Another "odd man out" cylinder scenario

[airguy]

On my IO-360 with SDS, I have my #3 cylinder behaving a bit "out of family" with the other 3. This is not a new occurrence, it's been this way for over a thousand hours now, and it's not changing - quite stable - but I've run out of easy answers on what the story is. The engineer in me says there's a reason and I can't quit picking at it, like a kid with a scab on his knee.

Symptoms are that the #3 runs about 40-50 degrees cooler on CHT, and about 150 degrees warmer on EGT. I don't claim to be an engine expert but I do know my way around them pretty well - I know that it's NOT a spark plug, wire, or coil issue (having swapped them around back and forth), it's NOT a temp sensor issue (swapped those around too), and it's NOT a mixture issue as I have a GAMI spread between all the cylinders of either 0.0 or 0.1, depending on which day of the week I run the test and how carefully I do it. The SDS system allows me to individually trim the fuel flow to each cylinder, so dialing in the GAMI spread is easy. I see this temp spread on both 100LL and 93E10, both ROP and LOP, cruise and climb. The exhaust valve is not leaking as far as I can tell - 78/80 on that cylinder during my recent condition inspection and the borescope pictures of the exhaust valve look great. I burn one quart of oil in about 20 hours - so the rings are all good. I even lapped the exhaust valve in-situ a couple hundred hours ago, with no change.

The obvious answers have already been covered, which means it has to be something odd - which is where I turn to the brain trust here.

It's apparent that less combustion is occurring IN the cylinder, and more OUTSIDE the cylinder in the exhaust valve and pipe as compared to the others - which points to either an exhaust valve leak (none found) or perhaps the exhaust valve is opening too soon - which would only be possible if my exhaust valve pushrod is too long by some small amount and opening before the others. I have not pulled the rods and measured them but that's getting up pretty high on my "to-do" list now. Equally possible, and just as unlikely, is the the cam lobes for the other 3 cylinders are all worn down almost identically, but the engine makes no metal at 1640 hours so that's not supported by evidence.

Some of you gray-headed guys (like me) got any good ideas?

 

[Paul 5r4]

It sounds like you've pretty much covered every possibility Greg! FWIW... (I was a nurse)... could it be possible that the cam lobe for that cylinder is clocked off just a bit at time of manufacture. Advanced I think. I really can't imagine this scenario. I agree, it sure seems the exhaust valve is opening just a nano second before the others.

 

[riseric]

HI Greg,
Not an engine expert but I do qualify as "grey-haired"...
I too believe that you covered/checked the most evident culprits.
Before I finished reading your post, I also thought about a cam lobe issue.
Doesn't mean it's wearing bad, just that it might be off a little since new.
Difficult to confirm outside of changing the cam...
If and when you measure the exhaust valves stems, try swapping ??
Seems there's a physical discrepency causing the temperature spread. Not injection/ignition.
Intake tube/port maybe also?
A carbon deposit in the exhaust header igniting unburnt fuel ?? (doubt it)
While I understand you trying to get all 4 as close as possible, (I'm guilty of that also) and thanks to the goodies we have on board we can be kind of maniac about this.
I hope for you that you resolve this, would be satisfying for sure.
Meanwhile, remember the "good old days" of flying with one CHT needle and no EGT... LOL
Good luck !!

 

[Roadjunkie1]

On my IO-360 with SDS, I have my #3 cylinder behaving a bit "out of family" with the other 3. This is not a new occurrence, it's been this way for over a thousand hours now, and it's not changing - quite stable - but I've run out of easy answers on what the story is. The engineer in me says there's a reason and I can't quit picking at it, like a kid with a scab on his knee.

Some of you gray-headed guys (like me) got any good ideas?

Grey-hair yes; engine expert: no. But it looks like you have covered all bases......except the one that is causing the aberrant readings. Might be something simple. Just for laughs, look at the intake manifold 'hose' condition and tightness of the clamps; intake manifold gaskets. Would a small air leak make the cylinder run leaner? Huh..... Still thinking; looking at that scab.......

 

[walkman]

On my IO-360 with SDS, I have my #3 cylinder behaving a bit "out of family" with the other 3. This is not a new occurrence, it's been this way for over a thousand hours now, and it's not changing - quite stable - but I've run out of easy answers on what the story is. The engineer in me says there's a reason and I can't quit picking at it, like a kid with a scab on his knee.

