Reasons for head gasket failure.
Discussion
Melted plug tip and both banks affected equally sounds like very high peak cylinder pressures and temps and if it only starts above 4000 rpm I would suspect the mapping of the Motec or a faulty MAP sensor.
Have a look at the ignition timing maps just to make sure its not massivly over advancing the ignition.
Did the Motec come already mapped ?
Can you download a known good map from a similar engine thats been dynoed ( must be plenty about running with Motecs ).
Or make up a patch lead and test run it with the DTA.
Have a look at the ignition timing maps just to make sure its not massivly over advancing the ignition.
Did the Motec come already mapped ?
Can you download a known good map from a similar engine thats been dynoed ( must be plenty about running with Motecs ).
Or make up a patch lead and test run it with the DTA.
Hello Steve
Many of these have probably been mentioned, and I don't want to sound like I'm teaching people to suck eggs (so sorry if I do). Sounds like fairly intense early detonation to me, the in-cylinder pressures and temperatures would be astronomical. Has the crank position sensor or wheel definitely correctly set relative to TDC? If its overboosting then effective CR maybe too high, is the boost gauge been calibrated recently. The inherent over fuelling used in tuning turbos may be damping usual detonation effects (now I really am clutching at straws). Maybe it could be a blockage in the cylinder head waterjackets (I know unlikely because there are 2 heads), maybe some coolant passages on the head gasket are missing.
Regarding torque up on the ARP studs, you can roughly measure the stretch by running the bolt down with the slightest amount of torque then measuring the angular displacement upto the torque-up point. That angle can be translated to a known displacement.
However this maybe a mute point because if it is just a torque value then the fastener is not designed to go into yield. A torque & angle spec is used to try and get the bolt close to the yield point using the torque value then into yield and beyond the UTS using a displacement value (from the angle). The idea behind this is so the assembled clamp load is consistent (by getting round friction effects of torque versus clamp load) and when stretched by cyl-head expansion (a big issue with Aluminium heads) it returns to the original clamp value (see load/disp graph for these bolts to see how it rides up and down the curve). I digress but I hope it explains the limitations of elastic torque only fasteners.
All the best
Phil
>> Edited by wheeljack888 on Friday 20th May 16:20
Many of these have probably been mentioned, and I don't want to sound like I'm teaching people to suck eggs (so sorry if I do). Sounds like fairly intense early detonation to me, the in-cylinder pressures and temperatures would be astronomical. Has the crank position sensor or wheel definitely correctly set relative to TDC? If its overboosting then effective CR maybe too high, is the boost gauge been calibrated recently. The inherent over fuelling used in tuning turbos may be damping usual detonation effects (now I really am clutching at straws). Maybe it could be a blockage in the cylinder head waterjackets (I know unlikely because there are 2 heads), maybe some coolant passages on the head gasket are missing.
Regarding torque up on the ARP studs, you can roughly measure the stretch by running the bolt down with the slightest amount of torque then measuring the angular displacement upto the torque-up point. That angle can be translated to a known displacement.
However this maybe a mute point because if it is just a torque value then the fastener is not designed to go into yield. A torque & angle spec is used to try and get the bolt close to the yield point using the torque value then into yield and beyond the UTS using a displacement value (from the angle). The idea behind this is so the assembled clamp load is consistent (by getting round friction effects of torque versus clamp load) and when stretched by cyl-head expansion (a big issue with Aluminium heads) it returns to the original clamp value (see load/disp graph for these bolts to see how it rides up and down the curve). I digress but I hope it explains the limitations of elastic torque only fasteners.
All the best
Phil
>> Edited by wheeljack888 on Friday 20th May 16:20
Yes Mike, first 2 builds had flat heads only. They were wire ringed after that, partly bacause they were blowing in a n almost blind search for a fix.
There isnt actually a fire ring as such on MLS gaskets. They are just thin metal shims along lines of old RV8 gaskets, but thicker. ( or indeed LS6 MLS gaskets )
Cam and crank triggers are as standard, albeit with JUN vernier camwheels for cam timing.
No overboosting. Its only really ever seen 1 bar.
Thats being read via 3 different sources.
Mechanical guage
Ecu Map sensor
Boost controllers map sensor.
Fuelling is a bit messy, but not over rich, and certainly not weak. No blockages, as again both block and heads are brand new.
I do agree that is must be massive detonation causing it....but whats causing the massive detonation ?
Engine will probably be removed, inspected and re-built over the next week or 2. It will be re-fitted with a normal std igniton system and ign timing verified throughout rpm range ( impossible at present due to MSD multispark, which cannot be turned off )
May consider new ARP studs too.
The first time it blew, with the first engine, it actually took some of the aluminium from the block face with it. Almost like a plasma cutter.
Pics from very first blow. Block destroyed, subsequently replaced. Heads luckily undamaged.
