Packard Forum Index » Post-War (1946-54)

Here's a truly frustrating technical problem so let's see what hypotheses our skilled mechanics can come up with to explain it. I think you can assume that at least one observation or test must be faulty, otherwise there'd be no problem. The car in question is a very low mileage 1954 Cavalier, 327 engine with 4/bbl of course. Prior history includes a valve job, and the cylinder head was resurfaced so the compression ratio is a bit higher than stock. The problem did not arise with this work, but several years and miles later and not connected to any known maintenance or repairs.

The problem manifests itself as throwing coolant out of the overflow. At worst at Interstate speeds, it can be as much as a quart every 15-20 miles. Yes, the head gasket has been replaced (several times). Yes, it does it with several cylinder heads, all of which have been tanked, magnafluxed, checked for straightness and surface ground if required. Yes, the head was properly torqued and retorqued each time it was removed. Yes, the studs have been checked and are not stretching. Yes, it does it with or without a thermostat present. Yes, the radiator has been removed, boiled, checked, and has a gravity flow rate at least equal to the original specification. It has done this with a variety of radiator caps and the neck seat and cap gaskets seem perfect. Compression values are normal enough and quite even, though #1 cylinder is the highest. No, the water pump has not been rebuilt but is silent and does not leak. Never any evidence of coolant in the crankcase.

Now the latest information: Putting each cylinder on TDC on compression and pressurizing the combustion chamber shows no leakdown except for #1 cylinder, which has been examined scrupulously under magnification and with intense light by at least half-dozen mechanics and there is absolutely no evidence of a crack. Although the leakdown was with the piston at TDC, we have also examined the cylinder walls and again, no evidence of a crack. And remember, the several heads were checked and are OK.

Now here's the last teaser: when pressurizing #1 cylinder while at TDC on compression, the coolant level rises in the radiator (cap off) and ultimately reaches the overflow, but no air bubbles are actually seen (perhaps didn't wait long enough). BUT maintaining the pressure on the cylinder and spraying oil on the head studs and nuts, bubbling of air escaping at random studs at the threaded end as far back as #6 cylinder was observed.

Definitely an interesting problem. But logic dictates that air is getting into the water jacket SOMEHOW. I know that doesn't help but for right now that's all I can offer.

The head gasket is defective. Put on a new one and spray it with copper coat when installing.

OR, there IS a crack in the block, possibly in the valve seat area, in a spot that's not too conspicuous.

And the block desk has been checked for straightness? No warpage? No deep scratches or anything of that nature?

I sure sounds like there is a leak between cylinder 1 and the water jacket. When running the hot cylinder gases pressurize the coolent and in blows past the cap over time. I had another car that did the same thing, but it never showed any bubbles in at the coolant at the filler neck. They do have test strips that you can put in the coolant that checks for the presence of hydrocarbons. Also I belive they have a tail-pipe snifer that can check for coolant in the exhaust. I also found residue on the spark plug in the affected cylinder.

I am making the assumption that yu changed head gaskets when installing new/different heads. Therefore, my guess is that there is a crack in the valve seat area. Have you had new seats installed? If so,it could be that they were not seated/sealed properly. Many of the earlier eights had problems with cracks in the valve areas. I have a 38 that had cracks in all but two valve seats.

As I'm sure you know the head studs do enter the water jacket. So, any leak from a cylinder compression would try to force water out the least restricted way.

Having said all that, I once bought a 40 limousine that supposedly had a cracked block. The owner said the water ran out almost as fast as he put it in. I bought it cheap. I found that the only thing wrong with the engine was a wrong head gasket.

Good luck and keep us informed as to what you find. I love/hate these puzzles.

Thanks for the comments thus far but nothing we haven't thought of or checked so please keep thinking about it. Of course we've changed head gaskets, each time the head was removed, probably at least 6 or 8 times by now, always a new gasket, from several sources. I've replaced Packard headgaskets probably hundreds of times so it's not like I have no experience doing it. No scratches or obvious things in the deck of the block. To repeat, so signs of any cracks in the valve seats, though we haven't removed the valves to check deep within the seats. However, if leakage occurs by pressurizing the cylinder with both valves closed, that kind of (?) eliminates that. And no coolant thrown from the exhaust that's noticeable, and we've looked! We haven't checked the block deck for flatness, but when you lay a true flat cylinder head on it w/o a gasket, it seems perfectly flat and true.

And no, the engine hasn't had any valve seats installed.

EDIT: And I should have added that we used a test kit to check for combustion byproducts in the coolant and the test was negative.

But there could be a slight warpage on the deck near a cylinder wall. So the outer edges of the deck and head are nice and square, but then there is a "depression" near the cylinder itself. So the gasket doesnt seal there.

Perhaps run a straighedge over the deck and look for any gaps near the cylinders.

Thanks Kev, I suppose that's possible though you have to wonder what could have caused that to happen after 54 years?

Not sure what would cause that on a Packard Eight, but I know on some modern motor that use sleeved cylinders that the sleeve can move and cause this issue. A honda motor did this to me one time. But it had 200,000+ miles on it.

I figured that if you had tied multiple heads, and gaskets, then the only thing left would be a block crack, or deck warpage.

Had this problem with a 226 Continental engine.I pulled my hair out with all the same test.I ran the engine with my cooling system pressure tester and found it would build over 15 psi only after the engine warmed up to normal operating temp.I let the engine cool and pulled the spark plugs and pumped up the cooling system to 15 psi and found no coolant leaks,even looking into the cylinders with a scope.So I repeated the test with the engine warm and found a hair line crack in the #6 cylinder about 3/8" from the top of that cylinder,on the valve side.The engine was rebuilt with that cylinder getting a sleeve.No ploblems after that. Huge PITA !!!