The preload has to be high enough to overcome peak cylinder pressure or the stud would have to be fatigue rated rather than designed for a static load.

I think it is unlikely that the increase in stress caused by expansion will cause yielding in a conventional sense.

The stresses caused by peak cylinder pressure does, I am sure cause studs to pull out on magnesium cases.

The traditional technique of trying to solve problems by overtightening fasteners will make this more likley.

I am not sure that using a much higher strength stud and therefore increasing the static load is a great idea as this will also mean that the sueprimposed cylic stresses will be much more likely to damage the case and cause stress relaxation which will loosen the bolt.

Whether this chould be called yielding or creep is arguable but it is certainly inelastic.

I would have thought that the 'best' engineering approach would be to use the minimum load needed to avoid fatigue but then the issue would be trying to either estimate or measure peak cylinder pressure which will vary significantly between atmospheric and Turbo charged motors.

The original design was for a 2.0 litre motor which even in race trim was about 200HP.

Now we have 700HP+ 3.8 litre engines so is it surprising there can be issues. IMHO the problem is really with the case not the stud accepting that early uncoated Dilavar studs could suffer from SCC problems and break for no apparent reason.

I am sure case savers which spread loads and reduce local stresses in the case will mean that unloading of the studs is less likely and the engine should be more secure.

Things are complicated, too, by the inaccuracies in measuring bolt preload with torque rather than stretch.

The Supertec stud has far more than just a superior (debatable for sure) material.
The fine thread 12 point nut is a far superior design both in material and functionality. No designer of high performance hardware calls out coarse thread unless threading into soft material. Rarely/never with a nut.
The extra thread length for more thread engagement in the case. Thread engagement to resist pulling
Spark plug clearance for twin plug application. Barrel nuts used from 64-89 will interfere with the spark plug in an twin plug conversion.
Corrosion resistant material that will never break (any fair jurist will admit that even new Dilavar studs break) and the nuts will never seize on the stud (common with Dilavar)
No installed length issues that slow down assemble time. Ask engine builders if they ever had an issue with stud install height. With the install length issue resolved you can install the piston and cylinder without the studs in place making pin and clip installation far easier. Without the studs, you can install the piston in the cylinder on the bench and install the P&C as a unit.

http://forums.pelicanparts.com/uploa...1324062473.jpg
http://forums.pelicanparts.com/uploa...1324062496.jpg


BTW: the complete set (24 studs, 24 12-point nuts, 24 hardened, ground washers, Loctite and anti seize) sells for nearly half (when all hardware is included) as much as the Dilavar stud option. $660 from our host.

After pondering the advantages of installing cylinders without the studs, I think I would want the two studs opposite the side I was working on to be there to keep the whole assembly from squirming around. Is there really inough more clearance laying the rod and assembly over on one side to make a difference?

I am recalling that the studs on the working side did, to some extent, get in my way on my most recent build. Because I don't know if I have ever been as frustrated as I was trying to get the wrist pin C clips installed on my J&E's.

And, armed with tales from others, I had the Stomski tool, too, which I never needed with the Mahle clips. I still had a lot of trouble. After various issues, including buying more clips because I had mangled some on installation (plus I had to pull the pistons to have valve pockets deepened a bit more), I finally started lubricating the insertion tool. That seemed to make quite a difference. The air color reverted from blue to a more neutral tone after that.

What I don't understand is why doesn't J&E cut their pistons for circlips with those nice little holes in them which allow the use of a tool to compress them easily, both for installation and for removal?

On my first twin plug motor we turned the barrel nuts down to a cone shape to gain plug socket clearance. And I screwed up the stud height, which led to other issues and eventually the demise of the case. Then I learned about flange nuts, which ever after I had used. Wondered why Porsche didn't use them as a matter of course. Or weren't they around in 1964?

Henry has thought this all through pretty well, so I think he is entitled to toot his own horn. I liked the package, and it gave me the opportunity to buy a moderately big ticket item from our host.

Here is what you get with Dilavar studs after 65000 miles.
This is a street driven turbo. So much for the constant sealing of Dilavar studs. BTW: no broken studs. The Pistons and cylinders showed little/no wear indicating that this engine was not abused.

http://forums.pelicanparts.com/uploa...1324166054.jpg
http://forums.pelicanparts.com/uploa...1324166064.jpg

Improperly installed? (Preload) If not broken then what happened?

The engine only had 65000 original miles. Factory installed Dilavar.
What you are seeing is what I've been seeing for 30+ years. Dilavar failure.
Even when they don't break they offer horrible head to cylinder stability.
If they actually worked to my standards I wouldn't have had to seek a remedy.

Are you suggesting that the studs are permanently deforming?

This would seem to suggest that there is a significant amount of stress relaxation taking place at around 200degC which from a metallurgical point of view seems a bit unlikely if you look at the basic composition of Dilavar.

It seems unlikely that Dilavar would creep at the preload being applied at 200-ish degC as this type of alloy would have time to rupture in the order of 10 000 hours at stress levels in the order of 100 ksi even at temperature of 1000degF.

Austenitic Stainless Steels re generally considered to have excellent creep properties and are commonly used for the manufacture of fasteners used in the high pressure stages of Gas Turbine Engines.

17-4PH on the other hand tends to be rated in terms of stress to rupture life of about 1000 hours at 70 ksi at temperatures of 1000degF.

This sort of says that if Dilavar causes problems due the a creep mechanism then martensitic stainless steel and conventional steels would be even worse.

A real puzzle.

My experience is in the field of Air cooled Porsche engine building (over 400 to date) and all the theory goes out the window when real life (practical) experience contradicts theory.
I am not suggesting anything other than under real life applications, my experience tells me that Dilavar is a horrible material.

