ShaunSG

iTrader: (1)

Squish is different from quench.
Squish involves displacement and turbulent mixing of gas and fuel.
Quench involves drawing heat out of the nearby gasses in the end gas area, raising detonation threshold.

Swirl is more often confused with tumble.
Swirl and tumble are totally largely separate from squish or quench, though swirl can be intensified by squish in some cases.

From a power standpoint..
It is easy to have too much swirl and/or tumble. It is possible to have too little.
It is possible to have too much or too little squish and/or quench.

SaabTuner

It's odd that they differentiate between the two. I'd imagine it rather difficult to generate squish, using the piston and head, without the associated cool boundary layers of both cool mechanical parts creating a cooler area of gasses. Odder still because the gas near the outer edge of the piston, and behind the intake valves, will always be cooler than the bulk of the gas in the center of the combustion chamber due to the boundary layers present over the metal.

I'd be interested in reading a qualitative study on the functional difference between squish and quench as I'm not convinced the cooling effect associated with quench has any measurable affect on detonation, rather that it is always associated with squish and squish does have a measurable affect on detonation. I find it extremely puzzling that the slight change in temperature of the end-gas, caused by the extremely brief period during which the piston is close to the head, would so dramatically affect the formation of a detonation wave which is known, and has been seen (photographed at length by NASA, 912, 857, and 761), to originate within the already partially-burnt gasses. Lowering the bulk temperature of the mixture, on the other hand, would have the obvious benefit of reducing combustion temperatures and pressures. But why would quench over such a small area have any significant effect? There have been a number of attempts to explain how end-gas temperature in general might be able to affect detonation, most notably that certain fuel species may decompose in the end-gas, as it is heated, then react at a very late point during the combustion process to setup a detonation wave. But I've found very little evidence to support those explanations. The most obvious evidence against that explanation is that chemicals like butane do not decompose prior to autoignition. If that theory were correct, chemicals like butane would not be capable of sustaining knock (it is), or at least would have the highest octane resistance of any other chemical (meta-xylene is notably higher), so that theory cannot be correct. In fact, most SAE papers I've seen which "study" knock have a characteristic sentence, similar to this one, somewhere within the first two paragraphs: "Knock is generally agreed to originate within the end-gas/unburntmixture/etc...." I have yet to see an SAE paper which successfully photographed the detonation wave itself in the manner that NASA did, so I find those paper's assumptions somewhat unsubstantiated. But perhaps I am missing something?

I appreciate the correction, though I still tend to refer to them both as swril since tumble is merely a swirling motion perpendicular to the cyllinder's axis while swirl is the same motion parrallel to the cyllinder's axis. They both have a somewhat similar effect on detonation and air-fuel mixing. Never hurts to have a single word to differentiate the two swirling axis though, so thanks.

Also the reference to the SRT-4's piston blurrs those lines to an extreme extent. The "hump" on one end of the piston uses squish to generate tumble and simultaneously quench the mixture in that area. At that point, the distinction between the three phenomena becomes more of an intellectual excercise.

-Adrian

strikethree

you guys kick ***. thank you for all of the responses. i appreciate it greatly.

strike (Dave)

ShaunSG

iTrader: (1)

Reciprocating piston motion doesn't happen without piston speed or acceleration, but differentiation is necessary to better understand what goes on and build on top of that.

Existing temperature gradients can always be augmented.

Again differentiation is necessary because how each affects port discharge (think charge coming back up at the intake valve in tumble), how effectively it mixes the cylinder contents (think setting up circular motion in a rectangular tumble plane), crank angle over which the motion lasts before breaking down, are all different (think changing aspect ratio of rectangular tumble plane across intake and compression strokes).

I would like to see some photos of these pistons. At the speed the piston is moving and the dwell time involved, I don't see a squish initiated movement sustaining itself especially in tumble. Swirl and tumble involve large vortices that last over a considerable crank angle range. To start them large and have volume reduction and still keep the vortice is much more usual than starting small and very violent motion with enough energy to keep inertia as volume expands considerably.


