turbo coolant and overflow
06-21-2004, 06:51 PM
#91
Again though, it seems like something like a blockage in the system would be a systematic problem, rather than 'it only happens until the coolant reaches the minimum mark then seems to abate' kind of issue. The head, valve, warpage, etc issue seems like it would fall into the same category - would it not be systematic and always a problem when running under boost or am I missing a piece of the puzzle here. I'm not trying to be a know it all, I'm seriously stabbing here, but perhaps my misconceptions about the cooling system in principal are holding me back...
06-22-2004, 12:06 AM
#92
Instructor
Thread Starter
Since I started this thread back in April there have been plenty of folks chiming in with similar symptoms, and in addition, I've received plenty of private emails describing the same symptoms. So there are lots of cars out there with this problem.
First some comments on some of the more recent posts:
eclou:
When you remove the expansion tank cap, the fluid level rises because you relieved the pressure above the coolant in the tank. That pressure above the coolant compresses any air (or exhaust gas) bubbles in the system, as well as expands the rubber hoses. When you relieve it, the bubbles expand and the rubber constricts, forcing the coolant level to rise.
Regarding the HC test - The first few times I did this test it was negative too, but that's because the tester isn't really designed to be used for what we're trying to detect. If you just use the tester as directed and bubble air drawn from the expansion tank through the dye, it will test negative. That's because the head gasket only leaks under high boost, so you have the drive the car real hard for a while in order to build up exhaust gasses in the coolant lines. The problem is that after you do this, when you go to remove the exp. tank cap, the drop in system pressure will immediately cause the coolant to overflow. By running the coolant level a little low, and driving for not too long, not too short, I was finally able to get the cap off and the tester in place quickly enough to bubble some air through the dye before the coolant level rose too high. The dye turned color almost immediately.
MachSchnell: in response to your 4 conclusions:
"(1) 1) A HUGE number of people have warped or uneven heads (does this really seem likely?)"
Yes, I think its very likely (more below)
"2) Upping the boost requires either people accept that lower levels are going to be the norm"
True, but this probably works only because it allows for more expansion volume before the relief valve opens. For a street car, you may be good for months. For a track car, maybe a few minutes.
"3) we find a higher pressure rad cap and see if that helps"
Another "treat the symptoms, not the cause" approach. My 23psi cap gave way not noticably different than a 17 psi cap. (I pressure tested them both, their ratings are accurate).
"4) o-ringing the head is a must"
Maybe.
My conclusion is that most of us are experiencing a block / head mating problem. The leakage has nothing to do with the type of gasket or head studs, nothing to do with modest increases in boost, and nothing to do with overly lean mixtures. Sure, any of these can cause gasket failure, but these are not the mechanisms at play here, at least not for me. I have 3 perfectly good gaskets hanging on my garage wall, 1 MLS and 2 WFRs, that show no signs of burn-through.
I've solved the problem the hard way. I swapped my 83K mile '86 motor for an 56K mile '89 motor. And I had the head re-surfaced and o-ringed.
Two days at the Glen - not a drop. I run the overflow tube into a catch bottle, so if there was literaly a drop of overflow, I would see it. Coolant level is mid-way between Min/Max marks when cold.
The last few times I removed the head, I checked it with a precision straight and could not detect any low spots. The machinist said it had a slight twist to it, that would be very difficult to catch with a straightedge. I also could not detect any low spots on the old block's deck, but I suspect that is where the real problem is. Those who have seen what the cylinders look like in these motors can understand why this is likely. The cylinders are only supported at their bases - and are otherwise completely free standing. No lateral support at all. This allows them to vibrate independent of the surrounding block, rubbing their aluminum top surface against the head gasket's steel fire rings. Something is going to wear.
This is my theory. Won't know for sure until the block is stripped and checked by the machine shop. This won't happen till the winter though. For now, I'm sick of being under the hood and just want to drive it!
Jon
First some comments on some of the more recent posts:
eclou:
When you remove the expansion tank cap, the fluid level rises because you relieved the pressure above the coolant in the tank. That pressure above the coolant compresses any air (or exhaust gas) bubbles in the system, as well as expands the rubber hoses. When you relieve it, the bubbles expand and the rubber constricts, forcing the coolant level to rise.
Regarding the HC test - The first few times I did this test it was negative too, but that's because the tester isn't really designed to be used for what we're trying to detect. If you just use the tester as directed and bubble air drawn from the expansion tank through the dye, it will test negative. That's because the head gasket only leaks under high boost, so you have the drive the car real hard for a while in order to build up exhaust gasses in the coolant lines. The problem is that after you do this, when you go to remove the exp. tank cap, the drop in system pressure will immediately cause the coolant to overflow. By running the coolant level a little low, and driving for not too long, not too short, I was finally able to get the cap off and the tester in place quickly enough to bubble some air through the dye before the coolant level rose too high. The dye turned color almost immediately.
