Piston failure analisis

[danl]

I lost a piston the other day before an event and I wanted to run my "theory" and I use that term loosely, by others that are smarter than I. Since this failure I've done a lot of reading on piston failures. One of the best articles that I've read describing it is written by what seems to be an OEM R&D tester for GM's Northstart V8 project. http://www.streetrodstuff.com/Articles/Engine/Detonation/ .

The failure went like so. I had done about 10 pulls earlier in the day checking AFR's trying to lean it out from bottom 9's to somewhere in the mid 10's. I did that but kept my timing the same. The car was knocking but I didn't have the logger on, I just watched the wideband out of the corner of my eye (which is both dumb and unsafe but I was in a hurry). Things were going great, the last pull I looked at my knock gauge and it was pegged (stock ecu boost gauge reads knock). I drove it back home and burned a new chip. Timing came in at about 3 deg at 4000rpm and ramped up to 8 deg by 7500rpm. My first pull on the new chip everything was great and the car had tons of power. On the return pull with no cool down I banged 4th and smoke out the back. It was thin at first so I thought I'd get to the top of a mountain (1 mile) where I could kill the engine and coast home. I only made it halfway up.

Important details. It is a stock motor with A1 headstuds on an OEM gasket. The car runs a 50 trim turbo with a hot pipe. Boost spikes to 38-40psi (guess can't watch the gauge) and tapers from there, probably 35psi or so to redline. Fuel is E85 and I throw 2000cc/min of methanol at it. 1000cc/min injectors are maxed out. The car had fuel, it was about mid 10's give or take for the whole pull.

My theory is that I cracked a second ringland or compromised the top ringland when I was detonating earlier in the day. During my shift to fourth it did knock but it only took about 1/4 second to go from no smoke to smoke. I think that if the second ringland was cracked that I took it out on the shift and torched the piston, the more I drove it the bigger the hole got till the motor shut off. If the top ring land was compromised it allowed heat to get to the rings and/or second ringland and torched it that way. Signs of detonation are the scuffing of the piston skirt due to a hot piston during the detonation events. However the cylinder head and plug looks clean. Also the piston top looks good as well. Also headgasket held go A1's!

For those interested yes the dipstick blew out, got oil all over the hot turbo and caught on fire. :) Luckily I carry a fire extinguisher but the engine bay was a mess.




http://i298.photobucket.com/albums/mm280/dsmdano/piston.jpg

[Marco]

That piston went lean and burned for whatever reason. Pistons smoked like that are from a lean condition. Everytime I've seen a piston torched like that there usually might be evidence on some of the other pistons of a lean condition, like pitting on the crown above the top ring land, and edges.

If you are injecting methanol in the pipe, it can not mix correctly as I'm sure you know, another thing to check is to have your injectors flowed. Your Wideband may read rich, but that's an average between 4 cylinders, and something to consider.

Last thing is you guys should really read your plugs, widebands may lie, plugs tell the tale immediately.

[Fathouse]

Looks like the pistons I pulled out of my turbo busa after to much boost and 200mph highway runs. :)

Fathouse

[Marco]

Although I don't agree entirely with that guys article you posted, so far I find it for the most part 90 percent accurate.

I did find this in his article though which explains in one paragraph what I think happened to your piston from what I can see.

" Some engines, such as liquid cooled 2-stroke engines found in snowmobiles, watercraft and motorcycles, have a very common detonation failure mode. What typically happens is that when detonation occurs the piston expands excessively, scurfs in the bore along those four spots and wipes material into the ring grooves. The rings seize so that they can't conform to the cylinder walls. Engine compression is lost and the engine either stops running, or you start getting blow-by past the rings. That torches out an area. Then the engine quits."

[danl]

Thanks for the input Marco. So do you think its detonation from the article or going lean thus causing detonation? Maybe something happened on the shift (stuck injector) on that cylinder. It knocked horribly on the shift and quickly went south.

I wasn't being fair, here are some more pictures. My numbering scheme is timing belt piston #1 and tranny side piston #4.