=====


I still haven't read the papers that you provided a long time ago. I get discouraged when I look at the photos and can't even make anything out to begin with.

I think it is good that you are questioning things at such micro level and perhaps may be able to re-write generally accepted truth if you uncover something new. At the same time considering the extreme complexity of these types of things (fluid, thermo, chemical) you could well have missed something. Fluid, thermo, and chem are tough areas even individually but interacting they are exponentially difficult to calculate and test. Combustion, tire, and aero dynamics are the 3 main areas even F1 has few solid answers on, but their focus is biased almost completely to the latter 2 because they are more 'tacklable'.

Absolutely no offense intended, but if you can miss something as simple and calculable as inertial relation to engine speed and displacement, then it is almost certain you are missing something here. I'm not talking down because I myself have made too many simple mistakes of my own to even want to attempt something like this alone and start from the very bottom in the micro realm. I just don't have that brains, time, energy or confidence right now, to do that. I have other more pressing macro problems I'm trying to solve or other macro concepts to grasp, macro dynamics to test.

I wish you the very best of luck in establishing the truth in that area. You need to take your questions to an open professional forum. You know where they are. Please let me know if, when, and where you do. I want to hear what people more familiar with combustion microdynamics have to say and gain some perspective before trying to wade though all the papers. It makes things go quicker.


Best regards

trinydex

iTrader: (12)

mmm i don't know why no one has looked at it the way i have... but here's a simpler picture.

much of car modding is extracting lost horsepower... very few things actually ADD horsepower (think removing exhaust restrictions etc.)

so on this note i've always seen timing advance as a way to extract lost horsepower... it's free. so think of the most simple boosted scenario... you have an ideal supercharger that pumps the exact same boost from idle all the way to redline, or you have a high comp engine (it would be the same thing, almost). given this scenario if you ran 0 timing all the way through you'd be losing power, or not making very much. by the very nature of rpm increase you need to add timing in order to not LOSE horsepower.

increasing boost is one of the few things that actually makes MORE power, so with the ideal supercharger you can increase boost, but you'd still have to advance timing as you go up in rpms.

now tha last point that everyone has been trying to make is that with more boost you NEED LESS timing, and you can only increase boost up to a point where your timing reduction gets stupid.

SlowCar

iTrader: (4)

Hi Dave, i myself am trying to find the limits of timing too. I had a chance to chat with Dynoflash-Al while he was in phx and was told that the stock ecu will not allow/start pulling timing >11 degrees(i'm on stock ecu w/ ecutek flash). I've seen another person on stock ecu + ecutek with 12 - 14 degrees.
I will get retuned in a few days time. My plan is to set boost @26psi, AFR = 12.5, and start cranking up timing till i see it either plateau/knock then back it off 1%. I'm at 11 degrees now@23psi boost.
No one could give me a definite answer and every car having a different knock threshold, i just have to experiment and see how it goes.

SaabTuner

Fair enough. Provided the piston can do enough thermal dissipation of the end gas, there is one loophole that would allow both quench and NASA's modified knock theories to coexist. If the end-gas is cooler, it is less likely to undergo pin-point or homogenous autoignition (not knock), either of which would exacerbate the sharp rise in the temperature of combustion as a whole which could then cause knock.

I've been reading over this interesting paper: http://powerlab.mech.okayama-u.ac.jp...2/C90_P309.pdf

If one looks closely at the piston thermal boundary conditions based on zone, one notices that the temperature of the air immediately against the piston surface is more closely related to the distance from the spark plug than from the proximity to the cyllinder wall, which would be the only distinction for the quench pad. That fact is very apparent when comparing zone 23 with zone 1. Zone 23 is a quench pad and is hotter than zone 1, which is in the middle of the piston. Zone 23, however, is closer to the spark plug. I don't think zone 23 will be doing any special measure of cooling, despite being a quench pad on the piston, if it is not any cooler than certain parts of the middle of the piston.

