Preignition

[PT20J]

There is an interesting thread on Beechtalk about a runaway preignition event caught on an engine monitor with borescope pix. https://www.beechtalk.com/forums/viewtopic.php?f=37&t=193567&start=0

A lot of us worry about detonation and preignition. There is some confusion in the literature about the two, but they are really related. Preignition is the same as overly advanced ignition timing and the effect is very high peak pressure and temperature which causes the end gas (the part of the mixture that has not yet burned) to all ignite spontaneously resulting in severe detonation. The resulting pressure waves and extreme heat damage cylinder components, especially pistons (pistons are made of aluminum which melts at lower temperatures than valves or cylinder walls and pistons have poorer heat rejection than the heads which have large cooling surfaces).

I think it's pretty hard to get a normally aspirated engine to detonate (let alone get hot enough for preignition) unless you abuse it severely. The APS folks commissioned a study years ago looking at accident reports and SDRs and the like and couldn't find any reports of detonation in NA engines. That doesn't mean it's impossible, but it means you really have to work at it. Turbocharged and especially turbonormalized engines are a different matter (the subject engine of the thread was turbonormalized) and it is not hard to mismanage them and get into the danger zone. The problem with preignition is that once it starts, it runs away meaning that the ignition happens earlier with each cycle increasing the detonation intensity.

Skip

[carusoam]

That is interesting... TN’d IO550...

Just one cylinder CHT losing control...

I didn’t pick up any other anomaly like an odd EGT or FF...

The valves look like they have no hours on them at all... shiny clean...

I didn’t spend much time there either... getting too late, still have miles to go...  

-a-

[larrynimmo]

That happened to me one time I was cruising LOP....I switched to ROP and issue went away and LOP works fine ever since....

IO360A3B6D

[DAVIDWH]

And where is that $35.00 digital CO monitor when you might need it??

 

 

[A64Pilot]

Pre-ignition is NOT the same, nor is it similar to detonation, but it can lead to detonation.

‘The difference is normal ignition whether it be pre or on time consists of an ignition source where a flame front proceeds at a set speed across the combustion chamber, ignition does not necessarily come from the electrical system, it can for instance be a hot piece of carbon etc.

The speed at which the flame front proceeds is set, it’s why on engines that have a much wider RPM range that ignition timing is advanced with RPM due to the shorter period of time there is for the charge outreach peak pressure, and why set timing works well on aircraft engines, because we operate at a narrow RPM band. 

‘Now where detonation differs is that there is no ignition source,  the entire fuel / air charge goes off at once, there is no source that a flame front proceeds from, it’s an instantaneous ignition of the entire charge. 

‘That’s why detonation is so destructive, it’s because there is a huge nearly instantaneous increase in pressure, well above any pressure you will get from an ignition source, and why it often blows holes in pistons, heat follows pressure, and or can cause detonation. Chicken or egg?

‘You certainly can get an NA engine into detonation, Fly full throttle at sea level and start leaning it out, it’s likely you may get detonation and or pre-ignition.

‘It’s what makes me nervous when I read the dissertations on how to run LOP where people are advised to simply increase Manifold pressure to recover any airspeed lost from being LOP, but aircraft haven’t been blowing up engines so it’s apparent that it works, but I suspect that in reality there just isn’t any throttle left to increase, most seeking efficiency go higher, and if your N/A above 6 or 7 thousand it’s likely your full throttle anyway.

Edited March 17, 2021 by A64Pilot

[N201MKTurbo]

11 minutes ago, A64Pilot said:

Pre-ignition is NOT the same, nor is it similar to detonation, but it can lead to detonation.

‘The difference is normal ignition whether it be pre or on time consists of an ignition source where a flame front proceeds at a set speed across the combustion chamber, ignition does not necessarily come from the electrical system, it can for instance be a hot piece of carbon etc.

The speed at which the flame front proceeds is set, it’s why on engines that have a much wider RPM range that ignition timing is advanced with RPM due to the shorter period of time there is for the charge outreach peak pressure, and why set timing works well on aircraft engines, because we operate at a narrow RPM band. 

