Stewartb
Final Approach
Those tests induce detonation using “severe” conditions. That makes perfect sense for testing but does not apply across the board to standard ops in the GA fleet.
What not efi like cars
- Thread starter Piper Dreaming
- Start date
Eric Pauley
Pre-takeoff checklist
I didn't say it wasn't. My point is that this is still not consensus on what detonation is. GAMI's tests show that the FAA (consensus?) threshold does not characterize detonation in a useful way.
The question is did you graduate from any of them?
Regardless, it seems every time you post you dig your hole deeper. Take this topic. You state detonation happens all the time which seems to imply you think that it is acceptable. No where do you post any reference that states detonation (of any level) is acceptable.
Well guess what. Only no engine detonation is acceptable. Its actually been a certification requirement since day one. But don’t take my word on it:
CAR 13.153
FAR 33.47
And here is what Lycoming states as well. You’ll notice those pesky little key words again: maintained and operated… properly.
And this is hardly a one-off on your part. Perhaps another recent example on your rivet thread, and I quote:
Not quite. Rivet substitution would fall under an alteration, not a repair. However, if you were to use an E series 6-6 rivet it would not be a major alteration either as the E 6-6 is an acceptable substitution for a DD 6-6. This would be basic mx 101 especially for a
an experienced mechanic like yourself.
Should I go on? There are others as well. So you see why myself and other mechanics can’t take you seriously. Perhaps take the time to educate yourself a bit more before posting on the next topic.
Silvaire
En-Route
- Joined
- Oct 10, 2012
- Messages
- 4,679
- Display Name
Display name:
Silvaire
As in you walked "through"? I can't imagine why anyone would go to A&P school twice.
PaulMillner
Line Up and Wait
>> The engine can function without detonation
That kind of begs the question... what constitutes detonation? If you look at the advisory circulars from the 1940s and the 1950s, detonation was detected using water-cooled Sperry accelerometers which flashed when detecting events. The FAA standard for "without detonation" was less than so many flashes per minute. Yet, the flashes were coming from single detonation events, out of the thousand+ combustion events per minute in the cylinder.
Water cooled accelerometers are no longer made, so the industry (auto and aviation) moved to fiber optic pressure sensors. But, how to analyze that data and compare it to the Sperry baseline that most of our engines and fuels were approved by? The FAA took a not-very-good science approach, and defined a new standard, immortalized in an ASTM specification. Unfortunately, they made math errors (!) and their assumptions were not good. The ASTM-approved method suggests that existing engines using 100/130 fuel don't meet the detonation standard. But... those engines and fuels have been operating fine for decades. Something is wrong.
Others took a different approach... they resurrected those Sperry water cooled accelerometers, and ran them in parallel to the new fiber-optic technology. In fact, the company that makes the fiber optic intra-cylinder pressure sensors did much the same thing (for the automotive world). This results in a different standard than the ASTM detonation spec. It all correlates back to those allowable flashes-per-minute... which are random, single detonation events. It's like the difference between medicine and poison... it's all in the dose!
The other aha was that the distribution of detonating events, both in the time domain and the severity domain, is different between leaded and unleaded fuels. The mogas world figured out it was different back in the 80's, but aviation engine operating conditions are different, so we had to come to our own understanding. In any case, looking at both frequency of detonations and their severity, and comparing to the Sperry standard of the past 100 years, leads to different correlations defining acceptable detonation... or rather how many events you can have and still call it certifiable "without detonation."
If you run any engine long enough, given the laws of probability, it's going to have detonation events... but one can still have those random events *and* deem the engine as operating without detonation... it's all in the definitions and the mathematics. Cool, eh?
Paul
That kind of begs the question... what constitutes detonation? If you look at the advisory circulars from the 1940s and the 1950s, detonation was detected using water-cooled Sperry accelerometers which flashed when detecting events. The FAA standard for "without detonation" was less than so many flashes per minute. Yet, the flashes were coming from single detonation events, out of the thousand+ combustion events per minute in the cylinder.