I'm having a hard time finding actual data to support the idea that adding quench pads has any measurable effect on the combustion temperatures around that region of the piston compared to the temperatures which would have been there without the quench pads. The benefits of the "squish" generated by the quench pads is obvious, but if there is no good evidence to a drop in temperature due only to the quench pads, there's nothing to support the idea that they do any "quenching". But if you have any good papers with evidence, I'd definitely be game on reading!

I still disagree, here's why: though swirl parrallel to the cyllinder's axis may have different propogation and turbulence properties when compared to a swirl perpendicular to that axis, so too would a swirl at a 20* inclination, 30* inclination and so on. Then, of course, different combustion chamber shapes, as well as swirl-ratios, would also affect both propogation and turbulence.

In the following paper, examining various swirl inclination angles on a lean-burn engine, even small variations in inclination could change the turbulence by almost a full order of magnitude: http://powerlab.mech.okayama-u.ac.jp...2/C90_P437.pdf

Since there are a theoretically infinite number of axis upon which one could generate swirl, all of which may have drastically different propogation and turbulence properties, I find differentiating between them based on relative position to the cyllinder's axis somewhat less meaningfull. Worse still, if it turns out that the majority of swirling in most engines is not perfectly parrallel, nor perfectly perpendicular, to the cyllinder's axis, the majority of swirling would neither be "swirl" nor "tumble". (maybe we should call it "stumble" )



And so it shal be done. Taken from here: http://www.sportcompactcarweb.com/pr...0scc_projneon/




=====


I agree. Schliren photography is more complicated than even regular photography. I had to stare at them for quite some time to see what they try to describe in the text. I think one of the references in NACA 912 involves a method for reading the schliren images and explains it a bit.

I admit to often missing things, but it generally has to do with things I don't know, rather than things I don't understand. The piston ineria problem was an unfortunate exception.

The problem here is I assumed that, because it was based on a sinusoidal motion, it would behave like a typical sin wave, or even a typical combination of sin and cosine waves; it doesn't. If I double the stroke (amplitude) of a sin wave, its first and second derivatives are also doubled. With a piston, at some rod-ratios, doubling the stroke less than doubles the inertial forces, but at other rod ratios it more than doubles the inertial forces.

Fortunately, it only more than doubles the inertial forces when the rod is less than the length of the stroke, which isn't possible anyway. But I haven't checked the boundary for 10-20-30% increases in stroke and requisite reductions in rod ratio from deck height.

In light of all that, I hardly think it's fair to say I made a mistake in something as "simple" as piston inertial forces; piston inertial forces are hardly simple! Rather, instead, that I made a "simple" mistake in not bothering to run the numbers before I ran my mouth.

Are you kidding?! Engineers bicker more than tuners, and just look at NASIOC!

Although I do find some perverse pleasure in watching other people bicker over things at which any pragmatist would balk, and they'd all forget about it after two or three .

Anyway ... back to our regularly scheduled programming ...

ShaunSG

iTrader: (1)

Adrian, re: zone 23.. convective gas temps in that zone were higher so the temperature gradient is either similar to, or augmented, vs other zones. It doesn't contradict squish or quench.
========

Re: quench. Yes papers on it are difficult to find especially in shareable pdf. I can only share what I have learnt from tests carried out in the industry and what a whole host of experienced people/engineers say and have presented on. Not every test or concept is documented. In simple terms, in most engines you have heavily stratified charge and pockets of fuel and air sitting around. When the piston speeds up and squeezes them out, it both evenly cools the area over the air-fuel is squished (quench pads and head) and as it is displaced at speed and mixed violently it both cools the mix and other surfaces that fuel might have been ejected onto. The temperature gradient is augmented by lowering the temperature in the end gas areas via these now cool pads. There is overall reduction in the cylinder, but it is not as great as at the quench areas where fuel spread has been perfect. Neither squish nor quench happens without the other. The effects of both are reduced in lower power and more efficiently designed / run engines.
==========


Re: inclination. Of course there are in betweens! There are marked differences between the two points at either end, but no one is arguing there are no in betweens. The differences I mentioned are very real. Major manufacturers in the US and in Japan have all investigated this to death, or done collaborative research and presented on it.
========


No I'm not kidding. The good guys when discussing technical issues don't let anything get in the way of good discussion and there is lots of perspective shared on either side of any issue that is really good to have in mind when researching.