‘Now where detonation differs is that there is no ignition source,  the entire fuel / air charge goes off at once, there is no source that a flame front proceeds from, it’s an instantaneous ignition of the entire charge. 

‘That’s why detonation is so destructive, it’s because there is a huge nearly instantaneous increase in pressure, well above any pressure you will get from an ignition source, and why it often blows holes in pistons, heat follows pressure, and or can cause detonation. Chicken or egg?

‘You certainly can get an NA engine into detonation, Fly full throttle at sea level and start leaning it out, it’s likely you may get detonation and or pre-ignition.

‘It’s what makes me nervous when I read the dissertations on how to run LOP where people are advised to simply increase Manifold pressure to recover any airspeed lost from being LOP, but aircraft haven’t been blowing up engines so it’s apparent that it works, but I suspect that in reality there just isn’t any throttle left to increase, most seeking efficiency go higher, and if your N/A above 6 or 7 thousand it’s likely your full throttle anyway.

At sea level, full throttle, lean to peak and pull the prop back all the way. It will knock.

[A64Pilot]

My Brother detonated an IO-520 in a Cessna 210 to death trying to keep up with me in a S2R-T65HG, Somehow he thought you got more power when you leaned one out, it took awhile as he called me on the radio saying it sounded like the landing gear were “thumping” the belly, and only one cylinder lost compression so there was no complete loss of power.

I didn’t take the engine apart, it was exchanged for a zero time motor, but I’m sure at least one piston had a dime sized hole right it’s center

[PT20J]

1 hour ago, A64Pilot said:

Now where detonation differs is that there is no ignition source,  the entire fuel / air charge goes off at once, there is no source that a flame front proceeds from, it’s an instantaneous ignition of the entire charge. 

It's only the end gas that "detonates." There isn't enough pressure and temperature to cause the entire fuel air charge to go off all at once in a spark ignition engine. There are two sources of compression of the end gas: the compression caused by the piston and the compression caused by the advancing flame front. In addition to increasing the pressure and density of the end gas, the advancing flame front also heats it by radiation. If the end gas gets hot enough to reach the autoignition temperature, it detonates because at the elevated pressure the molecules of reactants are pressed so close together that the burning is extremely rapid.

1 hour ago, A64Pilot said:

It’s what makes me nervous when I read the dissertations on how to run LOP where people are advised to simply increase Manifold pressure to recover any airspeed lost from being LOP, but aircraft haven’t been blowing up engines so it’s apparent that it works, but I suspect that in reality there just isn’t any throttle left to increase, most seeking efficiency go higher, and if your N/A above 6 or 7 thousand it’s likely your full throttle anyway.

I've been curious about this for a long time. At WOT, the amount of air the engine can pump is limited by displacement and density altitude so as the mixture is leaned, power necessarily decreases. Somewhere along the line someone got the idea that you could get the power back by increasing manifold pressure. John Deakin wrote that he cruised his turbonormalized Bonanza at greater that 80% power LOP with no ill effects. Maybe that's true and works fine. But power is power and the BMEP would have to be the same regardless of the mixture/MAP combination used to achieve it. I know there are claims that it works because the peak pressure is lower, and later, but that should reduce efficiency. I do know that this procedure is not approved by Continental or Lycoming. Continental specifically says to set power to 65% or below ROP and then lean to wherever you want and leave the throttle and rpm alone. Their reasoning is that you don't really know exactly where you are operating but can only infer it from the CHT and fuel flow. The pilot of the Bonanza thought he was safely 75F LOP.

Skip

[A64Pilot]

This does a decent job of explaining detonation, but it should be noted that detonation can occur at pretty much anytime there is fuel, air and compression, to include even after the spark plug has fired, however that detonation is not usually as destructive as often the piston is on its way down and combustion chamber volume is increasing. It’s most destructive if max pressure occurs prior to TDC as that blows pressure and heat through the roof, because there is no where for it to go.

https://www.procharger.com/what-detonation-and-how-can-it-be-controlled

Many people think an engine runs off of “explosions” of the air fuel mixture, and that’s incorrect, it’s actually when compared with piston speed actually quite a slow burn, however detonation is the explosion many think of.

