JET951

We have a customer with a 1994 928 GTS that is a bit of a collector car and as such has sat for quite some time, Last coolant change would be approx 15 years ago but had done almost no kms in this time. 75,000kms on the odometer. (46,000 miles)

Now we had a call about 3 weeks ago that the car would not start and that the owner could not turn the engine over by hand clockwise but would go anticlockwise a little bit. We told him to halt what he was doing to get the car towed to us. Fast forward to Monday and we removed the spark plugs to find cylinder 6 with coolant in it. We explained to the customer that unfortunately his engine has a breached head gasket and the worse case scenario is that hydrolock has occurred and don't be surprised to find a cracked cylinder. I would mention that we have seen quite a few S4 engines and a few GTS engines have the same cracked cylinder outcome.

This is what occurs, the customer takes the car for a Sunday drive and returns home and parks the car up in the garage. What he doesn't know is that the head gasket has been breached just a tiny bit and just enough so that when the car is parked and hot small amounts of coolant will now start to drip into the cylinder as the cooling system is at 14.7psi of pressure for the coming hour or two. Fast forward a week or two and the customer is going to take it out for another run, he goes to crank it over and if he is unlucky the car will start on 7 cylinders, the odd cylinder won't fire as there is coolant on the spark plug and in the cylinder. By this time it's too late and the damage has now occurred and in almost every instance there is a nice big crack or D chunk in the cylinder. Now if you are really lucky like our customer with this GTS it would seem that when the engine was turned off cylinder 6 looks to be just before TDC and when he went to start the car the starter motor just clicked as the engine could not be turned over due to the hydrolock and no other cylinder got a chance to fire. We are bloody amazed by the chances of this.

On this GTS you can see the corrosion which has taken place right under the fire ring on cylinder 6. This is also a bit of a warning to those cars that still have their original head gaskets to consider changing them out as the consequence are rarely as lucky as this customers 928 and this engine block will live on

I would make note that we don't see this with the 944's even though they do get breached head gaskets just like the 928 and it would seem that the starter motor and firing of only 3 cylinders is just not enough to brake the cylinders. Im not going to say it hasn't happened to someone but for our customer cars they don't seem to.

Regards
Sean

Darklands

The coolant was a lifetime filling? The channels looks very muddy!
Lucky Guy, GTS blocks are a rare beast today.

FredR

Sean,

Thanks for posting- very informative!

This is completely consistent with the corrosion process I have previously defined and no one else has an answer for.

Happy to offer you my analysis of what I see based on your photos published above however much appreciate if you can post pics for both sides of the gasket retrieved from bank 1/4 and also for cylinder No8.

Rgds

Fred

SwayBar

Thanks for sharing!

deutschmick

Wow! Lucky indeed.

Schocki

Yes, I agree this owner is a lucky guy.

However, not changing the coolant for 15 years and not really driving it is a setup for disaster.
Typical garage queen problem and preventable. I feel sorry for the car, not for the owner.

GregBBRD

Indeed, he is super lucky.
We've seen many of these engines with water in the cylinders from head gasket deterioration.
Over half of the time, this results in a split cylinder, which requires some major effort to repair.

Amazingly enough, this engine has very little of the pitting in the aluminum head and cylinders, caused from whatever theory one has.

karl ruiter

The justification for changing all head gaskets now.

JET951

Yes we do agree some more regular servicing would have been a great thing but what is done is done and we are just glad this block has no damage. We actually knew the previous owner of the car before the current owner who does want to bring the car back up to a drivable standard. If I can offer one statement is that the current owner just didn't know what was needed to keep them up to drivable standards and may have been under the impression that if it wasn't being driven then it may not have needed the regular maintenance. Its a tough lesson to learn.

Hi Greg,
Im quite surprised by the lack of corrosion on the heads, I thought there was going to be more.
Have you been able to have small cracks welded up and then dry sleeves installed? Or would you go the MID way of repair with the freestanding wet sleeves? So far we have been lucky to get 2nd hand engines and make one good one from 2 but im sure in the future with engines getting rare that repairs may be needed to keep the engines going

Absolutely Fred,
Here are the pics from one gasket, the other gasket had just deteriorated into pieces as seen in the pics above
What's interesting is that you can see the Block side of the gasket had deteriorated quite a lot where the coolant can flow along it as you can see the brown outline of the block imposed on the gasket as seen in pic 1,2 and 3.
Pictures 4,5 and 6 are from the head side of the gasket.

Again this really is an indication that the gaskets should be changed between the 15-20 years time frame somewhat irrespective of KM's travelled.

GregBBRD

Notes, in your text, above.

Zirconocene

Thank you very much for sharing (all of you, I really appreciate the combined wisdom on display here). All the same, this is giving me heartburn, to imagine what my HGs might be like. I know that pulling the engine is in my future, but this is really driving some paranoia in me.