Symptoms are that the #3 runs about 40-50 degrees cooler on CHT, and about 150 degrees warmer on EGT. I don't claim to be an engine expert but I do know my way around them pretty well - I know that it's NOT a spark plug, wire, or coil issue (having swapped them around back and forth), it's NOT a temp sensor issue (swapped those around too), and it's NOT a mixture issue as I have a GAMI spread between all the cylinders of either 0.0 or 0.1, depending on which day of the week I run the test and how carefully I do it. The SDS system allows me to individually trim the fuel flow to each cylinder, so dialing in the GAMI spread is easy. I see this temp spread on both 100LL and 93E10, both ROP and LOP, cruise and climb. The exhaust valve is not leaking as far as I can tell - 78/80 on that cylinder during my recent condition inspection and the borescope pictures of the exhaust valve look great. I burn one quart of oil in about 20 hours - so the rings are all good. I even lapped the exhaust valve in-situ a couple hundred hours ago, with no change.

The obvious answers have already been covered, which means it has to be something odd - which is where I turn to the brain trust here.

It's apparent that less combustion is occurring IN the cylinder, and more OUTSIDE the cylinder in the exhaust valve and pipe as compared to the others - which points to either an exhaust valve leak (none found) or perhaps the exhaust valve is opening too soon - which would only be possible if my exhaust valve pushrod is too long by some small amount and opening before the others. I have not pulled the rods and measured them but that's getting up pretty high on my "to-do" list now. Equally possible, and just as unlikely, is the the cam lobes for the other 3 cylinders are all worn down almost identically, but the engine makes no metal at 1640 hours so that's not supported by evidence.

Some of you gray-headed guys (like me) got any good ideas?

There is another possible cause, which I think is highly unlikely, and basically impossible to check or change without tearing down the engine, but that is there is a difference in the lifter.

 

[skirting_virga]

I assume that this could be a volumetric efficiency or charge homogenization problem relating to either intake flow, exhaust flow, or both. Intake and exhaust have a lot of weird dynamics (that I honestly only grasp in a basic way) that could potentially explain some or all of the difference. A cylinder with less turbulence and a less homogenous charge (i.e. a stratified charge) could have slower combustion. More swirl or tumble could also stratify the charge. Advancing spark on that cylinder only might correct for the problem. Lower V.E. on a single cylinder is not as easily solved unless the problem is mechanical (valve lift from the cam to the lifter). If your #3 cylinder simply ingests less air than the other cylinders across all RPM and manifold pressure settings, you can only balance that with fuel trim and accept the imbalance.

Are the heads ported similarly? Are the valve bowls polished? Are the intake runners through the valve seat flow benched and matched? Does the discrepancy occur equally throughout the RPM range, or is it amplified at some RPM (suggesting a dynamic flow issue in the form of intake or exhaust resonance)?

 

[Carl Froehlich]

I suggest an experiment (I assume your SDS can do this).

Do whatever SDS calls a mag check. Observe the EGT rise in each cylinder. Does #3 rise as much as the other cylinders?

Carl

 

[Toobuilder]

Grasping at straws a bit here, but if you have an updraft sump there may be some manufacturing anomaly that drives the behavior. The updraft sumps are truly awful designs and the amount of “awfulness” may magnify defects. One can look to the troubles getting even mixture distribution on some of the O-320 examples, and the resulting “fixes” for those. Some are application of carb heat, throttle blade angle, or even the installation of internal airflow director devices. Pulling at that thread, maybe install an “airflow modifier” device in the sump and see if you get ANY form of change that can inform you if it’s an airflow (and therefore VE) issue.

 

[DanH]

It's apparent that less combustion is occurring IN the cylinder, and more OUTSIDE the cylinder in the exhaust valve and pipe as compared to the others

No. Although there are ongoing reactions, 100% mass fraction burned is reached well before exhaust valve opening. A good reference like Haywood's Internal Combustion Engine Fundamentals has typical burning duration at 40 to 60 crank degrees, starting at ignition BTDC. The exhaust valve doesn't open until roughly 135 ATDC.

