Head Studs

[VFR750]

Henry,

Good question on stud characteristics. The unthreaded portion of your design is a good thing. You may find, that reducing the diameter of the unthreaded portion will further "soften" the springrate of the stud. What this will do is decrease the load variation for a given thermal displacement.

[VFR750]

If we can all agree on:
1) the weakest link in this whole assembly is the machined threads in the case (mag or Al)
2) Steel, Nickel, Titanium threads on a head stud are stronger than the case

Then, you could further reduce the shank diameter, since it is not the limiting feature in the assembly. Here is what I mean:


An undercut shank will lower the spring-rate of the shank because the cross sectional area is less. Notice this looks a lot like rod bolts. In fact, the best head studs should probably look just like rod bolts. I think Henry's are pretty close.

My 2 cents...

[Henry Schmidt]

Originally Posted by VFR750

Henry,

Good question on stud characteristics. The unthreaded portion of your design is a good thing. You may find, that reducing the diameter of the unthreaded portion will further "soften" the springrate of the stud. What this will do is decrease the load variation for a given thermal displacement.

Since you asked:
Early on as part of our on-going development (seven different iterations, 20+ test engines) we tried replicating the original stud configuration and discovered that by reducing the stud diameter we reduced the clamping force for a given preload and increase clamping force within the constraints of the case material improved head to cylinder stability and increased reliability(read longevity). wow, sorry about the loooong sentence.

Our studs also utilizes a finer thread pitch at the nut end. This allows for a more precise torgue measurement.
Torque for torque, the clamping force is higher with a fine threaded bolt. This is due to the more gradual angle formed by the fine thread helix.
Fine threads will, in general, have higher load carrying capacities and higher resistance to vibration loosening. A problem we saw with the ARP stud. That may be the reason ARP recommends such a crazy (high) torgue setting, 36 lb according to the ARP tech department.

We also use a 12 point, serrated flanged nut of dissimilar material to the stud reducing galling and seizing with time and exposure to heat and the elements.
If you build enough of these engines you are bound to see a plethora of stripped allen drives and seized stock nuts .
The serration also reduces vibration and heat cycle loosening.
Another benefit of the small nut is better plug connector clearance when twin plugging.

The design of our nut and stud combination eliminated the need for a specific stud install height so we make the stud longer for better/ deeper thread engagement and improved flexibility for difference engine dimensioning. IE: long rod engines

Eliminating the specific stud height means no measuring. This lack of measuring saves time but more importantly, allows the cylinders to be installed before the studs. With slight procedure changes, the assembly of these engines becomes easier and more precise.

[VFR750]

I think we agree. A good design. Without knowing your dimensions, and just looking at the pictures on your site, it appears you have a reduced shank head stud.

My way of looking at it for a given material:


Now, Titanium could offer a interesting option.

Ti has a Young's Modulus (YM) about 1/2 that of steel/nickels alloys. Perhaps, Ti would be a better choice, even though the coefficient of thermal expansion (CTE) is lower. Ti CTE is about 3/4 that of steel. So it might be interesting to see if that would trade well. Just based on the ratio of YM to CTE I think Ti would be better than a high-strength steel stud.

Anyone try Titanium head studs??

[Henry Schmidt]

Originally Posted by VFR750

If we can all agree on:
1) the weakest link in this whole assembly is the machined threads in the case (mag or Al)
2) Steel, Nickel, Titanium threads on a head stud are stronger than the case

Then, you could further reduce the shank diameter, since it is not the limiting feature in the assembly. Here is what I mean:


An undercut shank will lower the spring-rate of the shank because the cross sectional area is less. Notice this looks a lot like rod bolts. In fact, the best head studs should probably look just like rod bolts. I think Henry's are pretty close.

My 2 cents...

Thanks for your analysis but your description of our stud is slightly inaccurate.
Our stud is heat treated (aged) and ground three times to achieve the exact ( + or - .0005') size we desire.
Because our threads are rolled and not cut, the stud body is ground on each end to the proper size for each specific thread pitch. Then the center is reduced to a compromise size. The thread root on the fine thread end is .364" and center is .356" and thread root at the course end is .348". The thread outer diameter of both threads .387".

