Technical help - combustion and C/R
- Thread starter Barry Hart
- Start date
Barry Hart
New member
This is a subject that I don't know enough about this myself - so I might be talking cr*p here and exposing a lack of knowledge (something I never mind admitting to when appropriate as I want to understand everything as best I can and never want to mislead anyone if I am not sure) - but what little I do know is that there are a few different causes of ignition noises and I am wondering if some may make a slightly different noise to others - I don't know.
If an engine ran on such bad fuel or had flaky bits of glowing carbon stuck on the valve or piston, or ran on such a high boost that the C/R was too high, or had an inlet manifold leak making it too weak, etc - then - if any of that caused the mixture to self ignite/explode before the point of ignition the whole calorific value of the fuel is released almost instantaneously rather than expanding while it continues to burn as the piston passes over TDC to release most of its pressure rise as the piston is conveniently starting to move down the cylinder - when the crank angle can best exploit the rising pressure and turn it into useful rotational torque through the con-rod to crank angle.
Under normal ignition the gradual expansion of burning gas while the piston is also moving down the cylinder and increasing the volume/space that the gas can expands into is a happy coincidence of burning petrol based fuels and smooths the resulting forces on the piston that is twisting the crankshaft, the pressure rising significantly just as the piston starts its movement down the bore and until it has reached a point after which - although the pressure is running out - so to is the useful angle of the crank to turn it into useful torque through the con-rod.
If it detonates instead, then this more sudden release of energy operates on the piston much earlier and either before it reaches TDC or so soon afterwards that the crank is in a poor angle with the rod to move it rotationally in a smooth manner and the high pressure simply presses the con rod down onto the crankshaft journal with little chance of turning into useful rotation (as the con-rod to crank angle is too much in one
straight line). This shocks the rise (or early fall) of the piston - slowing down the crankshaft momentarily and this could I suppose be heard as a knock.
If on the other hand the problem was marginal and caused by poor squish, pooor swirl, or fuel which is not mixing well so only a very small part of the area under the piston in a small pocket detonates but almost imediately the spark still ignites the remaining fuel - so there are effectively two flame paths accross the piston/head - is it possible that this would result in a slightly less severe shock - perhaps more of a tinkling noise (which I have heard many years ago when old engines are struggling up hill in a high gear) - I don't know - but this could possibly result in two descriptions emerging - or then again perhaps it is just the different way it is expressed and the noise is actually the same.
Most modern cars have anti-knock sensors that retard ignition so perhaps that is why it is a rare occurence these days - although this preventative measure - in itself - shows that there is yet another phenomenon - which is that different fuels and different mixtures of fuels burn at slighly different speeds and so there must be another pre-ignition scenario in which the fuel does not explode until after it is ignited by the spark but then just reaches its high pressure release point too early - either because it has been sparked too soon or because it is a mixture that burns quicker than expected (otherwise retarding it would make no difference) but it then burns too quickly - so delaying the spark recovers the error? I guess detonation is still possible in this scenario if the fuel ignited too early then is still compressed some more by the continuing rise of the piston and added to the pressure rise from the fuel burning then results in a delayed detonation of the remaining fuel that hasn't burnt yet?
So it seems to me that you can have (1) detonation before the spark, (2) detonation in small pockets but still the main burn resulting from the spark and (3) normal burning but sparked too soon so the timing of the pressure rise of the fuel is too early for the crankshaft position or delayed detonation. Whether or not these all have the same noise or not - I am honestly not sure - but I imagine there is every reason to assume they might be slightly different?
Please put me right if any of that is wrong!
Baz
If an engine ran on such bad fuel or had flaky bits of glowing carbon stuck on the valve or piston, or ran on such a high boost that the C/R was too high, or had an inlet manifold leak making it too weak, etc - then - if any of that caused the mixture to self ignite/explode before the point of ignition the whole calorific value of the fuel is released almost instantaneously rather than expanding while it continues to burn as the piston passes over TDC to release most of its pressure rise as the piston is conveniently starting to move down the cylinder - when the crank angle can best exploit the rising pressure and turn it into useful rotational torque through the con-rod to crank angle.
Under normal ignition the gradual expansion of burning gas while the piston is also moving down the cylinder and increasing the volume/space that the gas can expands into is a happy coincidence of burning petrol based fuels and smooths the resulting forces on the piston that is twisting the crankshaft, the pressure rising significantly just as the piston starts its movement down the bore and until it has reached a point after which - although the pressure is running out - so to is the useful angle of the crank to turn it into useful torque through the con-rod.
If it detonates instead, then this more sudden release of energy operates on the piston much earlier and either before it reaches TDC or so soon afterwards that the crank is in a poor angle with the rod to move it rotationally in a smooth manner and the high pressure simply presses the con rod down onto the crankshaft journal with little chance of turning into useful rotation (as the con-rod to crank angle is too much in one
straight line). This shocks the rise (or early fall) of the piston - slowing down the crankshaft momentarily and this could I suppose be heard as a knock.
If on the other hand the problem was marginal and caused by poor squish, pooor swirl, or fuel which is not mixing well so only a very small part of the area under the piston in a small pocket detonates but almost imediately the spark still ignites the remaining fuel - so there are effectively two flame paths accross the piston/head - is it possible that this would result in a slightly less severe shock - perhaps more of a tinkling noise (which I have heard many years ago when old engines are struggling up hill in a high gear) - I don't know - but this could possibly result in two descriptions emerging - or then again perhaps it is just the different way it is expressed and the noise is actually the same.
