Head nut retightening... so many loose?
 

Hi,

In another post I mentioned that I found two loose nuts on the head studs. I bought a 10mm allen head wrench and retightened the two nuts using ample amounts of anti-seize. I then checked the torque on the other nuts on the left cylinder bank and noticed that all but 2 were fairly loose. Once again, I loosened all the nuts (one at a time) applied ample amounts of anti-seize and retorque them all to 24lbs.
I am going to check the other side tomorrow.

I have a couple of questions. I had a valve adjustment performed by a very reputable shop near Dallas about 500 miles before an accident sent my car into "rustoration". Is it standard procedure to check the torque of head nuts during a valve adjustment? Should I be dissapointed that they didn't even check the nuts out?

Another thing. The PO had the engine rebuilt about 30k miles ago. I want to think that no one ever brought the car back to have the heads retorqued. I don't feel like I should be worried as so many were loose that is seems like the followup service was performed.

I guess I need to report back with my findings on the right cylinder bank.

IMO once the studs are torqued properly, they should not loosen up and should not need to be retorqued unless something is wrong.

My guess is that you are experiencing the "magnesium case pulled stud" syndrome. drive it for a while and check them again. If they loosen up again you need to rebuild the engine.

Its interesting, and completely meaningless, that american iron type engines may have 75-100 ft-lbs of torque on the head bolts.

the mating surfaces of the head, to cylinder, to case, get pretty worn over the years, from running with loose head nuts, and the constant thermal cycling, and they are not on the same plane as when they were new. freshly machined engines remain tight, no retorqueing necessary, and old leakers need to be retorqued every time the valves are adjusted. if they retorque without pulling a stud, then it's the uneven mating surfaces that are causing the loose nuts, not the studs.

Yes, the cases bend and settle over time. If you can tighten each one to a specific value, consider yourself lucky and call it a day. When the nut spins and spins, you have a pulled head stud...

-Wayne

Well I pulled the top valve cover on the other bank and most were definitely less than the required 24lbs. I do have one problem. None of the top studs are broken, but I can't tighten one of the nuts completely because the stud goes too far into the nut and pushes out the allen head wrench. I tightened as much as I could but I am sure that I couldn't get the full 24lbs on it. The uncertainty of this sucks. Should I leave it or try double nuting the head and seeing if the stud is loose?

I will check the lower studs tomorrow if time permits.

Thanks for all the help. I will monitor the engine closely once I put it back in the car.

that one stud may be pulling. stack a couple of washers under it and see if it does the same thing.

Did I read correctly that you loosened all the nuts and then re-tightened? I am confused as to why you did this. Did I forget something?

David,

I took off a valve cover on the left cylinder bank and found one loose head nut and one that was finger tight. So I decided to check the torque of the other nuts. I only loosened them one at a time and retorqued. I am hoping that they were loose because no one ever checked them after the engine was rebuilt.

I can hope, can't I?? Please....

One of the problems with torque is the friction. if you torque a fastener with clean threads and good lubrication it will take less torque to achieve the proper stretch. If it has ran for some time, the lubricant will be less effective and the same amount of tyorque will result in less stretch because it will have additional friction.
If you really want to be as accurate as possible you would remove the fastener, clean it as well as possible, and re-apply the recommended lubricant. Lubrication on the washer and face of the fastener is just as important as on the threads, maybe more so.
Easy to say, harder to do.

I have a spread sheet I got from a company called Hytorq that has functions built into it for percentage of yield and also type of lubricant. You input which type of lubricant and what yield factor and it does all the calculations.
the differences of the friction factor for different lubricants is surprising, even more so when compared to non-lubricated threads.
I can e-mail the file to anyone interested but it is for very large fasteners and up to very high torque values (over 10,000 ft. lbs.) so it would only be useful for the lubricating factors when it comes to 911 engines. Basically a curiosity thing.

cool. I would love to have a copy.. thanks

RoninLB@optonline.net

Sam,
This it totally OT but what sort of application requires torque values of 10,000 foot pounds.
Just curious...

I've seen bolts tightened with a large neumatic wrench the size of a street repair air hammer. The 180MW turbine case had walls super thick. Don't know the torque, but they measured the case bolt stretch.

Ron, The spreadsheet is on it's way.

I actually left off a zero or two on that post, the chart goes up to 350,000 ft. lbs. for 6" bolts (depending on % of yield) but I don't work with studs that big very often.

I've spent most my working life either rebuilding, designing, or upgrading very large industrial machinery, some of it takes some unbelievable torque values. The bigger the machine the larger the fasteners or larger number of fasteners required. Depends on who designed it.

The biggest machine overhaul job I've run is unit #3 at San Onofre nuke plant, steam turbine rated at over 1 1/4 million horsepower. Rotating mass of over 3 million pounds.

Nowadays I'm only involved in centrifugal pumps. Some of em are pretty big but nothing you can't load onto a semi truck. We usually don't go much over 10,000 ft. lbs. of torque on the head studs.

We use hydraulic torque wrenches to do all the work. Some of ther wrench heads cost over $9000 each. Yikes, that comes straight out of my budget.