Persistent coolant leak

[1n5ane]

I have a 2006 TJ 4.0 with a persistent coolant leak from the radiator.

The cooling system has been well maintained, flushed, and filled with Zerex G05 every 4 years. I overhauled the cooling system 5 years ago, replacing the tstat + housing, hoses, and water pump, but kept the OEM radiator since it was fine.

The problem started 2 years ago when the 17-year-old OEM radiator split the top tank.

I bought an aftermarket radiator (CSF 3244) based on reviews, but it developed a leak between the tank and core within 4 months. I got a new radiator under warranty, but it also sprung a pinhole leak in the core 6 months later.

I tested the coolant for electrolysis and found no stray voltage; I've also pressure tested the coolant cap and it holds pressure.

I bought a higher-quality radiator (Denso 2219234) assuming it was the faulty CSF radiators, but it also developed a pinhole leak in the core 6 months later.

I pressure test every new radiator installation, and they all held 18 PSI without any apparent leaks.



Before I blindly throw a 4th radiator at the issue, is there something I should be checking for?

 

[Chris]

You should do a leak down test. There's a good chance you've got a crack in your cylinder head or head gasket. I'd want to rule that out first.

 

[1n5ane]

Thanks for the suggestion. I've never seen signs of coolant consumption (aside from the leak) so this is a bit unexpected.

I don't have a leakdown tester but I should be able to borrow a combustion gas tester from a buddy.

 

[Chris]

Thanks for the suggestion. I've never seen signs of coolant consumption (aside from the leak) so this is a bit unexpected.

I don't have a leakdown tester but I should be able to borrow a combustion gas tester from a buddy.

Let us know what you find. Often times it starts out very small.

 

[Gilaguy23]

I may be wrong but I think what Chris is getting at is leaked compression into the cooling system causing seam failure. A crap shoot as it may be, when you pressure test the cooling system id guess its cold and holds the 18psi. Once warmed up cracks or leaks can expand and overpressure the cooling system. You'd think a hose would blow first but who knows these days with our supply chain issues and lack of quality parts.

 

[1n5ane]

Wouldn't a head gasket leak also cause coolant consumption (ie. Coolant being burned)? I'd assume there's plenty of pressure in the cooling system to push coolant back into the cylinders after shutting off a hot engine.

Initially, I considered the head gasket a possibility. However, the absence of obvious symptoms, such as white exhaust on startup, and the fact that the coolant cap should be able to vent a small amount of pressure before the radiator bursts, led me to reconsider.

I'm also not loosing much coolant, less than half a quart every month.

Either way, I'll run a combustion gas test just to rule it out. Will keep you guys posted.

 

[Gilaguy23]

1/2 a quart a month is something Id be concerned about myself. You shouldn't lose any IMO. Still doesn't account for the radiator failures though.

 

[Link]

Yep you got a bigger issue than just bad radiators. Do some diagnostics. Sounds like you are headed in the right direction.

 

[MikeE024]

Initially, I considered the head gasket a possibility. However, the absence of obvious symptoms, such as white exhaust on startup, and the fact that the coolant cap should be able to vent a small amount of pressure before the radiator bursts, led me to reconsider.

A head gasket problem can send coolant to the exhaust, engine oil, or out the coolant reservoir. I experienced the latter on my race quad out in the dunes many years ago.

Do you have any coolant splatter on the underside of your hood or along the fender that might indicated fluid loss at the reservoir?

As you mentioned, it’s definitely worth testing your system.

 

[1n5ane]

Just an update for anyone interested.

I finally had some time yesterday to investigate this coolant leak. I re-ran my basic tests and the results make even less sense.

Pressure tested the cooling system again and the rad core pinhole is still there, but a new leak showed up in the top tank seam.
Coolant cap failed pressure test, will hold up to ~10psi before it mysteriously vents all the pressure.
Coolant now tests positive for electrolysis, I'm getting -0.24V between the coolant and chassis.
Combustion gas test is negative (as I had original suspected), so that rules out head gasket leaks.