Yet another Dilivar head stud question

[lindy 911]

Studs are on the way. Don't worry about the postage.

Lindy

[Flieger]

I am very interested to hear the results.

[Henry Schmidt]

If you need studs for your test, PM or email me and I'll supply some.
I think I have:
ARP
Raceware
Supertec (of course)
993 TT
Super Stud
Coated Delivar
MM
Factory steel

[AlfonsoR]

I have some dilivar studs as well with about 100k miles on them. I would be happy to send some to you for your test.

Please tell more about how you will test and what the goal is. Do you have John Bickford's books? Perhaps you could run the test at different temperatures to simulate the temp they are normally exposed to.

I think Charles N or maybe Henry might have info on thermal expansion coefficient of the cylinders. I wonder if there is any difference between the Nikasil vs Alusil?

I would be looking to determine the bolt loading cold vs at temp, and you might have to develop some correlation between cylinder temp and stud temp...

[vytenis]

Working on a plan... I have access to a shop to build some fixtures.

No doubt the Porsche case/cylinder/head/stud assembly is complex given the materials and loading situations.

We'll start simple and work our way up. I would like to understand the mechanical design aspects of this bolted joint. Long term effects such as corrosion and hydrogen embrittlement, which have been mentioned as causes of failure are out of my league.

My first curiosity is why 993 Dilivar studs would fail after initial pre-tensioning, without any other loads.

"V"

[AlfonsoR]

To understand that, I think you need to analyze the specific stud that failed. You would need to know if the material in the stud was to spec, and of course, you need to understand the details of how the spec material is forged. You would also need to know the conditions: was it previously used, number of heat cycles, the max temperature differential between cylinder and head stud.

I've read that in some alloys, higher silicone content causes mis-uniformities through-out the forging. I'm not saying Dilivar has high silicone content, just giving an example of needing to know the make-up of the material and understanding certain nuances about the forging process.

From what I understand , Porsche has changed the specification (make-up) of Dilivar at last one time. This is why Steve W and others recommend the 993tt studs, it's the latest recipe and has had very low failure rates.

Anyway, don't mean to give you an ear full, just really excited about your test as you can see.

[tom1394racing]

The problem with the Divalars is corrosion related fatigue failure.

Testing a new stud may not be indicative of the results you'll see in service.

[lindy 911]

To be clear, the studs I am sending V are the coated Dilivar bought from Porsche in 1992. They are not 993tt fully threaded studs but are the latest and greatest used in non-turbo motors. These studs were installed in a '78 SC case but never heat cycled. One stud broke an hour after torquing. We're talking less than 30 ft-lbs here fellas. It broke at the top of the thread down at the case.

This experience, for me at least, is enough to conclude that no matter how much you test or theorize these studs are absolute crap and unpredictable.

An interesting point: As a kid I put together lots of air-cooled motorcycle engines. The front of the cylinders faced the airflow and the back side did not. The exhaust also ran right next to the cylinders on one side and the other side was open to the air. The cylinder studs had exactly the same mechanical requirements as do our Porsche studs: clamp a single head to a single barrel. I NEVER saw a single stud failure!

Lindy

[AlfonsoR]

I think the point of the Porsche designed stud is to maintain the same clamping load through the whole temperature cycle as the engine goes from cold to warm and back to cold.

To do this, theoretically, you have to have a stud that has nearly the same thermal expansion rate as the cylinder. I say "nearly" because in the real world the temperature of the 4 studs and the cylinder will not be the same. With testing data, you may be able to compensate for this.

Anyway, what I am saying is that if you want to factor in thermal-rate-of-expansion, then your choice of alloys is probably pretty limited. I'm pretty sure Porsche chose Dilivar for that reason.

If on the other hand you want to throw that factor out the window, you could easily find material that is not prone to corrosion stress cracking (what it sounds like is happening to the old style dilivar) and can handle high temperatures. You could go with ASTM A193 (B16), with 125Ksi tensile strength or a duplex stainless like 17-4 with 160Ksi tensile. And I'm sure there are many more, those are just some that I have worked with on other type of equipment that come to mind.