[Henry Schmidt]

Originally Posted by VFR750

At some point, the load in the head stud will overcome the threads in the case

Absolutely correct.
The trick in this exercise is to create enough clamping force in the head to cylinder joint to promote desired stability within the limitations of the case material.
In the aluminum case this is a relatively easy proposition. In the mag case slightly harder. Increase the thread strength of the magnesium case requires an insert properly designed for this specific application. The "Case Saver" is and has been that insert for over 20 years. Still, working with the limitations of each material is key.
Although the engineering input we got from the initial design engineer was invaluable, what it boiled down to was trial and error. What we found was, on paper the things that should work, were observed to work differently. That was the reason for 7 different studs.

Originally Posted by VFR750

Anyone try Titanium head studs??

I looked at titanium nuts but the head studs would be cost prohibited for most applications.

[Lapkritis]

Originally Posted by VFR750

My way of looking at it for a given material:

Nicely done. It is interesting to note the air-cooled vw camp replaces the 10mm head studs with 8mm because the 10mm pull from the case more often (explained by your graph). However when they get to high hp applications the 10mm stud is then recommended for the clamping force.

[VFR750]

Yes, Ti studs would be very expensive. Probably 5-10x more. Would anyone pay if one stud cost ~$100 each? That gets expensive when it's (24) required.

I understand what you mean on ground + rolled design. Hard to show with my simple picture. An excellent method for high strength fasteners.

[chris_seven]

The issues being discussed are always basically the same.

The damage to cases is simply caused by the forces being reacted by the threads in the case.

The preload required to stop the head lifting is a given - it shouldn't really change whether the engine is cold or hot??

By minimising the increase in force produced by expansion the case is more likely to survive.

Does it matter if this reduction in force is a result of the expansion of the stud material or by reducing the spring rate of the stud?

We can change the spring rate of the stud by either reducing diameter or reducing E by switching to Titanium.

This means we have 3 methods of effectively controlling the force increase.

Standard steel studs in early engines had a 'shank' diameter of 7.7mm and is clearly a reduced shank design.

This is much different to the current stud designs whch seem to be closer to 9.0mm diameter. (0.356") which is greater than the minor diamater of the thread.

This is an increase in area of about 35% and hence an increase in stiffness of about 35%.

Is it accurate to consider some of the current studs to be reduced shank?

If we use Dilavar as a material then the difference in coefficient of expansion between steel and Dilavar for a given diameter results in a change of around 50% .

This means that a Dilavar Stud is around 50% less stiff in terms of increase in force.

If we switch to Titanium then the modulus is around 50% of the value of steel and the expansion coefficient is 75%.

If we use a design of stud that is similar to the original steel stud then the Titanium part would be effectively be about mid way between Dilavar and Steel.

If the current fully threaded Dilavar stud is considered its effective stiffness is based on a minor diameter of about 8.1mm which results in a stud which has a very low effective spring rate.

????

[Henry Schmidt]

Originally Posted by chris_seven

The preload required to stop the head lifting is a given - it shouldn't really change whether the engine is cold or hot??

This is where we disagree.
The clamping force on the heads when cold, is not adequate to prevent significant head movement.
I have posted many examples of head movement from insignificant clamping force, primarily from Dilivar studs both old and new versions. Now that Porsche is recommending a higher initial torque spec, perhaps that movement will be reduced.
Your suggestion that Porsche engineers did not expect an increase in clamping force as the engines heats up seems inconsistent with what we observe.

One last question: If Porsche engineers did not see a need for increased clamping force, why did they increase the torque requirements from the 964 3.6 to the 993 3.6 using essentially the same heads and cylinders?

[chris_seven]

Originally Posted by Henry Schmidt

This is where we disagree.
The clamping force on the heads when cold, is not adequate to prevent significant head movement.
I have posted many examples of head movement from insignificant clamping force, primarily from Dilivar studs both old and new versions. Now that Porsche is recommending a higher initial torque spec, perhaps that movement will be reduced.
Your suggestion that Porsche engineers did not expect an increase in clamping force as the engines heats up seems inconsistent with what we observe.

