Detonation, Running too lean, or high EGT's?

[pac1085]

I'm trying to figure out what caused my cylinder #4 piston to melt, I figured this could use it's own thread.

A friend of mine doesn't think it detonated, he thinks I would have bent a rod or actually broke (not melted) the piston.

The spark plug was fine and there was no damage to the head or turbo, I'm not sure where the chunk of piston went...maybe melted over on itself?

After looking at the datalog, my AFR was steady around 11.5 - 11.8:1 during the run. My coolant temp rose from 200 to 240 during the run. My IAT hit its peak of 160F right when the motor blew. 19 psi of boost, 20 degrees timing @ 6900rpm. I don't have any sort of knock sensor log or EGT. I didn't hear anything when it went, it just lost all power.

I have a Walbro pump, with a 4 bar FPR (OEM) running the Siemens 630's. Based on my trap speeds and my injector duty cycle I figure the car was making just under 400 crank hp.

I know cylinder 4 is at the end of the rail, maybe I need a bigger rail just to be safe...anyone know the limit of the stocker? I'll be installing a bigger intercooler and meth before taking it to the track again with the 2.0.

added details:
93 octane
BKR7E plugs

[solowb5]

Does that seem like to much timing uptop to anyone else? I would also say that you should be runing a larger rail @ 400 hp w/630cc. I'd say the metal obsorbed into the piston. Don't know if you've ever welded but, if you have. You would understand when you heat metal up and add filler rod it will just suck right into it if it's too hot.

[pac1085]

Yes, I'm pretty sure it was too much timing as well by about 4 degrees. I'm going to get the new motor dynotuned this time around, but just looking for additional opinions on what caused this.

[solowb5]

Then that is probally your answer. To much timing destroys alot of shit. I've melted all for plugs completely gone once and blew a headgasket the other time. Luckly nothing happened to my engine but, I had a built bottom end. Don't know if that made any differance but, I'd assume it had a part.

[pac1085]

I just looked back at the log and it actually was @ 20 degrees that run. I'm positive that's what did it - too much timing + high MAT and coolant temp. It looks like it exploded right up top too.

Looking back at the log I see two lean spikes, the first right after my launch in 1st gear while I was spinning, it went from 11.7 to 14.5 (just one record) then back into the 11's. The second lean spike was at around 6400 rpm in 2nd, went from 11.5 to 13.5 (for just one record) then back into the 11's. My other 5 logs were clean with no lean spikes...I'm assuming those were from knock?

[maxspeed]

I just looked back at the log and it actually was @ 20 degrees that run. I'm positive that's what did it - too much timing + high MAT and coolant temp. It looks like it exploded right up top too.

Looking back at the log I see two lean spikes, the first right after my launch in 1st gear while I was spinning, it went from 11.7 to 14.5 (just one record) then back into the 11's. The second lean spike was at around 6400 rpm in 2nd, went from 11.5 to 13.5 (for just one record) then back into the 11's. My other 5 logs were clean with no lean spikes...I'm assuming those were from knock?

some kinda miss thats for sure

[djwimbo]

One of the things I've noticed about everybody's datalogs is that cyl #4 always pulls timing first. I've seen flow data on the intakes, and it seems that Cyl 4 gets the least amount of flow, but that doesn't make sense as to why timing is usually pulled from that cylinder first.

The piston you burnt up is on the exhaust side correct? That would lead me to believe that high EGT's are involved.

IAT of 160 + 19psi + ECT~240 + 20* of timing = #4 burnt up. The self ignition point of fuel is 636 degrees for 87 octane, and I can't remember what 93 octane is.
Basically, your engine would've misfired under minimal load if ECT hit 240 anyways. Add the timing and the IAT's and she got real pissed off, real quick.

Have you removed that piston from the engine?
I am curious b/c my instructor awhile back was talking about piston rings walking around the piston over time. Obviously when installing a piston, it's rings have to be aligned correctly, but in some cases the rings themselves have spun around.
If you were getting too much blow-by on that cylinder through the ring end gap at that location, that would be another cause of the burnt out(and apparently missing) portion of that piston.
I just noticed that the burnt portion of the piston is right in line w/ a coolant passage as well... Just another thing to think about.

BTW 20* of timing advance wouldn't be that bad in a N/A application, but in the boosted world, I don't see that being a good thing

[Don Supreme]

20 on some race gas would have probably been OK...

[maxspeed]

20 on some race gas would have probably been OK...

dosent he run e85? so he should have had all the octane he needed...

[mike-2ptzero]

dosent he run e85? so he should have had all the octane he needed...