The effects of having drastically different rates of thermal expansion or of having drastically different temperatures between the cylinder and the studs may be:

1. If the cylinder grows much faster than the studs, then studs will see an increase in clamping load. The increased stresses will transmit to the cylinder walls, which if not stout enough, can result in a wall that is not perfectly round as far as the piston rings see it. This effect can cause abnormal wear on the piston rings.

2. If the cylinder cools much faster than the studs, then the studs will see a decrease in the clamping load and will have the effect of studs which are not tightened to the correct torque specs. The possible effect could be a cylinder which will want to "walk". This may result loss of compression and possibly a lot more damage depending on how bad the issue is.

The holly grail is to not have stud failures and to maintain a consistent clamping load through the entire temperature cycle. I think it's a pretty tough trick, really.

[brads911sc]

Buy Supertec and call it a day... proven, cheaper than tt by 50%, better performance, testing already done, no failures... While I understand the desire to understand why something might fail, there is no way that you can duplicate the real world application on a consistent basis with all the real world variables... even Porsche couldn't duplicate the real world Head Stud performance in a test environment... if they did.. they would have never released these studs to the factory floor... how many sets did perfectly fine in a test environment... only to have a nightmare in the real world... we will never know but im sure if they had head studs snapping in the test environment they would have went back to the drawing board...

[lindy 911]

You hit the nail on the head with #2. The head walks and leaks on cool down if restarted too soon. Most of this had to do with the proximity of thermal reactors and their contribution to cooling, or lack thereof. Without reactors, dilivar simply complicates the issue. If it were necessary to match thermal expansions of clamped parts and bolts, they would need to use the dilivar on all four corners, not just the bottom.

Lindy

[vytenis]

Still figuring...

Thanks for the stud offers. I will try the 993 studs and use my steel and 1987 Dilivars first.

I want to test preload and possibly some thermal effects. I'm thinking I can measure coefficient of thermal expansionm of studs and cylinders in my (wife's) oven to compare to published values.

Don't know and can't test stud properties and don't know actual temperatures.

Thinking about strain gages and a dial indicator to monitor preload. Instron machine I was looking at won't handle the loads needed for a stud.

I'll be back here in a few days with a plan to be shot at.

"V"

[vytenis]

Originally Posted by brads911sc

Buy Supertec and call it a day... proven, cheaper than tt by 50%, better performance, testing already done, no failures... While I understand the desire to understand why something might fail, there is no way that you can duplicate the real world application on a consistent basis with all the real world variables... even Porsche couldn't duplicate the real world Head Stud performance in a test environment... if they did.. they would have never released these studs to the factory floor... how many sets did perfectly fine in a test environment... only to have a nightmare in the real world... we will never know but im sure if they had head studs snapping in the test environment they would have went back to the drawing board...

You are absolutely right. Nobody is going to come close to the body of knowledge or expertise amassed by Porsche with this I might say embarrasing head stud problem.

I don't plan to to a dissertation, but as a mechanical engineer, there is a certain curiosity.

"V"

[wdfifteen]

Originally Posted by Carrerax

you can also just use the same steel studs as the top row and be fine. They are availailable here at PP or I know Tom1394racing will sell you used ones for very inexpensive. They are not known to break.

To be clear on this, Porsche used Dilivar on the bottom only?
My engine is an '88 Carrera. I'm trying not to break the bank on this valve job/while you are in there project. Should I replace the lower studs or all of them?

[tom1394racing]

You only need to replace the lower studs. The tops are steel and will be fine.

PM me if you would like to purchase a set of lower steel studs.

[safe]

Originally Posted by wdfifteen

Should I replace the lower studs or all of them?

Yes, I had one break on my 84 Carrera engine a few years ago. Just one and that stud had no plating left.