When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Hi Fred,
Thank you for your comments and I do appreciate your analysis but I don't think it is correct. I'll agree to disagree. I've addressed a couple of issues you raised below in CAPITALS.
It's most likely that several mechanisms are at play to cause corrosion, but given the localization/proximity to the gasket/head interface it appears to me that galvanic corrosion caused by a compromised head gasket is a major contributor.
All the best,
Joe
Hi Joe,
Initially you stated that crevice corrosion can only occur in saline solutions and I take it you have moved on from that. It is a 100% certainty that crevice corrosion is taking place in the 928 engine - the evidence is the pitting corrosion phenomena taking place within a flanged joint- no other corrosion phenomena can do this. The discussion therefore can only be about how the crevice attack occurs not whether it occurs. If we cannot agree on that then anything else is a waste of time. Pitting just does not occur in a galvanic attack.
I have theorised as to how this form of crevice corrosion occurs and I am far from certain that I have got this spot on. If the mechanism proposed is not 100% correct some other variant is in play.
You state that if there is no oxygen depletion there can be no crevice corrosion. Crevice corrosion is defined by electrochemical potential developed within the crevice- that has to be sufficient to sustain electron flow and also create an anodic environment - how it achieves such is comparatively unimportant. The small volume inside the crevice could not sustain a major acid attack without depleting itself rapidly yet alone continue for years on end.
Prevent this circuit from forming and the problem will not occur and therein lies the one and only solution. Clearly Porsche never foresaw this problem.
What we also see is that somehow the rubberised binder of the gasket material breaks down and then the evil brew eventually gets to the galvanised steel mesh and that duly corrodes as indicated by the orange staining [ferrous oxide] that we invariably see. If galvanic corrosion were in play how could both the cylinder head and the wire mesh both corrode simultaneously? I am not at all sure as to how the gasket is attacked but it would seem to be an acid attack.
The fire ring is in contact with the block but they are not simultaneously wetted. The gasket design ensures this cannot happen but after the ravages of the crevice corrosion attack the gasket can be eaten away until the fire rings are wetted and then a galvanic attack can be initiated- I have seen this happen but it is rare and ends up with the top of the cylinder being chewed away to the point it can destroy an engine.
Whatever is going on, it must also have to do with the head/block material also, I assume. My first project I ever worked on was a 1979 Fiat X1/9. We completely disassembled and rebuilt its engine with a friend. I don't remember seeing anything of the sort on it and it is older than these engines and Italian, mind you. These were the surfaces (after minor re-surfacing by a shop). They seem smooth. Sorry for the shoddy pics, these are 15 years old and I have no idea where the original size ones are. This is all I have.
Hi Fred,
Thanks for your comments. I was thinking about the formation of iron hydroxide seen in your photos along with the corroded aluminum in your earlier post and agree with your comment that it should not form if only GC is at work. I have a theory that might shed additional light on the topic. As mentioned in my last post, there are likely many things going on to wind up with corrosion as we see it on the cylinder head but they seem to be localized around the head gasket. It very well may be that the process starts with acid attack of the gasket material causing it to loose its integrity and exposing the reinforcing steel fibers and fire ring . As you pointed out in an earlier post, the ethylene glycol can/does breakdown forming a series of organic acids including oxalic, formic and glycolic acids so this makes sense. My contention is that this leads to conditions favorable to GC - with the aluminum head alloy being the anode and exposed areas of the fire ring or steel reinforcement being the cathode. It is entirely possible that at other locations, the exposed fire ring and/or reinforcing strands of steel are not in contact with aluminum and are attacked by the organic acids to form the iron hydroxide. It look like there are two different corrosion mechanisms at work in the pictures - GC where there is no rust and corrosion of iron by organic acid where we see the rust stains.