Higher indicated EGT (assuming it's not the usual case of a different probe placement in the gas stream) says higher pressure at valve opening. Pressure and temperature (in Kelvin) are proportional, so here pressure would be about 8% higher.

- which points to either an exhaust valve leak (none found) or perhaps the exhaust valve is opening too soon - which would only be possible if my exhaust valve pushrod is too long by some small amount and opening before the others.

Pushrod length doesn't change when the valve opens. The hydraulic unit makes pushrod length effectively the same for all, zero clearance at the cam. Zero dry tappet clearance (pushrod really too long) wouldn't allow the valve to seat, and it should have overheated long ago.

Back to lower CHT and higher EGT. The question to ask is "What might reduce peak pressure, and raise pressure at exhaust valve opening?"

Has this been going on since 0 hours?

 

[3bythelake]

I am no authority on this, but have you considered the possibility of a cooling airflow difference causing the CHT difference? And as long as all cylinders are peaking together I thought we didn’t get overly worried about absolute EGT differences due to differences in combustion airflow and EGT probe locations between cylinders. Mostly asking to expand my knowledge if you have already considered this.

 

[hgerhardt]

Looking at your EFIS screenshot, I'd say 1&3 are acting similarly and 2&4 are acting similarly. It's not just cyl 3.

CHT could be explained by better cooling on the RH side, but not EGT, unless probe placement is different left cylinders to right cylinders.
So, what else could be causing the RH cylinders to behave differently than the LH side?

Edit: I'm going to go out on a limb here, but it seems the compression ratio is different L to R sides. You could test for this by doing a cranking compression test. Not a differential test.

 

[DanH]

Looking at your EFIS screenshot, I'd say 1&3 are acting similarly and 2&4 are acting similarly. It's not just cyl 3.

Good catch. You're right. Nothing like the description.

 

[BobTurner]

I am totally unfamiliar with SDS. On a mag, half the sparks come from the N pole, half from the S. Anything like that on SDS, where half the sparks might have a different intensity or timing than the other half?

 

[skirting_virga]

Looking at your EFIS screenshot, I'd say 1&3 are acting similarly and 2&4 are acting similarly. It's not just cyl 3.

I noticed this also but kind of figured he may have picked a less representative screenshot and the trends he is noticing are specific to #3. Certainly worth clarification...

 

[akschu]

Are you certain you don't have an induction leak?

 

[akschu]

Does it behave differently at low altitude?

 

[BillL]

What in interesting condition. It indeed looks like 1 and 3 are doing the same. Since the CHT and EGT and both acting consistently, it is worthwhile to believe it is real. Although the "flame" is not going out the exhaust the heat energy sure is. Consult the PV diagram. Assuming perfect instrumentation, then this means the expansion work on the piston is less . More work removed would put the energy into the crankshaft not into the head and exhaust. Possibilities are real ignition timing (not the same peak pressure point) lower compression ratio (slower burning) . . . ok kinda the same but different. And something funny with the valve timing. Something mechanical . . . . . affecting the events, al low probability I think. It is almost like the cylinders on that side are farther from the crank - Cr would absolutely do this primarily affecting heat release rate. If the timing could individually be adjusted, it could be tested.

One way to know what the power produced by each individual cylinder is to cut it out. Since the SDS can apparently do this it would be one of my favorites. The measurement is at a steady cruise condition and measure the airspeed change with the cutout. If that is ignition the FF would not change, but if it is an injector, then the FF could also me measured. Personally doing cruise mag checks, it takes about 15 sec for the airspeed to drop. I would hope it is a fuel cutout.

The ground suggestion to do a spinning compression test might be helpful, but only to test the Cr theory. I have seen a dead cam but all the pressures were equal - fool me once.

AirGuy, these should make some sense to aid in troubleshooting.

 

[airguy]

Lots of good points made here, and you've given me some points to follow up on, thanks.

It is a Superior cold air sump, so it's possible this is an artifact of the intake runners I suppose. You are correct that the data indicates #1 and #3 are behaving together, #3 seems to be more affected but they are together. I have my oil cooler pulling air off the back of #4 so in theory there may be more pressure available on #1/#3 side for cooling which would explain the lower CHT - but that would have no effect on EGT. I have looked at all my induction lines and they are good - but this condition is present at wide-open throttle (my normal run configuration) so there would be negligible pressure drop across any leak in the first place - and the GAMI spread test would highlight a leak if it were present.

