[3_Dog_Knight]

About the car:
1996 Subaru Outback 2.5
Dual overhead cam phase 1 engine
Old factory gasket was not layered steel
120,000 miles



Hi, I just did a HG repair and there is no change in how it acts. I am starting to wonder if I have a cracked head or some other problem but I really want input and ideas if you please.

History: A few weeks ago I got in my outback and drove to the mail box at the end of the road. By the time I got to the mail box I could barely see behind me from all the white smoke coming out the exhaust. I'm not sure if it did this the last time I drove it or not because it was during a storm. Anyway, I immediately shut it off and pulled the radiator cap off to see if the coolant was low. Sure enough I couldn't even see any in there. For many months before that I've had a steady stream of bubbles coming up through the reservior tank but no smoking till then and the car has never overheated to this day. So I trailered it to a friends shop and replaced the head gaskets.

Finished replacing the head gaskets today and here is what happened after starting it. At first it did not smoke at all, then as the exhaust pipes heated up it was not only smoking like a freight train but the exhaust was practically blowing water out of the tail pipe. I wasn't completely alarmed because I figured there could be quite a bit of coolant still pooled up somewhere in the exhaust that it needed to burn out. I drove it 5 miles and by the time I got to my destination there was zero smoke coming out out of the exhaust and it was dry. Shut it off, let it cool, then started it back up and it did the same thing all over again. No smoke, then smoked like a freight train as the exhaust got hot, then zero smoke.

I also had a slight exhaust leak so I lifted it back up in the shop and pulled the exhuast manifold away from the cylinder heads. I looked up into the exhaust ports on the drivers side and it seemed a bit damp and passenger side was dry. This is also how it was before I put the new head gaskets in. I should also mention it has never acted the slightest bit hyrdo locked or ever sputtered on starting it up. It always runs smooth.

Anyway I decided I might as well be optimistic so I sealed up the exhaust leak with ultra copper and drove the car for 35 minutes hoping all residual liquids would burn out of the exhaust and it would be good from then on. Well I let it cool for 15 minutes and it did the same thing all over again next time I started it. No smoke, freight train, no smoke. There are however no longer any bubbles coming up through the reservior. But the smoking problem does not seem to be getting the least bit better.

Anyone every run into anything like this or have any ideas of what is going on?

 

[RobLog]

Have you checked the radiator to see if you are losing coolant? If not, it has to be as you said: residual coolant in the exhaust. Try drilling a small hole in the lowest part of the muffler to see if any liquid drains out ... it can hold a lot.

 

[1 Lucky Texan]

if the exhaust smells like coolant, and you are not now losing it from the radiator, i gotta go with Rob's idea.

but, I'm wondering if you could have some weird PCV or leaking valve seal - type issue and you're seeing smoke from oil (should be bluish and not smell like toasted marshmallows)

 

[3_Dog_Knight]

if the exhaust smells like coolant, and you are not now losing it from the radiator, i gotta go with Rob's idea.

but, I'm wondering if you could have some weird PCV or leaking valve seal - type issue and you're seeing smoke from oil (should be bluish and not smell like toasted marshmallows)

It's so funny you mentioned a smell like toasted marshmallows. That's exactly what I told my friend it reminds me of. So what are you saying that smell indicates, I'm confused.

 

[1 Lucky Texan]

that is coolant. Typical coolant is glycol based, it has a syrupy sweet taste and smell.. (PLEASE - never leave any where a dog cat or kid could drink ethylene glycol based coolant, even a little destroys kidney cells. deadly stuff)

so, if it still smells like that, you're still burning coolant. And, very important, if you're not losing it from the radiator or overflow bottle, it must still be leftover in the exhaust system as roblog mentioned.

 

[3_Dog_Knight]

Have you checked the radiator to see if you are losing coolant? If not, it has to be as you said: residual coolant in the exhaust. Try drilling a small hole in the lowest part of the muffler to see if any liquid drains out ... it can hold a lot.

