What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Engine Roughness Immediately After a Roll.

R

RVSpike

Active Member
Patron
I had just departed a nearby airstrip in my O-360 A1A CS RV-4, leveled off at 2500' MSL and set power to 23" x 2400 and leaned per normal.
I then executed a roll to the left followed by one to the right. My engine immediately began to run very rough! I reduced power, enriched the mixture, switched tanks, pump on, changed prop rpm, no change.
I have a P-mag on the left and a LSE Plasma II w/hall effect on the right. I pulled the circuit breakers for those one at a time, still no change. I then noticed my CHT on #2 cyl was past redline, as in 535 degrees! The other cylinder temp were normal.
By then, I was just about back in the pattern and landed. After a 30 minute cool down, I restarted the engine, performed a run-up and it ran great?!
I tied her down, installed the canopy cover and got a ride home.

I had a similar episode several weeks ago where I pulled a few G's and experienced engine roughness that lasted only a few seconds. I dismissed it as sump water ingestion.
Tomorrow I will de-cowl, visually inspect, bore scope #2 and ground run to estimate whether I can ferry it home 14 miles north for more involved diagnostics?
I have no idea how doing a positive G roll could affect the engine?
I did recently complete the Condition Inspection and switched to Autolite 386 plugs. Additionally, #2 cylinder was pulled for re-hone a re-ring 75 hrs ago. It's been running great except #2 almost always ~20 degrees higher in cruise.
I am asking the VAF brain trust for any ideas?

Thanks in advance.
 
well, the idea that springs to my little brain (...) is a question Carbed to injected?

If carb equipped, you could have problems with your float giving you either too much, or not enough, dino juice. There have been quite a few float changes over the years...

If you're injected... :confused:
 
The engine is a narrow deck with a factory fresh MA4-5 with the "Mooney Mod'
installed.

Would a sunk float cause a high CHT in 1 cylinder? I'm at a loss.
 
Would a sunk float cause a high CHT in 1 cylinder?
Click to expand...

Probably not, but rather an increase in all of them before giving you the overtemp warning on one. But since you stated that the others were "normal"...
The same reasoning applies if we were to consider a problem in the ignition system.

No more ideas right now my side, but I'm sure all the engine gurus here will chime in :)
 
Did the engine lose power with the roughness or no power loss and just roughness?
Good Luck,
Mahlon
 
550-600 on the EGT is a classic symptom of running on one plug instead of two (effectively reduces ign timing and EGT rises accordingly with a CHT drop). This creates a perceived roughness that is generally considered minor and overall power is not really affected much. First thing to check is with issues on the plug wires for #2. Will need to get an ohmmeter on them and move the wires around, as it could be the G's that moved the wire a bit and created the open (depends upon how well supported the wires are). If that all checks out, the second likely scenario for roughness with high EGT is a sticking exh valve, but less likely to appear due to G's and often first presents with sicking when cold and disappearing after a few minutes. Not common to just happen for a few seconds mid flight then disappear; However, they do not follow a rule book, so anything can happen. The fact that you recently replaced plugs, further points to this being loss of one ignition source on #2. WIth classic auto plug wires, it is not that hard to create an issue in the plug end terminal of a plug wire when removing it. We have all done it when the rubber fuses to the ceramic and this is why they make special removers that grip the base of the boot. I recently pulled the 8 COP coils off the BMW for timing chain replacement and they were all fused pretty well. Thought I got them all off successfully, but are required to pull on the coil to get them off, as the rubber part is buried in a tube. After the work was done, was getting intermittent misfires on 3 cylinders.

We need a better explanation of roughness to narrow this down better.

Larry
 
Last edited:
RVSpike said:
The engine is a narrow deck with a factory fresh MA4-5 with the "Mooney Mod'
installed.

Would a sunk float cause a high CHT in 1 cylinder? I'm at a loss.
Click to expand...

No. float issues will make ALL of the cyl's either rich or lean, depending upon the float issue. You are looking for an issue that can be isolated to #2 and given the symptoms is MOST likely to be ignition or exh valve, due to the intermittent nature. WIth a carb, it is quite difficult to suddenly get one cyl to intermittently mis-behave.
 
