head gasket replacement

[rcweston]

Head gasket replacement
I’ve had 2 major water leaks. One with the timing cover gasket and a second with the lower intake manifold gasket (both corrected), now I have another.
I have conflicting results. 1. Water vapor in the exhaust (on a hot day). 2 . Bubbles in the radiator neck when full. (Should have tested for combustion gases before I took it apart) 3. Loosing coolant continually, and 4. Matching compression at all cylinders (very close to 120 psi on all cylinders) .
I shouldn't see white exhaust and I shouldn't have bubbles in the radiator, both could indicate a head gasket failure from all the studding I have done. But conflicting results, I should also see bad compression in at least one of the cylinder if not two of them. I see good compression on them all. I'm deep enough in this engine with all that has been done to this point, So I'm taking it apart again, I'm doing the head gasket job.
Upon disassembly, Head gaskets looked great. No warp age detected with .002 gauge on ether head.
At this point I'm still waiting for parts, but have a couple of questions.
The head bolts were extremely hard to break free with a 28 inch breaker bar, but removed ok. It seemed like they had some sort of thread locker on them, sort of a sticky thread all the way coming out. I cleaned the end of one with a wire wheel and could see a big difference.
Now, at re installation;
Some info states that you need to lubricate the head bolt threads with a little oil, other say to use thread locker, while still others say to put them in dry. My Gm manual doesn't say one way or the other, it just says to torque them to 44 lbs plus 95 degrees. Which way is correct, or does it depend on the manufacturer of the bolts? I have ordered new bolts.
Also I would like to make sure the blot holes and threads are clean before the new ones are installed. I measured the tread as a metric 1.5, but I can't find the diameter measurement. It is too big to fit a 10 x 1.5 nut and wobbles in a 12 x 1.5 nut. I have no reference for an 11 and can't seem to find one. I would like to chase the threads but need the exact size before I would even attempt it.
Any other tips from your experience would be greatly appreciated.
Thanks for any help.
Rex

[Tech II]

Need to check cylinder walls for possible cracks....may even have to have the heads sent out and checked, especially if you see nothing wrong with the head gaskets....on this engine, I usually find gasket problems on #1 cylinder......look at your piston tops for a "washed off" look, for the problem cylinder...

As far as head bolts, make sure you "chase" the threads in the block....use new head bolts....I would oil the threads, and lay the bolts on a rag, letting the excess drip off, then a quick wipe before installation....they are torqued in a specific sequence from the inside, outward........I torque them at half the torque on the first pass, full torque on the second, recheck torque on third, then the final degree turn.......

[Schurkey]

What engine? Some engine families have known issues with cracked castings--heads, usually.

The wet cylinder should be free of carbon, like it was steam-cleaned (which it was.)

The service manual will tell you what to use on the bolt threads--oil is most common, some engines need a thread sealing compound. Aftermarket (high-performance) bolt kits may suggest a different torque and special moly-based lubricant, or specific brand of thread sealer.

[rcweston]

Originally Posted by Tech II

Need to check cylinder walls for possible cracks.

Thanks for the info.
the cylinders look remarkably good. no apparent scratches or markings, barely a mark at the top of the piston travel. Nothing at all that looks or feels like a crack using a bright light.

Originally Posted by Tech II

. on this engine, I usually find gasket problems on #1 cylinder......look at your piston tops for a "washed off" look, for the problem cylinder.

..
None of the cylinder looked as if they had any washed marks. all looked like they had a small amount of deposites on the heads and on the piston tops. the intake valve was lighter than the exaust valve and the piston heads had a light carbon (black) look to them. maybe just as important is that they all looked the same

Originally Posted by Tech II

As far as head bolts, make sure you "chase" the threads in the block....use new head bolts....I would oil the threads, and lay the bolts on a rag, letting the excess drip off, then a quick wipe before installation....they are torqued in a specific sequence from the inside, outward........I torque them at half the torque on the first pass, full torque on the second, recheck torque on third, then the final degree turn......

I've chased the thread bolt holes in the block. They looked much cleaner than the old bolts looked. I first blew them out with compressed air. got a little debris from each hole but nothing that looked gooey or sticky. very small particles less than 1/4 the size of a head of a pin and smaller.
After chasing them (11mm x 1.5 mm) I got a powdery substance. but very very little, almost undetectable in some holes.
One last thought that I had over the week end. could the water be coming from the neck of the air intake were the water is connected to the area around the mas air flow connection?

[rcweston]

Originally Posted by Schurkey

What engine? Some engine families have known issues with cracked castings--heads, usually.

The wet cylinder should be free of carbon, like it was steam-cleaned (which it was.)

The service manual will tell you what to use on the bolt threads--oil is most common, some engines need a thread sealing compound. Aftermarket (high-performance) bolt kits may suggest a different torque and special moly-based lubricant, or specific brand of thread sealer.

Thanks Schurkey for the reply.
This engine is a 3.1L.
It has about 190K on it and has preformed great over the years. The only real problem I have seen has been the coolant leaks from the different gaskets. 6 or 7 years ago when I replace the first intake manifold gasket set, I changed from Dexcool to regular green antifreeze.
Then just last month or so I changed a leaking timing chain gasket and re-chainged the Intake manifold gasket for a second time.
I do set a code almost every time I drive over a mountain (steep inclines for a long ways) that takes more than five minuets. I don't remember the code number, but it indicates the Mass Air flow. After a couple of short drives, it always goes away and doesn't return until the next trip over the mountains.
This has been the only problems with this engine.
As for the head and cylinders, none of them looked washed. They all look the same.
It's frustrating not being able to find a real problem that can be pinpointed!
I've received all the parts now, have cleaned the engine up well and will start to put it back together today, despite the manual saying nothing of lock sealant or coatings on the head bolts. I'll slightly oil them and torque them as has been suggested.
Thanks for your help.
I'll return in a few hours for any additional responses.
Rex

[rcweston]

Originally Posted by Tech II

I usually find gasket problems on #1 cylinder......