I've had to add a little coolant every time it has cooled down but so far I've just attributed this to air pockets in the cooling system when I refilled it. I just have a really bad feeling that the pressure in the cooling system is slowly bleeding into the exhaust after shut down somehow and then burning out of the exhaust manifold when it gets hot enough. But I can't imagine how that could happen unless the head is cracked. But I never hear anyone mention cracked heads on these things, only bad HG's and warped heads. Anyway I guess I shouldn't jump ahead of myself. I like the drilling the hole idea thank you. I'll do that and drive it a few more times keeping a close eye on the coolant level and engine temp and report my findings.

 

[1 Lucky Texan]

I've had to add a little coolant every time it has cooled down but so far I've just attributed this to air pockets in the cooling system when I refilled it. I just have a really bad feeling that the pressure in the cooling system is slowly bleeding into the exhaust after shut down somehow and then burning out of the exhaust manifold when it gets hot enough. But I can't imagine how that could happen unless the head is cracked. But I never hear anyone mention cracked heads on these things, only bad HG's and warped heads. Anyway I guess I shouldn't jump ahead of myself. I like the drilling the hole idea thank you. I'll do that and drive it a few more times keeping a close eye on the coolant level and engine temp and report my findings.

you could pull and compare spark plugs. The 'wet' cylinder's plug would probably look different - if you're still burning coolant.

I suppose you could drop the axle-back and try to dump it out or shake it to determine if it's holding any coolant.

 

[3_Dog_Knight]

Have you checked the radiator to see if you are losing coolant? If not, it has to be as you said: residual coolant in the exhaust. Try drilling a small hole in the lowest part of the muffler to see if any liquid drains out ... it can hold a lot.

I know my explanations are a little long so thanks in advance to anyone who powers through them. I just want to be clear on what I'm seeing and thinking.

It's loosing a lot of coolant from the radiator and doesn't even begin to draw coolant back from the reservoir. Last time I had the exhaust manifolds off and the engine was still cooling down I heard a strange sound coming from inside the damp exhaust port. Reminded me of the sound a one way reed valve makes when drawing air threw it slowly in reverse. Kind of a buzzing sound, like maybe layers of the gasket were vibrating from air being sucked through them as the engine coolant contracted. The reservoir hose is not leaking and is properly positioned so it's obvious the system has found a path of less resistance.

It still smokes like a freight train shortly after start up. But it kind of baffles me how it could loose so much coolant and not smoke the slightest bit after driving it 5 or 10 minutes. Maybe expansion of the hot engine parts is temporarily closing up a gap caused by a warped head? I wish I had gone ahead and machined them like many people here have said, but I just thought since it never seemed to overheat, that there should not be any warpage and they seemed straight when using a straight edge on it.

It seems like with this engine design there are not many ways coolant can end up in the exhaust? I mean there are no coolant passages in the intake manifold like a small block v8 would have except for the small passage at the throttle body for preventing frost. I can't imagine that could be related to my problem. Cracked or porous head or block seems feasible but not as likely as a warped head. Is there any other potential paths for coolant to get into the cylinders or exhaust?

This job is really starting to cut into my budget badly. I'm even toying with the idea of machining the heads but reusing the new Subaru head gaskets I just put on it. They are OEM multilayer steel. I didn't even see anything on them that looked like a sealant that would be damaged in instalation, just layers of spring steel so I'm wondering if they might still be good as new with only 45 minutes of driving on them and the temp gauge never deviating from the half way mark. I mean the tightening process itself practically has you installing the gasket twice when you torque it, loosen it all the way, then retorque/angle it back down. I'm kind of thinking out loud here and hoping for some good ideas and advice. Obviously dumping another $100 into head gaskets would be the safer bet, but with being out of work right now I have more time than money. But if someone could give me a good explaination of why this is not likely to work, or how in their own experiences it didn't, then you could be doing me a favor in talking me out of it.