Some think that Diesels detonate as a normal operation, but excepting for initially starting, they don’t, the fuel burn in a Diesel is initiated the moment fuel becomes present, there is so much leftover heat from the combustion event that occurred fractions of a second ago that it ignites the fuel. But cold diesels will knock, and that knock is sometimes detonation of a partial fuel charge.

 

[A64Pilot]

I’ll run LOP myself often but I do so only at lower power.

What bothers me is when I read that you can run high power LOP.

You can go fast, or you can run LOP but you can’t go fast, LOP. Much of the fuel saving from running LOP is actually from slowing down, but who cares why, so long as you burn less to cover the same distance.

 

Detonation destroys engines, so avoid operations that may allow it, I won’t bump my timing up either, although I’m sure you can get away with it, it reduces the margin.

 

Lycoming is a different combustion chamber design than Continental, and I suspect just from listening to some of the Continental’s idle they have a more radical a profile too.

The 210’s 520 loped like a dragster. I used to love to hear it idle.

‘Point is I think Continental engines work better LOP than Lycoming, or at least a 520 works better than a 540, I can’t speak for others, but my 360 without Gami’s runs smoother leaner than my 540 did, maybe it was because the 540 was a parallel valve motor? 

[N201MKTurbo]

My Lycoming runs great LOP, but I can recover the power with more MP.

[PT20J]

13 minutes ago, A64Pilot said:

‘Point is I think Continental engines work better LOP than Lycoming, or at least a 520 works better than a 540, I can’t speak for others, but my 360 without Gami’s runs smoother leaner than my 540 did, maybe it was because the 540 was a parallel valve motor? 

Could be. The roughness comes from cycle to cycle variations in combustion rate. Combustion, even ROP, has has cycle to cycle variation (not to be confused with cylinder to cylinder variation) -- the peak pressure point moves around quite a bit. The leaner you go, the more variation in combustion timing. The amount of turbulence in the combustion chamber has a significant effect on the combustion rate and a lot of the turbulence is introduced by the air flowing past the restriction of the intake valve. So, an engine that creates more turbulence (up to a point) should run smoother LOP.

We all know that Mike Busch is a data-driven kind of guy (he's an engineer after all). I asked him at MooneyMax if he had any data on the long term effects of running high power settings at LOP mixtures. He didn't. I asked him what he does in his turbo C-310 and he said "65% or less -- I'm a longevity guy, not a speed guy." I would like to see some long term info on this (like running several engines to TBO at high powers and LOP in a test cell), but most everything seems to be anecdotal. 

Skip

 

[N201MKTurbo]

30 minutes ago, PT20J said:

Could be. The roughness comes from cycle to cycle variations in combustion rate. Combustion, even ROP, has has cycle to cycle variation (not to be confused with cylinder to cylinder variation) -- the peak pressure point moves around quite a bit. The leaner you go, the more variation in combustion timing. The amount of turbulence in the combustion chamber has a significant effect on the combustion rate and a lot of the turbulence is introduced by the air flowing past the restriction of the intake valve. So, an engine that creates more turbulence (up to a point) should run smoother LOP.

We all know that Mike Busch is a data-driven kind of guy (he's an engineer after all). I asked him at MooneyMax if he had any data on the long term effects of running high power settings at LOP mixtures. He didn't. I asked him what he does in his turbo C-310 and he said "65% or less -- I'm a longevity guy, not a speed guy." I would like to see some long term info on this (like running several engines to TBO at high powers and LOP in a test cell), but most everything seems to be anecdotal. 

Skip

 

I can't imagine what harm it does. The cylinders run cooler LOP. The exhaust runs hotter, I suppose it could be detrimental to the exhaust valves. I have 4 Lycoming cylinders sitting in my hangar that were run LOP for most of their 1600 hours. They look quite good and were running perfectly when I pulled them. 

