Detonation myths

[Steve W]

Great article on engine detonation:

http://www.avweb.com/news/columns/182132-1.html

[IROC]

Great article. I actually read once that when choosing which octane rating for your engine, you should pick the one that causes slight detonation under load.

Anyway, it is interesting that the author brings up the "old wore out beater car" example. I'm sure we've all seen it - the car that pings like mad but yet we keep seeing that same car on the way home from work every day.

Mike

[Dantilla]

Very informative. Thanks. But it made me really miss my old Bonanza.

Maybe someday I will be able to say that both my car and airplane have air-cooled flat sixes.

[banjomike]

What the?

TBO, TBO, LOP!?!? what the hell are these TLAs?

[TimT]

"rich of peak" - "lean of peak"

ROP- LOP

that avweb site has some great reading

[Lorenfb]

So, for those of you with performance chips that have noticed pinging lately
in these hot summer months, just consider the pinging as a problematic myth
as some would have you believe by reading the linked article. The article
should be forwarded to Porsche R/D as they may have been too concerned
about pinging by their conservative ECU timing maps and the use of knock
sensors in all later model Porsche engines.

If only the Internet had been available earlier for the Porsche engineers at
Weissach, they would not have been concerned about the effects of pinging
in causing broken rings and damaged pistons and as such would have
"pushed" the timing maps for greater performance and not used knock
sensors. How could they have been so naive in their research and resulting
engine designs?

Myths???????????

The Internet conundrum; What's trash info or not?.

[RoninLB]

The author stressed that CHT is the leading indicator to pinging. So if you have good CHT the imperfections of complete combustion are not a detrement. We don't adjust the mix, maniflld psi, etc that a pilot does. We control the rpm and loading. The outside ambiant temp and altitude are fixed.

fwiw, a good method for controling CHT on the street is to adjust your load and rpm based on the direction of the CHT.. huh!
Let's say we were just N of LA heading south into the hills on I-5 at 75mph in 5th. It's summer time ambiant daytime temps. Before we hit the hills our CHT was 250-260deg for the past 4hrs. Now all of a sudden we glance and see the CHT creeping up as we enter the foothills. As the hill gets steeper we're still doing 75mph in 5th but our CHT just increased 50deg in 2mi. So we downshift to 4th, decrease speed load, and run the rpm that lowers the CHT to a more reasonable level to our liking. The CHT is a reading that registers almost immediately on the CHT ga.

If we are only monitoring the oil temps the lag can be 5-8min.+ before we figure out the engine is heating up. That's to dam long at high CHTs for my act. So now you not only have an infered high CHT to bring down but you also have high oil temps to deal with. Oil temps don't immediately drop because we just decided to deal with it. It takes awhile, it takes a long while.. So you probably created a situation where the engine is vulnerable to damaging detonation for a long time.

[ChrisBennet]

Excellent info Steve and Ronin. I wonder how fast the Motronic head temp sensor reacts.
-Chris

[Mark McClure]

On that topic what is the typical relationship between CHT and Oil temp. I understand that that may be a hard questions since the oil temp depends upon the ability of the cooling system but there must be a relationship between the two. I have heard that the optimum CHT for a 911 is 109deg C

[Randy Webb]

Good one, Loren.

Engineers who really want to get on the edge but control pinging use an acoustic snesor in the cylinder most likely to ping and back off based on that readout - it is done in real time. Saab has used such systems, and may have pioneered them.

[ChrisBennet]

I realize it's inconvenient to read such a long article so here are a couple of excepts from it concerning the FAA's proposed detonation standards. Look at the comments.

A few years back, some of the research done by General Aviation Modifications Inc. (GAMI) in Ada, Okla., began to raise further questions in my mind about detonation. George Braly, the founding genius and chief engineer, started running a highly instrumented engine deep into detonation, and recording data that no one had ever seen before. Pelican's Perch #43: Detonation Myths

Light Detonation:


Medium Detonation:

[mtelliott]

The author noted that "the spark fires at about 20 to 25 degrees before top dead center (TDC)". Is this the case for Porsche engines? I always thought that spark occurred closer to TDC?

[ChrisBennet]

Originally posted by mtelliott
The author noted that "the spark fires at about 20 to 25 degrees before top dead center (TDC)". Is this the case for Porsche engines? I always thought that spark occurred closer to TDC?

Spark advance varies of course but for a 3.2 it peaks at over 24 degrees at full throttle. Under some part throttle conditions it can be closer to 40 degrees of advance.
-Chris

[RoninLB]

Originally posted by Mark McClure
what is the typical relationship between CHT and Oil temp.

I have heard that the optimum CHT for a 911 is 109deg C

afaik it's a direct relationship that may need some interpertation under variable conditions.. meaning if we're driving in cold or hot ambiant temps that can affect oil temp. So a desert cruise that generates high oil temp may only slightly affect CHT. A winter cruise thru the mountains starts with low oil temps till the cylinder loading generates high CHT which will slowly raise the oil temps. The idea imo is to monitor the CHT to prevent overheating or even overstressing.

109deg C equals 228deg F is cold CHT for our cars imo

and afaik knock sensors come in various flavors because they are all poor indicators.. better that nothing, but not great. I guess 3 flavors would be better that 1 being used.

[stlrj]

Mixture is a major player in all this. Changing the mixture changes the speed of combustion (the speed of the flame front), as well as the temperature and pressure of combustion. For any given conditions, moving away from about 50 ROP in EITHER direction (rich or lean) will typically increase the margin from detonation. This does NOT mean that detonation can occur only at 50 ROP, only that it is most likely there.

You can see this for yourself, for all the old radial charts show it, and both Lycoming and TCM produce charts that show CHT peaks at about 30 ROP, while maximum power occurs at about 80 ROP. The 15:1 ratio occurs, essentially at what we all know as our familiar "peak" EGT on our engine monitors.