Last edited:
Thanks for the responses.

There was a loss of power associated with the roughness. If felt and sounded like one cylinder was dead.
For clarification, the #2 cylinder spiked to 550 degrees, not the EGT.
I pulled the power back significantly once the roughness started to control the temp spike and to descend to pattern altitude, so an accurate estimate of power loss is only a guess.
There was no real change when individually switching off the ignitions.
I hope to find something obvious once I examine the engine this afternoon, until then, I’m stumped!
 
RVSpike said:
Thanks for the responses.

For clarification, the #2 cylinder spiked to 550 degrees, not the EGT.
I pulled the power back significantly once the roughness started to control the temp spike and to descend to pattern altitude, so an accurate estimate of power loss is only a guess.
There was no real change when individually switching off the ignitions.
I hope to find something obvious once I examine the engine this afternoon, until then, I’m stumped!
Click to expand...

My apologies for mis reading that. Definitely a whole different kettle of fish. How fast did it get to 550 and where did it start? Assuming the CHT sensor is accurate, about the only thing that can take a cyl from 400 to 550 in a few seconds is detonation or pre ignition and both are bad. In that case, I would not run it again without a borescope analysis. If it slowly rose to 550 since take off, then I would be looking for a birds nest or other air flow blockage. Once you get to 550, it is GOING to detonate and that would start the roughness.
 
Last edited:
lr172 said:
the only thing that can take a cyl from 400 to 550 in a few seconds is detonation or pre ignition and both are bad.
Click to expand...

Second this^. Mike Busch's engine book talks to this subject and I think they had a customer with a Cirrus eat its engine due to detonation during takeoff when Temps skyrocketed immediately. My guess is something internal with one of your Electronic Ignitions caused improper timing, resulting in detonation. Once your cylinder was scorching hot the detonation continued while shutting down either ignition at that moment, but then ran smooth again once engine was shut down and cooled. Please share borescope findings.

https://resources.savvyaviation.com/detonation-and-pre-ignition/
 
Last edited:
If your engine monitor records data, stick the flight into Savvy's app (free) and post the graphs. It might help shed some light.
 
post #1
"I have a P-mag on the left and a LSE Plasma II w/hall effect on the right. I pulled the circuit breakers for those one at a time, still no change."

post#8
"There was no real change when individually switching off the ignitions."

I don't think your ignitions are the problem, but your statements above in attempting to alternately check the ignitions by pulling the breakers one at a time may not be having the effect you wanted.
The Lightspeed ignition will turn off if you pull the breaker as it needs external power to run - but the P-Mag will keep on running (assuming you are over 900rpm) even though you pulled the breaker as it has it's own power source. So as stated, when you attempted to check the Lightspeed is running OK, the P-Mag was still running & would mask what ever the Lightspeed was doing (or not doing).
Both ignition systems have the option to have a 'P-Lead' wired to separate toggle switches (or ACS ign switch) that when grounded, would turn each ignition off just like a mag.

Back to the +G induced dead cylinder - this is a wild suggestion, but have a real close look for cracks in the exhaust flange on that cylinder. If maybe a crack is opening up on that pipe because of twisting under +G loading, this might direct exhaust gases to the CHT probe (high temp reading) & would kill the plugs in that cylinder as well if the back pressure was lost to the exhaust port.
 
Last edited:
Question and a thought

Question, do you have flop tube pickup in the tank being used

Thought: look for loose intake pipe on that cylinder.

I suspect the cylinder went lean (for some reason), was being operated in what Mike B calls the 'red fin', and went into detonation / preignition.

Borescope.
Also look for melted aluminum on the plugs...
 