MachSchnell: in response to your 4 conclusions:
"(1) 1) A HUGE number of people have warped or uneven heads (does this really seem likely?)"
Yes, I think its very likely (more below)
"2) Upping the boost requires either people accept that lower levels are going to be the norm"
True, but this probably works only because it allows for more expansion volume before the relief valve opens. For a street car, you may be good for months. For a track car, maybe a few minutes.
"3) we find a higher pressure rad cap and see if that helps"
Another "treat the symptoms, not the cause" approach. My 23psi cap gave way not noticably different than a 17 psi cap. (I pressure tested them both, their ratings are accurate).
"4) o-ringing the head is a must"
Maybe.
My conclusion is that most of us are experiencing a block / head mating problem. The leakage has nothing to do with the type of gasket or head studs, nothing to do with modest increases in boost, and nothing to do with overly lean mixtures. Sure, any of these can cause gasket failure, but these are not the mechanisms at play here, at least not for me. I have 3 perfectly good gaskets hanging on my garage wall, 1 MLS and 2 WFRs, that show no signs of burn-through.
I've solved the problem the hard way. I swapped my 83K mile '86 motor for an 56K mile '89 motor. And I had the head re-surfaced and o-ringed.
Two days at the Glen - not a drop. I run the overflow tube into a catch bottle, so if there was literaly a drop of overflow, I would see it. Coolant level is mid-way between Min/Max marks when cold.
The last few times I removed the head, I checked it with a precision straight and could not detect any low spots. The machinist said it had a slight twist to it, that would be very difficult to catch with a straightedge. I also could not detect any low spots on the old block's deck, but I suspect that is where the real problem is. Those who have seen what the cylinders look like in these motors can understand why this is likely. The cylinders are only supported at their bases - and are otherwise completely free standing. No lateral support at all. This allows them to vibrate independent of the surrounding block, rubbing their aluminum top surface against the head gasket's steel fire rings. Something is going to wear.
This is my theory. Won't know for sure until the block is stripped and checked by the machine shop. This won't happen till the winter though. For now, I'm sick of being under the hood and just want to drive it!
Jon
06-23-2004, 08:01 PM
#93
Why would our cars be so much more prone to warped heads than others, though? I understand what you're saying about the freestanding piston sleeves potentially being a problem, but there are plenty of other cars with bottom attachment only sleeves who do not experience warped heads to this degree. I personally know that my head was resurfaced by a very reputable head shop...I'm by no means suggesting that they did a perfect job and that it's not possible that it's still warped, but for so many to have this problem? It just seems far fetched to me, but then again, it's a Porsche - whatever it can do to make you spend money, it will. Again it seems like we're back to o-ringing as the only realistic solution, since I think that the likelihood of finding a bunch of 50k mile 89 motors or heads is pretty slim for all of us experiencing the problem. How many times can you deck and level a head before the C/R goes up to the point that you have more problems than just the slight overflow of coolant. Heck, I know I've put in 2 4 hour trips where I was on the highway doing a lot of passing and slowing down (read - boosting frequently, and boosting high) in the same day and the coolant level never dropped...I'm just a bit dubious (I guess it's the optimist in me hoping that these aren't cars that will have to be forever known for warped heads). I also was always under the impression that warped heads were more frequent under iron block/aluminum head cars where the temp difference between materials caused something similar, but don't we have alu block and heads?
06-23-2004, 10:32 PM
#94
Pro
Join Date: Sep 2002
Location: Montreal and Texas
Posts: 554
Likes: 0
Received 0 Likes
on
0 Posts
Well I recently found out why I was overflowing with Tony G's car. The o-ring in the 4th cylinder was completely collapsed, right above the coolant passage. So the o-ring worked perfectly, everywhere, except where it was squished. Now that I have replaced all of my o-rings, in the head and resurfaced the block, I have no leaks and no overflow.
Flatness of the block, is obviously as important as the head.
Now lets just see, how long and how much boost it will withstand!!!!
Flatness of the block, is obviously as important as the head.
Now lets just see, how long and how much boost it will withstand!!!!
06-24-2004, 01:13 AM
#95
Instructor
Thread Starter
Re-surfacing the head is the easy part.
Decking the block is a big deal. I think this is where the problems are.
There is a specification in the factory manual that shows the minimum dimension between the surface of the cylinder head and a reference plane. I've heard that you're good for about 3 shavings before you hit this minimum, depending of course on the condition of the surface.
I'm not convinced that o-ringing is a necessity. I did it to one my heads just because I really wanted to do everything I could while everything was apart. But I think for reasonable boost levels (17psi or less), it shouldn't be a necessity.
