Head gasket failure discussion - coolant/exhaust/oil exchange

JHZR2

Staff member
Joined
Dec 14, 2002
Messages
52,631
Location
New Jersey
Have some general questions regarding head gaskets. As some may know I have a few diesels which need HG work, and I’d like to take some of these on.

My 1991 350SD (call it car #1) is known to have HG and rod issues. Mine had oil in the coolant. Never did I find a situation where the cooling system retained pressure afterwards. It’s a pristine, low mileage example. I’m trying to get the HG done but my Indy had issues with the job and keeps getting a misfire at idle. It’s a side job so it’s slow.

I found another 91 (call it car #2) that has the same telltale oil in the coolant overflow. It cranks and starts beautifully when cold soaked in 30 degree weather. Overall car is very nice. It’s not as low mileage as my first one, but I can get it REALLY cheap. The difference with this one is that after sitting for some time (as ai understand it, days), the cooking system still had pressure. The first thing I did was open the radiator cap and it had pressure from who knows when. I started it and ran it (no boost, car was parked in), and no pressure or bubbles from the overflow.

Neither car had milkshake oil.

So:
Car #1 oil in coolant, no retained pressure in system.

Car #2 oil in coolant, sustained pressure many days.

I get it that it could be as simple as differences in the seal quality on the radiator cap (there is no overflow, the cap is on the bottle). But could one imply cracked head or some other issue over the other?

More general question - how could the system hold pressure and not push coolant back into the cylinder or exhaust? Every cooling system I’ve had stays pressurized for a bit as the engine cools down. But of course it goes away to nothing. This engine on car #2 was stone cold. I assume this means that the way exhaust gets into the system must only be under boost.

That question holds for my car #1. How is it
That oil gets into the coolant, but after the car shuts down, and oil pressure goes away, that water doesn’t make its way back in? I understand why it wouldn’t happen when the car is operating with >>16 psi of oil pressure.

I have a 1993 (W140 body style) with a perfect factory rebuild engine installed by Mercedes of Manhattan. The car is perfect on the inside except for the driver seat. Paint needs some help. But the engine and transmission are more or less perfect when running, and NO signs of an HG issue. This car may be a flip, or a beater, but I really bought it for the value of the hood engine and AT, being that they are the same model as my 91.

Of course my one 96 Ram also has the known overboard coolant leak from the HG. It’s very slight, I knew about it when I bought it, and didn’t make any issue driving across country. It was common. But it’s another project. Thus I’m interested in learning about head gaskets. If I can know how to deal with this, I think I can make a variety of nice diesels and other nice cars come back to life as a hobby. I have the tools and space. Just not the first go at it and a few learning points along the way.

Anyway, any insight would be appreciated.

Thanks!!
 
It is possible the leak requires a higher pressure than the cooling system pressure, the oil pump on these cars make quite a bit of pressure off idle where the cooling system is 20 psi or less.
For example of the point of leakage needs more than 20 psi to push oil through it the cooling system will not push coolant back through that gap.
It is common to see small coolant leaks at joints that cease quickly once the engine is shut down for another example.

This is going to be a hard leak to locate being internal, the oil cooler IIRC is oil to air (not coolant) so I would guess HG, or a crack in the head or block.
The engine holding pressure for days is most likely a bad cap on the bottle. Just thinking out loud, I will think this over a bit more.
 
Is it possible to do a lead down test similar to what is done in gasoline engine??

If possible this would be a good strting point.
 
Is it possible to do a lead down test similar to what is done in gasoline engine??

If possible this would be a good strting point.
Anything is possible.

I own car #1.

I can get car #2 for a song.

If I get car #2, I can do tests, but clearly it needs a HG. It would be my first DIY HG, which would be why I’d try it.

If I pulled the head, I’d check piston heights and get the head checked for cracks. At that point there isn’t much need for other tests, I’d think.

Key is I don’t want to start on a project to run into a cracked head… thus curious if the presence of lingering pressure indicates a different issue, like cracking instead of just pushing oil in a HG fault.
 
Do you have good physical mobility and the ability to lift a diesel head off the engine? Doing the replacement is not "hard" but requires quite a bit of physical labor or the proper lifting equipment.
 
Anything is possible.

I own car #1.

I can get car #2 for a song.

If I get car #2, I can do tests, but clearly it needs a HG. It would be my first DIY HG, which would be why I’d try it.

If I pulled the head, I’d check piston heights and get the head checked for cracks. At that point there isn’t much need for other tests, I’d think.

Key is I don’t want to start on a project to run into a cracked head… thus curious if the presence of lingering pressure indicates a different issue, like cracking instead of just pushing oil in a HG fault.
How would a cracked head maintain pressure in the cooling system for days? Cylinder compression dissipates quickly once the engine is shut down.
I don't think I am fully understanding the issues.
 
It's a 1991 vehicle.
71Ze1Q4avfL._AC_SL1477_.jpg
 
How would a cracked head maintain pressure in the cooling system for days? Cylinder compression dissipates quickly once the engine is shut down.
I don't think I am fully understanding the issues.
Same here.

The car was stone cold. Yet when I opened the pressure cap on the coolant reservoir, I heard and saw coolant escape under pressure when I moved the cap to the first position. So my only assessment then has to be that it had been sitting that way for a long time. The car was parked in, at a locked location, so it’s not like it is daily driven.

I’ve read about how if your radiator hose is hard the next morning, then it means you have a cracked head or bad head gasket, since somehow exhaust got in there and pressurized the system. It has never made sense to me how such a leak would be unidirectional.
 
