Reasons for head gasket failure.
Discussion
You could try re-building with standard studs and gaskets. Head gaskets are funny things, and what you might think is a better aftermarket gasket may be inappropriate for the block/head stiffness. It could be that the clamping load exerted by the ARP studs via the gasket is relaxing after a few heat cycles and/or producing plastic deformation in stud,head,block or gasket.
Go back to standard to generate a baseline.
Go back to standard to generate a baseline.
Sounds facile to say it, but are you ABSOLUTELY sure that the CR calcs are now correct and that the boost gauge is reading accurately? Wastegate functioning as required to limit peak OP?
The way you describe the failure, it just sounds like the pressure regimes being generated in the cylinders are above the limits of the gasket design.
Oh - one other thing. Was the same torque wrench used each time to torque down the head bolts? If it is not reading correctly then it might explain the failures. Silly maybe, but a thing to cross check if I were you.
>> Edited by nel on Wednesday 18th May 09:10
The way you describe the failure, it just sounds like the pressure regimes being generated in the cylinders are above the limits of the gasket design.
Oh - one other thing. Was the same torque wrench used each time to torque down the head bolts? If it is not reading correctly then it might explain the failures. Silly maybe, but a thing to cross check if I were you.
>> Edited by nel on Wednesday 18th May 09:10
I was advised that it is fairly common for blown motors to lift the heads slightly, and to make allowance for this in the design of the breather and water system. So it may be that in some situations even your correctly specced and fitted head gasket is running with relatively little/no clamping loads?
Perhaps a compression test would show whether there is something unusual going on in terms of cam timing or compression ratio?
Grasping at straws, supposing there was a massive restriction in the exhaust, massively wrong exhaust lift or timing, whatever that prevented the exhaust gasses from being expelled on the exhaust stroke. Wouldn't it be possible in theory for the re-compressed exhaust gases to reach a very high pressure indeed, significantly higher than at TDC on the compression stroke? I'm pretty sure you would have spotted the funny noises and total lack of power, like I said I'm grasping at straws here ...
Grasping at straws, supposing there was a massive restriction in the exhaust, massively wrong exhaust lift or timing, whatever that prevented the exhaust gasses from being expelled on the exhaust stroke. Wouldn't it be possible in theory for the re-compressed exhaust gases to reach a very high pressure indeed, significantly higher than at TDC on the compression stroke? I'm pretty sure you would have spotted the funny noises and total lack of power, like I said I'm grasping at straws here ...
Not that familiar with the subaru engine but a few questions.
Are both head gaskets affected or just one ? If only one could the head be cracked ?
If both are affected and its coping with normal driving then as soon as you give some welly they both give up and detonation has been ruled out (ecu probably doing what its supposed to be doing unless its retarding the ignition so far the cylinder temps are going way up - more likly to affect exhaust valves before gasket) then the heads must be lifting, therefore either the bolts/studs are stretching or they arn't torqued correctly.
No offence here but you're not using the wrong readings on the torque wrench ( N/m not lb/ft with it being a new one ) ?
Thought not, so maybe do a check with another torque wrench.
If they are both the same (ish) then then I would seriously suspect the bolts/studs ARP or not. If they are fairly long and the heat treatment wasn't quite right it wouldn't take a lot to just lift the heads a bit and hot combustion gasses sqirting from one cylinder to the other would soon destroy the gasket.
Are both head gaskets affected or just one ? If only one could the head be cracked ?
If both are affected and its coping with normal driving then as soon as you give some welly they both give up and detonation has been ruled out (ecu probably doing what its supposed to be doing unless its retarding the ignition so far the cylinder temps are going way up - more likly to affect exhaust valves before gasket) then the heads must be lifting, therefore either the bolts/studs are stretching or they arn't torqued correctly.
No offence here but you're not using the wrong readings on the torque wrench ( N/m not lb/ft with it being a new one ) ?
Thought not, so maybe do a check with another torque wrench.
If they are both the same (ish) then then I would seriously suspect the bolts/studs ARP or not. If they are fairly long and the heat treatment wasn't quite right it wouldn't take a lot to just lift the heads a bit and hot combustion gasses sqirting from one cylinder to the other would soon destroy the gasket.
For the joint to work correctly you need to apply a sufficient clamp load to locally yield the sealing rings on the gasket and have them with a compressive residual stress even when your cylinder pressures are maxed out.
I think you need to look at your torques on the head fasteners. You really want to find the yield point on a spare fastener, and do a torque to angle graph. By knowing the dynamic cylinder pressures you can apply a torque, then an angle that keeps the compressive residual stress in the gasket, whilst not reaching the point where the fasteners yield (which in turn loses the compressive residual stress and blows the gasket).
You may well have a batch of soft bolts, you might want to get them checked for hardness.
I think you need to look at your torques on the head fasteners. You really want to find the yield point on a spare fastener, and do a torque to angle graph. By knowing the dynamic cylinder pressures you can apply a torque, then an angle that keeps the compressive residual stress in the gasket, whilst not reaching the point where the fasteners yield (which in turn loses the compressive residual stress and blows the gasket).