There is other experienced engine builders that still use dilavar (993 turbo studs) with good results. Why don't they have the same experience as you? Very strange I think.

Perhaps I'm misreading your statement.
Your statement leaves me wondering how you know they (the other builders) don't have the same experiences as I do. The engine I posted is one of many engines that I've disassembled that show the same instability.
Are you suggesting that this factory built engine with leaking heads and Dilavar studs is unique to my experience?

BTW: Everyone gets to choose the parts and building process they deem appropriate for their requirements.
Dilavar studs are the safe choice because builders can blame Porsche for any failure. "I chose Dilavar because it's what the factory does". I choose something different and offer that option to others.
Like many other engine builders, I choose the push the envelope when building these engines because I want to build something special. The factory never cross drilled production cranks and yet most would have to admit that the improved lubrication offered by cross drilling is beneficial. We cross drill every crank. This is not to suggest that builders who don't cross drill every crank are doing something wrong, I'm just suggesting that there may be ways to improve on the factory process.

The point I was trying to make is that the well established behaviour of the family of materials that is being debated would indicate that they are extremely unlikely to suffer from stress relaxation.

If Dilavar suffers from creep why doesn't an alloy with lower properties exhibit the same behaviour?

There is some reasonably sound Science behind my comments about material selection which is why I provided some basic data. This is not a theory it is a practical measure.


It is quite clear that there are bolting issues with 911 engines I just don't accept that swapping stud material is guaranteed to solve the problem unless there is a sound reason.

..

I think I can provide another example of Henry's experience with Dilivar studs. My '79's engine was built with all Dilivar studs. It exhibited the same condition that Henry's picture shows. Leakage above and below the cylinders. Engine was very damp at many of the sealing surfaces. Cyls-to-case leaked, cyls-to-heads leaked, heads-to-cam towers not so bad.

The leaks weren't constant drips that would leave a puddle. Only a few drops- some at startup, some after cool-down. However the engine was always VERY wet. Upon tearing this engine down, I found the head stud nuts were inconsistently tight. Some came loose very easy, some required only a slight amount of effort. Other engines i've done, the nuts were typically a bit resistant and came free with a loud squeak. Maybe the oil leakage made its way along the studs into the threads, making the nuts easier to remove?

Not the greatest picture, but does show similar oil residue above and below the cylinders.

http://forums.pelicanparts.com/uploa...1324308012.jpg

The only problem with the Porsche studs was corrosion and stress risers (product finishing) and it has long ago been addressed.

regards

Also, it should be noted that my engine shown here is in a dedicated race car. It was rebuilt in Feb of 2005. So it is driven hard pretty much all the time once up to operating temp. I suspect that would aggravate the problem more so than a street driven engine.

I'm not trying to be argumentative here but come on!
If the stud issue is a thing of the past, how do you explain inconsistent head to cylinder sealing (leaking and movement) and broken studs. Even broken 993 TT Dilavar?

From an earlier thread on this topic:

As an mechanical engineer I am very interested in an explination of the modern engineering practices that is used.
Maybe the Supertec stud is an ok product, but I get worried when you say the best available solution from Porsche is horrible and yours is "perfect".

The arguement about Dilavar will, it seems, never end so lets try again to see if we cam make sense of the issues.

The fact that there is a bolting problem is indisputable as there are just too many engine with cylinder heads leaking which is very easy to observe.

The interesting point must be the cause of the problem and then finding a good solution.

It is also clear that the early Dilavar Studs suffered from breakages particularly in Continental USA.

Precipitation Hardening Stainless Steel does suffer from Stress Corosion Cracking particulalty in the presence of Salt. (Chlorides). The resin coating helps in this regard unless the coating becomes damaged.

Obviously if studs break then they allow the heads to leak.

If, however the stud doesn't break the only way that the head can leak is if the bolt stretches.

The only mechanism that can be responsible for a material deforming at constant load is known as creep.

I don't really believe that any rengineering steel will suffer from creep at the temperatures applied to 911 head studs.

Dilavar is basically a Precipitation Hardening Steel and its metallurgy means it is very unlikley to deform permanently at the temperatures and stresses involved in 911 head studs.

A basic estimate of the stress in a 911 head stud shows that for a torque of 24 lbsft the axial force would be about 3700lbs which would result is a stress of around 48ksi - not really very high.

My point about 17-4PH which is a Precipitation Hardening Martensitic Stainless Steel is that its creep properties are just not as good as the Dilavar family but still significantly better than standard steel bolts.

If Dilavar suffers from creep elongation so will the other materials being discussed. This is not a theoretical consideration it would be part of a considered approach to material selection using data established by significant test programmes.

As I don't think any of the bolts used are suffering from creep elongation there must be another explanation and as Magnesium and Aluminium Alloys are prone to stress relaxation at temperatures around 200degC it is possible that this is a part of the problem.

I also questioned the idea of using higher loads to 'solve' this problem and if there is an issue with stress relaxation occuring in the casting a higher load will only make it worse.

The last engine we had with leaking heads had conventional steel studs on the top and Dilavar on the bottom. Two heads were leaking quite badly (50-60psi compression) but the Dilavar studs were still tight it was the upper studs that were loose.

I am also not 100% convinced that a $10 CE ring is the best device available and there must be better ways to seal the heads.

My point was not to criticise any specific stud but to try to establish the real reasons for these problems as I think blaming Dilavar is too easy unless they have broken.

I have a lot of respect for the aftermarket and its capabilities. I tend to agree with Weissach as to the CE ring and its abilities as well as the integrity of the Barrel in the larger displacement engines. With the current design you are going to see evidence at some mileage of oil and the dirt it attracts with any of the current bolt and joint arrangements and except for the appearance the motors seam to run relatively well. Hence the cost effectiveness of replacing one for the other.

regards