2002 Legacy Overheating After New Radiator

[legacyofbrutality]

So here's the scoop, I just replaced the radiator in a 2002 Subaru Legacy base model with AT. The original OEM radiator cracked just below the top neck, causing the car to start overheating, hence the replacement. I replaced it with a parts store replacement radiator, not an OEM one. I didn't change any hoses, thermostat, etc., just the radiator and a coolant fill. I know, I probably should've done a new thermostat too, but needed the car back on the road asap, which now hasn't happened.

 

I filled the radiator through the fill neck, topped off the overflow tank and let the car idle up to normal temp. I made sure to fill it until it didn't draw anymore in. First test drive, within a mile or so the gauge starts climbing to 3/4 or so. Pulled over, let it cool down and limped it back home. I did some research, thinking that there might be air locks, so I tried various methods of burping - squeezing the upper and lower hoses while running with the radiator cap off, letting the heat run full blast while do so... nothing seems to help. I'm not sure if there is still air in there or if I have a problem with the thermostat possibly?

 

I'm solely basing all of this off the stock temp gauge... At idle, the car never goes over the 1/2 mark. It only starts to soar when you take it for a drive, and it goes up within a mile give or take. The heat is very hot and working great, the lower radiator hose never warms up at all??? Not sure if it should? Take into consideration that it's in the lower 30's outside. Both cooling fans kick on and off at idle as well. I've not seen it draw anything from or overflow into the tank at all. The coolant is not boiling, it's not shooting out of the neck or doing much of anything with the rad cap off. I've tried just about everything to make certain there is no air trapped. So I'm kinda stumped at this point... My next plan of action is to drain it all and replace the thermostat, but I thought I'd seek some advice before doing so.

Edited March 3, 2016 by legacyofbrutality

[CNY_Dave]

The lower hose will be cold until the thermostat opens, then it gets hot- so could be t-stat.

 

Heater core connected and not blocked or blocked off? If coolant can not flow through these lines the t-stat might never open.

 

Try getting the front end high and burping it.

 

Radiator fans come on?

[legacyofbrutality]

The lower hose will be cold until the thermostat opens, then it gets hot- so could be t-stat.

 

Heater core connected and not blocked or blocked off? If coolant can not flow through these lines the t-stat might never open.

 

Try getting the front end high and burping it.

 

Radiator fans come on?

 

Nothing is modified from the factory, so no hoses or anything are blocked off. Both fans come on like they should.

[Numbchux]

Mileage on the car? Head gasket history? Any issues with it overheating before the failure?

 

Thermostats virtually never fail shut (I say "virtually" to give the benefit of the doubt. I've tested many, and never seen it happen, and never heard of or seen a situation where a new thermostat solved an overheating issue and been convinced that it wasn't just the fresh coolant/burping). IMO, replacing your thermostat would be a waste of energy and money.

 

Good heat means you've got good coolant circulation, which pretty much guarantees that you don't have an air bubble.

 

 

First thing I would do, is get my hands on an OBDII scanner, just to get a second opinion on the temperature.

 

But, the fact that it doesn't run warm at idle, and it does under load sort of screams head gasket to me. It's possible the overheating from the failed radiator caused head gasket failure, or even head warpage. Next thing I would do is get a chemical "block tester" kit (a little clear plastic container that you put in the radiator neck, and fill with a blue chemical, and draw air through it from the half-empty cooling system, and if there are combustion gasses in the coolant, the chemical will change yellowish. I know when I worked at AutoZone, we had the tool for rent for free, and the chemical was like $7 for a bottle) and see if it that is the issue.

[legacyofbrutality]

Mileage on the car? Head gasket history? Any issues with it overheating before the failure?

 

Thermostats virtually never fail shut (I say "virtually" to give the benefit of the doubt. I've tested many, and never seen it happen, and never heard of or seen a situation where a new thermostat solved an overheating issue and been convinced that it wasn't just the fresh coolant/burping). IMO, replacing your thermostat would be a waste of energy and money.

 

Good heat means you've got good coolant circulation, which pretty much guarantees that you don't have an air bubble.

 

 

First thing I would do, is get my hands on an OBDII scanner, just to get a second opinion on the temperature.

