phasing leads to a higher acceleration of the crank train than a slow combustion with retarded combustion phasing. This investigation is using the engine speed during the working cycle in which the blip takes place between 0 and 540 [degrees]ca (leaving out the compression phase) as a reference. This is therefore an average engine speed from the instantaneous engine speeds derived from the 54 tooth times of the trigger wheel. Figure 12 contains the number of blips and stalls as well as the stall probability as a function of this engine speed. As the engine speed was not varied actively during the measurement of the different blip start positions the average speed over the working cycle during the blip has almost a normal distribution. It is clearly visible that the stall probability is increasing with lower speeds as there is not enough kinetic energy for the following compression stroke. This shows how important a single combustion event can be in the context of stalling during high transients in idling. The worst case would be a slow combustion with low torque production in the working cycle right before the blip event (WC1) followed by a fast combustion with early pressure rise before top dead center in the working cycle during the blip event (WC2). Especially for motorcycle engines with high power density, the idle stability is a critical situation as they are designed for high volumetric efficiency at high engine speeds rather than high turbulence at idle operation. This trade-off is a challenge for the combustion stability .
Investigation of operation parameters in idling
As a result of the aforesaid findings the average idle speed as well as the ignition angle will be investigated in detail. Therefore these two parameters are varied within the engine management system as shown in Table 3. The investigations are carried out at a blip start position of 170 [degrees]ca aTDC.
The results are depicted in Figure 13. It is clearly visible that for low engine speeds the stall probability is increased significantly. The highest stall probabilities are measured in combination with an early ignition angle leading to a worst case stall probability of over 50 %. The early ignition angle leads to a stronger pressure rise before TDC and therefore a higher demand in work of compression. For very late ignition angles therefore late combustion phasing the stability of the combustion gets worse leading to higher engine speed fluctuations and an increasing stall probability. According to the boundary line the best behavior is reached with an ignition timing close to top dead center. The calibration point for the idle speed and ignition timing is always a trade-off between fuel consumption, idle stability, noise level and the stability against stalling. To keep a high stability against stalling the operating point should always be calibrated on the right side of the drawn boundary line in the Figure 13. In general higher engine speeds have a positive influence on stability against stalling, a reduction of engine speed in idling improves the fuel consumption , .
In order to give a hint for the optimization of the calibration data in idling the fuel consumption for this operation range is shown in Figure 14 together with the determined boundary line. Obviously the fuel consumption is increasing with rising engine speed. Furthermore the influence of the retarded ignition and therefore the reduced combustion efficiency is visible. In this case a significant reduction of fuel consumption at a constant stall probability could be achieved via an earlier ignition angle and an increased idle speed. The applicability of this measure is limited due to an increasing noise level with rising engine speed. As there is no objective measure to evaluate this based on this investigation, the current series calibration point marked in Figure 13 for ignition angle and average engine speed is considered as an optimum.
Focus of this work is the definition of a methodology to investigate the stability of a small single cylinder motor cycle engine towards stalling because of sudden throttle opening in idling. Therefore an automation procedure to apply sudden throttle opening is introduced and the resulting blip events are compared to manual worst case blipping events. The system is able to apply realistic and reproducible throttle blips and allows therefore an optimization of the idle conditions.
The significance of the instantaneous engine speed in the working cycle in which the blip occurs is proven. It is shown that a single retarded combustion process or misfire can lead to a stalling of the engine at a sudden throttle opening. The only counter measure to reduce the stall probability in such a situation is an increase of the average engine speed in idling to gain a certain safety margin towards under speed right before the blip.
Furthermore the significance of the ignition angle has been investigated in detail. Early ignition timing leads to an increased pressure build up inside of the combustion chamber before top dead center and therefore an increased risk of stalling.
It is shown that the developed methodology can be used to optimize the stability against stalling by providing an objective measure for the stall probability. The defined process is used exemplarily to optimize ignition timing and average engine speed in idling for the demonstrator vehicle.
[1.] Jost, F., "Technische Analyse von Motorsteuerungssystemen fur Motorrader mit Fokus auf das Transient-Verhalten im Leerlauf bei schneller Drosselklappenoffnung," Master's Thesis, Karlsruhe Institute of Technology, 2016.