Today I un-cowled, visually inspected and bore-scoped #2 cylinder. Everything looked good. Run-up was normal. Flew it home at 22” and 2400 rpm however, the #2 CHT crept up to 420 degrees, which is hotter than normal.
At the home hangar I re bore-scoped #2 with a couple of other experienced builders/mechanics as a second set of eyes. Everything was normal; good crosshatch, minimal scoring, usual carbon and lead deposits, valves and seats all normal. A compression test came back ok, but not stellar at 64 to 72 over 80, with #2 being the strongest.
The plan so far is to ditch the Autolite 386 plugs for IR27 and bushings, order a fresh set of wires for the LightSpeed, remove the valve springs and check for guide binding or wobble.

Unfortunately my VM1000 doesn’t record data other than max values, so a Savvy solution isn’t an option. I really don’t want to pull a cylinder unless I can find a reason to. Please keep the great troubleshooting ideas coming because I’m still stumped.

Ralph, good point on the C/B not killing the P-mag. And I like your wild suggestion, but we’ve been looking for anything using a high power light, any leak, streak, stain, crack, bird nest, and still, no joy.
 
RVSpike said:
Today I un-cowled, visually inspected and bore-scoped #2 cylinder. Everything looked good. Run-up was normal. Flew it home at 22” and 2400 rpm however, the #2 CHT crept up to 420 degrees, which is hotter than normal.
Click to expand...

Given winter temps in WA...
At those temps 420F CHT at that level of power is astonishing... and not good at all.
What did the top of the piston look like?
 
One more random thought... bad CHT sensor.
if nothing else, swap them and see if the problem follows the sensor.
maybe G's flexed the wire... some kind of connection issue... I'm not convinced, but it is a thought.
 
Water in gascolator?

Taltruda said:
When is the last time you pulled the drain plug out of the carb and ran the electric fuel pump about 30 seconds?
Click to expand...

Perhaps water in a gascolator bowl.
Or in a wing that managed to find it's way into the fuel intake tube.
 
Don’t think this is your problem but since you said you don’t know how positive g maneuvers can impact the engine I wanted to give you an example.

You can easily starve an engine in multiple rolls in a row even if they are positive (hoover type) rolls unless you have the trapdoor installed in the fuel-tank.

The centrifugal force will push the fuel to the outside of the tank away from the intake. What happens exactly (roughness, engine quits) and after how many rolls it happens depends on the amount of fuel in the tank and the details of your fuel system.

Typically though if that is the cause the engine will stop being rough or restart within a couple of seconds and operate normally thereafter.

So that doesn’t match all of your other symptoms.

Same can happen in an extended spin > 7-10 turns makes my engine stop.

Oliver
 
RVSpike said:
Today I un-cowled, visually inspected and bore-scoped #2 cylinder. Everything looked good. Run-up was normal. Flew it home at 22” and 2400 rpm however, the #2 CHT crept up to 420 degrees, which is hotter than normal.
At the home hangar I re bore-scoped #2 with a couple of other experienced builders/mechanics as a second set of eyes. Everything was normal; good crosshatch, minimal scoring, usual carbon and lead deposits, valves and seats all normal. A compression test came back ok, but not stellar at 64 to 72 over 80, with #2 being the strongest.
The plan so far is to ditch the Autolite 386 plugs for IR27 and bushings, order a fresh set of wires for the LightSpeed, remove the valve springs and check for guide binding or wobble.

Unfortunately my VM1000 doesn’t record data other than max values, so a Savvy solution isn’t an option. I really don’t want to pull a cylinder unless I can find a reason to. Please keep the great troubleshooting ideas coming because I’m still stumped.

Ralph, good point on the C/B not killing the P-mag. And I like your wild suggestion, but we’ve been looking for anything using a high power light, any leak, streak, stain, crack, bird nest, and still, no joy.
Click to expand...

IMHO, you need to find out why it got to 550. The fact that roughness appeared with th high CHT, kind of points away from a bad probe, though I agree that a strategically placed crack in the exh flange area could cause it. I have run 386 plugs for almost a 1000 of hours and have all confidence that those, on their own, did not cause this. Very high advance doesn't mak sense either, as ALL of the CHTs would rise (ignition advance is system wide, not cyl to cyl - though a spark jumping across plug wires could cause that). It is very difficult to get a PV lyc with stock compression to detonate with CHTs below 450, so pretty confident that leaning was not the cause of this regardless of that redfin ****. Sorry but you just can't make a stock lycoming rise to 550 CHT via leaning, something more insidious is happening, assuming it is not a probe fault.