Cheers (no sarcasm intended with the sign off, I'm really glad to be able to learn about these things)

GregBBRD

Doubtfully anyone at Porsche would have dreamed that a paper head gasket (glorified paper, but still paper) would have lasted for 35 years.
....Just like they never thought that the rubber O-rings and piston seals in an automatic transmission would have lasted that long.

FredR

Some thoughts based on your photos/dialogue:



1. The water bath and internal passage ways of the cylinder head show no signs of corrosion whatsoever. This tells one that the coolant was in perfectly serviceable condition despite the many years it was resident in the motor.

2. Coolant is not corrosive to the alloy elements of the system- this is a myth. Coolant based corrosion was more a problem for engines built with ferritic materials [i.e. cast iron] and that was due to oxygen in the water. Corrosion problems only happen in alloy engines when the glycol thermally degrades to form organic acids of about pH4.

3. Aluminium alloys are protected by a passive film of aluminium oxide and as long as this film remains intact the alloy will not corrode. However aluminium based alloys become extremely vulnerable to corrosion when exposed to pH values outside the range of pH6 to pH 9.

4. If the coolant were to acidise all the alloy components in the flow circuit would be attacked and the alloy radiator core would fail first due to the fact that it has the thinnest sections. To date I have never seen or heard of this happening to a 928- I have asked for photos of such but no takers so far.

5. The chemical package in the coolant has components such as anti-foaming agent and seal lubrication but the most important components are the corrosion prevention chemicals [the Hybrid OAT component in 2nd generation coolants] and [even more important] the buffering package. In an alloy engine the HOAT component most probably is not consumed and thus the buffering agent is the single most critical component- this prevents the glycol from thermal degradation by keeping the pH alkali hence the soapy feel of the coolant.

6. If the coolant buffering agent were to get exhausted the ethylene glycol will thermally decompose at temperatures in excess of about 55C to form organic acids. The coolant is buffered to about pH8.5- when it drops to pH7 it is exhausted- I use fish tank water test strips to monitor the pH. This is why we see examples of coolant sitting in the motor for long periods of time and nothing happens to the wetted alloy surfaces of the engine and the radiators.

7. The problems with the head gasket and the cylinder head are a completely different matter. The root cause of the corrosion problem we see on the 928, 944 and 968 engines is the ingress of coolant between the gasket and the head in the areas that are not counter supported on the block side- the buffering agent depletes and the glycol decomposes to form organic acids. In the examples I have seen serious problems can gestate in as little as 3 years and yet some examples can and do last over 30 years with little to no corrosion issues and the determining factor as I am concerned is whether the coolant has worked its way into the crevice bounded by the gasket and the head. Why this happens on some examples and not on others remains a mystery.

8. In the case you have posted the somewhat unusual feature is the extensive damage that is seen to the gasket and yet the head corrosion, although of serious consequence in one small location on No6 cylinder is surprisingly minimal given the two consequences tend to go hand in hand. When serious corrosion damage occurs it is characterised by a pitting attack and when such happens in a gasketed joint it is caused by crevice corrosion- the damage at No6 cylinder seemingly has pits in the damaged field.

9. The corrosion process clearly undermined the fire ring hence the leak. Once the electrolyte contacts the stainless steel fire ring the corrosion process probably accelerated due to galvanic corrosion adding to the grief.

10. There is an irony in the gasket design. Each cylinder is sealed by two concentric rings. The inner fire ring is the combustion chamber pressure seal- the outer sealing ring formed by the fibrous gasket material is intended to ensure that coolant cannot get to the fire ring and thus create a galvanic corrosion attack. This tells me that the gasket designers expected coolant to get there and clearly did not realise or appreciate the possibility of what seemingly can and does transpire when this happens thus a design flaw or inherent weakness.



11. The gasket damage is predictable and typical of what is commonly seen. As one can see in your latest photos the ink design on the block side of the gasket is invariably bleached by hot coolant but that is all that happens from exposure to coolant on the block side. The important thing to note is that the gasket is not aging- it is being attacked destructively from the head side. Once the block side is breached the problem stops due to exposure with stable coolant.

12. The gasket is chemically attacked from the head side and initially the acidic gunge breaks down the Buna N binder. The gasket is formed from two sheets of Klingerite pressed onto a galvanised steel mesh that is barbed. As you can see sections of the gasket have been completely wiped out. When the gunge hits the mesh the galvanising does not last too long and then it takes out the steel - this forms ferric oxide- the orange/brown coloured residue. The volume of ferric oxide formed is greater than the volume of steel consumed and the gasket starts to delaminate. You might want to peel back a section of the 1-4 bank gasket in a moderately damaged section to take a look at what was going on in there.

13. Given the extensive damage to the gasket this probably explains the orange staining seen in the water bath area- presumably ferric oxide washed out from the gasket. I would suggest you give the radiator a good flush before re-use.

