Thermostat housing gasket won't seal!

Hi All,

I just replaced my thermostat. Last time I did this, no problems. This time I found a slow drip on the bottom of the housing. So took off the t-stat housing again, replaced the gasket, put RTV on, and closed it up. Torqued bolts to 15 ft-lbs (Haynes says 13, but just to be on the safe side). Result, same drip is back within a day (not right away).

What's up with this? I cleaned the surfaces very carefully, removed all old gasket, finished with fine sanding paper, and de-greased. I put RTV very even and thin on the gasket, and torqued the bolts very evenly.

The only thing I can think of is that my RTV sucks. I bought a tube of Permatex Form A gasket sealant (fast drying, hard setting) about 6 months ago. I used it before also (when I did waterpump and t-stat, that's when I bought it) and then it worked fine. Can that stuff get old? Am I not using it right?

Or should I just put more torque on the bolts? (I don't have an FSM)

I've taken the serp belt off already and waiting for the engine to cool off before I take the housing off again. Don't mind replacing the gasket again, but I want to know what's gone wrong the other time so that I get it right this time.

Any input would be great!

Tim

Did you clean the outside of the housing where the hose fits on, clean the inside of the hose where the rust from the housing hardens to the hose, retighten the hose clamp?
I had this same problem recently. Drives you nuts finding the leak.

Yep, cleaned that and hoses are new. It's really leaking from the bottom of the housing. Around the hoses it's dry. It looks like it's really the gasket that's giving me trouble and I just don't understand why.

you RTV both sides of the gasket?

Yes, did that the first time. The second gasket I got at Nappa and that was adhesive so I only did RTV on the non-adhesive side. The guys at Nappa said that the adhesive would work as an RTV too.

Dumb question - you didn't crack the thing, did you? That aluminum casting is thin, and easy enough to crack...

If you want torque specs, I have them (most of them, anyhow - if you're not there, find a year that's close to yours) on my site - just for people like you who haven't gotten FSMs yet. I do recall the torque on those screws to be fairly low tho - somewhere around 12-13 pound-feet? Anyhow, I'm reasonably sure it's not 15.

As far as the adhesive on the NAPA gasket being like RTV - I'm inclined to think it's not. I've not seen that gasket specifically, mind, but I've seen similar setups, and it's just a tacky bit to help hold the gasket in place while you're working - not for filling in cracks, pits, and other surface irregularities. There's never been any "build" to the sticky gaskets I've seen, so I don't use them.

For RTV, you can use RTV Black (general-purpose stuff) or #518 Gasket Eliminator (for tougher jobs, where you need more build.) Don't bother with the blue stuff, and I've not use the "cheez whiz" - since the regular toothpaste-tube stuff works just fine for me. I've tried to use the blue stuff - and it works about as well as screen doors on a submarine. No, thanks.

I had the same issue. I even went to NAPA a bought a new cover. I then, as per 5-90, used the black rtv and a new gasket, had not had a problem since.

Prior to this, I NEVERED used RTV, only aircraft gasket sealant for my cars. That's what I used the first three times, and it leaked. Experience pays, and that's why I like this site. I never would of thought to use a good RTV, but now I know better!!

At least the gasket is cheap!

Enjoy,
Bob

Thanks guys, I didn't crack the thing (checked because I expected something like that the first time). The torque in the Haynes says 13, just made it 15 the second time, just to be on the safe side (I don't have a terribly accurate torque wrench....and I will buy a good one now also when I get my new gasket, since I'm getting pretty sick of that thing anyhow....pretty useless piece of equipment if it aint accurate imho).

I will get the good RTV and a new gasket. Will RTV the adhesive side also. Just wonder why they make that thing adhesive, the RTV will make it stick anyway right?

Just another thought, Could I have over-tightened (I didn't put excessive force on it but was certainly a bit tighter than 13)? Can that ruin the gasket?

The stat can slip out of position when the housing is placed over it,it can be tightened down in that position and NOT have the housing break,but it will leak.Be sure to clean the recess in the head to allow the stat to seat fully.A couple small dabs of RTV are helpfull in holding the stat in place until the housing is bolted on.15 ft lbs is the correct torque ,be sure the bolts are not too corroded or they may break off.
Wayne

dutchjeep said:

Thanks guys, I didn't crack the thing (checked because I expected something like that the first time). The torque in the Haynes says 13, just made it 15 the second time, just to be on the safe side (I don't have a terribly accurate torque wrench....and I will buy a good one now also when I get my new gasket, since I'm getting pretty sick of that thing anyhow....pretty useless piece of equipment if it aint accurate imho).

I will get the good RTV and a new gasket. Will RTV the adhesive side also. Just wonder why they make that thing adhesive, the RTV will make it stick anyway right?

Just another thought, Could I have over-tightened (I didn't put excessive force on it but was certainly a bit tighter than 13)? Can that ruin the gasket?

Click to expand...

Paper gaskets are far more difficult to crush than rubber or cork - you're more likely to crack the housing than you are to crush the gasket.

Check torque on the tables I've posted - they're directly from FSMs. I use my Haynes manual to even up the odd leg on my workbench - they're decent for imports, but the ones for domestic vehicles are best used as firelighters.

Considering that there is a fairly wide inherent inaccuracy involved in using turning force to measure fastener preload, you really do want an expensive torque wrench. The "torque angle" method would be far more accurate - where you either turn the fastener to "finger tight" or you use a smaller (and rather more reliable) torque figure, then turn the fastener a specified angle to finish. One of these days, I plan to see about calculating starting torque and angles for the fasteners used in our applications - but there are a number of factors involved. However, the angle would be a more accurate method, since torque specs are usually for "clean, dry" threads - and friction becomes highly variable on dry metal surfaces...