All the best,
Joe
Originally Posted by FredR
Hi Joe,
Initially you stated that crevice corrosion can only occur in saline solutions and I take it you have moved on from that. It is a 100% certainty that crevice corrosion is taking place in the 928 engine - the evidence is the pitting corrosion phenomena taking place within a flanged joint- no other corrosion phenomena can do this. The discussion therefore can only be about how the crevice attack occurs not whether it occurs. If we cannot agree on that then anything else is a waste of time. Pitting just does not occur in a galvanic attack.
I have theorised as to how this form of crevice corrosion occurs and I am far from certain that I have got this spot on. If the mechanism proposed is not 100% correct some other variant is in play.
You state that if there is no oxygen depletion there can be no crevice corrosion. Crevice corrosion is defined by electrochemical potential developed within the crevice- that has to be sufficient to sustain electron flow and also create an anodic environment - how it achieves such is comparatively unimportant. The small volume inside the crevice could not sustain a major acid attack without depleting itself rapidly yet alone continue for years on end.
Prevent this circuit from forming and the problem will not occur and therein lies the one and only solution. Clearly Porsche never foresaw this problem.
What we also see is that somehow the rubberised binder of the gasket material breaks down and then the evil brew eventually gets to the galvanised steel mesh and that duly corrodes as indicated by the orange staining [ferrous oxide] that we invariably see. If galvanic corrosion were in play how could both the cylinder head and the wire mesh both corrode simultaneously? I am not at all sure as to how the gasket is attacked but it would seem to be an acid attack.
The fire ring is in contact with the block but they are not simultaneously wetted. The gasket design ensures this cannot happen but after the ravages of the crevice corrosion attack the gasket can be eaten away until the fire rings are wetted and then a galvanic attack can be initiated- I have seen this happen but it is rare and ends up with the top of the cylinder being chewed away to the point it can destroy an engine.
Whatever is going on, it must also have to do with the head/block material also, I assume. My first project I ever worked on was a 1979 Fiat X1/9. We completely disassembled and rebuilt its engine with a friend. I don't remember seeing anything of the sort on it and it is older than these engines and Italian, mind you. These were the surfaces (after minor re-surfacing by a shop). They seem smooth. Sorry for the shoddy pics, these are 15 years old and I have no idea where the original size ones are. This is all I have.
Your photo goes a long way to explaining why the 928 engine has this affliction- the block on that motor is of a closed deck construction therefore where the gasket envelops the coolant transfer passages there is hard contact on both sides of the gasket and it seals.
The 928 block is an open deck design and worse, it uses the head gasket to block off or keep open the transfer ports as needed. There is nothing on the block side to compress against and sooner or later coolant seeps into minute gap [the crevice] between the gasket and the head and then the fun starts.
Your photo goes a long way to explaining why the 928 engine has this affliction- the block on that motor is of a closed deck construction therefore where the gasket envelops the coolant transfer passages there is hard contact on both sides of the gasket and it seals.
The 928 block is an open deck design and worse, it uses the head gasket to block off or keep open the transfer ports as needed. There is nothing on the block side to compress against and sooner or later coolant seeps into minute gap [the crevice] between the gasket and the head and then the fun starts.
I see. There are coolant channels outside of pressure pressed areas. Blocking off passages with the head gasket does sound like a terrible idea also.
I see. There are coolant channels outside of pressure pressed areas. Blocking off passages with the head gasket does sound like a terrible idea also.
The gasket is semi rigid- it is formed by pressing two sheets of Klingerit sheet gasket material over a galvanised steel barbed mesh with the fire rings pressed into place. The gasket is counter held in the fire ring area and also on the peripheral sealing surfaces but not beneath the transfer passages. It will seal to some extent and therein lies the problem- once the coolant seeps in it lies stagnant with no flow and thus no hope of additive replenishment. Thereafter it is a crapshoot as to how it degrades.