14. GB has already covered the possible consequences of coolant maldistribution. As the coolant flows through the engine it enters the water bath from the front and leaves the block at the rear passing into the heads and then back to the front of the engine and into the radiator. As the coolant passes along the water bath it heats up picking up about 8 degrees C on top of the 90C it left the radiator at thus the rear most cylinders see hotter coolant. As the coolant passes close to the cylinder head there will be more drag and a bit less cooling at the hottest spot. Presumably that is why Porsche put some flow metering holes in the gasket that get progressively bigger towards the rear of the motor- this because the differential pressure will vary slightly. Thus when corrosion takes out chunks of the gasket and permits more flow through some of the originally blanked off transfer ports it is not difficult to see why the cooling process can get significantly disrupted.

15. With respect to the hydro-lock situation that really is a crab shoot and by the sound of it the owner is indeed a very fortunate chap . When the motor heats up the coolant expands by approximately 0.5 litres and that expansion pressurises the 1 litre of air space in the expansion tank and the pressure in round terms can double from 15 psia [atmospheric pressure] to 30 psia max before the cap blows. The volume of each cylinder displacement is 625cm3 so conceivable that the cylinder leaking can get filled with coolant once the motor stops running as the static pressure of the hot system drives coolant through the leak path.

16. I suspect if the cylinder is completely filled the starter motor simply stalls. The real serious damage is probably caused when the cylinder is partially filled and the piston gets up to full cranking speed on a compression stroke and then locks up when the non compressible coolant hits the head – a scary scenario unfortunately. As GB says check that cylinder carefully for any signs of mechanical deformation. You might consider a dye-pen examination to check for any cracks that might be present mid stroke in the bore that are not visible .



Appreciate you will likely be aware of most if not all of the above but some of our friends looking in are probably just not aware of such ramifications.

JET951

Hi Fred,
I agree with almost all that you have said but in this case there is absolutely deterioration to the gasket on the block side that is not evident on the head side of the gasket.

In quite a few spots the block side of the head gasket has been deteriorated quite severely and yet the head side where you say this is where the process occurs is not evident at all. This would indicate that the coolant has been able to penetrate the gasket from the block side of the gasket, possible because of spent very old coolant, seeing as this engine only had 75,000kms on it would seem that it wasn't heat cycles that have finished off these gaskets.

4 pictures below show the same area on both sides of the gasket

first picture is block side, second picture is head side same location on head gasket
Regards
Sean
third picture is block side, forth picture is head side same location on head gasket

FredR

Sean,

Good to see you putting some thought into what you are seeing in front of you and your logic is apparent however I would politely suggest the logic is flawed and I will expand on that.

The major logic flaw in your assessment is that the main body of the coolant can eat away in discrete locations on the block side but if such really were happening the entire wetted surface of the exposed surfaces would be chewed away and that is just not the case. Just take a look at the first photo you posted of the block side - one can see very clearly that the gasket has been delaminated because of what was going on inside the gasket. As I stated in point 13 of my analysis once the gasket is breached the gunge can migrate along the tiny gaps formed by the sheet material where there is no compression, the reinforcing mesh corrodes, the resulting volume of corroded product increases, the gasket delaminates and the gunge can and does migrate along the inside of the gasket. When the gasket is compressed either side by the head and the block this does not seem to happen and thus why the periphery of the gasket is generally not breached and comes out intact with the original print colouration intact as can be seen in your photos. Now perhaps I can suggest a simple test- take a section of the peripheral gasket residue that is still intact and try flexing it gently- does it remain intact or does it crack?

The photos you published show without a shadow of doubt that coolant is penetrating between the gasket and the head and this is indicated by the bleaching of the ink in areas where there is no obvious surface damage head side and one can also see signs of the orangy discolouration from the mesh underneath being corroded and thus breaking through to some extent. When one sees serious damage to the head like the path that cuases the leak one invariably sees the gasket damaged adjacent to that location. The problem in these locations is that once it starts there is nothing to stop it short of dropping the coolant and thus disrupting the electrolytic curent flow. On the other hand is the requisite potential is not reached there will be no localised severe damage from crevice corrosion.

Indeed if one looks carefully at the undamaged surfaces on the head side of the gasket one can see lots of little dots which is an indication of something going on below the surface- the dots being the location of the barbs. This of course tells one that once the degraded coolant has breached the gasket material it can and does migrate along the length and breadth of the gasket and once that happens it is somewhat random where it breaks through. In the case of the 5/8 gasket this process was extensive thus why for cylinders 5, 6 and 7 there was literally nothing left of the gasket material other than the outer periphery and the fire rings.

I also believe there is another issue with this design in that where there are openings in the gasket for coolant to flow through, sooner or later the coolant simply seeps between the two gasket sheet layers that are not in compression, stagnates and eventually acidises thus why for instance we sometimes see the gasket material eaten away local to those metering holes in the gasket .

As I indicated in a previous discussion modern Porsches post the 928 generation have the coolant passages in compression from both sides of the flow conduit and this I suspect is no coincidence.