Pics from 2nd blow up, gasket badly buckled
There isnt actually a fire ring as such on MLS gaskets. They are just thin metal shims along lines of old RV8 gaskets, but thicker. ( or indeed LS6 MLS gaskets )
Cam and crank triggers are as standard, albeit with JUN vernier camwheels for cam timing.
No overboosting. Its only really ever seen 1 bar.
Thats being read via 3 different sources.
Mechanical guage
Ecu Map sensor
Boost controllers map sensor.
Fuelling is a bit messy, but not over rich, and certainly not weak. No blockages, as again both block and heads are brand new.
I do agree that is must be massive detonation causing it....but whats causing the massive detonation ?
Engine will probably be removed, inspected and re-built over the next week or 2. It will be re-fitted with a normal std igniton system and ign timing verified throughout rpm range ( impossible at present due to MSD multispark, which cannot be turned off )
May consider new ARP studs too.
The first time it blew, with the first engine, it actually took some of the aluminium from the block face with it. Almost like a plasma cutter.
Pics from very first blow. Block destroyed, subsequently replaced. Heads luckily undamaged.
Pics from 2nd blow up, gasket badly buckled
Both events were in the same place by the water jacket, is the mls getting to hot in this area and could it be the cause of the detonation? Is the stock gasket composite and do others use the mls gasket without issues?
There seems to be a lot of heat by what is presumably the exhaust side of the chamber, excessive backpressure from the turbo, to little overlap? How do the exhaust valves look?
There seems to be a lot of heat by what is presumably the exhaust side of the chamber, excessive backpressure from the turbo, to little overlap? How do the exhaust valves look?
Thats from the first 2 blows.
First time seemed to be limited to that side, although with severe block damage ( LHS ). 2nd blow was both sides, with LHS being much worse, although not as severe as blow 1.
1s and 2nd blow burst radiator ( and 1st heater core, although it was corroded anyway )
3rd and 4th times were blown on all cylinders. Not so much water loss thru engine when it happened. 3rd time saw alloy radiator totally burst with obvious water loss.
Pic from first blow.
You cant really get an idea from colour, as the amount of water that pased thru was too much, and as I said, engine wasnt set up or running properly. It would have been rich, into the 10's or so, and also running rich during running in.
It is the axhaust side, but exhaust valves are clean and sound.
First time seemed to be limited to that side, although with severe block damage ( LHS ). 2nd blow was both sides, with LHS being much worse, although not as severe as blow 1.
1s and 2nd blow burst radiator ( and 1st heater core, although it was corroded anyway )
3rd and 4th times were blown on all cylinders. Not so much water loss thru engine when it happened. 3rd time saw alloy radiator totally burst with obvious water loss.
Pic from first blow.
You cant really get an idea from colour, as the amount of water that pased thru was too much, and as I said, engine wasnt set up or running properly. It would have been rich, into the 10's or so, and also running rich during running in.
It is the axhaust side, but exhaust valves are clean and sound.
railtracksport said:
looking at the photos the head gasket looks in correct as some of the feeds in the water jacket are blocked
Look more closely. The slots in the casting allow sand to fall out of the cores after the casting process, in other words most of the slots aren't there to allow coolant flow. Looking at the pics I can see some casting inside a slot which looks intentional but it's defo casting and not part of the gasket. One of the pics shows a big coolant passage which looks like this is where coolant enters the head.
Any progress with this Stevie?
Boosted.
Stevie, other ideas/angles possibly.
1. Difference in compressed thickness of stock gaskets vs. MLS gaskets.
2. Compression was 8.8... changed to 8.25. How?
3. Could engine builder had the blocks deck height changed, or had the heads skimmed? How did he obtain the 8.8 cr? How did you lower it to 8.25?
What all of the above could do is effectivly change the length of the studs. Shorter block... shallower heads... thinner gaskets. Add any/all up and you could be tightening the nuts down onto the studs shoulder. Actually running out of threads, so to speak. Can't tell or see from pics if this might have happened. Heads could be "just" tight enough to hold under little/no load conditions, but when you add in the boost, head walks/raises and gasket says goodbye.
Alan
1. Difference in compressed thickness of stock gaskets vs. MLS gaskets.
2. Compression was 8.8... changed to 8.25. How?
3. Could engine builder had the blocks deck height changed, or had the heads skimmed? How did he obtain the 8.8 cr? How did you lower it to 8.25?
What all of the above could do is effectivly change the length of the studs. Shorter block... shallower heads... thinner gaskets. Add any/all up and you could be tightening the nuts down onto the studs shoulder. Actually running out of threads, so to speak. Can't tell or see from pics if this might have happened. Heads could be "just" tight enough to hold under little/no load conditions, but when you add in the boost, head walks/raises and gasket says goodbye.
Alan
The first pic of liner damage does look like detonation, although I'm surprised it didn't take out the plug too.
As we're now clutching at straws, could it be a case of the large turbo generating 0,8 - 1,0 bar at low rpm at a very inefficient part of the compressor map, and putting too much heat into the air ? Too much advance, and the engine starts to det, controller winds back the ign causing the chamber to heat up. All happens too fast to reduce boost before the gasket blows ... ?