Did you look for poc marks on the piston crown? THis is a typical byproduct of severe detonation. Pre ignition typically produces holes or fractures.

Larry
 
Last edited:
Is it possible…?

I was involved with the assembly of a friends O-320 today. While checking and adjusting piston ring end gaps, it occurred to me that I may not have checked the gap at full depth (top of stroke) on my offending #2 cylinder. I verified proper gap at the normal 4” depth, but I didn’t confirm the minimum .0075 gap at full choke depth. Could this cause the hot cylinder conditions I’ve seen?

After this epiphany I bore-scoped #2, yet again, focusing on the upper combustion chamber area, but still saw nothing obvious.

This theory doesn’t align with pulling G’s as the cause, but instead with power reduction after period of max power and the resulting relative rapid cooling that could be causing the rings to bind at the top of stroke??
It’s the one idea that’s confined to the offending cylinder #2 and could explain slightly higher CHT after settling into a cruise config.

Am I on to something, or am I out in left field?
 
Seems possible that maneuvering could simultaneously stir up some water and shake a poor connection of the CHT probe. Have you tried wiggling the probe wires and see if this reproduces jumps in CHT?
 
RVSpike said:
I was involved with the assembly of a friends O-320 today. While checking and adjusting piston ring end gaps, it occurred to me that I may not have checked the gap at full depth (top of stroke) on my offending #2 cylinder. I verified proper gap at the normal 4” depth, but I didn’t confirm the minimum .0075 gap at full choke depth. Could this cause the hot cylinder conditions I’ve seen?
Click to expand...

No. When a proper ring gap does not exist, excessive heat is not the typical symptom. When the gap is too small, once the ring gets hot enough, it binds and breaks the ring land on the piston and often cracks the ring. This tends to drop compression, so no real opportunity to get that hot. You will often see scratching on the cyl walls from the loose broken parts. The fact that your symptoms abated after a while kind of removes this as a possibility. Gap size doesn't really matter until the point that it becomes 0 and then it binds. It is in there to deal with expansion from heat and leave a margin to insure it doesn't go to 0.

The critical measurement for ring gap on a choke bore cylinder is the one you missed. The gap down at the bottom is somewhat irrelevant, it is the one in the choke area that matters. It is difficult for the machinist to have repeatable choke tapers (they use stones to do it) and this is why the top measurement is so critical to deal with variances and keep an adequate margin.

I would stop focusing on the G's as the cause. You have no real evidence to put the blame on the G's; Could just as easily be coincidence. You still haven't answered how fast the CHT rose from 400 to 550 and that is a critical piece of info to ferret out a cause. A slow rise and a rapid rise point to very different things.

I would move away from fuel issues. You simply can't make the CHTs go to 550 due to too much or too little fuel without something else being out of whack or a cooling airflow problem, especially at 2400 RPM.

Larry
 
Last edited:
kaa said:
Seems possible that maneuvering could simultaneously stir up some water and shake a poor connection of the CHT probe. Have you tried wiggling the probe wires and see if this reproduces jumps in CHT?
Click to expand...

I struggle to see this as a probe issue, as roughness appeared with the high reading and this would be expected with such a high CHT. I don't see how it doesn't detonate at that high of a CHT. Not saying it can't b the probe, but the roughness that appeared with the high CHT would seem to indicate that it was real. However, that could be coincidental as well.
 
The engine roughness points to a real problem; not a just a CHT probe problem.

This does not rule out a CHT probe problem accounting for the 420 CHT reading at low power cruise. Swapping probes might be helpful. Alternatively, using thermo crayons on all the cylinders might be helpful. If you can keep three cylinders well below 400 on takeoff and climbout, a 400 degree thermo crayon would tell you if #2 is actually going to 420. Another possibility is Temperature Indicating Brake Paint. It goes through nine distinct color changes starting at 360 with the next change at 430.