Water cooled accelerometers are no longer made, so the industry (auto and aviation) moved to fiber optic pressure sensors. But, how to analyze that data and compare it to the Sperry baseline that most of our engines and fuels were approved by? The FAA took a not-very-good science approach, and defined a new standard, immortalized in an ASTM specification. Unfortunately, they made math errors (!) and their assumptions were not good. The ASTM-approved method suggests that existing engines using 100/130 fuel don't meet the detonation standard. But... those engines and fuels have been operating fine for decades. Something is wrong.
Others took a different approach... they resurrected those Sperry water cooled accelerometers, and ran them in parallel to the new fiber-optic technology. In fact, the company that makes the fiber optic intra-cylinder pressure sensors did much the same thing (for the automotive world). This results in a different standard than the ASTM detonation spec. It all correlates back to those allowable flashes-per-minute... which are random, single detonation events. It's like the difference between medicine and poison... it's all in the dose!
The other aha was that the distribution of detonating events, both in the time domain and the severity domain, is different between leaded and unleaded fuels. The mogas world figured out it was different back in the 80's, but aviation engine operating conditions are different, so we had to come to our own understanding. In any case, looking at both frequency of detonations and their severity, and comparing to the Sperry standard of the past 100 years, leads to different correlations defining acceptable detonation... or rather how many events you can have and still call it certifiable "without detonation."
If you run any engine long enough, given the laws of probability, it's going to have detonation events... but one can still have those random events *and* deem the engine as operating without detonation... it's all in the definitions and the mathematics. Cool, eh?
Paul
Last edited:
PaulMillner
Line Up and Wait
That's true, but not the definition. Like your reference says, pre-ignition occurs BEFORE the spark plug fires. Detonation occurs as the burning mixture compresses pockets of end gas (that haven't yet combusted) and causes them to explode rather than burn... combust at greater than the speed of sound in the medium. That sharp pressure rise stresses the engine mechanically, and it also disrupts the boundary layer of gas at the cylinder head, greatly increasing heat transfer. That's why we see the CHT rise, and because the heat it transferring from the gasses to the metal, we see the EGT decline, the heat's been lost to the cylinder. Unfortunately, the cylinder isn't designed to reject that much heat, and can attain disastrous temperatures, melting the piston, etc.
Dan Thomas
Touchdown! Greaser!
- Joined
- Jun 16, 2008
- Messages
- 11,360
- Display Name
Display name:
Dan Thomas
Just opening the throttle too fast can cause detonation briefly. High cylinder pressures at low RPM give time for the molecules to break down and autoignite.
Unless its changed in the last 7 years, the FAA certification point is based on a specific % margin from when a defined number of “flashes per minute” is detected during operation at the max and min mixture fuel flows. Its not just based on the number of flashes alone. The standard has a built-in cushion when operating the aircraft within its certified power limitations. It was my understanding that one of the UL problems was being able to meet these existing detonation margins which I believe are 10 or 12% before the defined detonation occurs. I’ll look for the guidance once I get home.
True. But it depends. The problem arises when you bring your aircraft to the shop with CHT spikes and intermittent rough running and its determined to be detonation. Even though you believe these maybe “random” events, all the airworthiness guidance points to no detonation is acceptable if the aircraft is maintained and operated within limitations. It’s a slippery slope when releasing aircraft from mx which has caught several, both owner and mechanic, off guard with undesirable results on both sides.
In the context of my reply, I was simply showing pre-ignition and detonation were separate items, not provide a definition which is complex as you and others have noted.
Last edited:
PaulMillner
Line Up and Wait
yeah... But no. Not in any common understanding of the English language. The FAA defines 'no' detonation as less than a certain number of detonation events per minute. We have to be a little bit clever to discuss this in a way that has engineering meaning. Saying no detonation without referencing those definitions is meaningless... you're not shedding light, just heat!
Agree. So lets define them. One reference, AC 33-47-1, defines two types of detonation: incipient (5-9 flashes/min) and limiting (10-20 flashes/min) with limiting detonation the trigger point for the certification detonation margins. And those margins are defined as 12% outside of the maximum and minimum fuel flow limits.
So in the context of my post above, how would you, a field mechanic, or another pilot determine the difference between incipient (FAA “no detonation”) and limiting (certified detonation)?