[2.] Manz, P., "Influence of a Rapid Throttle Opening on the Transient Behaviour of an Otto Engine," SAE Technical Paper 922234, 1992, doi:10.4271/922234.
[3.] Willand, J., Nieberding, R., Vent, G., and Enderle, C., "The Knocking Syndrome - Its Cure and Its Potential," SAE Technical Paper 982483, 1998, doi:10.4271/982483.
[4.] Saito, S. and Adachi, M., "STUDY ON THE PROCESS OF BACKFIRING," SAE Technical Paper 2001-01-1881, 2001, doi:10.4271/2001-01-1881.
[5.] Corti, E. and Forte, C., "Idle Stalling Phenomena in High Performance Spark Ignition PFI Engines: an Experimental Analysis," SAE Technical Paper 2011-24-0158, 2011, doi:10.4271/2011-24-0158.
[6.] Kajitani, M. and Nonami, K., "High Performance Idle Speed Control Applying the Sliding Mode Control with H Robust Hyperplane," SAE Technical Paper 2001-01-0263, 2001, doi:10.4271/2001-01-0263.
[7.] Kerns, J. and Surnilla, G., "Fuel Assisted Idle Speed Control for Lean Burn Gasoline Engines," SAE Technical Paper 2006-32-0009, 2006, doi:10.4271/2006-32-0009.
COMPRESSION, COMP, CO - compression stroke of working
crank - voltage signal from inductive crank position sensor [V]
[DELTA]WC - Number of working cycles between WC2 and WC with stall
ECU - Engine Control Unit
EXHAUST, EX - exhaust stroke of working cycle
INTAKE, IN - intake stroke of working cycle
[lambda] - air-fuel ratio [-]
[n.sub.i] - target engine speed for grid measurement [rpm] i = -2....+2
[n.sub.mot] - engine speed [rpm]
p - pressure [mbar]
[p.sub.0] - ambient pressure [mbar]
[p.sub.int] - pressure inside of the intake manifold [mbar]
[p.sub.int,idle] - pressure at intake valve closing with completely closed
throttle valve [mbar]
[p.sub.int,IVC] - pressure at intake valve closing [mbar]
POWER, PO - power stroke of working cycle
syncro - crank position synchronous reference point
TPS - Throttle Position Sensor value [%]
[U.sub.ign] - ignition voltage signal [V]
[U.sub.inj] - injection voltage signal [V]
WC - working cycle defined by a length of 720[degrees]ca consisting of
power, exhaust, intake and compression stroke
WC1 - working cycle before the blip event starts
WC2 - working cycle in which the blip event starts
WC3 - working cycle after the working cycle where the blip event starts
WC4 - second working cycle after the working cycle where the blip event starts
[zw.sub.out] - ignition angle [[degrees]ca bTDC]
[zw.sub.i] - target ignition angle for grid measurement [[degrees]ca bTDC] i = -8....+5
Henning Heikes and Felix Jost
Robert Bosch GmbH
Table 1. Technical data for demonstrator vehicle Displaced volume 400 cc Number of cylinders 1 Compression ratio 13:1 Number of valves 4 Trigger wheel 36-2 teeth Cooling System liquid Table 2. Count of all applied blips, all stalls and respective stall modes Applied blips Stalls Mode A Mode B Mode C Mode D 20800 757 26 721 0 10 Table 3. Values for variation of calibration parameters in idling Value Minimum Delta Maximum Unit Target for average 1530 75 1830 rpm engine speed Target for ignition angle 28.5 3 -10.5 [degrees]ca bTDC
|Printer friendly Cite/link Email Feedback|
|Author:||Heikes, Henning; Jost, Felix|
|Publication:||SAE International Journal of Engines|
|Date:||Feb 1, 2017|
|Previous Article:||Development of an ammonia reduction aftertreatment systems for stoichiometric natural gas engines.|
|Next Article:||A Numerical Investigation of Ignition of Ultra-Lean Premixed [H.sub.2]/Air Mixtures by Pre-Chamber Supersonic Hot Jet.|