Initially I had reason to believe the engines that had the most problems were ones that sat for weeks or months on end without being used as I saw three such examples over here around the same time and the main agents were baffled as to what was going on and their owners were pissed because the heads were written off- Porsche do not sanction repairs . When we transplanted my late 90S4 motor into my current GTS chassis some 16 years ago it was in perfect condition - the motor was [and still is] run regularly. I also use the original IAT coolant formulation. When Jim posted about the degradation in his GB built motor after three years of service and 30k miles and saw I this problem- clearly in the early stages of development- that tended to kick my initial suspicions into touch. My late S4 motor has now been in the GTS chassis for some 16 years and I wonder if it is in equally god condition or is festering away.
Initially I had reason to believe the engines that had the most problems were ones that sat for weeks or months on end without being used as I saw three such examples over here around the same time and the main agents were baffled as to what was going on and their owners were pissed because the heads were written off- Porsche do not sanction repairs . When we transplanted my late 90S4 motor into my current GTS chassis some 16 years ago it was in perfect condition - the motor was [and still is] run regularly. I also use the original IAT coolant formulation. When Jim posted about the degradation in his GB built motor after three years of service and 30k miles and saw I this problem- clearly in the early stages of development- that tended to kick my initial suspicions into touch. My late S4 motor has now been in the GTS chassis for some 16 years and I wonder if it is in equally god condition or is festering away.
So, it's a crapshoot? Depends on the exact mating and individual quality of the gasket, block, and heads? My S3 has 37k mi on it and wasn't used much in the past 13 years. I just got it. I will be doing a TB/WP service on it soon. I guess I'll remove the thermostat housing as well and see if I see any hints. Apparently you can see it there also?
So, it's a crapshoot? Depends on the exact mating and individual quality of the gasket, block, and heads? My S3 has 37k mi on it and wasn't used much in the past 13 years. I just got it. I will be doing a TB/WP service on it soon. I guess I'll remove the thermostat housing as well and see if I see any hints. Apparently you can see it there also?
The thermostat housing is an excellent port to see whether there is a general corrosion problem going on. That is what one might expect to see if the coolant additive package was exhausted and the operating temperature then degraded the entire 14 litres of coolant inventory. Then you would expect to see corrosion all over every coolant wetted surface. What we see in practice is pitting corrosion damage on the surface of the head adjacent to the unsupported areas of the gasket with the most vulnerable areas being the real estate adjacent to and outboard of the fire ring area.
Much as I have looked at every example posted on Rennlist that I could find I have yet to determine a clear and obvious pattern other than the corrosion we typically see.
The one thing you can be absolutely sure of is when looking at a head that is damaged you then look inside the transfer ports and they are as clean as a whistle then the coolant is not the root cause of the problem.
The thermostat housing is an excellent port to see whether there is a general corrosion problem going on. That is what one might expect to see if the coolant additive package was exhausted and the operating temperature then degraded the entire 14 litres of coolant inventory. Then you would expect to see corrosion all over every coolant wetted surface. What we see in practice is pitting corrosion damage on the surface of the head adjacent to the unsupported areas of the gasket with the most vulnerable areas being the real estate adjacent to and outboard of the fire ring area.
Much as I have looked at every example posted on Rennlist that I could find I have yet to determine a clear and obvious pattern other than the corrosion we typically see.
The one thing you can be absolutely sure of is when looking at a head that is damaged you then look inside the transfer ports and they are as clean as a whistle then the coolant is not the root cause of the problem.
If one lives in an area where heavy freezes are rare (it doesn't go much below 25 even in the harshest winters here in the morning) would it make sense to use a higher dilution solution?
If one lives in an area where heavy freezes are rare (it doesn't go much below 25 even in the harshest winters here in the morning) would it make sense to use a higher dilution solution?
Rolls Royce figured out that the lowest freezing point they could get with their new liquid cooled world beating Merlin engines prior to WW2 was something like minus 52C with a 70:30 water glycol ratio. The stock Porsche brew at 50:50 is something in the region of minus 35C as I recall. 70:30 also has better heat carrying capability.
In the UK it gets pretty cold at 47,000 feet especially during winter time.