Dans question about how long it has been going on is a good one - the lower CHT was present at least a thousand hours ago, but my early Phase I data doesn't show it. At that time I was operating on Bendix injection and one mag, one Plasma EI. That got me thinking and I started digging back into my records - I converted the engine to SDS in January of 2020, at 547 Hobbs time. The pic below is just after that point, and is the first saved screenshot I have that clearly shows the discrepancy in CHT, but the EGT's are all still pretty close. A changing EGT over time could be a probe or a wearing valve, but the CHT changing with the installation of the SDS does point to something involved with the SDS install, anecdotally at least. Now I'm going to have to carefully comb through the few hundred screenshots I've saved over the last thousand hours and try to find where the EGT started moving, and correlate that with maintenance going on. Gee, thanks a lot for that Dan

The symptoms are there for both ROP and LOP, high and low altitude, so that points in the direction of a real hardware issue unrelated to mixture. My timing does vary between sea level and about 7000' for detonation margin running 93E10 via the SDS map, but I spend so little time at low altitude (and stabilized, where a screenshot would make sense) that I don't know if I'll have any data for that - I'll have to comb through the screenshots. The SDS coil packs have 4 wire outputs driven by 2 coils on each pack. The #3 top plug is fired by the left pack and the bottom plug is fired by the right pack - my next task along those lines would be to swap the packs left for right and see if that makes a difference. In theory I suppose a weak coil on the left would result in a weak spark on top, which might make a difference in the flame front and resulting burn pattern, but I have replaced that left coil pack once in the last 500 hours due to a cracked case, I've replaced all the wires once, and the plugs never survive an annual.

Lower Cr is unlikely - but not impossible - I did buy the engine used with 105 SMOH, rebuilt by an individual, not a certified shop - so a hardware discrepancy there is marginally more likely than a new or "conforming" engine. Lower Cr would have shown up in the CHT in the first 500 hours though when I was running Bendix and mags - it would not correlate with the SDS install and my screenshots do seem to indicate that correlation is real.

 

[blaplante]

Pushrod length doesn't change when the valve opens. The hydraulic unit makes pushrod length effectively the same for all, zero clearance at the cam. Zero dry tappet clearance (pushrod really too long) wouldn't allow the valve to seat, and it should have overheated long ago.

Sticky lifter? Which isn't fully making up the valve lash? In a Corvair (car) easily heard as the "Click and Clack" (as in the radio show)... but I wonder if it would be audible in an unmuffled Lycoming.

 

[blaplante]

Put your injectors on a flow bench. Or swap injectors left to right sides.

 

[Finley Atherton]

Mogas? Exhaust valve seat recession??

 

[airguy]

Mogas? Exhaust valve seat recession??

I'm running 93E10 - but the old wives tales about valve recession are just not supported by data. I'm about 99% springloaded to call BS on that idea just on face value - but that 1% of me is going to wait until I have a chance to actually pull fresh borescope pictures and post them here.

 

[rv8ch]

Nothing to add on the cause of the CHT/EGT difference, but I sure hope you always have those kinds of tailwinds. Impressive!

 

[Finley Atherton]

I'm running 93E10 - but the old wives tales about valve recession are just not supported by data. I'm about 99% springloaded to call BS on that idea just on face value - but that 1% of me is going to wait until I have a chance to actually pull fresh borescope pictures and post them here.

A sample of one. A friend here in Australia had rapid seat recession in his new Lycoming RV 10 in all 6 cylinders using mogas. Could check the dry lifter clearance to rule it out.

Fin 9A

 

[1949MGTC]

I had a similar issue with #3 cylinder. I swapped injectors between 1 and 3 and it followed the injector. After multiple cleanings and using a loupe to check the injector I ultimately bought a new injector and that resolved the issue

 

[rockyfatcat]

I had a similar issue with #3 cylinder. I swapped injectors between 1 and 3 and it followed the injector. After multiple cleanings and using a loupe to check the injector I ultimately bought a new injector and that resolved the issue

Swapping stuff and watching if the problem follows is always a good troubleshooting technique. Have you swapped the CHT or EGT sensors to other locations? You might find that you are chasing the wrong issue.