 

But, the fact that it doesn't run warm at idle, and it does under load sort of screams head gasket to me. It's possible the overheating from the failed radiator caused head gasket failure, or even head warpage. Next thing I would do is get a chemical "block tester" kit (a little clear plastic container that you put in the radiator neck, and fill with a blue chemical, and draw air through it from the half-empty cooling system, and if there are combustion gasses in the coolant, the chemical will change yellowish. I know when I worked at AutoZone, we had the tool for rent for free, and the chemical was like $7 for a bottle) and see if it that is the issue.

 

Mileage is 170K. I haven't ruled that out, but there are no real signs of a head gasket problem, at least from what I've gathered. Old coolant had no oil residue, definitely none in the new that I put in (yet). Same with oil, no coolant mixed in it. Visual inspection at the gaskets, there is no signs of seepage. There's no pressure build up with the rad cap off when you rev the motor or anything.

[1 Lucky Texan]

if you get a t'stat, get the OEM 'style'.

 

or, pull it and test on a stove top - do NOT let anyone drink even a drop of coolant. Animals will lap it up if you leave some outside - it will kill them.

[legacyofbrutality]

if you get a t'stat, get the OEM 'style'.

 

or, pull it and test on a stove top - do NOT let anyone drink even a drop of coolant. Animals will lap it up if you leave some outside - it will kill them.

 

Yes, I plan to buy an OEM if I do decide to replace it. The more research I do, the more I think that it is stuck closed. As I mentioned, the lower radiator hose never warms up at all. I mean it stays completely cold after  idling for 20+ mins, even after the gauge indicates that it's overheating coming back from a test drive. It's cold as can be.

[idosubaru]

Replace tstat and radiator cap with Subaru OEM.

 

we can't conclusively determine what happened yet:

1.  if it was the radiator - and no additiona damage.  then it needs burped, keep trying. 

2.  if it was something else that caused it to overheat - which compromised the radiator - then  that opens up a few other possibilites. 

3.  it was the radiator but the overheating damaged something else then there could be a bad radiator and an additional failure

that's unlikely but if other things aren't adding up then we can't consider the "radiator" diagnosis foolproof. 

 

hopefully it's burping or the tstat but it does sound close to a headgasket failure.  i've yet to see an actual failed EJ Subaru thermostat (though i've seen them mis-identified as the culprit many times).  a lot of thermostat "fixes" woudl end up being some other issue if people/posters woudl follow up with threads more often.  they do fail so certainly test it but i've seen far more blown headgaskets than bad tstats - like an order of magnitude diffrenence.  they fail so rarely i'm not in the habit of changing them preemptively like i used to years ago.  all of my 250k and 200k Subaru's have the original Tstat as far as i know.  not worried about them one bit. 

 

your two observations are not failure modes of the engine in question:

EJ engines do not mix oil and coolant   (in the rare cases they do - they're abused to that point - it's already known/obvious the HG's are bad)

when those gaskets fail internally they don't show external leak signs - they push exhaust gases into the coolant and cavitate the flow. 

 

the typical diagnosis of this type of Subaru headgasket failure - is that about a minute or two after it starts overheating your cabin heat will grow cold.  if you're not seeing this, you may not be letting it overheat enough for it to blow the latent heat out and start to cool down in the cabin. 

[legacyofbrutality]

Replace tstat and radiator cap with Subaru OEM.

 

we can't conclusively determine what happened yet:

1.  if it was the radiator - and no additiona damage.  then it needs burped, keep trying. 

2.  if it was something else that caused it to overheat - which compromised the radiator - then  that opens up a few other possibilites. 

3.  it was the radiator but the overheating damaged something else then there could be a bad radiator and an additional failure

that's unlikely but if other things aren't adding up then we can't consider the "radiator" diagnosis foolproof. 

 

hopefully it's burping or the tstat but it does sound close to a headgasket failure.  i've yet to see an actual failed EJ Subaru thermostat (though i've seen them mis-identified as the culprit many times).  a lot of thermostat "fixes" woudl end up being some other issue if people/posters woudl follow up with threads more often.  they do fail so certainly test it but i've seen far more blown headgaskets than bad tstats - like an order of magnitude diffrenence.  they fail so rarely i'm not in the habit of changing them preemptively like i used to years ago.  all of my 250k and 200k Subaru's have the original Tstat as far as i know.  not worried about them one bit. 