Tops of the other 3 pistons. Notice the rings dangling, they are still in great shape:
http://i298.photobucket.com/albums/mm280/dsmdano/000_1288.jpg?t=1244919389

Exhaust side of the three pistons:
http://i298.photobucket.com/albums/mm280/dsmdano/000_1289.jpg?t=1244919445

Intake side of the three pistons:
http://i298.photobucket.com/albums/mm280/dsmdano/000_1290.jpg?t=1244919471

Piston #1 again top/side (exhaust) view:
http://i298.photobucket.com/albums/mm280/dsmdano/000_1282.jpg?t=1244919536

Piston #1 thus again. Side view:
http://i298.photobucket.com/albums/mm280/dsmdano/000_1283.jpg?t=1244919586

Also all the plugs (even the holed cylinder) look good. My camera is horrible at taking those pictures but I can try if you wish. Other than oil on the plug in the cylinder that let go nothing out of the ordinary. Perhaps a little bit of aluminum speckling but that is a chicken or the egg ordeal. The plugs are good enough that they will start up and run a car just fine.

[jrohner]

The failure went like so. I had done about 10 pulls earlier in the day checking AFR's trying to lean it out from bottom 9's to somewhere in the mid 10's. I did that but kept my timing the same. The car was knocking but I didn't have the logger on, I just watched the wideband out of the corner of my eye (which is both dumb and unsafe but I was in a hurry). Things were going great, the last pull I looked at my knock gauge and it was pegged (stock ecu boost gauge reads knock). I drove it back home and burned a new chip. Timing came in at about 3 deg at 4000rpm and ramped up to 8 deg by 7500rpm. My first pull on the new chip everything was great and the car had tons of power. On the return pull with no cool down I banged 4th and smoke out the back. It was thin at first so I thought I'd get to the top of a mountain (1 mile) where I could kill the engine and coast home. I only made it halfway up.

Important details. It is a stock motor with A1 headstuds on an OEM gasket. The car runs a 50 trim turbo with a hot pipe. Boost spikes to 38-40psi (guess can't watch the gauge) and tapers from there, probably 35psi or so to redline. Fuel is E85 and I throw 2000cc/min of methanol at it. 1000cc/min injectors are maxed out. The car had fuel, it was about mid 10's give or take for the whole pull.


If you mean 9's & 10's for a gas AFR running E85, how can your car even run decent being that rich? If you meant E85 AFR, I can see where you're problems came from being that lean.

You're using a LOT of fuel, I don't know what hp you're at but just the 1000's alone should be easily good for easily over 500awhp on E85 (unless your fuel pump is not keeping up), plus you're giving a decent amount of meth.

From the description of you're timing map, your timing is awfully low, and it's probably contributed to being so rich.

[Marco]

It depends, if its been running lean for a while, you will usually see pitting all around the edges of the pistons. When its holed that bad its hard to put a finger on exactly what the hell it was, that initiated the destruction, so we go back and check everything. First thing we do is number the injectors and get them flowed.

[danl]

Point noted Marco. I'm going to junk these injectors and go with a set of Bosch's like I'm running in my evo.

[The Wookie]

Pits on intake side- detonation.

Pits on exhaust side- lean.

If the stock piston ring land gets broken from detonation its very very very possible to torch the piston in that area, all that heat and pressure has to go somewhere.

[danl]

That is good to know Wookie and backs up what happened to my last motor. I'm going to check the injector wiring on #1 and replace all the injectors with new units while its apart.

[Marco]

Replacement doesnt get to the root of the problem. Checking the old ones getting the data and then backtracking from there will. You will also be that much smarter in the end :D

[danl]

Yeah, I guess I got to start learning from my mistakes huh? Brute force only gets you so far. :)

EDIT:

Also does anybody have any good information on where to look to learn about reading plugs in a performance turbocharged application? The only things I've been able to find are written towards carbed normally aspirated V8's and deal with how to jet with respect to reading plugs. I was wondering what a lean plugs looks like. Currently I just look for peppering on the white portion and/or silvery splatter to indicate detonation and combine this from knock sensor and wideband readings. Being able to properly decipher plugs will help me on a cylinder to cylinder basis.

[Blaine]

Yeah, I guess I got to start learning from my mistakes huh? Brute force only gets you so far. :)

EDIT:

Also does anybody have any good information on where to look to learn about reading plugs in a performance turbocharged application? The only things I've been able to find are written towards carbed normally aspirated V8's and deal with how to jet with respect to reading plugs. I was wondering what a lean plugs looks like. Currently I just look for peppering on the white portion and/or silvery splatter to indicate detonation and combine this from knock sensor and wideband readings. Being able to properly decipher plugs will help me on a cylinder to cylinder basis.