Back to the roughness and high CHT. The problem seems induced by maneuvering. Maneuvering can cause wires to move around, possibly causing a crossfire between adjacent plug wires. You could examine the plug wires for signs of damage due to abrasion or crossfire. Putting separators between the wires where they are close could help as well. Where they pass through the baffles could be a trouble spot.

That's all I have for now except a couple of links. Good Luck.

https://www.amazon.com/Markal-Thermomelt-Temperature-Indicator-Fahrenheit/dp/B001YYBIPC?ref_=ast_sto_dp&th=1
https://www.pegasusautoracing.com/productdetails.asp?RecId=4974&utm_source=google&utm_medium=cpc&utm_campaign=3156&gclid=CjwKCAiAlp2fBhBPEiwA2Q10D9-y0Tz3fXbYHXETnCiCtEW8JaNszUSBYZTfLpWibJAhVIhsZbMVTxoCRm0QAvD_BwE
 
Thank you Larry and others for your analysis and ideas.
I have new bore-scope pictures that show a series of radial gouging or cracking along the bottom of the cylinder. The cylinder top shows only minor scuffing. Tell me what you think. The piston top and edges appear clean with no signs of thermal damage. There is no external evidence of cracking, etc. I am going to proceed with cylinder removal unless there is some reason I should hold off?
 

Attachments

  • 96130B4E-14CE-475F-B107-B4BC0DA2DB36.jpeg
    96130B4E-14CE-475F-B107-B4BC0DA2DB36.jpeg
    105.2 KB · Views: 52
  • B8368127-429E-4794-8E04-39FD6C079366.jpeg
    B8368127-429E-4794-8E04-39FD6C079366.jpeg
    103 KB · Views: 65
I am going to proceed with cylinder removal unless there is some reason I should hold off
Click to expand...

The main reason to hold off is that apparently there’s no reason to remove the cylinder… I’d wish mine, on a perfectly running motor, looked akin :)
Many hypotheses on this problem by most experts now, and we are all holding our breath as to the very cause of your symptoms.
Always eager to learn…
 
The cylinder is off and as is often the case, it looks better than what’s seen in the bore-scope pictures. But due to my undiagnosed symptoms, every detail is suspect. There is a small bluish area at the very top of the cylinder, a small hue of brown here and there, and of course a few vertical piston scuff marks, none of which can be felt. The locals on the field wished their cylinders looked as good.:rolleyes:
I plan to give a few passes with the ball hone, and then reinstall with the same rings then continue troubleshooting. FYI, this engine has only 106 SMOH a 63 since #2 was re-honed and re-ringed.
 

Attachments

  • 8454B002-7E43-4F1D-8E35-8E1DEA6FD099.jpeg
    8454B002-7E43-4F1D-8E35-8E1DEA6FD099.jpeg
    150.6 KB · Views: 78
  • 867BFC62-B146-4F77-B29F-EBE3BB788813.jpeg
    867BFC62-B146-4F77-B29F-EBE3BB788813.jpeg
    154.5 KB · Views: 78
  • 9B887FA2-5F79-48F0-8D0E-0CD8E0D0C327.jpeg
    9B887FA2-5F79-48F0-8D0E-0CD8E0D0C327.jpeg
    94.6 KB · Views: 91
Did you take a really close look at the Autolite plug tip for that cyl compared to the other three?
 
I've only seen one post mentioning spark plus cross firing, I would look into this. What condition are the pug wires in, and how are they routed.
 
lr172 said:
I struggle to see this as a probe issue, as roughness appeared with the high reading and this would be expected with such a high CHT. I don't see how it doesn't detonate at that high of a CHT. Not saying it can't b the probe, but the roughness that appeared with the high CHT would seem to indicate that it was real. However, that could be coincidental as well.
Click to expand...
Right, but the mechanism by which such high CHTs would cause engine roughness would probably leave the piston at least a bit melted. It's hard to see what would cause roughness, such high CHTs and leave the piston and cylinder unscathed.
 
