My 617.912 began to overheat, unexpectedly. Luckily, I only had to drive it 2 miles
back to the barn.

Tore out the water pump, but it was fine. So the engine had to come out.

I split the head last night, expecting I'd be able to find a breach, somewhere. Here
are the pictures. The yellow highlights point to my suspected leak. The threaded
boss in the block has rust at the location, unlike any other boss.
<click pic large versions>

What think all? I plan to inspect the head for warp over the weekend.

__________________
--frankb

1982 300TD 617.912 <NLA> . 1975 2002 E10-EFI . 1976 914c6 . 1983 MG LMIII <NLA>. 1988 Montero / 616<NLA> . 2001 TLS . 2005 Saab 9-5 Wagon

Seems a bit extreme to tear the head off so quickly. Did you check the obvious first I hope? Did you make sure that the coolant level was high enough, or that there wasn't a clog in the system somewhere before you tore the head off? Or that you temp gauge was reading correctly? Corroded senor connections gave give false readings.

__________________

Ah well the head's off now, so rework it

__________________
$60 OM617 Blank Exhaust Flanges
$110 OM606 Blank Exhaust Flanges
No merc at the moment

When coolant gets into a cylinder it steam cleans the carbon from the combustion chamber. That doesn't look like a leak to me.

Blindly ripping things apart before a proper diagnosis is performed is a sure route to more problems / failure.

The proper route would have been to perform a leak down then you would have known for sure if there was a leak from combustion to coolant. This would have also identified the exact cylinder.

The leak area you have high lighted leads to a dowel pin and not coolant. It is possible that slight combustion gasses were making their way to the dowel pin but that isn't the cause of the overheat.

The car overheated. Not a fantasy, as I could smell it. I pulled in into the garage,
started it after cooldown; and it was running over 120C in a matter of minutes. Not
"normal" behavior. Coolant level was correct, and appeared to be flowing. Temp
gauge had been reading fine for years, so a "corrosion" change in a matter of
minutes is highly unlikely, dont'cha think?

__________________
--frankb

1982 300TD 617.912 <NLA> . 1975 2002 E10-EFI . 1976 914c6 . 1983 MG LMIII <NLA>. 1988 Montero / 616<NLA> . 2001 TLS . 2005 Saab 9-5 Wagon

I'd agree on both counts, except we're talking an overheating cycle of a few minutes.
Still think it would be enough to clean a combustion chamber?

__________________
--frankb

1982 300TD 617.912 <NLA> . 1975 2002 E10-EFI . 1976 914c6 . 1983 MG LMIII <NLA>. 1988 Montero / 616<NLA> . 2001 TLS . 2005 Saab 9-5 Wagon

I'm perhaps not as blind as you think.

The transmission had partially failed ;
and the engine had shown exhaust guide wear from a leakdown test and chain measurement done several months ago.

Doing a proper leakdown --with known leaking valve guides-- is a bit of a crap shoot, yes?

And the "threaded boss" is just that; NOT the dowel pin next to it, which is also
rusted. And its the only boss with rust, as previously stated.

In particular, I am concerned by the SPALLED tracks between the cylinder sealing area, the dowel pin perimeter and the boss
... which might/might not show combustion gas mixing with something else, like coolant from the port to the left of the boss --

-- IF I'm not "blind". Sarcasm noted.

__________________
--frankb

1982 300TD 617.912 <NLA> . 1975 2002 E10-EFI . 1976 914c6 . 1983 MG LMIII <NLA>. 1988 Montero / 616<NLA> . 2001 TLS . 2005 Saab 9-5 Wagon

I don't see anything in your pics that would concern me.
The black streak in the first pic looks like oil from the head bolt.
Did you try to just wipe that off?
What does the head gasket look like aroud that location?
Rust in the head bolt threads is common on this engine.
Just run a tap through the thread and replace the head bolt if it is rusted.