Just gonna throw this out there...

I replaced the water pump, rad, hoses and t stat on mine a few years back and last year I noticed a leak in the same spot you described. I took it apart again and went and got a new gasket from NAPA and it had the small bit of RTV on it too. I don't know why but I read the tube of RTV I had used (on both sides of the previou gasket) and it said it was corrosive. I went back to NAPA and looked at the RTV and found both corrosive and non-corrosive types. I went back home and looked at the t stat housing and it was all pitted. I do not know if the RTV caused this but I think that's why it would not seal. It said it was corrosive to aluminum. I got a new housing and used the new gasket and non corrosive RTV and it's been good.

5-90...what are your thoughts on what I said?

Well, I use Permatex sealers pretty much exclusively, and I've used both black and copper when sealing aluminum against iron (copper on manifold gaskets - wet and dry - and black on water gaskets,) and I've not noted any corrosion.

What brand were you looking at? I don't know about Permatex having two different "lines" separated by potential corrosion - so I want to make sure I'm not missing something...

FWIW, I had that problem too. Turned out my housing had warped ever so slightly. A new housing from the parts place and no more problem. Didn't use any sealer either. Just a new housing and gasket.

Remove hose, remove housing, take out thermostat.
Clean housing, open bottle of indian head gasket cememt, coat one side and let it get a bit tacky, stick it on housing. Coat other side, install thermostat, tighten bolts to spec.
Take indian head and do a quick wipe around the outside of the housing where the hose goes. Stick hose on.
Top off system, run engine, let it get hot and go from there.
No leaks after that.
RTV bites, don't wast your time with it unless you get a perfect bead it can cause the housing to go on cockeyed, just a hair but enough to seep.

I would get a new housing, they are cheap, and the cracks can be very difficult to detect. I replaced mine 3 times before I found the small hairline crack.....

Thanks guys. Having read all this I will replace the housing, along with the gasket. You're right, they're inexpensive and it's just not worth having to drain the cooling system again. Since most of you had had no problem with RTV, I'm just gonna go with that. If it still leaks after that I will get me some of that Indian stuff. Will check for corroding properties of the RTV (never thought of that before, they should put that on in BIG letters). I expect that I may have a hair crack in the housing after all since even without RTV it should not have leaked.

5-90, did a check on those torques on your website, just to make sure I had it right. I Checked this page

http://www.geocities.com/JeepI6Power/1992XJMJ.html

Maybe I'm all wrong but it says 22 lbs-ft??!! Did I look wrong, is this the wrong torque, or should I now definitely light my Haynes which says 13 (was gonna do that one of these days anyway, just need to save up for the FSM)?

Tim

Just reporting back. Put on the new t-stat housing. The replacement housing was made of cast iron, not alu, so I decided I could go with the higher torques (22 lbs*ft). Finally it's closed, no more dripping. I checked the old housing but can't find any cracks, still expect them to be there, but probably too small to see without pressure on there.

Anyway, I am richer with experience, understanding of RTV's and gasket makers, and a new torque wrench, so everything is good!

Thanks guys for all the help!

Tim

Glad it's working. Just remembered what happend to me replacing the t-stat on a 96. I used RTV with no gasket and the long bolt (top one) bottomed out in the hole. The housing slid back and forth a bit even though the bolt was tight. I ended up putting a washer under both bolts and it's been fine. Contributing factor: I also ran the gasket side of the housing back and forth a bit on a file to smooth it out from some gasket removal dings so the combination of no gasket + the filing was enough. I don't think I removed much (but also didn't measure what I removed) but it at least indicates that there is/was not a lot of spare depth in the bolt hole.

Anyone ever o-ring the housing? When I have my heads cleaned up, I plan to have them o-ring the block/head and the housing. Seems like a good idea?

dutchjeep said:

Thanks guys. Having read all this I will replace the housing, along with the gasket. You're right, they're inexpensive and it's just not worth having to drain the cooling system again. Since most of you had had no problem with RTV, I'm just gonna go with that. If it still leaks after that I will get me some of that Indian stuff. Will check for corroding properties of the RTV (never thought of that before, they should put that on in BIG letters). I expect that I may have a hair crack in the housing after all since even without RTV it should not have leaked.

5-90, did a check on those torques on your website, just to make sure I had it right. I Checked this page

http://www.geocities.com/JeepI6Power/1992XJMJ.html

Maybe I'm all wrong but it says 22 lbs-ft??!! Did I look wrong, is this the wrong torque, or should I now definitely light my Haynes which says 13 (was gonna do that one of these days anyway, just need to save up for the FSM)?

Tim

Click to expand...

I'd also answered your PM, but didn't know you'd posted in the open.

I rechecked that figure in the FSM, and it is correct. It still seems a bit high to me, but it's correct.

I've probably just been using 12-13 pound-feet out of sheer force of habit - since the pressure involved is relatively low, and the sealing area relatively large, it's never been a problem for me.

That low figure may just be something I picked up when I was doing mainly SBChevvy work, and I just kept it around in my head.

One thing I'd forgotten to say - all figures posted on my site are from manufacturer literature (OEM or aftermarket) and are rigorously checked before the page "goes live." Any mistakes are faithfully reproduced from the original...