RVSpike said:
The cylinder is off and as is often the case, it looks better than what’s seen in the bore-scope pictures. But due to my undiagnosed symptoms, every detail is suspect. There is a small bluish area at the very top of the cylinder, a small hue of brown here and there, and of course a few vertical piston scuff marks, none of which can be felt. The locals on the field wished their cylinders looked as good.:rolleyes:
I plan to give a few passes with the ball hone, and then reinstall with the same rings then continue troubleshooting. FYI, this engine has only 106 SMOH a 63 since #2 was re-honed and re-ringed.
Click to expand...

Everything looks pretty normal except the worm shaped line on the cylinder wall and that is very irregular. Is it deep or superficial? Makes me think something may have been injested and burned causing the heat. I would closely examine the blue area. Blue is sign of overheated steel and is one of the first color change on the heat spectrum (550-600*). That may help you understand how the overheat event started. DO NOT Forget to gap the ring while in there. The little tan spots lower on the wall are possibly glazing during break in. Though straw/yellow brown is also a steel heat color (400-500), so that is also possibly an overheat signature.

With a re-hone, you need new rings. You can put the old rings back in the existing cylinder without much issue, but if you hone the cyl, you MUST put in new rings otherwise it will never break in correctly.
 
Last edited:
D Weisgerber said:
I've only seen one post mentioning spark plus cross firing, I would look into this. What condition are the pug wires in, and how are they routed.
Click to expand...

I mentioned that several days ago. It is not a high probability item due to the fact that 4 bangers only fire every 180 degrees and therefore a jumped spark would not occur in a dangerous portion of the combustion stroke. That is a much bigger issue on 8 cyl engines that fire every 90* which is late enough to actually ignite a charge and cause pre-ignition. While a misfire at BDC could cause roughness, it would not account for the high CHT, as at BDC the charge is uncompressed and therefore cannot be ignited by a spark.
 
Last edited:
kaa said:
Right, but the mechanism by which such high CHTs would cause engine roughness would probably leave the piston at least a bit melted. It's hard to see what would cause roughness, such high CHTs and leave the piston and cylinder unscathed.
Click to expand...

Not necessarily.

IF the CHTs rose to 550 and then mild to moderate detonation occurred due that heat, then there would not likely be much evidence, as the OP took action as soon as the roughness occurred. On the other hand if severe detonation caused the overheat, then yes there would likely be evidence of that. At a minimum we would see poc marks in the carbon coating on the crown area.

The OP found blue staining on the steel, confirming the overheat, so it could be deduced that the detonation was due to the overheat and NOT detonation causing the overheat, given no evidence of pre-ignition or severe detonation.
 
Last edited:
lr172 said:
While a misfire at BDC could cause roughness, it would not account for the high CHT, as at BDC the charge is uncompressed and therefore cannot be ignited by a spark.
Click to expand...


Am wondering why an otherwise ignitable fuel/air mixture cannot be ignited if uncompressed.

The P-MAG Installation & Operating Guide mentions that it uses a wasted spark system. This fires two spark plugs every 180 degrees; increasing the chances for crossfire.

Compressing an already ignited cylinder charge would seem to produce very high cylinder pressure with no detonation possible because the fuel has already been burned.

In my reasoning, crossfire cannot be ruled out as a cause of the rough running and high CHT. Further reasoned discussion welcomed.
 
koupster said:
Am wondering why an otherwise ignitable fuel/air mixture cannot be ignited if uncompressed.

.
Click to expand...

ignitability is directly related to level of compression. The more you compress it, the easier it is to ignite, all the way up to diesel level compressions where it self-ignites solely from the comression / no spark needed. From my research a simple spark from a sparkplug will not ignite an uncompressed fuel/air charge. It must be compressed first. It probably can be ignited by a flame or more aggressive means, but not a tiny spark. If you are curious, pull one plug out of each cylinder (i.e. no compression) and see if it kicks over and runs. My money is on the fact that it won't. Adequate CR is a key principle in the otto cycle engine. It can ignite with partial compression (90* BTDC in my explanation above), but not 0 compression. It may happen occasionally by chance, but not on a regular basis.
 