[carusoam]

Things about flame fronts and explosions...

 

Talking to a chemical engineer one day...

1) An explosion is a very fast flame front...

2) Things that slow flame fronts down... excess molecule of gasoline, or oxygen...

3) Excess gasoline molecules don’t burn when they bump into other gasoline molecules...

4) Same thing happens when two O2 molecules bump into each other...

5) So... having the exact fuel mixture for peak EGT.... generates a very fast flame front...

6) Both LOP and ROP will be slower...

7) Having excess gasoline works better as a coolant than having excess air...

8) Taking heat of vaporization into consideration... along with the different heat capacities of the two molecules being considered...

9) If... we had a red box/fin gauge on our instrument panel... Running high power LOP would be pretty easy... I don’t think many people would like the power output...  it is to hard to get excess airflow... compared to tuning up the excess fuel flow... 

10) We can really get to know our own engines and set the MP, and FF to maintain a high level of LOP... but a simple miscalculation, things can go awry...

11) Maximum climb leaving the ground in my plane... the mixture is set-up to be about 300° ROP.... doing something similar LOP is technically interesting and can be followed using a chart for adjusting LOP °F vs altitude...

12) this only becomes interesting for saving a gallon or two... 

13) fortunately, the higher you climb LOP, the more LOP you become... not like you would accidentally enter the red box if you fell asleep while climbing and were getting your IR clearance....

 

One question... what caused the one cylinder of this Brand B plane to decide to go amok?

The other five were happily going along peacefully...

Was there a known Fuel injector blockage or something to cause a change to its localized fuel ratio?

PP thoughts only, I should invite Ross( @Shadrach)... he likes This stuff...

Best regards,

-a-

[A64Pilot]

19 hours ago, N201MKTurbo said:

I can't imagine what harm it does. The cylinders run cooler LOP. The exhaust runs hotter, I suppose it could be detrimental to the exhaust valves. I have 4 Lycoming cylinders sitting in my hangar that were run LOP for most of their 1600 hours. They look quite good and were running perfectly when I pulled them. 

The cylinder heads do run cooler, often much cooler, but it’s largely due to the lower power output as cyl head temp follows power output pretty closely.

‘LOP does work, however it’s for low power, don’t believe that it’s not, no gasoline piston engine that I’m aware of doesn’t go into “power enrichment” when run at high power, even the newer computerized fuel injected outboards, that excess fuel does help cool, it’s not needed to make more power, it’s needed to keep the temps in control. So I’m thinking the marketing dept made that decision, because run one LOP and it’s not the “Hot Rod” it is ROP.

Lycoming doesn’t recommend LOP, in my opinion that’s not because it doesn’t work, but because the power loss in a Lycoming is quite large, that may be due to lower swirl in a Lycoming combustion chamber. Swirl is something that’s pretty new, it wasn’t thought about back in the day when Hemi combustion chambers were the hot ticket.

One day maybe airplanes will have Pentroof combustion chambers and four valve heads, but I’m not holding my breath.

 

So if LOP is used as an economy measure, which also means lower power and slower speed, it won’t do any harm, and in fact I believe a solid argument could be made that an engine operated that way will outlive one run at high power, rich and higher cyl head temps, of course that guy is going much faster and a counter argument could be made he’s running his motor for less hours I guess.

‘It’s when people try to recover the power lost from running LOP that bad things can happen, face it, yes it can be done with modern computer controlled engines with knock sensors in the manifold and the ability to independently adjust each fuel injector and vary ignition timing independently for each cylinder, but for our maybe 1940’s engines? Maybe not. We think with engine monitors we have all the data to do it, but compared to a microprocessor hooked to a knock sensor and an O2 sensor in the exhaust and learning ability to build injection and timing “maps”, we just can’t possibly run at the edge like those motors can.

Edited March 18, 2021 by A64Pilot

[Shadrach]

23 hours ago, A64Pilot said:

‘You certainly can get an NA engine into detonation, Fly full throttle at sea level and start leaning it out, it’s likely you may get detonation and or pre-ignition.

