: testing for bad HG



Harry Yarnell
02-16-09, 01:18 PM
I have four Northstar engines sitting on the floor in the garage. These engines have come out of parts cars, and have not been started. What's the best way to determine whether the HG's are any good.
I've filled the engine with coolant (as best I could without a radiator), and charged each cylinder at TDC with 100# of shop air and observed the coolant for any bubbles, but I'm not sure that's a good test.

dkozloski
02-16-09, 01:40 PM
I have four Northstar engines sitting on the floor in the garage. These engines have come out of parts cars, and have not been started. What's the best way to determine whether the HG's are any good.
I've filled the engine with coolant (as best I could without a radiator), and charged each cylinder at TDC with 100# of shop air and observed the coolant for any bubbles, but I'm not sure that's a good test.
That is the Cadillac of definitive tests. The Cadillac engineer that designed the cooling system swore by it.

97EldoCoupe
02-16-09, 03:54 PM
It's not a good enough test.

100#'s of air does not nearly equate to the force of combustion. Not even close.

The combustion force when that fuel/air mixture ignites can literally flex the head, if some, or even only one or two bolts have released their clamping force on the head. Yet there's enough clamping force to handle 100 psi of shop air without blowby, and even light throttle conditions/steady highway cruising.

The real test is out on the highway, WOT runs.

Sorry to say but if they're not installed in the car, there's no way to correctly and accurately test those engines. The best bet would be to do the HGs on which ever engines you plan on using or selling. Then you're sure- anything else and you're taking chances.

97EldoCoupe
02-16-09, 03:57 PM
Northstars have their own testing process- I know what works and what doesn't, so I don't follow any normal compression/pressure tests, even though some of the time, you can get an accurate diagnosis by the standard testing procedures. That's just it: SOME of the time.

dkozloski
02-16-09, 04:30 PM
It's not a good enough test.

100#'s of air does not nearly equate to the force of combustion. Not even close.

The combustion force when that fuel/air mixture ignites can literally flex the head, if some, or even only one or two bolts have released their clamping force on the head. Yet there's enough clamping force to handle 100 psi of shop air without blowby, and even light throttle conditions/steady highway cruising.

The real test is out on the highway, WOT runs.

Sorry to say but if they're not installed in the car, there's no way to correctly and accurately test those engines. The best bet would be to do the HGs on which ever engines you plan on using or selling. Then you're sure- anything else and you're taking chances.
FWIW, Al Cline who was a lead Cadillac engineer on the design of the Northstar engine said many times on this forum that if the headgaskets were leaking that shop air pressure would show it. He stated that the air test was more definitive than the combustion gas check of the coolant. His experience involved hundreds if not thousands of Northstar engines. His statements came from hard engineering data and not opinion.

Ranger
02-16-09, 05:07 PM
Yup, he swore that was the ONLY way to test them.

97EldoCoupe
02-16-09, 06:55 PM
So hard engineering data doesn't prove that combustion pressure exceeds 100PSI? when that spark ignites the mixture, there's a hell of a lot more force in the combustion chamber than 100 pounds. Plenty enough to raise a head on one end if there's a pulled bolt or two- depending on where those bolts are located.

dkozloski
02-16-09, 07:23 PM
I'm not arguing that combustion pressure does not exceed 100PSI. I'm reporting that a senior Cadillac engineer that was a member of the original design team stated many times that 100PSI shop air was enough to reveal any blown head gasket. If combustion pressure lifted the head enough for combustion gases to shoot by the gasket it ain't going to heal up and stop leaking by the clamping pressure of adjacent fastners and reappear another day.

Harry Yarnell
02-16-09, 07:39 PM
It has been my experience that I've never found a loose headbolt upon disassembly. The clamping force, in my opinion, was not a factor in the gasket failure. But I've only done three...

dkozloski
02-16-09, 07:59 PM
It has been my experience that I've never found a loose headbolt upon disassembly. The clamping force, in my opinion, was not a factor in the gasket failure. But I've only done three...
Al Cline said the same thing. He maintained that the headgasket is a sandwich with a metalic core. Because of poor cooling system maintenance the core corrodes and is eaten away. This relieves the clamping pressure in this area. The gasket then leaks coolant into the headbolt pocket which corrodes the aluminum threads and the threads disintegrate. He said that Cadillac engineering has disassembled and inspected countless engines and this was the failure mode nearly every time. The gasket fails before the headbolts fail. Others report just the opposite with the bolt/threads failing first. I suspect the difference observations has to do with the generation of the engine with the oldest being gasket failure first and newest being thread/block failure.