 

your two observations are not failure modes of the engine in question:

EJ engines do not mix oil and coolant   (in the rare cases they do - they're abused to that point - it's already known/obvious the HG's are bad)

when those gaskets fail internally they don't show external leak signs - they push exhaust gases into the coolant and cavitate the flow. 

 

the typical diagnosis of this type of Subaru headgasket failure - is that about a minute or two after it starts overheating your cabin heat will grow cold.  if you're not seeing this, you may not be letting it overheat enough for it to blow the latent heat out and start to cool down in the cabin. 

 

I understand about how the ex. gases get pushed into the coolant, but wouldn't this cause an overflow, at least into the tank or pressure forcing coolant out of the neck with the rad cap off? If so, I have none of that to speak of. As far as heat inside the cabin, it's great. I've not experienced a loss of or any change really. Once it's warmed up, the heat blows warm as it should. Even on the test drives where the temp raised to 3/4, at one time almost to the red as I pulled off, the heat inside never changed. In fact, I kept it running full blast the whole time driving, hoping that might keep the temps down a bit and/or circulate the coolant more in case of air.

[ocei77]

Seems like classic air in the system.

You need to fill via the upper radiator hose first, then when it starts to overflow, clamp it and fill with the engine idling.

Do a search for burping the system.

 

O.

[legacyofbrutality]

Seems like classic air in the system.

You need to fill via the upper radiator hose first, then when it starts to overflow, clamp it and fill with the engine idling.

Do a search for burping the system.

 

O.

 

Trust me, I've read so much on this it's driving me crazy. I registered here as a last resort for possible help, because of so much conflicting stories, solutions, etc. I can easily drain the coolant and refill it that way, BUT is that necessary? Can it not be burped/bled after the fact if I didn't fill it that way to begin with?

[idosubaru]

no need to drain the system, you're fine to burp without completely draining the system.   i've never filled via the upper radiator hose, definitely not necessary. 

 

replace tstat, rad cap, and heck while it's apart i might even test the radiator for flow just to make sure it's not clogged internally due to some rodent or bugs while it was on the shelf or something weird if you're symtpoms are that confusing. 

 

so far the 3/4 on the guage has me pausing any diagnosis and i don't think it's the headgaskets - but i wouldnt' completely rule them out either. exhaust gases pushing into coolant wont' cause an overflow/bubbling - that's only on really bad ones.  it can be very tiny sporadic bubbles and on initial onset may even be hard to spot or only happen ocassionally, not continuously. i've seen them only do it at highway speeds, only with the A/C on, or only every few months be symptomatic.  they often slowly get worse over time.

 

but again - the 3/4 of the way up the gauge makes me hesitate on any diagnosis without testing the tstat and radiator flow (in general - i realize yours is new/presumably good)

[legacyofbrutality]

no need to drain the system, you're fine to burp without completely draining the system.   i've never filled via the upper radiator hose, definitely not necessary. 

 

replace tstat, rad cap, and heck while it's apart i might even test the radiator for flow just to make sure it's not clogged internally due to some rodent or bugs while it was on the shelf or something weird if you're symtpoms are that confusing. 

 

so far the 3/4 on the guage has me pausing any diagnosis and i don't think it's the headgaskets - but i wouldnt' completely rule them out either. exhaust gases pushing into coolant wont' cause an overflow/bubbling - that's only on really bad ones.  it can be very tiny sporadic bubbles and on initial onset may even be hard to spot or only happen ocassionally, not continuously. i've seen them only do it at highway speeds, only with the A/C on, or only every few months be symptomatic.  they often slowly get worse over time.

 

but again - the 3/4 of the way up the gauge makes me hesitate on any diagnosis without testing the tstat and radiator flow (in general - i realize yours is new/presumably good)

 

That's kinda the game plan at this point... change the thermostat, in which case I'll have to drain the coolant anyway. Yes, radiator arrived new yesterday and in a rush I installed it without thinking to check it for anything obstructing. Definitely possible. I guess after draining to install the thermostat, I can then check it by running some water through it. From there, I will try the radiator cap... I just want to try and isolate the problem by doing 1 thing at a time.