What this is telling me is that we are making a big mistake by disconnecting our 02 sensors and setting our mixtures richer than peak EGT with the idea that richer is cooler when just the opposite is happening. At 30 degrees rich of peak, cylinder head temps peak, moving richer to 50 degrees ROP is a danger zone for detonation, and richer yet you finally get max power at about 80 degrees ROP.

What a price to pay for a little more power, a risk of more detonation and abnormally high cylinder head temps.

Cheers,

Joe

[RoninLB]

Originally posted by stlrj
What this is telling me is that we are making a big mistake by disconnecting our 02 sensors and setting our mixtures richer than peak EGT with the idea that richer is cooler when just the opposite is happening.

what this is telling me is that aircraft is a different game.

[RoninLB]

Originally posted by RoninLB
what this is telling me is that aircraft is a different game.

this is about a similar game.

[stlrj]

If you were at peak EGT where your engine was designed to operate, you would never experience the problems encountered when leaning your engine from 80 degrees ROP to 50 ROP where detonation is likely to occur down to 30 ROP where CHT's top out.

Cheers,

Joe

[beepbeep]

Originally posted by Randy Webb
Good one, Loren.

Engineers who really want to get on the edge but control pinging use an acoustic snesor in the cylinder most likely to ping and back off based on that readout - it is done in real time. Saab has used such systems, and may have pioneered them.

Acoustic sensors are old-school and are not used by SAAB anymore. Best way to control knock is by using so called "ION-knock sensing". That way, ECU can sense detonation after each combustion cycle, for each separate cylinder and adjust timing in real time before next cycle even begins. If knock occurs again, fuzzy logic will wind back timing even more and richen the mixture until it stops.

Here is block-shematics of how it works:


Here's some more info:
http://www.fs.isy.liu.se/~larer/Projects/main.html

Acoustic sensors are unreliable, cannot quite distiguish wich cylinder is pinging and don't work at high RPM's. They are better than no knock cntrol at all, of course.

While article is very informative, it is somewhat hard to implement in real-life driving beacuse number of reasons:

1. Most cars have closed-loop ECU's.
2. You are not in control of mixture/ignition in the car
3. Most cars have knock-sensors or better
4. Most cars have much better combustion chamber designs than dinosaur aircooled engines used in airplanes.
5. Airplane engines have fix ignition timing with magnetos (!)
6. Transient loads on car engines are much higher.

Loren: you are always picking on aftermarket chips for Porsches. I believe most are aware that tradeoff for extra power is less knock resistance in case of Carrera engines. That being said, there are perfectly knock-resistant motor/ECU combinations that are software limited to certain power-level for other reasons.

SAAB Trionic (which I'm generally quite accustomed to) is limiting torque for example. With other words, it will vary ignition timing, boost pressure, mixture and even opening of its FBW-throttle co achieve same torque troughout the range regardless of height/heat/engine wear (and degrading that if usning less octane). It's not uncommon to see ECU open the throttle only 70% at WOT in cool weather, just beacuse it knows it achieved it's programmed mass-flow.

There is some real extra power that can be mustered from engines equipped with similar types of ECU's by only software.

Of course, mechanical modifications will make it easier/more reliable but I'm quite annoyed by your "all chips are fraud, they don't make power and/or are dangerous for the engine"

World is not black and white. Chips do work but your mileage may vary from implementation to implementation, depending on architecture beneath.

[klatinn]

World is not black and white. Chips do work but your mileage may vary from implementation to implementation, depending on architecture beneath.

Correct IMHO. OEM ECU's are designed for the worst case conditon the engine will likely encounter in regards to fuel, climate and maintenance. Plus an additional safety margin. Porsche used to test 911's in the summer in the Sahara desert in Tunesia for example.
In addition they have to be designed for the worst case production tolerances.

The hp gains in production cars in the last years have come from 3 reasons:

1. Better understanding of the combustion processes.
2. Therefore better cyl. head shapes
3. Better compensation for engine environment situations, therefore running closer to optimum.

A correctly designed aftermarket chip can optimize the ignition timing and fueling for a particular engine and fuel quality. The tradeoff is of course that the engine no longer can tolerate as wide a range of fuel qualities and climates as the OEM version.

I agree also with this though for aftermarket tuning:
The relationship between ignition timing, AFR and knock is often very poorly understood even by many called tuners. As a manuf. of AFR meters I get many messages like this:

"My engine develops max power at 9.5 AFR without pinging. I'm running 16# of boost, and I am running 34 deg. advance at 5500 RPM."

This is of course a waste of fuel, power and extremely polluting.
What happened is that the extremely rich mixture causes a slow burn, this slow burn put the peak pressure point (ppp) in the cylinder into the reasonable area of 15-18 ATDC where he created some power without knocking.
Instead of running that much advance, this guy could just retard a lot more and run much leaner, getting more power, knock resistance, and lower consumption in the process.

Aircraft engines as mentioned in the article run at fixed advance because of the reliability issues of variable advance. They run magnetos also because of reliability (works even with elektr. failure). The advance in aircraft engines is fixed to the optimum for max takeoff power at takeoff RPM.
They also run double plugs for reliability and power reasons. Double plugging, like in an 911, reduces the flame-front travel time/distance. Specially for the big-bore, short stroke engines used. This reduces the required advance to get to the correct ppp and reduces the likelyhood of knock because of lower pressure rise before and shortly after TDC where chamber volume decreases or stays nearly constant while the pressure rises.
Modern combustion chambers put the plug smack in the center and are optimized to get the shortest flame-front travel possible. Short travel means less advance required for optimum ppp.

Just my $0.02 of ramblings.

Klaus