Why would the type of induction change the way you read plugs?

[Marco]

Reading plugs requires the motor to be shut off immediately when it is in the operating range that you are tuning. Also works easier if you have alot of plugs, especially clean ones.
It worked for Carburetors because they were a bit of simpler metering device than EFI. problem is you can be way off on your tune on an EFI car and not know it because you will pass over that cell fairly quickly, but it may be enough to get you into trouble. I've seen 'tooned' cars that wouldn't hold a candle to even an untuned carburetor. Wookie know more about carbs than I'll ever know, but I know how to tune them. As a matter of fact anybody that tunes any car should know the 5 metering circuits of a carburetor off by heart, understand these principles and you will understand tuning EFI much better.

I tuned JR motorsports motor entirely by plug color on the dyno because we were getting screwy readings from the wideband and the egt's. Sometimes you need to get back to basics.

Plug tuning on different fuels will give you different colors as well, so the best thing to do is start by checking your plugs frequently and learning the information as you go on a good running motor.

Heat range selection should be chosen by counting the number of heat effected threads on the plug.

[brucer]

danl,

everything thats been said is right,

here is a link if you want it
http://www.dragstuff.com/techarticles/plug-pictures.html

my problem is as stated, you need to run a fresh plug and kill it as soon as the pull is over. so the turbo is baking. that is to get fuel, timing marks might last through some general driving.

[danl]

Thanks guys.

Marco, the plugs are boxed up and ready to ship out. I called FIC and they will take delivery. I didn't tell them what cylinder let go, but the plugs are numbered and we shall see the results.

Brucer, thanks for the link. Looks great.

[Lolli DSM]

So did you get any results?

[danl]

All the injectors tested out fine. I also inspected my injector wiring harness as well as the clips (which I replaced a number of years ago). They were fine as well. I then replaced the injector clips to a newer lower profile style and bought some nice ratchet crimpers to do a proper crimp job as I've had problems with solder joints in the past.

[Lolli DSM]

Ok.. what car is this you have? Looks like dsm 7.8/1 comp pistins.

[Marco]

So, now you have to narrow it down to the tune.

[ShaneZ]

Single nozzle meth injection claims another?

Too low of timing?

I run 17 degrees of timing on a stock motor 1G at 34psi on E85 with a 75 shot of nitrous on it.

[danl]

So, now you have to narrow it down to the tune.


Point taken. The new motor is up and running, stock pistons still :)

[Lolli DSM]

Single nozzle meth injection claims another?

Too low of timing?

I run 17 degrees of timing on a stock motor 1G at 34psi on E85 with a 75 shot of nitrous on it.

Haha I blew 2 sets of stock 1G pistons on 30psi. But it was a stupid tune :D

[D.AsyMn]

Can anyone tell me if a piston is put on a rod backwards would there be any problems if it was installed in the engine with the piston with the arrow pointing towards the timing belt. I had the 2g pistons put on the 1g rods and the guy put one rod on backwards. Please help he said it shouldnt cause any problems but before I put it together Id like to know for sure.

[Aaron@English]

Can anyone tell me if a piston is put on a rod backwards would there be any problems if it was installed in the engine with the piston with the arrow pointing towards the timing belt. I had the 2g pistons put on the 1g rods and the guy put one rod on backwards. Please help he said it shouldnt cause any problems but before I put it together Id like to know for sure.

He's wrong, offset pin. Take it back and make him do it right.

[Marco]

Can anyone tell me if a piston is put on a rod backwards would there be any problems if it was installed in the engine with the piston with the arrow pointing towards the timing belt. I had the 2g pistons put on the 1g rods and the guy put one rod on backwards. Please help he said it shouldnt cause any problems but before I put it together Id like to know for sure.


It will cause big problems, the offset changes rod angle at TDC, it sould be like advanceing your timing 6 degrees

[D.AsyMn]

There used to be a forged flat top available, so keep surfing Ebay. Yes, any of the piston manufacterers will make any piston you want, just be prepared to wait.