Dan Thomas
Touchdown! Greaser!
- Joined
- Jun 16, 2008
- Messages
- 11,360
- Display Name
Display name:
Dan Thomas
A four-banger at 2500 RPM fires 5000 times per minute. A six-cylinder, 7500 times per minute. 10 to 20 detonation events in 5000 firings is 0.2 to 0.4%. In the six it is between 0.13 and 0.27%.
Small numbers. Certainly not "constantly detonating." Moreover, these are necessarily small end-gas events, not whole-cylinder detonations.
The percentages get even smaller if we use the typical redlines of up to 2850 RPM instead of cruise settings.
As Bell says, the published data for engine operation keeps one well clear of detonation margins.
Small numbers. Certainly not "constantly detonating." Moreover, these are necessarily small end-gas events, not whole-cylinder detonations.
The percentages get even smaller if we use the typical redlines of up to 2850 RPM instead of cruise settings.
As Bell says, the published data for engine operation keeps one well clear of detonation margins.
PaulMillner
Line Up and Wait
The worst case cylinder is the one monitored... there's no summing of all four or six accelerometers. So divide your numbers by four or six accordingly. But you've got the general idea.
Paul
PaulMillner
Line Up and Wait
That's a little muddled, but OK... Reduce takeoff fuel flow by 12% and monitor for detonation. My airplane doesn't have a minimum fuel flow limit... it does have a fuel flow redline which I strive to always achieve (or slightly exceed).
That's what engine monitors are good for. If CHT is stable and below a metallurgically significant temperature, say 380 F or below, then there's no operationally significant detonation occurring. If whatever is happening is not operationally insignificant, then why do I as pilot, or you as mechanic, care if it's incipient or limiting?
What GAMI’s novel work demonstrated, to the FAA’s satisfaction, is that not all detonation events are alike, and that unleaded gasoline doesn’t have the same distribution of different kinds of detonation events as leaded gasoline does. So the trick is in setting the thresholds for detonation events of interest. The water-cooled accelerometer didn’t give us that kind of resolution. But, today’s fiber-optic intracylinder pressure sensors do. Then, it’s a matter of math to define the detonation events of interest, and correlate the incipient and limiting detonation limits using the more complete today’s data to the older technology blinking light data.
Paul
Last edited:
pfarber
Line Up and Wait
- Joined
- Jun 20, 2021
- Messages
- 650
- Display Name
Display name:
pfarber
wheels (the part i am working in) are explicitly listed as a repair. what is your acceptable/approved data for substitutions when the part is readily available?
as for detonation you are moving the goal posts. thr statement is that engines detonate all the time... you just don't hear them. my best guess is that people (oh you bastions of knowledge on poa) seem to think the only meaning must be literal vs colloquial use if 'all the time'. ive cited multiple sources that concur. the faa detonation tests are just the inability to agree with someone you think is wrong, but cannot come up with any sources to justify your statement.. to you either put your head in the sand or fingers in your ears or some other wacky permutation of 'waaahhhh!!!'
as for my anp schools there was a 30 year gap. my time in the navy, IT and life allowed me to take the course again. its a no brainer to let the govt pay for my school and living expenses vs working in a shop for 6 months for less.
now nod your head up and down and pretend you understand what i said
I should have been clearer. The 12% margins are only determined during certification tests. You would never enter these margins unless you purposely operated your engine in excess of its certified maximum and minimum fuel flow operating limitations.
For example, if your mixture control at full rich equated to 100% and 60% at the minimum operating lean fuel flow, in order to pass the detonation margin test you would have to increase the mixture to 112% and reduce it to 48% before any limiting detonation was detected. The AC I mention gives one version of this requirement.
Operationally insignificant per an FAA acceptable reference or…?
Keep in mind the 1st part of the definition of airworthy is the aircraft conforms to its type design or certificate. So in the case of detonation the conformity guidance in the CAR or FAR points to “no detonation.” So it doesn’t conform to the type design and is not airworthy.
Or you could take it further and determine via the AC margin test procedure it is only limiting detonation that is the trigger provided you can differentiate between incipient and limiting detonation. Regardless, its how the system works but not everyone is a fan of the system. From the mx side I prefer the system.