Thanks again for responding. I hope whatever I learn from this experience will help others as well.

 

[RobLog]

I doubt if anyone will tell you it's OK to reuse the HGs ... but if I was really short on money, I'd do it and take my chances. If it didn't work out, then the only thing lost is time. There has been discussion of whether or not to coat the HGs with the copper spray made for this purpose with people both for and against. I wouldn't normally advise using it, but in the case of trying to reuse a HG, I'd probably use the spray sealer.

If only one side is leaking, then only redo that side ... saves both time and money.

For future reference for anyone who may read this thread, I always recommend having the heads checked for flatness by a machine shop or other competent person. The shops I've worked with will do this for a minimal charge and usually for free.

When my HGs went, I cooked the crap out of my engine. It was so bad, the engine died on me three times because it was so hot (I was in a semi-emergency situation and had to keep going). Both heads were warped but not cracked, and I paid the shop extra to run tests that really check for cracks. Besides planing, two exhaust valves needed a little work. 25,000 miles later, the engine runs great.

 

[3_Dog_Knight]

UPDATE: Cause has been discovered.......

It turns out the problem has to do with a cracked or porous cylinder head. I was preparing to pull the heads back off to have them pressure tested and resurfaced. This time when I pulled the exhaust manifolds away from the heads, it wasn't just a little damp, it actually had a constant drip of coolant out of the exhaust port for the front cylinder on the drivers side. I think it may have been flowing more this time because I actually had the radiator cap off. Anyway, upon closer inspection I saw a very strange sight. As I looked up into the port there appeared to be something shiny wedged between the valve stem and the port that looked like part of a sewing pin. It was shiny like one and the same diameter. Well I took a long screw driver and put it up to the 'object' in question and miraculously it turned out to be a tiny stream of coolant squirting out of the wall of the port and directly into the side of the valve stem. I can't get over how much it looked like a shiny metal pin. Anyway, now with my screwdriver blocking the path of the tiny stream, the coolant was now running down my screwdriver.

Because the stream of coolant was so perfectly cylindrical, I'm guessing there is a pin hole there due to a poorly manufactured head having porous aluminum in that area. It seems like a crack would not produce such a perfect little stream but I'm not certain. Also, could be my imagination but the surface metal didn't seem as smooth in the area of the leak.

I suspect this is not the first time this little pin hole has sprung a leak judging by a thin layer of plastic looking material I noticed when changing the water pump. I think at some point in this car's life it has done this before and someone fixed it with a sealant. Well, my budget has been wrecked by this project all ready so I'm giving Bar's sealant a try. It already seems to be working. I ran the car for half an hour and pulled the ex manifold back off and there was no sign of leakage. Also the smoking and coolant usage has either stopped or slowed to a crawl. I'm going to continue to monitor the levels closely. I've owned the car for the last 20,000 miles of it's life, so if this was done to fix it before I'm hoping to get close to another 20,000 before I have to find a more permanent solution.

Anyone here ever have any similar experiences and used something like Bar's? If so how successful was it?

 

[3_Dog_Knight]

Just a quick update. I tried the Bar's for a while. As some of you said, it was not a fix, even for the crack/pores I experienced. It was however a good band aid to keep me going for a bit while adding water occasionally.

Then I found a new head and installed it. Smoking quit after the very first drive once it burned out of the exhaust. Have put a couble thousand miles on the car and all coolant consumption has stopped. The coolant reservoir level went down maybe an inch after 2 drives then stabilized once all the air purged itself from the system. Gave me a bit of a scare after the first 100 mile trip because I could smell a sweet smell when I stopped, but I think maybe it was still burning out of the muffler a little which didn't get hot enough to burn out before until this long trip. I went on a 200 mile trip after that with no smell and the level has not dropped from the full mark the slightest bit. So it's looking good!

 

[Storkiln]

Really great news that you were finally able to fix it. This was quite the adventure. Thanks for documenting and updating what seemed like a freak of nature issue.