Head stud nut loose AGAIN

[Jerome74911S]

My car went pfut, pfut, pfut on acceleration. Not good.

I discovered a loose nut on one exhaust head stud. The nut came right off, but the steel stud seemed to be tight and solid.

Since I had the nut in my hand, I simply reinstalled it and torqued it to spec, using anti-seize compound.

The pfut was gone. Good.

Several hundred miles later the pfut is back. #@!!?*

The same nut was loose again, so I torqued it up again.

Questions:

If the stud is pulling the case-saver out of the crankcase, how far does it have to move before the nut gets loose and the engine goes pfut? Half of a millimeter? Or, is this theory incorrect?

But, after the nut was-torqued, twice now, shouldn't the stud protrude into the nut if the stud is moving out? The stud does not appear to protrude into its nut any more than any other stud.

Thoughts?

[yelcab1]

Loctite, what have you got to lose?

[zedsn]

Has the stud appeard to be pulled out some? Is the end sticking out farther than the other studs? If not maybe buy a new nut and see.

[Dpmulvan]

+1 on loctite

[john walker's workshop]

If it has case savers, I assume it's a 2.7. The issue with those engines is uneven wear on the cylinder bases, cylinder tops and head sealing surfaces. If everything was machined perfectly parallel to the crank centerline, the heads generally stay tight. Was all this done to your engine?

[Jerome74911S]

Originally Posted by john walker's workshop

If it has case savers, I assume it's a 2.7. The issue with those engines is uneven wear on the cylinder bases, cylinder tops and head sealing surfaces. If everything was machined perfectly parallel to the crank centerline, the heads generally stay tight. Was all this done to your engine?

Wow, I wish I could answer this. I see evidence of machining having been done in various places, but I don't know about perfect parallel-to-the-crank work.

Could it be that this cylinder, #6, has imperfect machining done to it? Everything else is tight and leak free.

There are at least 20,000 miles+ since rebuild.

[Jerome74911S]

Originally Posted by zedsn

Has the stud appeard to be pulled out some? Is the end sticking out farther than the other studs? If not maybe buy a new nut and see.

The stud does not stick out more than the others. Good idea to buy a new nut - and put some locktite on it.

[fanaudical]

To answer your original question (how far does the stud have to move to lose torque on the nut?): A very short distance. Think of the studs as very stiff extension springs. Tightening the nut essentially stretches the spring and produces clamping force. That clamping force stretches the stud. If the threads pull in the case by the same amount of stretch, torque is gone.

Rough/approx numbers from "back-of-the-envelope" clamping calc: ~20 ft-lbf of headnut torque produces ~5000 lbf of clamping force in a head stud and imparts ~0.0006 inches of stretch (assuming steel stud). It doesn't take much relaxing to cause trouble.

[Bill Douglas]

JB Weld?

[chris_seven]

Originally Posted by fanaudical

Rough/approx numbers from "back-of-the-envelope" clamping calc: ~20 ft-lbf of headnut torque produces ~5000 lbf of clamping force in a head stud and imparts ~0.0006 inches of stretch (assuming steel stud). It doesn't take much relaxing to cause trouble.

I agree that the difference in height between the studs to cause a loss of preload is virtually immeasurable particularly if you don't have the starting height but the stretch produced by a preload of 5000lbs is much greater than 0.0006".

The stress produced in the shank of the head stud by a preload of 5000lbs is around 67000 psi. (I do agree that a preload figure of 5000lbs is a good estimate)

The calculates to approximately 450MPa.

If you take the Young's Modulus in tension as 200GPa for a steel stud then the resulting strain would be 0.00225. (Strain is always a non-dimensional number)

At 100 degC the Young's Modulus would reduce by 10% to around 180GPa and the strain for a given stress would increase but I have ignored this factor for the rough calculations we need to discuss the fundamentals.

If the stud is 140mm long - neglecting the ends -then this results in an elongation of 0.31mm - around 0.0012".

This is equivalent to just about 70 degrees of rotation of the nut. If the elongation were only 0.0006" then the angle between fully loose and fully tight would only be 3.6 degrees which would make it quite tricky to tighten head nuts consistently.

The relationship between preload and elongation is linear so every 0.001" that the stud pulls out the preload will reduce by around 10%.

Let's is not forget that a steel stud in a magnesium case with a Nikasil barrel sees an increase in preload of around 1500 lbs due to expansion.

