Does anybody know a "rule of thumb" for realistic cooling
system leak rates on 300-ish CI v8 engines?

Tested with compressed air, the cooling system drops about
a quarter PSI per minute at 9 PSI. There's an audible leak
at the water pump shaft, which changes when the shaft is
turned. Bearings feel good, no noise, bind or play.

The longer story is that the car very slowly lost coolant
through visible weeps in the (20 year old!) radiator. I
did a final pressure test before pulling the radiator,
finding no leaks except the radiator and an insignificant
leakdown rate, convincing me that radiator was the only
problem. A bench test confirmed the repaired radiator is
tight

Now that it's back together, a final (air, not coolant-filled)
system test shows what seems like a larger leak. Cooling systems
are far from perfect, but several PSI per hour seems like a lot.
The Stant pressure test instructions suggest a "steady" reading
for two minutes is acceptable, but that is with coolant, not air.

Thanks for reading, and any thoughts....

bob prohaska

Anybody been through this sort of drill?

On Sat, 14 Aug 2021 01:00:53 -0000 (UTC), bob prohaska
<bp@www.zefox.net> wrote:

>Does anybody know a "rule of thumb" for realistic cooling
>system leak rates on 300-ish CI v8 engines?
>
>Tested with compressed air, the cooling system drops about
>a quarter PSI per minute at 9 PSI. There's an audible leak
>at the water pump shaft, which changes when the shaft is
>turned. Bearings feel good, no noise, bind or play.
>
>The longer story is that the car very slowly lost coolant
>through visible weeps in the (20 year old!) radiator. I
>did a final pressure test before pulling the radiator,
>finding no leaks except the radiator and an insignificant
>leakdown rate, convincing me that radiator was the only
>problem. A bench test confirmed the repaired radiator is
>tight
>
>Now that it's back together, a final (air, not coolant-filled)
>system test shows what seems like a larger leak. Cooling systems
>are far from perfect, but several PSI per hour seems like a lot.
>The Stant pressure test instructions suggest a "steady" reading
>for two minutes is acceptable, but that is with coolant, not air.
>
>Thanks for reading, and any thoughts....
>
>bob prohaska
>
>
>Anybody been through this sort of drill?
>
The system MUST be tested wet as water pump seals, for example, depend
on fluid pressure to seal completely. I suspect ypou WILL end up
changing the water pump to stop the seapage but you also need to check
EVERY hose connection for corrosion of the metal parts. Vapour gets
out where liquid cannot - often leaving a white deposit behind in the
joint

Clare Snyder <clare@snyder.on.ca> wrote:
> On Sat, 14 Aug 2021 01:00:53 -0000 (UTC), bob prohaska
> <bp@www.zefox.net> wrote:
>
>>Does anybody know a "rule of thumb" for realistic cooling
>>system leak rates on 300-ish CI v8 engines?
>>
>>Tested with compressed air, the cooling system drops about
>>a quarter PSI per minute at 9 PSI. There's an audible leak
>>at the water pump shaft, which changes when the shaft is
>>turned. Bearings feel good, no noise, bind or play.
>>
>>The longer story is that the car very slowly lost coolant
>>through visible weeps in the (20 year old!) radiator. I
>>did a final pressure test before pulling the radiator,
>>finding no leaks except the radiator and an insignificant
>>leakdown rate, convincing me that radiator was the only
>>problem. A bench test confirmed the repaired radiator is
>>tight
>>
>>Now that it's back together, a final (air, not coolant-filled)
>>system test shows what seems like a larger leak. Cooling systems
>>are far from perfect, but several PSI per hour seems like a lot.
>>The Stant pressure test instructions suggest a "steady" reading
>>for two minutes is acceptable, but that is with coolant, not air.
>>
>>Thanks for reading, and any thoughts....
>>
>>bob prohaska
>>
>>
>>Anybody been through this sort of drill?
>>
> The system MUST be tested wet as water pump seals, for example, depend
> on fluid pressure to seal completely. I suspect ypou WILL end up
> changing the water pump to stop the seapage but you also need to check
> EVERY hose connection for corrosion of the metal parts. Vapour gets
> out where liquid cannot - often leaving a white deposit behind in the
> joint

Indeed, it leaked air in both directions, pressure and vacuum. With
coolant, it seems to seal just fine. No leaks at all. Holds pressure
and refills from the overflow tank on cooling, as it should.

I'm more than a little surprised.....

Thanks for writing,

bob prohaska