Jon.
Decking the block is a big deal. I think this is where the problems are.
There is a specification in the factory manual that shows the minimum dimension between the surface of the cylinder head and a reference plane. I've heard that you're good for about 3 shavings before you hit this minimum, depending of course on the condition of the surface.
I'm not convinced that o-ringing is a necessity. I did it to one my heads just because I really wanted to do everything I could while everything was apart. But I think for reasonable boost levels (17psi or less), it shouldn't be a necessity.
Jon.
06-24-2004, 07:06 PM
#96
Hmm, the decking point is a good one I hadn't put enough consideration into, and I suspect that where it was done, they probably did not deck, only planed the head. Perhaps that's where I need to look next.
06-24-2004, 11:47 PM
#97
Pro
Join Date: Sep 2002
Location: Montreal and Texas
Posts: 554
Likes: 0
Received 0 Likes
on
0 Posts
One note on shaving the ahead. The factory limits the shaving to 0.6mm, at which point the head is beyond spec.
A new stock head is at 24mm (thickness between the head surface and the front machined lip). The shaved limit is 23.6mm.
Now, we have to remember why the factory didn't go any further. Warranty for one, and ****y gas that people might use which can ultamitly cause detonation, due to the slight increase in compression.
I'm running mine way beyond that, at 23.3mm, without a problem. It came that way from Milledge, and I run good 94+ octane, to compensate for the slight increase in compression.
A new stock head is at 24mm (thickness between the head surface and the front machined lip). The shaved limit is 23.6mm.
Now, we have to remember why the factory didn't go any further. Warranty for one, and ****y gas that people might use which can ultamitly cause detonation, due to the slight increase in compression.
I'm running mine way beyond that, at 23.3mm, without a problem. It came that way from Milledge, and I run good 94+ octane, to compensate for the slight increase in compression.
07-06-2004, 10:40 PM
#98
Just an update - I decided to stop filling the overflow tank when it burped/puked coolant to see if it leveled off at a lower level. It did great for about a month, but today, after absolutely no on-boost driving (5 minutes from work to home never breaking 8=9psi) I parked it and it puked everything. I'd been waiting to replace the expansion tank because there was still a little coolant and I don't like to just drain if at all possible, so I replaced the tank and the cap (plus the old one was too dirty to really see the coolant level) and ran the heck out of it, and everything seems groovy thus far - lots of 18psi driving and no change in coolant level (though it was a little wierd - sat at the 'min' mark when warmed up until I shut the car off, then dropped a significant amount - to the very edge of the 'top' of the tank - then came back to min after maybe 15 seconds)...I'll keep an eye for the next week or two, I'm certainly not that optimistic after reading this whole thread, but maybe I got lucky and it was just an expansion tank and/or cap issue since it puked after absolutely no spirited or on-boost driving.
07-06-2004, 10:58 PM
#99
Instructor
Thread Starter
Here's another thread that lends credance to the block flatness (or lack there of) theory:
https://rennlist.com/forums/showthre...eferrerid=2948
Its called "Where to get sleeving kits" in this forum.
There's plenty of discussion about the open deck design leading to cylinder wear due to vibration of the bores.
Jon.
https://rennlist.com/forums/showthre...eferrerid=2948
Its called "Where to get sleeving kits" in this forum.
There's plenty of discussion about the open deck design leading to cylinder wear due to vibration of the bores.
Jon.
01-19-2011, 03:57 PM
#100
Resurrected from the dead 
Found this thread that is most like my own car problem
I have the same problems as MachSchnell and Jon Schepps
i´ve changed the headgasket twice and resurfaced the head but still it pushes out water.
The car is running cool all the time, the tempindicator is below the first mark (lower tempswitch and lower thermostat).
i´ve also changed the expansiontank and cap.
Now i´m going to test for CO2 in the cooling water and then pressure test the cooling system.
I know that the turbo waterpump is leaking a little and also the heater valve seems to leak a little.
Could these 2 problems be the cause of my overflow or do i still have problems with the surface of either the block or head or could it be a small crack in the turbo?
I hope the CO2 test will tell me the answer.
Found this thread that is most like my own car problem
I have the same problems as MachSchnell and Jon Schepps
i´ve changed the headgasket twice and resurfaced the head but still it pushes out water.
The car is running cool all the time, the tempindicator is below the first mark (lower tempswitch and lower thermostat).
i´ve also changed the expansiontank and cap.
Now i´m going to test for CO2 in the cooling water and then pressure test the cooling system.
I know that the turbo waterpump is leaking a little and also the heater valve seems to leak a little.
Could these 2 problems be the cause of my overflow or do i still have problems with the surface of either the block or head or could it be a small crack in the turbo?
I hope the CO2 test will tell me the answer.