I think my scruffy calculations are in the right ball park but would be quite happy to be proven wrong

Sadly, it looks as if the studs may be pulling out and if the engine already has case savers it may be a problem.

It may be worth fitting a Dilavar stud in this one location as this will reduce the increase in pull out force due to expansion by around 1100 lbs.

We have made Titanium studs for this application as the lower value of Young's Modulus reduces the pull out forces due to expansion compared to a steel stud but not by as much as for a Dilavar Stud.

[Bill Douglas]

Originally Posted by chris_seven

I think my scruffy calculations are in the right ball park but would be quite happy to be proven wrong

Exactly Chris; scruffy. I rest my case - JB Weld

[chris_seven]

What do you plan to do with the JB Weld ?

Glue the head onto the barrel?

If you try to use it to repair the stud boss I think it will just pull out.

Lets face it JB Weld just isn't strong. Its tensile strength is about 4000psi which is around a fifth of the strength of a cast magnesium alloy even after it has suffered from 40 years of stress relaxation.

[Bill Douglas]

Sorry Chris, the big grin was to show I was joking. JB Weld is the hicks way of fixing things and I was kinda making fun of myself. And Scruffy - obviously far from scruffy, you had done some really good maths there.

[chris_seven]

Bill,

I do appreciate that you weren't being serious but I have seen engines that have literally glued together in the past and it is sometimes scary to see some of the 'repairs' that are carried out. I once bought a Lotus Seven with a Twin Cam engine with a patch glued onto the side of the block after it had thrown a rod.

The same engine had the valve seat areas welded up and re-machined. The clown that did the job plugged a pinhole into the water jacket with a aluminium screw and Loctite and then filed off the head. It lasted about 20 minutes before it expired in a cloud of steam.

The problem with this engine is that if the case saver is pulling out I am not sure what can be done.

The only mechanism that explains this behaviour on mag cases will be due to stress relaxation which is when material plastically deforms due to being loaded for long periods of time.

Magnesium is known to suffer from this problem and the thermal reactors must create a sufficiently high temperature to activate this mechanism.

Military Helicopter gearboxes using magnesium cases are temperature monitored to ensure that this effect doesn't become an issue.

Unfortunately the only cure is to remove the damaged material and if the case saver hasn't done this then it is a problem hence my suggestion to try to reduce the applied tress with a Dilavar stud.

Not the best solution but it may work.

[Jerome74911S]

As the poor OP I am thankful for the detailed analysis, but my only comment is Yikes . . .

I did not pay attention to the number of degrees of rotation required to re-tighten the nut, but I would not be surprised if it exceeded 90 degrees.

[mreid]

Are these stock studs with the barrel nut?

[theiceman]

lets go back a long way ...
I am quite surprised you hear pfft pfft pfft with one loose head nut. I think there is more going on here and the others are on the brink also .

[chris_seven]

Originally Posted by Jerome74911S

As the poor OP I am thankful for the detailed analysis, but my only comment is Yikes . . .

I did not pay attention to the number of degrees of rotation required to re-tighten the nut, but I would not be surprised if it exceeded 90 degrees.

Lets hope they are not pulling out and that there is another cause.

[Rob 930]

I'm going to ask some basic and obvious questions that don't seem to be clear from this thread: What engine is this? Your name implies it's a 74 911S. True? If so, does it have the original magnesium case? If it has a magnesium case, does it have threaded inserts installed around the head studs ("case savers" as you have called them)? You can tell my looking at the engine from underneath. If it doesn't have threaded inserts, then it sounds likely that you're suffering from classic stud pulling, of the type that all these engines are well known for.

It's very unlikely that the nut is backing off and that Loctite or JB weld, a new nut, or any other "fix" at the nut end is going to help at all. The problem is most likely caused by a loss of bolt preload due to movement of the stud. And I don't know of any way to halt that or lessen it without a tear down and proper repair.

[Rawknees'Turbo]

Originally Posted by theiceman

lets go back a long way ...
I am quite surprised you hear pfft pfft pfft with one loose head nut. I think there is more going on here and the others are on the brink also .

Hearing a noise similar to an exhaust leak is the norm when there is a broken head stud/loose nut - guys who haven't reported hearing it, even though they found one during a valve lash adjustment, etc., either aren't very observant of noises or just got lucky that no combustion loss was happening despite the broken stud.