Most modern cars have anti-knock sensors that retard ignition so perhaps that is why it is a rare occurence these days - although this preventative measure - in itself - shows that there is yet another phenomenon - which is that different fuels and different mixtures of fuels burn at slighly different speeds and so there must be another pre-ignition scenario in which the fuel does not explode until after it is ignited by the spark but then just reaches its high pressure release point too early - either because it has been sparked too soon or because it is a mixture that burns quicker than expected (otherwise retarding it would make no difference) but it then burns too quickly - so delaying the spark recovers the error? I guess detonation is still possible in this scenario if the fuel ignited too early then is still compressed some more by the continuing rise of the piston and added to the pressure rise from the fuel burning then results in a delayed detonation of the remaining fuel that hasn't burnt yet?
So it seems to me that you can have (1) detonation before the spark, (2) detonation in small pockets but still the main burn resulting from the spark and (3) normal burning but sparked too soon so the timing of the pressure rise of the fuel is too early for the crankshaft position or delayed detonation. Whether or not these all have the same noise or not - I am honestly not sure - but I imagine there is every reason to assume they might be slightly different?
Please put me right if any of that is wrong!
Baz
Neil Haughey
New member
I am pretty sure you are correct Baz, this is what is described in the text books ISTR, which is why I prefer not to dwell too much on pre-ignition. Detonation and pre-ignition are not necesarily the same thing. Its not impossible for example for pre-ignition to occur just prior to the spark, then it would feel just like normal. Like you say part of the charge can detonate after normal spark ignition. According to the books as little as 5% of the fuel mix is enough to detonate, however I don't know if knock sensors can pick this up. In theory if a small amount of the charge hasn't mixed well or the cylinder has a hot spot and can result in detonation of a small part of the charge then this could be happening on a fairly regular basis I guess? I suppose if one was to attach a large number of sensors to a flywheel one could measure the acceleration of the crank at lots of different points, it would be interesting to know in reality how smooth the torque into the crank really is.
Barry Hart
New member
Neil, I think that the torque varies quite a lot during each cycle, but the mass of the flywheel and all the pistons, rods, cams etc and the transmission to some extent damps it down.
Our own 3 litre turbo is fitted with a 968 clutch (because I wanted to reduce effective flywheel weight (no dual mass flywheel etc) and didn't want to waste torque or invite snatch as a result of twisting or winding up a torsion damper) and the transmission feels uncannily smooth. However (remembering that all the gears are in contact with each other all the time - not just the ones in the gear you selected) in low gears and moderate acceleration you can (unusually for a 944) hear gearbox scrawl (thats what I call it anyway) the same as you can hear in a 3.2 911 carerra - and this occurs when the input shaft is speeded up and then slows down between each power pulse and in so doing speeds up and slows down the output shaft. If the speed is relatively low then the small gap between the teeth on each shaft of all the gears not in gear - first touches on the driven face and then the decelerating face - shaking backwards and forwards and making a fine jingling noise until the revs rise so there is not enough time between each pulse for the gear to slow down enough to hit the other face and the noise goes away.
This crankshaft torsion fluctuation cycle (or phenomenon) also has an effect at high speed under high loads and this is when pure torque is not the only influence upon acceleration and engines with more firing sequences do not waste as much time (or lose torque in strain energy) re-accelerating the mass of the engine, transmission and car between firing impulses and even though for comparable capacity the smaller cylinders are not quite as individually powerful, they may manage to creep up to a higher speed when the bigger cylinder engine of the same capacity hits a brick wall more predictably.
Finally - if an engine has a steep bhp curve - it must also be considered that running on beyond the peak bhp revs - even though the engine is less powerful and less torquey there - still picks up the next gear at a higher point in the rev band - which overall may result in greater acceleration overall.
I hope to soon be testing our engine at different gear change revs to establish exactly what power/torque curve achieved the best results.
Baz
Our own 3 litre turbo is fitted with a 968 clutch (because I wanted to reduce effective flywheel weight (no dual mass flywheel etc) and didn't want to waste torque or invite snatch as a result of twisting or winding up a torsion damper) and the transmission feels uncannily smooth. However (remembering that all the gears are in contact with each other all the time - not just the ones in the gear you selected) in low gears and moderate acceleration you can (unusually for a 944) hear gearbox scrawl (thats what I call it anyway) the same as you can hear in a 3.2 911 carerra - and this occurs when the input shaft is speeded up and then slows down between each power pulse and in so doing speeds up and slows down the output shaft. If the speed is relatively low then the small gap between the teeth on each shaft of all the gears not in gear - first touches on the driven face and then the decelerating face - shaking backwards and forwards and making a fine jingling noise until the revs rise so there is not enough time between each pulse for the gear to slow down enough to hit the other face and the noise goes away.
This crankshaft torsion fluctuation cycle (or phenomenon) also has an effect at high speed under high loads and this is when pure torque is not the only influence upon acceleration and engines with more firing sequences do not waste as much time (or lose torque in strain energy) re-accelerating the mass of the engine, transmission and car between firing impulses and even though for comparable capacity the smaller cylinders are not quite as individually powerful, they may manage to creep up to a higher speed when the bigger cylinder engine of the same capacity hits a brick wall more predictably.
Finally - if an engine has a steep bhp curve - it must also be considered that running on beyond the peak bhp revs - even though the engine is less powerful and less torquey there - still picks up the next gear at a higher point in the rev band - which overall may result in greater acceleration overall.
I hope to soon be testing our engine at different gear change revs to establish exactly what power/torque curve achieved the best results.
Baz
Posts made and opinions expressed are those of the individual forum members
Use of the Forum is subject to the Terms and Conditions