One may be able to get a NA aeroengine to detonate but I do not agree that it's anything near a certainty, especially in level flight.  I have done some pretty extensive high power experimentation at sea level DAs with my IO360 and have never seen engine data that would suggest anything nearing detonation in level flight.  I think that I could get a NA certified aeroengine to detonate if I tried really hard but it would require determination and deliberate abuse of more than just the mixture.  Perhaps something like sustained full throttle while leaned for max power at near SL with the airplane at high pitch behind the power curve and cowl flaps closed. Even then it would take more time to get there than any sane person would want to spend.

Also,  pre-ignition is not caused by mixture as you appear to suggest in the above quote. Pre ignition is an ignition source other than the ignition system that starts the combustion event in advance of the ignition system.  It could be caused by a damaged spark plug electrode or a detached thread from the spark plug or a helicoil that protrudes into the combustion chamber or a carbon deposit or any other object in the combustion chamber that has absorbed sufficient heat to ignite the mixture in advance of the ignition system.  As Skip suggests in the original post the two event are very often related.  Pre-ignition often leads to detonation and detonation can lead to cylinder damage which can cause pre-ignition. The good news is that one you've holed the piston they both tend to stop...which I learned first hand riding to San Francisco on an old Triumph single that started the journey from Santa Rosa as a twin.

Edited March 18, 2021 by Shadrach

[Shadrach]

2 hours ago, A64Pilot said:

The cylinder heads do run cooler, often much cooler, but it’s largely due to the lower power output as cyl head temp follows power output pretty closely.

‘LOP does work, however it’s for low power, don’t believe that it’s not, no gasoline piston engine that I’m aware of doesn’t go into “power enrichment” when run at high power, even the newer computerized fuel injected outboards, that excess fuel does help cool, it’s not needed to make more power, it’s needed to keep the temps in control. So I’m thinking the marketing dept made that decision, because run one LOP and it’s not the “Hot Rod” it is ROP.

Lycoming doesn’t recommend LOP, in my opinion that’s not because it doesn’t work, but because the power loss in a Lycoming is quite large, that may be due to lower swirl in a Lycoming combustion chamber. Swirl is something that’s pretty new, it wasn’t thought about back in the day when Hemi combustion chambers were the hot ticket.

One day maybe airplanes will have Pentroof combustion chambers and four valve heads, but I’m not holding my breath.

 

So if LOP is used as an economy measure, which also means lower power and slower speed, it won’t do any harm, and in fact I believe a solid argument could be made that an engine operated that way will outlive one run at high power, rich and higher cyl head temps, of course that guy is going much faster and a counter argument could be made he’s running his motor for less hours I guess.

‘It’s when people try to recover the power lost from running LOP that bad things can happen, face it, yes it can be done with modern computer controlled engines with knock sensors in the manifold and the ability to independently adjust each fuel injector and vary ignition timing independently for each cylinder, but for our maybe 1940’s engines? Maybe not. We think with engine monitors we have all the data to do it, but compared to a microprocessor hooked to a knock sensor and an O2 sensor in the exhaust and learning ability to build injection and timing “maps”, we just can’t possibly run at the edge like those motors can.

Sorry, but no.  LOP works very well at high power settings and is indeed cooler for a given output. This is easily seen in turbo engines but can also be demonstrated in NA engines at low altitude. Where I live in the mid atlantic we see DAs in the winter that are more then 3000' below a standard day.  It is not uncommon at all to see SL DA at 3000AGL.  When fighting headwinds I often fly at low altitudes with the engine WOT ram air open and LOP. Under said conditions the engine will make somewhere in the vicinity of 80% power on the lean side of peak which can be confirmed by IAS. The trouble is keeping the coolest CHTs above 300°. You are correct that CHTs track power output but that is not the full story.  Power is a function of mean cylinder pressure not peak. A cylinder can make the same power with lower peak pressure through the power stroke if the pressures are more even during the power stroke.  For a given output, ROP operations tend towards higher peak pressures and higher CHTs. LOP operations tend towards lower but more consistent pressure throughout the stroke and lower CHTs.  The late great Walter Atkinson described it to me as the difference between a hammer blow and a steady push.  Two different ways of doing the same amount of work.

