Strange heating problem

[Das Ninjabot]

On the way back home from school last night I realized that my heat had stopped working and my temp gauge was starting to creep north. Oddly enough it was on the same damned stretch of road that I blew my head gasket on last November.

 

Cue lots of cursing and general aggravation as I had just finished 11+ hours of studying for the MPRE that I had to take this morning but I digress. I let off the gas and coast for about 1/2 mile until I came to a stoplight that was red. The coasting got the car from 3/4 hot to it's normal 1/2 way point, at no time did it ever get hotter than 3/4 of the gauge. I turned the car off and wait until the light changes. Light turns green, I restart the car, and boom, full on heat like nothing is wrong. I drove about two miles, make a stop at the store, and then continue the additional mile home. At no time in my trek home after I shut off the car at the light was there any other issue, nor did the heat ever go north of the normal 1/2 way point on the gauge.

 

I took the gf's car to the test this morning and I'll look at the Subie when I get home but this is up there with the odder things I've seen in my wrenching. A buddy suggested it may be the thermostat but I've never seen one act like that. It's a brand new oem one with less than 5k on it, but it could be defective. I did just replace the coolant temp sensor so I was thinking maybe some air got in the system and created a cavitation problem that reared its head until I shut off the car.

 

I'm really hoping that it's something simple and not another head gasket, I used MLS gaskets, copper sealant, and did the work myself (save decking the heads). Anyone have any thoughts on this one?

[Whopper]

Probable issues. Air bubble in the system so try and burp it. Bad thermostat in the water pump.

[doublechaz]

+1 to needed it burped.

[Setnev]

I used MLS gaskets, copper sealant, and did the work myself (save decking the heads). Anyone have any thoughts on this one?

 

I have seen so many builds ruined by copper sealant. Heres why:

 

1. Its purported to seal the imperfections between the head gasket and block, but with Subarus, there is no room for error on the imperfections that copper sealant is meant to patch, which is why it's so common for Subaru blocks and heads to be milled during gasket replacement.

2. Subaru blocks and heads have a max tolerance of 0.05mm of warpage before they need to be milled. Comparable 4 cylinders are like .5mm, in fact I did a HG on a Volvo S40 that had a tolerance of .7mm.

3. MLS gaskets have a thin coating on them. Introducing a foreign material on that coating will cause them to not seal correctly. Copper spray is easily contaminated if the deck surface is not absolutely prepped perfectly. I'm talking every trace of water and oil off the sealing surface and prepped with isopropyl alcohol before application. Even then, a spray pattern can be uneven to the point where it creates artificial warpage.

 

I hate to throw this at you like that, but I've done countless Subaru engines and never once used copper spray. Remember, the early EJ25 motors suffer more from internal head gasket failure than external. My vote is, bleed the cooling system with a spillproof funnel and pray that's all it is. If you hear sloshing in the heater core after bleeding the system, that engine might be coming back out for a redo. It sucks, but it happened to me 1500 miles after I rebuilt the motor in my car when I didn't check the block deck and used the wrong gaskets on my rebuild. Lesson learned. 10,000 miles later and still going strong with a new block and heads and good MLS gaskets.

[Das Ninjabot]

Gross. I did rub down the block and heads with alcohol before I put the sealant on but I was not aware of how much smaller the Subaru tolerances are.

 

When I checked the next day I was down about 3/4 quart of coolant. It appears as if I may have a leak at the coolant crossover, thought I changed those O rings, but can't be sure. There is a small amount of puddling on the top of the block. Thankfully it was not in the crankcase because the oil was low as well. I'm not happy about either of those things, but at least it doesn't appear on the surface that the coolant made its way into the oil via internal I'll keep an eye on this, I really do appreciate the heads up Setnev, wish I had seen some mention of this before I finished the rebuild.

[Setnev]

Well it certainly looks like you dodged a bullet on that one. My brand new motor loses oil, but I think I have a leak on the oil separator plate. Even though you replace the o-rings for the crossover, I found they leak shortly after. This last time, I went to True Value and got similar diameter o-rings that were slightly thicker than stock and applied a very thin bead of RTV on the mating surface of the crossover pipe. No leakage yet.