[1 Lucky Texan]

could've had some paper toweleing, foam plugs or other material in the openings I guess.

[legacyofbrutality]

could've had some paper toweleing, foam plugs or other material in the openings I guess.

 

I'm going to check it over way more closely this time around, definitely not ruling it out. It came in a cardboard box, strapped with another piece around it, no foam, peanuts or anything like that in the box.

[Bushwick]

Conventional thermostats fail closed and they do eventually fail. If they somehow start sticking slightly open, the engine temp will be cooler and possibly fluctuate between 1/2 mark at idle, then get cooler as you move.

 

They make a fail-open style which will stay open and cause engine to run cooler (like 1/4 way) in event of failure, but in my experience they are junk, as I've had 2 of those fail almost exactly 1 year after buying, and too coincidental to have a year apart t-stat fail open from different build dates.

 

As others pointed out, trapped air can be nuisance. VW's 1.8t engine comes to mind of a lousy system that can be very difficult to get the air out. I had one that would creep past the 1/2 way mark and it just wouldn't get the air out in a timely fashion. Forget how I actually did it (I might have left cap loose and worked the upper and lower hoses with front end slightly on an incline until it let the stubborn bubbles come to surface.

 

In your case, $5 for a brand new t-stat is well worth the piece the mind and rules out the old one. Then it'll be a matter of burping the system if it still gets hot, unless there is an issue with the water pump or a clog.

[Fairtax4me]

I've had new radiators come in the box with black plastic plugs in the hose outlets. Make sure the outlets are clear.

 

Thermostat would be my next guess, especially since you have good heat.

[legacyofbrutality]

No restrictions in the radiator, but definitely a bad thermostat. I took the old one out, put it in a pot on the stove with a thermometer. Went well past the 172 degree point and never opened. I let the water fully boil beyond 200 and it doesn't open even the slightest. I'm going to pick up a new one today, reinstall and hopefully that will solve the problem.

[Numbchux]

No restrictions in the radiator, but definitely a bad thermostat. I took the old one out, put it in a pot on the stove with a thermometer. Went well past the 172 degree point and never opened. I let the water fully boil beyond 200 and it doesn't open even the slightest. I'm going to pick up a new one today, reinstall and hopefully that will solve the problem.

 

Holy crap. That's a first for me! Thanks for actually testing it, too.

 

Anything stuck in it that might be preventing it from opening? 

[legacyofbrutality]

Holy crap. That's a first for me! Thanks for actually testing it, too.

 

Anything stuck in it that might be preventing it from opening? 

 

I squeezed the spring with a pair of pliers and it released enough pressure for the valve to release, that's the only way it would move. I thought maybe getting it unstuck that way, it might help it to work again somewhat... but I put it back in the boiling water and nothing, still stuck closed. It didn't looked really corroded or anything, so I'm not exactly sure what the deal is with it.

[desertsubaru]

Get OEM or Stant super stat. Do not use cheapo stats.

[DaveT]

+1 to desertsubaru

[legacyofbrutality]

Installed the new thermostat, refilled coolant and burped it. Took it for a test drive, running it easy and hard a bit for about 10 miles. I then let it set and idle for a good 15 mins or more after the drive. I still have really good heat and noticed no issues at all. The gauge hasn't risen over the half mark at all, so hopefully the problem is resolved now.

Edited March 5, 2016 by legacyofbrutality

[Fairtax4me]

Thermostats go bad every now and then. They don't usually "look bad", they just decide they're either not gonna open, or not gonna close.

 

Glad you got it fixed!

[bmaness]

Like the chap above, I joined the forum because I'm give out on answers...

 

The machine overheats at idle or in traffic is the problem. Neither radiator fan seems to ever run. Put a new oem thermostat in today and no change. Have not direct-wired the fans to see if they work but feel sure they will. The 2 x 20 amp fuses (1 inside, 1 under hood) are good. A cursory check of the relays with a vom shows good. I'm not dropping any water, and there seems to be circulation through the system.

 

Additionally I should say the parking lights are blinking intermittent with the machine turned off. I think this is tied to the security system but am again unsure.

 

What am I missing?