He was for some time but was running normal pump gas and a pump tune when the motor let go.

The stock rail is good for well over 400whp, TJM was doing it on their A4 and a few others.


The wide band also needs to be setup different when using E85 vs normal gas. Not sure if he did that for his setup.

I just looked back at the log and it actually was @ 20 degrees that run. I'm positive that's what did it - too much timing + high MAT and coolant temp. It looks like it exploded right up top too.

Looking back at the log I see two lean spikes, the first right after my launch in 1st gear while I was spinning, it went from 11.7 to 14.5 (just one record) then back into the 11's. The second lean spike was at around 6400 rpm in 2nd, went from 11.5 to 13.5 (for just one record) then back into the 11's. My other 5 logs were clean with no lean spikes...I'm assuming those were from knock?

pac1085 - where is your wide band 02 sensor located?

[pac1085]

He was for some time but was running normal pump gas and a pump tune when the motor let go.

The stock rail is good for well over 400whp, TJM was doing it on their A4 and a few others.


The wide band also needs to be setup different when using E85 vs normal gas. Not sure if he did that for his setup.




pac1085 - where is your wide band 02 sensor located?

the wideband doesn't have to be setup different, only if you want it to. lambda 1= lambda 1 irregardless of which fuel you are using. I left the display calibrated to show gasoline AFR's to make it easier for me (this doesn't affect the tune any)

but yeah, this run was done in pump gas with a different tune.

the sensor is too close to the turbo, only about 10 inches away. it will be lower when I get the new downpipe made for sure!

[mike-2ptzero]

the wideband doesn't have to be setup different, only if you want it to. lambda 1= lambda 1 irregardless of which fuel you are using. I left the display calibrated to show gasoline AFR's to make it easier for me (this doesn't affect the tune any)

but yeah, this run was done in pump gas with a different tune.

the sensor is too close to the turbo, only about 10 inches away. it will be lower when I get the new downpipe made for sure!

The lean spikes were most likely misfires not lean at all since the fuel is burning closer to the 02 sensor because you have it too far up stream.


Someone brought up the AFR once already on another one of your threads, stating that when using a alcohol based fuel which E85 is you must use a different calculation for the wide band. Even my LM-1 states this in its manual and in its programming software.

Alcohol fuels such as methanol and ethanol, are partially oxidized fuels and need to be run at much richer mixtures than gasoline. As a consequence, the total volume of fuel burned per cycle counterbalances the lower energy per unit volume, and the net energy released per cycle is higher. If gasoline is run at its preferred maximum power air/fuel mixture of 12.5:1, it will release approximately 20 MJ (about 19,000 BTU) of energy, where ethanol run at its preferred maximum power mixture of 6.5:1 will liberate approximately 25.7 MJ (24,400 BTU), and methanol at a 4.5:1 AFR liberates about 29.1 MJ (27,650 BTU).[citation needed] To account for these differences, a measure called the fuel's specific energy is sometimes used. It is defined as the energy released per air/fuel ratio. For the case of gasoline compared to the alcohol fuels, the specific energies can be found at:

If you were running the E85 at 11:1-12:1 which you would normal pump gas the car was running way to lean since it should be run much richer if your still using the wide band settings for both E85 and pump gas.

So the question to ask is what was your AFR while running the E85? Because the damage to the piston was done over a long period of time not just while running the car at the track on pump gas that night.

[pac1085]

My car wasn't running lean on E85. It's hard to explain, but I'll try.

Forget about AFR. AFR = (Lambda * X) where X is the stoich AFR of that fuel (such as 14.7 for gasoline)

Lambda of 1 is always stoich, it doesn't matter what type of fuel is used.

Converting lambda to AFR just makes it easier for humans to read. If my gauge read 11.5:1 when running E85, it was actually 7.57:1 in e85 AFR.

[mike-2ptzero]

My car wasn't running lean on E85. It's hard to explain, but I'll try.

Forget about AFR. AFR = (Lambda * X) where X is the stoich AFR of that fuel (such as 14.7 for gasoline)

Lambda of 1 is always stoich, it doesn't matter what type of fuel is used.

Converting lambda to AFR just makes it easier for humans to read. If my gauge read 11.5:1 when running E85, it was actually 7.57:1 in e85 AFR.

Even if you are right isn't that 7.57:1 still too lean for E85 seeing that they say it should be run at 6.5:1? Either way it sounds like you were running a bit on the lean side for a period of time on the E85.

[pac1085]

Well technically I was running E70 winter blend so that changes it a little :)

Either way, I'm not ruling that out as a possibility.