You need to dive into areas that have the highest concentration of people that are qualified to get into the quench details and then gain perspective. NASIOC is not a professional engineering forum, neither is this one.
==========


Best of luck and best regards

ShaunSG

iTrader: (1)

I don't see those pistons generating swirl since there is no biasing visible in the pictures.

SaabTuner

But it does contradict the idea that quench pads will result in any measurable increase in cooling. If quench regions on the pistons do not run any cooler than the rest of the piston, you won't increase the combustion cooling without increasing the surface area of the piston.


I would urge you to ask them if they have conducted any tests specifically designed to seperate quench and squish. I have a feeling that they have only assumed quench when it occurs in combination with squish and have not actually measured the drop in temperature due to quench alone. The data I have found supports that idea as it indicates that the quench pad is not notably cooler than the rest of the piston. This, of course, doesn't contradict the idea that the surface of the piston cools the end-gas. Rather, it contradicts the idea that adding a "quench pad" somehow increases that cooling. And, if the quench pad does not add to the cooling of the end gas more than any other part of the piston, the term "quench" becomes somewhat superfluous.

Actually, by the time the piston reaches near TDC to create squish, the combustion mixture is already hotter than the piston surface by about 300*C due to the increase in heat from compression. Forcing the mixture out at high speed would heat that region of the piston further making the quench pads even less effective at cooling the end-gas later during the combustion process.

If squish and quench cannot be found seperately, then the only data on quench would be based on piston temperatures and thermal boundary conditions. As I noted in the paper before, both of those are not exceptionally low in the quench pad region. But if there have been tests to specifically measure the cooling effect quench pads have on combustion gasses near the pad, I'm game for a read.

No doubt. But at what inclination do you divide between swirl and tumble? Why bother referring to tumble as a seperate phenomenon from swirl when it is merely another axis of swirling? If some swirl generator generates a vortex 30 degrees off the cyllinder's axis, do we call it swirl or tumble? It just doesn't seem like a meaningfull distinction to me.

I'll look into it. Any particular forums you have in mind?

-Adrian

SaabTuner

Chrysler designed those pistons specifically to generate swirl, though you'd call it tumble. It's hard to describe. Basically it forces the mixture in the small volume side of the hump up and across the top of combustion chamber and down the other side, generating a swirl/tumble in the larger area.

Maybe I'll make some drawings in paint later or something. If you think about it for a while, I think you'll see what I mean.

Fourdoor

iTrader: (11)

Sean@SG and SaabTuner:

Please take the rest of your discusion to PM's. It is nice and civilized with no name calling, so I will not delete the posts already made, but they are WAY off topic and cluttering up this thread. It would actually be a good topic for a new thread if you want to continue a public discussion of the subject just say so and I will split this thread.

Thanks,

Keith

EVIL_EV0

iTrader: (6)

From my days of DSM tuning.....

I've found that you set a pig rich AFR.....and run base timing......

Then you set your boost where you want it to be.....

Then you start leaning out your AFR until its at the ratio you want.....(this actually will add overall engine timing in the process)

If you still dont see knock you then can begin to add timing......you can add timing to different RPM ranges....this makes fine tuning easier than fine tuning of boost. It should be noted that you may need to pull some timing to prevent knock in specific areas of the band as well.

So ..... set boost ..... set fuel .... adjust with timing.

EvilBlueEvo8

iTrader: (26)

Thats a pretty neat way of going about this.

SaabTuner

Okie doke. Sorreh. You can split it if you want so that people can read some of the documents if they search. But I doubt there's much more to add to that discussion until more factual evidence can be found for either side. Still fun while it lasted!