Catastrophic engine failure, suspect pre-ignition (Pics to follow)

[FWombat]

What was your ignition timing like around the point of failure?

Years ago, I tore a rod in half on my stock engine (stock rods, pistons, turbos, etc). I was using an AEM, full-blast 9B turbos, and spraying a TON of pure methanol (around 1000cc/min). The car was pulling smooth and clean when suddenly it felt like it hit a wall. Turns out that a rod had twisted 90 degrees and then sheared apart, splitting the piston in half. The ring lands on the piston were intact and the piston face didn't show signs of significant detonation.

Later looking at the log of the failure, the knock sensor didn't show even a hint of knock until it went nuts from the half-rod hitting the block over and over. Just before this, I saw a huge spike in MAP and AIT - it pegged the MAP sensor, and AIT went up to something like 100+ degrees at the same time. A harddrive crash lost the log.

My theory on the failure cause? The waste spark lit off the methanol charge in the intake manifold. If ignition timing is LESS than 16 degrees BTDC, then the waste spark will occur AFTER the intake valve on that cylinder has started to open, potentially exposing the intake charge to the spark. If it's open enough...BOOM.

Here's a diagram of the stock cam timing to illustrate (courtesy of Jeff Lucius's always excellent stealth316.com):



- Brian

[Dj4play]

Originally Posted by DoctorDex

I got working on pulling the motor today. The motor is _nearly_ ready to come out, just need to get my hands on a cherry picker, which won't be possible until this weekend.

One thing I was't sure what to make of was the fact that when I drained my oil, there was more coolant in my oil than there was oil in pan. In addition, there was fair sized chunk of metal that became lodged in the drain hole.

sounds like the broken piston piece was large enough to lodge into a coolant jacket into the head. If that is the case the head is junk

[DoctorDex]

Originally Posted by FWombat

What was your ignition timing like around the point of failure?

Ignition timing just before the event was 11*.

Originally Posted by FWombat

Years ago, I tore a rod in half on my stock engine (stock rods, pistons, turbos, etc). I was using an AEM, full-blast 9B turbos, and spraying a TON of pure methanol (around 1000cc/min). The car was pulling smooth and clean when suddenly it felt like it hit a wall. Turns out that a rod had twisted 90 degrees and then sheared apart, splitting the piston in half. The ring lands on the piston were intact and the piston face didn't show signs of significant detonation.

Later looking at the log of the failure, the knock sensor didn't show even a hint of knock until it went nuts from the half-rod hitting the block over and over. Just before this, I saw a huge spike in MAP and AIT - it pegged the MAP sensor, and AIT went up to something like 100+ degrees at the same time. A harddrive crash lost the log.

My theory on the failure cause? The waste spark lit off the methanol charge in the intake manifold. If ignition timing is LESS than 16 degrees BTDC, then the waste spark will occur AFTER the intake valve on that cylinder has started to open, potentially exposing the intake charge to the spark. If it's open enough...BOOM.

That is a very interesting theory and I think you are on to something. Believe it or not, I had a few similar occurrences for the first time ~ 2 weeks before this happend. It all started when I lowered my base timing from 10* to 7.5* and dropped my base fuel pressure, to pull as much timing as possible for a hot weather knock cushion. That evening I went out and tried some tuning pulls and when I went WOT it felt like I hit a wall/fuel cut. There was no knock during the event and my EBC recorded a 36 psi pressure spike which was triggering the "warn" function on the Blitz DSBC. When I reviewed the logs, it didn't look like fuel cut, as the ECU wasn't pulling fuel. I didn't know what to make of it, but I assumed it must be fuel cut, because I couldn't think of anything else that could cause that feeling, so I leaned it out a tick only to hit the wall again (this time it was a 32 psi spike). I pulled over and increased my base fuel pressure by 4 psi and leaned it out another notch and this time I didn't hit the wall. However, I only tested it in 1st and 2nd gears. By leaning it out and increasing base fuel pressure I would have been increasing timing, which was inadvertently fixing the problem. Only problem is, I didn't test it in all gears. In 5th gear, there would be more load @ WOT and even less timing than in 1st & 2nd which would mean it'd be much more likely to happen in 5th.

[DoctorDex]

I guess my next question is, is there any way to prevent this from happening, without having to crank up my base timing to 16 degrees?