As far as head bolts, make sure you "chase" the threads in the block....use new head bolts....they are torqued in a specific sequence from the inside, outward........I torque them at half the torque on the first pass, full torque on the second, recheck torque on third, then the final degree turn.......

The repair is completed!!
Thanks everyone for all the advice I received with the three water leak troubles in my post over this last 45 days. 1 timing chain cover 2. lower intake manifold, and 3 head gasket.
.... I did not find anything that even looked like a leak on the head gaskets but replacing them stopped all the with sweet smelling exhaust.
clean and prep work took 4 time as much time as the actuacl replacement work, but i feel it was worth it.
It took about four days to get the water to level out and stop needing additions, even after bleeding the system. It has been 10 days now and everything seems to be working well with short (less than 5 miles) to medium(less than 15 miles) drives. Time to take it out for an hour drive to regain confidence that all is well!
Thanks again for your help
Rex

[Schurkey]

If the engine has been swallowing coolant, you can about guess that the O2 sensor is poisoned, too. Would be worth testing with a scan tool.

[rcweston]

Originally Posted by Schurkey

If the engine has been swallowing coolant, you can about guess that the O2 sensor is poisoned, too. Would be worth testing with a scan tool.

Thanks Schurkey, it has swallowed a lot of coolant over the last 8 months, I would guess more than 4 or 5 gallons. The o2 sensor would have been a breeze to remove with everything apart, but that's not the case now.
I don't have a sacn tool but do have a very good multimeter and a code reader. Could it be done using these tools?

[Schurkey]

Originally Posted by rcweston

Thanks Schurkey, it has swallowed a lot of coolant over the last 8 months, I would guess more than 4 or 5 gallons. The o2 sensor would have been a breeze to remove with everything apart, but that's not the case now.
I don't have a sacn tool but do have a very good multimeter and a code reader. Could it be done using these tools?

The multimeter will work, the code-reader will be useless unless the sensor is so shot that it meets the on-board criteria for setting a code.

[rcweston]

Originally Posted by Schurkey

The multimeter will work, the code-reader will be useless unless the sensor is so shot that it meets the on-board criteria for setting a code.

what am i looking for with the voltage?
I know from past experience with my daughters lumina that it is a four wire unit. one lead is 12 volts. one is ground. these two operate the heater element of the sender. The other two leads must be the sensing leads, but I don't know what they should measure. are wee looking for a voltage signal, a ground signal, or could this be sophisticated enough to have a frequency signal of some type?

[Tech II]

What you are looking for is rapid voltage changes from 0.1 or 0.2 to 0.8v DC....you can go to AutoZone, and they can plot your O2 waveform(precat) to see that is it changing ok.......

[Blue Bowtie]

The best way to test is with the sensor removed. Get voltage readings from the sensor element while the tip is hot and immersed in a propane flame, then remove the flame and continue reading voltage while the tip is still hot. The propane flame should be fairly lean and contain oxygen, and that should produce a fairly high sensor output.

[Tech II]

Have never seen an O2 sensor tested like that......

But wouldn't it be easy to just back probe the O2 sensor harness, and create a rich/lean condition at the intake?

Of course your test only shows extremes......it would not show a "lazy" O2 sensor that is slow to respond.....

Rule of thumb......if the O2 sensor has a lot of mileage on it, and there has been a coolant leak, replace it....It may function, but if lazy or slow to respond, will effect fuel/air mixture efficiency.....

[Schurkey]

I learned the VERY hard way that sometimes it's just better to stuff a new O2 sensor in, than to dick with testing it.

I spent two years on and off, trying to diagnose a lean condition at highway speed. I thought I'd considered everything--fuel pressure, injector cleanliness (not plugged) "false air" in the exhaust from exhaust leaks or improper air injection operation; misfire, etc.

The problem continued to get--gradually--worse. At low speed, the O2 sensor seemed completely functional. Decent voltage swings, decent crosscounts. As the speed went up, I could watch the voltage trend towards lean, and the crosscounts progressively became fewer. The fuel trim would go richer and richer; and by about 60 mph, the cross-counts were zero, fuel trim was WAY high, the Check Engine light came on and the lean-code set. To me, this signaled a genuine lean condition--except--the truck still had enough power to bury the speedo even with the Check Engine light glowing, and it wasn't like I was getting great gas mileage from the lean-burning engine. In fact, the fuel mileage was even worse than usual.

Eventually, it was setting that lean code at 45 instead of 60.

NOTHING checked out as being defective, and I'd beat my head against the wall long enough. In desperation, I crammed a THIRTEEN DOLLAR O2 sensor in the manifold, and...Problem Solved. The old sensor was actually broken inside by the time I got it unscrewed from the manifold--whether it was already broken, or I broke it unthreading it, I'll never know.

To this day, I can't tell you why that sensor only appeared bad at higher speed. Got overheated and refused to work???