You may well have a batch of soft bolts, you might want to get them checked for hardness.
stevieturbo said:
It is being caused by external elements. I wouldnt say detonation si ruled out, but certainly the last time there was no audible detonation. Ive had engines detonating like mad before when testing and abusing, but never blown a head gasket. Pistons usually went first.
I tend to agree with you - I don't believe that mis-timed or mal-functioning ignition could cause instantaneous head gasket failure. Sure, may result in pre-ignition and detonation, ultimately melting piston crowns and so, but the failure of the head gasket would result from the overheating caused by pre-ignition, so not immediately.
Same with fuelling - running lean in the turbo regime would cause overheating that would knacker the head gasket, but not immediately. Too rich would coke her up and make her run badly, but shouldn't affect the head gasket.
A couple of other considerations - I nearly posted yesterday thinking of a crack somewhere through into the cooling circuit, but see someone else has proposed that already. Thing is, that would give symptoms even before boost is applied, and yet you say the engine runs fine until then.
I tend to agree with comments above on verifying whether adequate clamping force is being applied, but failing that I'm scratching around for other failure mechanisms.
You say that the gaskets are really cooked when you dissemble the engine after failure. I don't know the Subaru engine at all other than knowing its a basic boxer layout, but could you be getting an airlock in the cooling system such that the heads are not actually seeing any water so cook the moment that the heat regime rises with boost application? This high temeperatures could also lead to head distortion causing sealing ring failure? Are you sure that the water pump is doing its job and taking the heat away from this area? Definitely worth checking.
Finally, getting a bit tenuous now, but how about the crankcase breather system? Could turbo boost pressure be bleeding back and pressurising the crankcase? A bit far-fetched maybe, but if the crankcase was getting the boost pressure and forcing oil past the scraper rings into the cylinders, the incompressible liquid would effectively be increasing your compression ratio? Scratching at straws maybe...
Best of luck - must be incredibly frustrating!
nel said:
Finally, getting a bit tenuous now, but how about the crankcase breather system? Could turbo boost pressure be bleeding back and pressurising the crankcase? A bit far-fetched maybe, but if the crankcase was getting the boost pressure and forcing oil past the scraper rings into the cylinders, the incompressible liquid would effectively be increasing your compression ratio? Scratching at straws maybe...
A pressurized crankcase would account for the "lack of power" that was mentioned. When the spark plugs fire the pistons would have to fight against the pressure in the crankcase to get to bottom dead center.
eliot said:
Are you using the special molly lube that comes with ARP bolts? - you need to pull it down MUCH tighter if you use the supplied molly lube.
eliot.
Anybody feel free to correct me if I'm wrong but using any lubricant (which should be used on any threaded fastener) won't require applying more torque to the bolts. It just means that you will hopefully arrive at the torque that you need with out any problems, I.E.pulling the threads out of the block or the bolt.
Trooper2 said:
Are you using the special molly lube that comes with ARP bolts? - you need to pull it down MUCH tighter if you use the supplied molly lube.
eliot.
Anybody feel free to correct me if I'm wrong but using any lubricant (which should be used on any threaded fastener) won't require applying more torque to the bolts. It just means that you will hopefully arrive at the torque that you need with out any problems, I.E.pulling the threads out of the block or the bolt.
The stud in effect becomes part of the base metal. Arp's have a fine rolled thread and quite high torque settings. The lube allows a higher clamp load to be achieved and cuts down on frictional losses.
Stevie, I'm sure that some of the cometics won't seal properly to the head or deck faces so you need very smooth faces or maybe different gaskets (composite)?
Boosted.
stevieturbo said:
It sounds to me like we are having a totally random badly timed ignition event ( or series of events )which is causing the blow. But what is causing the ignition event ?
That's illogical captain. Still find it hard to believe that an ignition problem can lead straight head gasket failure without the engine overheating first.
At a loss really.
I'm struggling to think what other options are left, and all I can think to do is look at it from first principles and see if it triggers any bright ideas.
Things that would cause the head gasket to go are: clamping loads are too low compared to the cylinder pressure, or uneven (for example either the block or the head distorted), or a poorl designed/constructed head gasket. What other possible reasons are there?
So you are looking for a list of crazy ideas why any of these could be occuring, and looking for ways to prove they aren't.
Some random thoughts to start:
Could you do a dummy assembly with pressure sensitive paper under the heads to prove you are getting the clamping loads you think you are?
Can you measure whether the stretch bolts have actually stretched to show whether they have yielded?
Is there any way to measure the actual in-cylinder pressure?
Things that would cause the head gasket to go are: clamping loads are too low compared to the cylinder pressure, or uneven (for example either the block or the head distorted), or a poorl designed/constructed head gasket. What other possible reasons are there?
So you are looking for a list of crazy ideas why any of these could be occuring, and looking for ways to prove they aren't.
Some random thoughts to start:
Could you do a dummy assembly with pressure sensitive paper under the heads to prove you are getting the clamping loads you think you are?
Can you measure whether the stretch bolts have actually stretched to show whether they have yielded?
Is there any way to measure the actual in-cylinder pressure?
Gassing Station | Engines & Drivetrain | Top of Page | What's New | My Stuff