A little snippet that Stan may appreciate: believe it not when Boeing woke up and put the Merlin engine in their power starved Mustang to create the P51D and improved its performance sensationally, their engineers discovered that by clever [lucky?] design of the coolers the impact of heating up the air being used to cool the motors gave the aircraft a power boost as the air [when heated] expanded volumetrically and that produced more thrust thus creating an afterburner type effect that apparently was rather useful to its pilots.
Hi Fred,
Thanks for your comments. I was thinking about the formation of iron hydroxide seen in your photos along with the corroded aluminum in your earlier post and agree with your comment that it should not form if only GC is at work. I have a theory that might shed additional light on the topic. As mentioned in my last post, there are likely many things going on to wind up with corrosion as we see it on the cylinder head but they seem to be localized around the head gasket. It very well may be that the process starts with acid attack of the gasket material causing it to loose its integrity and exposing the reinforcing steel fibers and fire ring . As you pointed out in an earlier post, the ethylene glycol can/does breakdown forming a series of organic acids including oxalic, formic and glycolic acids so this makes sense. My contention is that this leads to conditions favorable to GC - with the aluminum head alloy being the anode and exposed areas of the fire ring or steel reinforcement being the cathode. It is entirely possible that at other locations, the exposed fire ring and/or reinforcing strands of steel are not in contact with aluminum and are attacked by the organic acids to form the iron hydroxide. It look like there are two different corrosion mechanisms at work in the pictures - GC where there is no rust and corrosion of iron by organic acid where we see the rust stains.
All the best,
Joe
Hi Joe,
Good to know that there are at least two souls on the planet who have some idea as to what is really on corrosion wise in the 928 cooling system!
For sure it is a somewhat complex corrosion mechanism and it took quite some thinking to fathom out what I published to date. What I do find astonishing is the dearth of input from folks on the list- either for or against the thinking process here. There again what seems straight forward to me with my particular background is probably mind boggling science to many.
Unless something is done to prevent this from happening it is simply going to recycle itself albeit the time constant is something in excess of 15 years.
For sure a galvanic attack could be initiated if and when the fibre part of the gasket is eaten away such that the attrition reaches the fire ring location. I have seen this happen over here but to date have no photos of such in my data base.
What I do find astonishing is the dearth of input from folks on the list- either for or against the thinking process here. There again what seems straight forward to me with my particular background is probably mind boggling science to many.
LoL. Or maybe some of us can’t justify the time right this instant. But, in the absence of the time to make a long coherent post:
Your conclusion based upon your observation of your hard coolant y-pipe is incorrect. I have seen no correlation between head condition and hard line condition.
“Pinto” motor (whenever it is) above is very-probably not Alusil. Remember that our engines are Alusil -> Gasket -> Cheap Aluminum with the cheapness inversely proportional to age. Don’t assume that because X happens in a cast iron block that the mechanics are identical in ours.
Why would coolant between the deck and gasket surface to get “refreshed” for small values of “refresh?” Diffusion.
Does the reaction yaw’ll posit result in new reactants? Or …
LoL. Or maybe some of us can’t justify the time right this instant. But, in the absence of the time to make a long coherent post:
Your conclusion based upon your observation of your hard coolant y-pipe is incorrect. I have seen no correlation between head condition and hard line condition.
“Pinto” motor (whenever it is) above is very-probably not Alusil. Remember that our engines are Alusil -> Gasket -> Cheap Aluminum with the cheapness inversely proportional to age. Don’t assume that because X happens in a cast iron block that the mechanics are identical in ours.
Why would coolant between the deck and gasket surface to get “refreshed” for small values of “refresh?” Diffusion.
Does the reaction yaw’ll posit result in new reactants? Or …
Dave,
You do not see a correlation and therefore my deductions are incorrect?