As we're now clutching at straws, could it be a case of the large turbo generating 0,8 - 1,0 bar at low rpm at a very inefficient part of the compressor map, and putting too much heat into the air ? Too much advance, and the engine starts to det, controller winds back the ign causing the chamber to heat up. All happens too fast to reduce boost before the gasket blows ... ?
stevieturbo said:
When coming on boost, the engine/turbo makes the strangest noise. We had put this down to surge, so, changed ex housing for a larger one, to slow spool up. We shouldnt have had to do that, as the engine/turbo combo has been built and run successfully in the US before by other guys.
Noise was still evident although built boost much later. Do not believe it was surge. Very loud fluttering noise on WOT, as revs climb even at 1 bar. Doesnt last long though, as engine always blows up.
I have no idea what that means, but it seems very significant to me. Can you tell whether the noise comes from the intake, exhaust or the engine itself? I'm trying to think of things that would stop the exhaust from breathing properly, insufficient valve lift, misaligned/obstructed joint somewhere, maybe just an awkward corner somewhere disrupting the gas flow and forming a choke?
From what you experts have said about the damage I get the impression that the problem is detonation or excessive pressure within the cylinders rather than insufficient clamping pressure, is that how you see it?
Have all the engine failures been in the same car? Maybe it's something about the surrounding systems that we haven't thought of.
The fluttering noise you describe from the intake makes me look for some part of the intake duct collapsing under the slight suction when you open the throttle. Or maybe it's blow-by increasing as you raise the boost, pressurising the crank case and pushing unexpected amounts of oil or combustion gas out the breather into the intake? I don't see yet why either of these would cause the failures, but I'm still inclined to try to track down the noise.
The fluttering noise you describe from the intake makes me look for some part of the intake duct collapsing under the slight suction when you open the throttle. Or maybe it's blow-by increasing as you raise the boost, pressurising the crank case and pushing unexpected amounts of oil or combustion gas out the breather into the intake? I don't see yet why either of these would cause the failures, but I'm still inclined to try to track down the noise.
i have just had a thought did you have the block machined as well as the heads skimmed if by doing this you move the head across slightly you have compensate for this with a thicker head gasket i know this because on v6 v8 and v12s we would only try to machine the head i will try to find out some more when i get to work on monday i will get some spec sheets out to look at and be intouch
Blocks have never been machined. First damaged one was replaced with a new short motor.
Heads were only machined very very lightly during the fitting of o-rings. Checked after every blow up to be totally flat.
Either way we are not running either enough boost, or timing ( as displayed on ecu, which may or may not be correct ) that it should be an issue.
Heads were only machined very very lightly during the fitting of o-rings. Checked after every blow up to be totally flat.
Either way we are not running either enough boost, or timing ( as displayed on ecu, which may or may not be correct ) that it should be an issue.
The way you are describing that noise, I'm wondering if you are experiencing some sort of resonance in the intake side, between (maybe involving) the compressor and the individual cylinders.
I don't know if I'd go as far as to say that it's the root, but it could potentially cause the charge pressure to be higher than any digital indication system that you may have.
Have you tried and old fashioned (bourdon tube)pressure guage?
Many of the electronic systems are heavily averaged, to remove electrical noise say from the ignition system. I reckon you might get a different picture with a mechanical guage.
Edited to add>>
Thinking further, it could be that the casting for the plenum chamber on the bad car is a different size to the one on the good car. I'm thinking that the casting might have more metal in it (greater wall thickness). Have you tried weighing them for comparison.
Any difference in volume would definately affect resonance.
Further edited to add>>
For a whole variety of reasons, I'd say that it wouldn't take much to become unstable under some circumstances.
(It's all about polynomials, transfer functions, poles and zeros. Too difficult for me to understand, let alone explain)
>> Edited by dilbert on Sunday 29th May 01:31
I don't know if I'd go as far as to say that it's the root, but it could potentially cause the charge pressure to be higher than any digital indication system that you may have.
Have you tried and old fashioned (bourdon tube)pressure guage?
Many of the electronic systems are heavily averaged, to remove electrical noise say from the ignition system. I reckon you might get a different picture with a mechanical guage.
Edited to add>>
Thinking further, it could be that the casting for the plenum chamber on the bad car is a different size to the one on the good car. I'm thinking that the casting might have more metal in it (greater wall thickness). Have you tried weighing them for comparison.
Any difference in volume would definately affect resonance.
Further edited to add>>
For a whole variety of reasons, I'd say that it wouldn't take much to become unstable under some circumstances.
(It's all about polynomials, transfer functions, poles and zeros. Too difficult for me to understand, let alone explain)
>> Edited by dilbert on Sunday 29th May 01:31
Gassing Station | Engines & Drivetrain | Top of Page | What's New | My Stuff