 

Edited March 18, 2021 by Shadrach

[AH-1 Cobra Pilot]

28 minutes ago, Shadrach said:

Power is a function of mean cylinder pressure not peak. A cylinder can make the same power with lower peak pressure through the powerstroke if the pressures are more even during the power stroke.  For a given output, ROP operations tend towards higher peak pressures and higher CHTs LOP operations tend towards lower but more consistent pressure through out the stroke.

So, what you are saying is that if I were to graph Pressure vs. Distance Across the Piston, you would see an initial sharp spike near the spark plug at the top of the stroke, with the ROP being higher; then, as the piston moves downward, the plot would become more rounded then flatter towards the bottom of the stroke, and when LOP, these pressures toward the bottom of the stroke would be higher?

[Shadrach]

Just now, AH-1 Cobra Pilot said:

So, what you are saying is that if I were to graph Pressure vs. Distance Across the Piston, you would see an initial sharp spike near the spark plug at the top of the stroke, with the ROP being higher; then, as the piston moves downward, the plot would become more rounded then flatter towards the bottom of the stroke, and when LOP, these pressures toward the bottom of the stroke would be higher?

Sort of.  As it has been explained to me by folks with access to far more sophisticate equipment and experience, if you were to graph the cylinder pressure across the power stroke you would see a sharp peak and decrease in pressure on the rich side and a less drastic peak and less drastic decrease on the lean side. Put simply, for the same mean pressure across the power stroke, ROP generates pressures far above and below the mean while LOP pressures cluster closer to mean throughout the power stroke.

One can see this in simple flight testing. Set up power for an attainable airspeed on either side of the mixture spectrum (all other things being equal). Record CHTs for that airspeed at ROP and LOP.  CHTs will be lower on the lean side. How much so depends on the power setting.

[N201MKTurbo]

I can go faster LOP than ROP in cruise. If ROP I hit the CHT limit (380F) if LOP I hit the TIT limit (1550). The plane goes faster at the LOP limit.

[Shadrach]

Just now, Freemasm said:

Great stuff just torque up the geek factor just a little bit for people like myself in a slow work day.

Deflagration = sub-sonic combustion. (BTW. a gasoline/air flame front can be surprising slow. 100 ft/sec in some conditions)

Detonation = supersonic combustion.  

Indeed I spent much of my life not knowing that the sound of pinging was a series of small sonic booms occurring in the cylinder.

[A64Pilot]

99 % of the “valve” rattle and knocking we hear isn’t detonation. it’s pre-ignition. 70’s cars were the worst.

pre-ignition isn’t usually all that harmful.

‘Thankfully detonation is rare, most likely you have never heard or experienced it, very often even short duration detonation is destructive

[A64Pilot]

Found an article specific to aircraft

https://blog.aopa.org/aopa/2015/04/13/destroy-your-engine/

[Shadrach]

5 minutes ago, A64Pilot said:

99 % of the “valve” rattle and knocking we hear isn’t detonation. it’s pre-ignition. 70’s cars were the worst.

pre-ignition isn’t usually all that harmful.

‘Thankfully detonation is rare, most likely you have never heard or experienced it, very often even short duration detonation is destructive

OK I'll bite. Tell me what makes the pinging sound.  Also intentionally induced light detonation for short periods of time in the GAMI/APS engine test cell has not shown any damage to the engine but it is not something I would want to try in the air.

[Shadrach]

7 minutes ago, A64Pilot said:

Found an article specific to aircraft

https://blog.aopa.org/aopa/2015/04/13/destroy-your-engine/

Interesting article. Mike does not really go much into combustion science detail.  My biggest takeaway from the article is that seemingly competent people can be oblivious to a situation that demands immediate action. There but by the grace of god go I...