But they still had to pass the detonation margin requirements, same as the PAFI fuels have to with the AC. What I’m curious about is if they had to follow the 12% benchmark or if they came up with their own margin value? Same with the limiting detonation definition. Would be interesting to read the approval document listing for the STC.
Here's one. Plus there are about a half- dozen other specifications. Readily available is not a concern with substitutions.
So the FAA guidance is not an acceptable source? Or would you prefer Mike Busch instead?
But the question remains did you pass one of those A&P schools and get your certificate? The jury is still out.
PaulMillner
Line Up and Wait
I know you have the concept, but you’re not expressing it quite correctly. Let’s talk about never below…
Nah, that’s not how that works! Let’s use your 100% full rich example. For my turbo-Cardinal, that’s 24 GPH at takeoff power, 31” manifold pressure. The FAA detonation test would require me to not exceed limiting detonation at full power, 31” manifold pressure, and 12% *less* fuel flow, or in other works, at 21.1 GPH.
Remember, at takeoff full rich, your engine is set up to flow *extra* fuel to suppress detonation. If, instead, you flow less fuel, you don’t have that extra-rich mixture to suppress detonation, and it can occur.
My airplane doesn’t have a “minimum operating lean fuel flow” specification. But in cruise, at 31” and 2500 RPM, I can come back to 11 GPH, about 60 degrees lean of peak, and the engine is very happy operating there for hours on end… cool CHTs, reasonable oil temperature, smooth power. If instead I decided (foolishly) to operate at 13.5 GPH, that would be about peak EGT. The CHTs are going to get very warm, and on an ISA+15 day, I might even see evidence of detonation. But only good operating practice (and CHT limits) influence that. There’s no FAA red line fuel flow to protect me. I have to think about and understand my engine’s operation.
Sure! Start with the FAA approved red line CHT for the engine. If you run at peak EGT on my turbo’d engine, you’re not going to stay below red line CHT under all circumstances, and certainly not on the FAA engine-cooling-limiting-case 100F OAT day. An even better FAA acceptable reference is Lycoming’s service instruction that advises that for longest cylinder life, keep CHT below 420F. That’s great guidance there, and Lycoming says that statements with airworthiness implications have been vetted by the FAA.
Again, the problem is what is “no detonation?” It’s not the engineering definition of no detonation. As we’ve just discussed, the FAA says both incipient and limiting detonation are just fine. So apparently, the airworthiness definition of no detonation means limiting detonation or less. I find that definitionally confusing and unhelpful. Apparently, with the FAA, one’s understanding of English mileage may vary.
Doesn’t it bother you at some level that observable detonation is defined as “no” detonation. Words don’t mean what we think they mean! It’s all useful from an engineering perspective, but it certainly is more than confusing enough to confound pilots and others trying to puzzle out just what’s being defined.
There’s a white paper (special FAA terminology for policy publications) that explains all that… but it may be considered GAMI confidential. In participating in the FAA’s GAMI certification review as the fuels’ expert, I got to see some parts of that white paper. GAMI did use the same 12% IIRC reduction of fuel flow to demonstrate adequate detonation margin. What GAMI disagreed with the FAA about is how to translate/interpret the observed fiber optic pressure events into incipient and limiting detonation as defined by the Sperry water-cooled accelerometers. GAMI’s white paper described what the FAA eventually deemed an acceptable alternative method that did NOT include the math errors of the FAA’s adopted method. Instead, I’d say GAMI’s method is superior to the FAA’s ASTM-blessed method… but that’s just me.
There could well be times that is a good idea. Let's say you're doing a maximum performance takeoff from a mountain strip, and an exhaust valve sticks. I've had this happen to me. With 25% less power on a four banger, you may not clear the trees at the end of the runway, or have sufficient room to land and stop before the tree line. By understanding how the engine is setup, you will realize that you can reduce fuel flow roughly 10%, increasing power output, though eating away at the detonation margin. In the 60 seconds or so it will take you to clear the trees, the thermal mass of the three hard-working cylinders will protect you from destructive effect... you can then return to full rich. That beats eating branches all to heck.
Paul