Dive in on whichever one is cheapest to start on, the only way you are going to learn is by doing it. These are not complex engines, you are intelligent, you'll figure it out. Take it slow, be careful and inspect everything. You can have the head checked once it is off.
 
Same here.

The car was stone cold. Yet when I opened the pressure cap on the coolant reservoir, I heard and saw coolant escape under pressure when I moved the cap to the first position. So my only assessment then has to be that it had been sitting that way for a long time. The car was parked in, at a locked location, so it’s not like it is daily driven.

I’ve read about how if your radiator hose is hard the next morning, then it means you have a cracked head or bad head gasket, since somehow exhaust got in there and pressurized the system. It has never made sense to me how such a leak would be unidirectional.
If the engine was not started there is no way exhaust gases would remain in the system. It is the job of the cap to relieve pressure if it reaches the caps rating. This has the symtoms of a bad cap, use a cap tester on your pressure tester, it takes 5 seconds.
 
If the engine was not started there is no way exhaust gases would remain in the system. It is the job of the cap to relieve pressure if it reaches the caps rating. This has the symtoms of a bad cap, use a cap tester on your pressure tester, it takes 5 seconds.
Yes the cap would vent if meeting rating, but if a hg caused an exhaust leak into the system, it could be pressurized indefinitely below the cap rating. It was an OE MB cap…. It only released the pressure when I opened the cap…
 
Dive in on whichever one is cheapest to start on, the only way you are going to learn is by doing it. These are not complex engines, you are intelligent, you'll figure it out. Take it slow, be careful and inspect everything. You can have the head checked once it is off.

OVERKILL has the idea here.... You're overthinking it. These things are obviously a hobby/pastime/interest to you AND you have the means to do it. Just dive in!
 
I think those IDI MB diesels run a really high compression ratio, are you sure it’s oil in the coolant and not exhaust gases? That would be a LOT of oil pressure to pressurize a cooling system for DAYS! I’m sure you know the 350SD isn’t called the “rod bender” for nothing-they seem prone to hydro lock.
 
If the engine was not started there is no way exhaust gases would remain in the system. It is the job of the cap to relieve pressure if it reaches the caps rating. This has the symtoms of a bad cap, use a cap tester on your pressure tester, it takes 5 seconds.
They run 22-1 compression, 18 bar MINIMUM test compression-over 260 PSI! That’s a lot if it works it’s way out somewhere.
 
I have seen many blown headgaskets/ cracked heads that did not put any visible water/ coolant into the oil. You would see it in a uoa.

My guess is that it takes 40 + psi of oil pressure to push past the bad spot be it a seal or a gasket.
 
Yes the cap would vent if meeting rating, but if a hg caused an exhaust leak into the system, it could be pressurized indefinitely below the cap rating. It was an OE MB cap…. It only released the pressure when I opened the cap…
Remember how simple overflow tanks worked? Hot and cold level, the cap had to open when the engine was cooling down to draw the coolant back into the system through a vent valve. Bad cap vent valve the system will not release pressure in the system as the coolant contracted. Just test the cap.

If anything the fact it is holding pressure for days points to a tight system with no internal leakage.

Edit: Is it possible someone used original bars leak in this one? That will leave an oily film in the overflow bottle.
 
Last edited:
Remember how simple overflow tanks worked? Hot and cold level, the cap had to open when the engine was cooling down to draw the coolant back into the system through a vent valve. Bad cap vent valve the system will not release pressure in the system as the coolant contracted. Just test the cap.

If anything the fact it is holding pressure for days points to a tight system with no internal leakage.

Edit: Is it possible someone used original bars leak in this one? That will leave an oily film in the overflow bottle.
It looks like my 1991 350sd - black diesel oil in the coolant tank because it’s the highest point.

My interpretation was that the cap holds a seal below say 15psi, but if the air in the headspace is pressurized to that, then when it’s released water will shoot out. I saw this when I used a pressure test system and there was air in the radiator…

I think the car is clean enough and cheap enough to make it worth a shot at a diy HG. It’s not as mice as my current 91, but could be fun…
 
It looks like my 1991 350sd - black diesel oil in the coolant tank because it’s the highest point.

My interpretation was that the cap holds a seal below say 15psi, but if the air in the headspace is pressurized to that, then when it’s released water will shoot out. I saw this when I used a pressure test system and there was air in the radiator…

I think the car is clean enough and cheap enough to make it worth a shot at a diy HG. It’s not as mice as my current 91, but could be fun…
My thoughts exactly, some of these sealers are really good, there used to be a product called 'LIQUID GLASS', only available at a pharmacy,
newer products are compatible with antifreeze, there is an interesting video on youtube and they remove spark plugs to determine which cylinder pair the gasket is blown. Good Luck
 
Somewhat related, I took my coolant cap off recently to take some old coolant out, and put some flush in, because I want better performance from my heater core, which, while it works, could work better, in my opinion.

I was almost sure the car was cold, I took my cap off, and I had a quick short "pop" of pressure and some coolant.

Might be nothing, maybe the engine wasnt fully cold, but maybe the coolant level was a touch low, letting pressure form and be held in there?

I would suggest the same test I mentioned in the Subaru thread that was posted a short time ago.. Engine cold, radiator cap off, either look for coolant shooting out on startup, or put a balloon over the opening and see if it inflates, it should not, at least not until the engine heats up, maybe. If at all, it is cold outside so pressure building vs ambient temp.. point is, if overpressurization is present, it will show itself.
 
Back
Top