Last edited:
lr172 said:
ignitability is directly related to level of compression. The more you compress it, the easier it is to ignite, all the way up to diesel level compressions where it self-ignites solely from the comression / no spark needed. From my research a simple spark from a sparkplug will not ignite an uncompressed fuel/air charge. It must be compressed first. It probably can be ignited by a flame or more aggressive means, but not a tiny spark. If you are curious, pull one plug out of each cylinder (i.e. no compression) and see if it kicks over and runs. My money is on the fact that it won't. Adequate CR is a key principle in the otto cycle engine. It can ignite with partial compression (90* BTDC in my explanation above), but not 0 compression. It may happen occasionally by chance, but not on a regular basis.
Click to expand...

Hey Bill, you'll need to get me smarter here please. We light-off our gas turbines (continuous flow systems but initially spark ignited) on both NG and atomized diesel at essentially ambient pressure (unloaded compressors). Higher pressures tend to require higher energy as the sparks will "quench" as pressure increases. Let me know.
 
Freemasm said:
Hey Bill, you'll need to get me smarter here please. We light-off our gas turbines (continuous flow systems but initially spark ignited) on both NG and atomized diesel at essentially ambient pressure (unloaded compressors). Higher pressures tend to require higher energy as the sparks will "quench" as pressure increases. Let me know.
Click to expand...

You are correct that any viable fuel air mixture can be lit. It is impossible to light liquid fuel though - only the gasses will achieve a viable mixture. We used to pour diesel on blocks of wood in a stove in the sawmill I worked at long ago. On a freezing morning it would take multiple stick matches held right on the fuel wet lumber to vaporize enough to get a light. The same is true of gasoline though it is a lot easier to vaporize.

I'm guessing that you spin your gas turbines up to 15% RPM or so before you introduce fuel and spark. That is going to provide some compression and more importantly flow. Once a light off occurs, and the turbine spins up under pressure more fuel is added to maintain the proper ratio with the additional air provided by the compressor and it accelerates. If too much fuel is added too early - overheat. Also, if you did manage to ignite fuel/air in the combustion chamber at near zero RPM, it would be just as likely to expand towards the compressor as it would towards the turbine and you'd melt stuff without achieving a start.

Ed Holyoke
 
Bicyclops said:
You are correct that any viable fuel air mixture can be lit. It is impossible to light liquid fuel though - only the gasses will achieve a viable mixture. We used to pour diesel on blocks of wood in a stove in the sawmill I worked at long ago. On a freezing morning it would take multiple stick matches held right on the fuel wet lumber to vaporize enough to get a light. The same is true of gasoline though it is a lot easier to vaporize.

I'm guessing that you spin your gas turbines up to 15% RPM or so before you introduce fuel and spark. That is going to provide some compression and more importantly flow. Once a light off occurs, and the turbine spins up under pressure more fuel is added to maintain the proper ratio with the additional air provided by the compressor and it accelerates. If too much fuel is added too early - overheat. Also, if you did manage to ignite fuel/air in the combustion chamber at near zero RPM, it would be just as likely to expand towards the compressor as it would towards the turbine and you'd melt stuff without achieving a start.

Ed Holyoke
Click to expand...

Mostly yes, some no. As mentioned, we unload our compressors with very large extraction valves/piping. Like most things, it's to save money. Weight isn't a real concern with these large industrial turbines. Yes, there's some flow but less than most would think (can't give real numbers without risk of getting in trouble). The Ps isn't too much above Pamb. With a loaded compressor, the speed would be even lower as it would be like trying to strike a match in a hurricane. Fuel flow is managed in each can to a very small portion of the overall injection points available. We have the earthbound advantage of large starting motors, many > 3000HP. We also have the advantage of motorizing X00 MW generators to start our newer machines. An astronomical amount of torque but keeping teh extractions is cheaper than upsizing motors/torque converters or the large variable frequency drives.