[FWombat]

Originally Posted by DoctorDex

Believe it or not, I had a few similar occurrences for the first time ~ 2 weeks before this happend.

There was no knock during the event and my EBC recorded a 36 psi pressure spike which was triggering the "warn" function on the Blitz DSBC.

Yeah, this sounds like the the intake charge lighting off. This is actually the first time I've talked with someone that had the data log capability in place to see the pressure spike.

You're exactly right that increasing your base fuel pressure and leaning out a notch would increase timing.

Originally Posted by DoctorDex

I guess my next question is, is there any way to prevent this from happening, without having to crank up my base timing to 16 degrees?

My solution has been to use a 50/50 water-meth mix. You still get fantastic cooling and some extra fueling, but the mixture doesn't seem to be ignitable. Reading the old WWII research papers, this is the mixture that they found to yield the best knock suppression results.

[DoctorDex]

If your theory is correct, then you`d think spark blow out would also be able to cause this sort of phenomenon. When the spark `blows out` it would leave a loaded cylinder, which would be burnt by the wasted spark. If this is the case, then I`m wondering why we don`t see this happening more often.

Originally Posted by FWombat

My solution has been to use a 50/50 water-meth mix. You still get fantastic cooling and some extra fueling, but the mixture doesn't seem to be ignitable. Reading the old WWII research papers, this is the mixture that they found to yield the best knock suppression results.

I read those papers as well, but I haven`t been able to replicate their results. I must admit that I found that knock control seemed to plateau somewhat at concentrations higher than 60:40 meth:h20. However, with that being said, I still had better results with 100% meth and it was easier to tune with (don`t have to worry about quench). However, after this incident, I`m a bit leary of ever running 100% meth again.

I apologize, my apostrophe key is broken, hence the ````

[Dj4play]

Originally Posted by DoctorDex

If your theory is correct, then you`d think spark blow out would also be able to cause this sort of phenomenon. When the spark `blows out` it would leave a loaded cylinder, which would be burnt by the wasted spark. If this is the case, then I`m wondering why we don`t see this happening more often.



I read those papers as well, but I haven`t been able to replicate their results. I must admit that I found that knock control seemed to plateau somewhat at concentrations higher than 60:40 meth:h20. However, with that being said, I still had better results with 100% meth and it was easier to tune with (don`t have to worry about quench). However, after this incident, I`m a bit leary of ever running 100% meth again.

I apologize, my apostrophe key is broken, hence the ````

your nozzle stuck. plain and simple. You cracked your piston before you engine failure.

[FWombat]

Originally Posted by DoctorDex

If your theory is correct, then you`d think spark blow out would also be able to cause this sort of phenomenon. When the spark `blows out` it would leave a loaded cylinder, which would be burnt by the wasted spark. If this is the case, then I`m wondering why we don`t see this happening more often.

A few things...

First, methanol ignites much more easily that gasoline, I believe.

Secondly, in your "blowout" scenario, by the time the intake valve opens most or all of the unburned fuel/air mixture will have already been pushed out the exhaust port.

Third, with methanol injection your entire intake plenum is filled with methanol mist. It's when the flame carries back from the intake valve into the plenum (and back down into other cylinders on their intake stroke!) that you get into trouble. With normal gasoline injection, the fuel is sprayed directly at the back of the intake valve, so this isn't a possibility.

[DoctorDex]

Originally Posted by Dj4play

your nozzle stuck. plain and simple. You cracked your piston before you engine failure.

I understand what you are saying, but I have a hard time believing it is just a coincidence that I had a number of these strange backfire type events in the days prior to the failure (I even had one occur the day before while in 1st gear). The way those backfire events felt, I`m sure they had enough punch to crack a piston. Similar to fuel cut they were able to stop my car dead in its tracks at WOT, except this definitely wasn`t fuel cut.

[keithmac]

From what I`ve found tuning for fixed rate methanol injection 2nd gear is useless, I logged my car on a few 1/4 mile runs and 3rd gear was leaner than 2nd for some reason.

I suggest the first time you get on it after a tune you keep a beady eye on the AFRs in the higher gears. I`m building an audible warning circuit for when afr`s drop leaner than 11.5:1 with meth active, this will also kill the boost controller.