__________________

76 240 D. Bought in 1998 for $25.
85 300D. Got it for free with a bad engine. ( Sold )
60 Unimog 404. What was left of it, was given to me. Now powerd by 617A.
88 560 SEL. Bought without engine and trans. Now powerd by 617A.
67 250 SE. Cuope. For resto or sale.
64 220SE. For resto.

A macro pic might be more informative. I'll post that and a head flatness inspection
later today.

The spall .. the series of little dots .. is corroded. It doesn't wipe off. I've shaved
the gasket off now, and will photo both sides under a 2X lens.

Thanks, I didn't know it was common on this design.
Yes, I can see where the water migrated to the bolt hole. Since its only that
single location -- all the others came out oily -- I'm suspicious.

__________________
--frankb

1982 300TD 617.912 <NLA> . 1975 2002 E10-EFI . 1976 914c6 . 1983 MG LMIII <NLA>. 1988 Montero / 616<NLA> . 2001 TLS . 2005 Saab 9-5 Wagon

Blind as in not having a diagnosis strategy and just randomly taking things apart.

This means nothing to the over heat situation.

A leak down test won't diagnose " leaking " ( ? ) valve guides.

Valve guides can only wear , and for all practical purposes , not leak. If they are worn they need to be measured with the head off with plug gauges or a dial indicator put against the valve stem for measurement. If by " leaking valve guides" , you mean " leaking valve stem seals / valve guide seals" A leak down test won't identify them either since a leak down test won't apply any pressure to the guides.

For completeness, a leakdown test is where one has both valves closed and air is pumped into the cylinder through the glow plug / injector or spark plug hole. Leakdown is expressed as a percentage. During a leak down test it is important to identify _where_ the air is being lost.

Most engines with any amount of time on them will have slightly leaky exhaust valves ( air out the tail pipe ) , slight leakage here is no cause for concern.

Mid time engines will have air leaking past the rings, this isn't any great cause for concern if oil usage isn't excessive.

Very high time engines will have leaking intake valves, this becomes a power issue because it allows combustion gasses to enter the intake manifold giving the effect of full time exhaust gas recirculation.

>>> If a head gasket fails or a head cracks from combustion to coolant, leak down air will enter the cooling system. It is good practice to remove the rad cap when doing a leakdown, any bubbles / geyser of coolant is a indicator that the cylinder you are testing is where the leak is. If this had been done prior to removing the head you would have known for sure there was a combustion / coolant leak in a specific cylinder.

Now, there are very isolated instances where a leak down won't identify loss of coolant into the exhaust but, identifying the source is very easy when the head is taken off. An example is a early 80's Ford Escort that would crack the floor of the exhaust port allowing coolant to escape in to the exhaust.

If there was a coolant leak to the head bolt or dowel pin, the coolant would have nowhere to go once the miniscule volume was filed.

( BTW, the yellow arrow in the head pic is pointing to a different cylinder than the block pic, I am referencing the block pic for the dowel pin. )

There does not appear to be a direct path from combustion to the coolant passage. However, if a pre tear down leak down test showed leakage to coolant in that cylinder _and_ the head tested crack free, I'd call the gasket bad in that location. And you would not go through the " what if ? " , " is this the bad part ? " series of gyrations.

I'd call it living up to your avatar sig line. .. .

That image doesnt look like a breach--more like some oil that was around that hole got drawn between the gasket and the head as it came off....

How is the thermostat/radiator. I think I would have replaced the thermostat first and checked the fan-belt/water-pump and for airlocks before tearing into the motor...

__________________
http://i190.photobucket.com/albums/z...0TDnoplate.jpg

Alastair AKA H.C.II South Wales, U.K. based member

W123, 1985 300TD Wagon, 256K,
-Most recent M.B. purchase, Cost-a-plenty, Gulps BioDiesel extravagantly, and I love it like an old dog.

W114, 1975 280E Custard Yellow,
-Great above decks needs chassis welding--Really will do it this year....