Firstly the Fiat engine block material is irrelevant to the discussion at hand. Mr Merope asked a simple query- "why did the heads on his Fiat engine not corrode like the 928 heads are doing? I gave him a simple answer- the Fiat engine is a closed deck construction and therefore does not have an open gasket design that allows the coolant to wet the minute gap between the gasket and the head ergo it does not corrode.
Same kind of thing viz the Y-piece and the head corrosion- the similarity is the fact that the same coolant gets trapped in a similarly small gap, the corrosion strikes from the outside of the pipe and only in the area adjacent to the trapped liquid. The trapped liquid does not get refreshed, the coolant degrades into various organic acids, electrons start flowing and little holes start appearing on the outside of the pipe that eventually break through- can you see the similarities now?
On the other hand given there are 4 different corrosion mechanisms to choose from and you reckon it is not crevice corrosion then be a good chap and advise which one of the remaining three possibilities it might be and then I will be happy to review your assessment and more importantly with a credible and thorough analysis.
The very reason that crevice corrosion takes place is because the main body of the coolant cannot refresh the depleted fluid in the crevice. You may have heard of the word "diffusion" but do you really understand it? Diffusion is very conditional- if you are interested in it then you might enjoy reading up on Fick's law of diffusion- it is rather complicated and not many folks can grasp it but therein lies the engineering logic as to why diffusion does not take place.
As to the mechanics of the process I had to sit down and fathom it out from first principles and even now I am still far from certain I have it spot on - maybe you would like to share your thinking as to why we see what we do?
Initially you stated that crevice corrosion can only occur in saline solutions and I take it you have moved on from that. It is a 100% certainty that crevice corrosion is taking place in the 928 engine - the evidence is the pitting corrosion phenomena taking place within a flanged joint- no other corrosion phenomena can do this. The discussion therefore can only be about how the crevice attack occurs not whether it occurs. If we cannot agree on that then anything else is a waste of time. Pitting just does not occur in a galvanic attack.
I have theorised as to how this form of crevice corrosion occurs and I am far from certain that I have got this spot on. If the mechanism proposed is not 100% correct some other variant is in play.
You state that if there is no oxygen depletion there can be no crevice corrosion. Crevice corrosion is defined by electrochemical potential developed within the crevice- that has to be sufficient to sustain electron flow and also create an anodic environment - how it achieves such is comparatively unimportant. The small volume inside the crevice could not sustain a major acid attack without depleting itself rapidly yet alone continue for years on end.
Prevent this circuit from forming and the problem will not occur and therein lies the one and only solution. Clearly Porsche never foresaw this problem.
What we also see is that somehow the rubberised binder of the gasket material breaks down and then the evil brew eventually gets to the galvanised steel mesh and that duly corrodes as indicated by the orange staining [ferrous oxide] that we invariably see. If galvanic corrosion were in play how could both the cylinder head and the wire mesh both corrode simultaneously? I am not at all sure as to how the gasket is attacked but it would seem to be an acid attack.
The fire ring is in contact with the block but they are not simultaneously wetted. The gasket design ensures this cannot happen but after the ravages of the crevice corrosion attack the gasket can be eaten away until the fire rings are wetted and then a galvanic attack can be initiated- I have seen this happen but it is rare and ends up with the top of the cylinder being chewed away to the point it can destroy an engine. This is not remotely true. The wide fire ring on the cylinder side of the gasket for an S4 engine is in constant contact with the coolant...it protrudes past the cylinder on every '87 to '95 engine. And worth mentioning, the early (original) head gaskets for the '85/'86 engine did not have a wide fire ring on the cylinder side of the gasket...and the heads in these engines are always in better condition than S4 heads.
I'm not involved in this discussion....way above my pay grade.
The above is offered simply as a "point of order".
Semi-retired, as of Feb 1, 2023.
The days of free technical advice are over.
Free consultations will no longer be available.
Will still be in the shop, isolated and exclusively working on project cars, developmental work and products, engines and transmissions.
Have fun with your 928's people!
09-05-2021, 06:47 PM