The large available starting torque helps with the disadvantage of emissions permitting that includes the starting cycle in many cases. The fuel flow is initially just enough to keep the flames lit as it accelerates (for emissions) then start package input torque and fuel are opposingly scaled until the former is null at seq speed (60Hz).

Hadn't thought about this old part of my life in quite a while.
 
Last edited:
I didn’t suspect a probe issue but I checked it anyway. I used boiling water verify the CHT probe is working and accurate.
The main reason for removing cylinder was to check ring gap at the choke and it’s per spec at approx. .025”
Also as can be seen in the photos, there is no visible damage to the piston crown, no pock marks or craters, no signs of melting. All eight sparkplugs look identical.
I don’t know the rate of CHT temp rise, but I assume it was rapid because it was 500 plus the first time I looked.
I’m planning to re-install the cylinder with the same rings and without re-honing, and focus troubleshooting elsewhere.

As Jordan suggested, could an internal fault with either the Light Speed or the P-Mag send a fully advanced spark to just one cylinder? …….. a fatigue cracked circuit board that flexes under a G load? I’m taking wild shots in the dark. I will call the vendors for ideas.
 
Follow-up on the mysterious 'Engine Roughness After a Roll':

After re-installing the #2 cylinder, I spoke to Klaus at Light Speed Engineering.
He is convinced the AutoLight 386 plugs with their incompatible heat range is causing detonation and roughness.
So, I re-installed the bushings and new Denso IK27 plugs per Light Speed documentation on the top plug locations and installed new NGK BR8ES on the bottom locations per E-Mag documentation. Timing on both systems is set to 2.5 degrees ATDC.

Now with three hours of flight time on the new plugs, I am unable to reproduce the engine roughness, doing a roll or otherwise.
Cylinder temps are tolerable in the high 300's, normal for my installation including #2 occasionally rising to ~404 degrees.
I seems a number of people are using AutoLite 386's with success, so I'm proceeding with cautious optimism the issue is solved.
I'm going to continue to monitor temps and fine tune baffles etc. and consider cutting another inch off the lower cowl exit.

Thanks for the responses.
 
RVSpike said:
F
He is convinced the AutoLight 386 plugs with their incompatible heat range is causing detonation and roughness.
.
Click to expand...

Glad to hear all appears well, though I would keep an eye on things. Yes, the 386 is a hotter plug than we should use, but only 2 or 3 jumps on the scale to the hotter side. As mentioned, I have used them for 800 hours with no issues and remain convinced that they are NOT hot enough to cause detonation on their own, assuming CHTs are kept at reasonable levels. That said, if one lets things get too hot, then yes, the hotter plug's neg electrode will get to a glowing state, causing pre-ignition, before a cooler plug does. So, a cooler plug offers more margin when things go downhill. In your case, I remain convinced that the 386 alone didn't cause the overheat. So the question is why did the heat go up so high. I am a little concerned that you didn't find and address a cause and therefore may see the event again. Then again, it could have been a unique anomaly never to be seen again.

You can research plug heat ranges and what effects they have on various different aspects of the cylinders functions. Over time, you can read the heat signature on the neg electrode to see how hot or cold it is in your particular engine. I am still able to see a band on my 386's, telling me they are not excessively hot. Though it's position is further inboard than an optimal heat range. I did recently go to the NGK 4LRU plugs that have a slightly colder heat range due to the frustration with hte 386's rusting.
 
Last edited:
RVSpike said:
As Jordan suggested, could an internal fault with either the Light Speed or the P-Mag send a fully advanced spark to just one cylinder? …….. a fatigue cracked circuit board that flexes under a G load? I’m taking wild shots in the dark. I will call the vendors for ideas.
Click to expand...

Pretty sure the LS uses two wasted spark coils. Therefore, the CPU sends a signal every 180 degrees of crank rotation to fire one of those two coils. I would not expect a fault that would send errors to only one cylinder. It could lose TDC and mistakenly send a trigger at the wrong point, but it wouldn't keep hitting only one cylinder over and over. It would very likely be causing problems across at least two cylinders and more likely all of them.
 
You must log in or register to reply here.
Back
Top