Ranger
02-16-09, 09:00 PM
I wasn't arguing with you either Jake. Just agreeing that that is what he had said many times.

97EldoCoupe
02-17-09, 08:13 AM
I know Ranger.

I'm one that will usually believe scientific data & research but sometimes you have to look at "how it is" and then figure out "how it got to that point" rather than saying "this is scientifically impossible from an engineering standpoint. Sometimes, what seems impossible, is actually possible under the right conditions.


If combustion pressure lifted the head enough for combustion gases to shoot by the gasket it ain't going to heal up and stop leaking by the clamping pressure of adjacent fastners and reappear another day.

Actually Dkoz, that's exactly what happens- for a little while. The rest of the bolts that still have their firm hold are doing the work of the ones that let loose. Under load, the head and those remaining bolts will flex just enough to let a bit of exhaust gasses past the fire ring (HG) and into the coolant passages. Then when you let off the throttle, the head flexing will stop and the bolts will shrink back to their normal size- the flex is so little but that's all it takes. It's because of this, that more bolts begin to fail VERY quickly. They can't handle the increased pressure. What's happening here is why you can drive with bad HGs for quite a while sometimes, if you keep your foot off the gas. It eventually gets to the point where you can't drive anywhere.

And yes many times it's strictly gasket related- the inner metal core does corrode away, and very badly. That's why more frequent coolant changes are critical.

tateos
02-17-09, 12:30 PM
Yes - I think Jake is correct on all counts - he usually is!

I think the gasket does fail first, then the threads. When I did my HG project, the left bank gasket was blown with 3 bad threads. i think they did not all go at exactly the same time - I think they went bad in succession - when enough went bad, the head gasket leaked combustion gases into the cooling system. My right bank had 1-2 holes that were mushy, with wet/bad thread, but that gasket did not show signs of leakage - not yet. Also, all of my bolt holes that were bad were wet, and the threads looked like they had anti-seize compound on them - that was the corroded aluminum. The bolt holes that were good - the ones that the bolt snapped when loosening - those holes were dry and the bolts cames out with no aluminum on them at all; I think those block threads were fine, but of course I inserted all 20 holes!

zonie77
02-19-09, 10:32 AM
I agree with the HG fails which causes bolt failure...back to Harry's original question...there was a member having HG symptoms that he could not confirm with the cyl pressure test. He did the test multiple times. The guru was still on at the time and held that it wasn't HG's. The member traded the car in but came back later and said the dealer mechanics told him it was HG's. I don't think he did the exhaust gas test.

I think there is a small (sweet?) range where the clamping power is sufficient to hold for the cyl pressure test. Jake just said about this same thing. If the engines are in that range you would get a false good test. I don't know what the chances are of being in that range...pretty small I would think.

Just more fuel on the fires!:D

dkozloski
02-19-09, 03:16 PM
I agree with the HG fails which causes bolt failure...back to Harry's original question...there was a member having HG symptoms that he could not confirm with the cyl pressure test. He did the test multiple times. The guru was still on at the time and held that it wasn't HG's. The member traded the car in but came back later and said the dealer mechanics told him it was HG's. I don't think he did the exhaust gas test.

I think there is a small (sweet?) range where the clamping power is sufficient to hold for the cyl pressure test. Jake just said about this same thing. If the engines are in that range you would get a false good test. I don't know what the chances are of being in that range...pretty small I would think.

Just more fuel on the fires!:D
I can't picture a head gasket withstanding more than a few seconds of fire shooting by it as the head lifts at WOT and then resealing when the throttle backs off. This situation has to be far more theoretical than actual and progress very rapidly.

bigtone
02-22-09, 03:59 PM
On my 98 deville, I never did see bubbles with 100 psi air in each cylinder. I heard what sounded like a burp from the expansion tank when I pressurized #1 cylinder. I was expecting more bubbles, but did not see any. And I was going thru a gallon of coolant in 50 miles, and could not pull a hill without overheating. All I could see on disassembly was a small area in the gasket on top of the coolant wall that looked like it was steam cut. So I'm not sure pressurizing is the best test out there.