Engine Detonation Effects re Knock Sensor Strategy 3

[PDMeat]

Hello all,

I have a question that I think fits in this forum as it's more related to engine health than purely sensors and electronic control. The question is: What are the long term effects of modern engine control closed loop knock detection strategies on engine structural integrity?

Put another way, as a modern automotive ECU listens for "onset of knock", just how harmful are these "mini-knocks".

The application is for modified modern turbo cars with aftermarket tuning (chips/piggyback/flash tunes). There are several internet forums where people (even tuners companies) argue whether or not the mini-knocks the knock sensor listens for are actually harmful or not, esp if the engine boost is raised from say 8psi to 15psi on 91/93 octane fuel.

The chief concern is that on completely stock turbo engines at 100HP/Ltr this is ok, but modified up to 150HP/Ltr or more the amount of small knocks (vs time) will increase and the knocks themselves will be larger, harmful and should be avoided via retarding timing.

TIA

 

[ivymike]

As a former piston designer, I wouldn't discount that concern off the cuff- it certainly sounds plausible enough. To really answer the question you'd have to know what cylinder pressure looks like during the "mini knock" events that you describe. If knock avoidance is based upon accurate measurement of crankshaft or block acceleration, and the "mini-knocks" are happening due to mid-stroke autoignition of a small amount of remaining mixture, and you're looking at a relatively short crankshaft, then there's a pretty good chance that there's no harmful effect. My primary concern would be whether peak cylinder pressure reaches excessive levels. Under certain conditions you can break piston rings by having persistent peculiarities in the cylinder pressure trace, but a slight blip midway through the expansion stroke seems unlikely to be problematic.

 

[mattsooty]

The knock control in modern EMS systems is a lot more intelligent than just a knock sensor detecting any old knocking event then the timing being retarded.

It is generally the case that the ECU will make a quantitive decision of the intensity of the knock with the spark being retarded a greater amount with a higher intensity.

The amount of spark retard is then stored as an adaption which corresponds to that particular cylinder for a particular speed/load window (the speed/load range is broken up into a number of different sites which can have different adaptions applied to them).

What this effectively means is that the 'mini knocks' that are heard are indeed that because the big knocks will have already caused a certain amount of retard when they first occured and this retard is then stored within the ECU memory.

Thus if the boost pressure is increased, or indeed a lower Octane fuel is used the damaging knocks only occur a few times before they are 'adapted out'.

Obviously increasing the boost pressure miles above the OEM intended may result in the spark being retarded to its clamp, whereby it cannt retard any further. This is where the MIL would be illuminated and I would expect serious engine damage to then occur.

There is also strategies that work in the opposite direction increasing the advance when the engine is run in a speed/load window without any knock present.

Just for interests sake a rule of thumb that is often used is that peak knocking pressure (ie the additional cylinder pressure caused, over & above normal, due to knock) is roughly the engine speed divided by a thousand - so for example at 3000rpm a 3 bar increase is tolerable without knock system invervention.

MS

 

[PDMeat]

Thanks IvyMike and MattSooty.

If we were to assume that the knock onset from increased boost etc was coming just from the spark timing be too advanced, what max cylinder pressures, heat, shock etc does one see with a knock control strategy running? I've read that cylinder pressure for a typical passenger car with normal WOT combustion may only rise to around 1200psi.

It seems MattSooty is saying that a 3bar increase in cylinder pressure from a "mini-knock" is near (but not past) the point where the ECU will begin to retard timing. So if this is correct, does that mean an engine that normally sees 1000psi cylinder pressure at WOT that mini-knocking up to around a total of 1050psi will be tolerated before the ECU steps in and starts to adapt/pull timing?

My concern is if we went from 100HP/Ltr @ 8psi to 150HP/Ltr @ 15psi, that knock onset detection from fine (under 1050psi) to flagged might go from 1050 (@8psi boost) to like 1300+PSI with significant heat, shockwaves etc before the ECU could correct it.

My unprofessional guess is the ECU would probably react fast enough and we wouldn't see 1300psi before it could act.

I'm assuming that most (quality) modern engines with block girdles, cross bolted main caps, etc are overbuilt at least 50% (prob 100%) and can take the added stress without any undue effects.

Does this sound reasonable?

 

[Surestick]

I've always been curious how the ECU handles a change from a low to a high octane fuel.
Is the timing advanced a bit every once in a while to keep track of where the detonation limit is?

 

[mattsooty]

Yes & No.....

Yes, the ignition retard applied to the base table, due to the action of the Knock Control System will diminsh to zero after a certain number of knock free cycles. However, when this reaches zero you are then left with the advance values populated in the base table. The spark wont keep advancing past these values until it reaches knock.

Having said that, if the spark is mapped very aggresively reliying heavily on the KCS in normal operation (especially in a forced induction application) it may be advanced further than knocking would usually allow, if a fuel with an octane higher is used.

Hope that helps

MS