One last question: If Porsche engineers did not see a need for increased clamping force, why did they increase the torque requirements from the 964 3.6 to the 993 3.6 using essentially the same heads and cylinders?

I don't really disagree about the need to increase clamping I am just trying to find some way of quantifying the problem.

I can certainly understand this requirement with larger and more powerful engines using the same basic thread sizes and surely we need to increase clamping when the engine is both hot and cold, which is consistent with Porsche practice.

Where my only concern lies is that when we increase clamping we are more likley
to pull the studs out of Mag cases - and I do completely agree with the use of Case Savers but I question if we need the same level of preload on a 1969S engine as we do on a 3.6 Turbo.

If the argument is to increase the clamping force on the engine when cold and then reduce the effects of temperature increase by using a stud with a lower stiffness then I could agree quite happily for Aluminum engines.

I just feel that there may be alternatives solutions that may better deal with mag cases and as I have to build myself a 1969S engine on an original case for an FIA rally Championship I dont' want to do the wrong thing.

I discussed Grade 5 Titanium Studs in a thread a few months ago and my concern was one of galling and the relationship between torque and preload.

I bought a suitable load cell to make some measurements about 4 months ago and we have found that by using a Boron Nitiride based Dag we have obtained very good and consistent results using some Ti studs that currently have machined rather than rolled threads. The next step is to try rolled threads.

I have yet to devise an experiment that measures the increase in load at a
Delta T of 125 degC but hope to have this done in the next few weeks.

[Henry Schmidt]

Originally Posted by chris_seven

I don't really disagree about the need to increase clamping I am just trying to find some way of quantifying the problem.

I'm confused; in an earlier thread you stated that you disagreed with my assertion that increased clamping under warm conditions was good.
What changed?

Originally Posted by chris_seven

Where my only concern lies is that when we increase clamping we are more likley
to pull the studs out of Mag cases - and I do completely agree with the use of Case Savers but I question if we need the same level of preload on a 1969S engine as we do on a 3.6 Turbo.

.......edit......

The clamping pressure should not be the same for a 2.0 liter mag case engine as a 3.6 turbo.
That is why our product is sold with different torque specs based on application.
I believe (based on hundreds of disassembled engines) that increased clamping force is desirable to a point. That point moves based on application. My argument regarding mag case engines has always been that you must control the heat production. It's not enough to cool the engine but you must not generate the heat. That means limiting the horse power produced. Next, we have to generate enough clamping force to produce a stable joint between the head and cylinder. To that goal we created a head stud that could have a low initial torque valve and still create a stable platform. The challenge was creating stud and nut combination that had a low initial torque that would not vibrate the nut off. Next we must reinforce the case to accept the additional force. With CaseSavers we accomplish just that.
Of course too much clamping force can create negative results but what is too much? We believe we have a functional balance and have 25+ years of data to draw on. There are many engine builders out there with years of success using varying formulas so by all means experiment but keep in mind many of the "new" paths have already been tried and rejected.

[blue72s]

Originally Posted by Lapkritis

It is interesting to note the air-cooled vw camp replaces the 10mm head studs with 8mm because the 10mm pull from the case more often

Despite using cast iron cylinders?

[suprxman]

Just dropped my 80 sc and heard a clinking sound as we wrasseled the engine and trans out from under the car.
I asumed clinking was bad and pulled the valve covers. ALL 12 of the dilivars were broken. This is a 67k mile car and po and original owner was a little old lady from Texas. Really.
I'm trying to come up with an art project to make from the studs.

[Lapkritis]

Originally Posted by blue72s

Despite using cast iron cylinders?

...and only making 36hp. Lots different over there.

[James Brown]

Henry, ever toy with using springs under the head stud nuts to compensate for the expansion of the heads/cylinders? just a thought...

[Walt Fricke]

I thought a stud could be analyzed as if it were a spring.

[tom1394racing]

I have never seen a head stud pull from an Aluminum case on a stock motor.

From all of this discussion, it seems to me that the best solution for a stock, Al case, '78-'89 motor with the lower Dilivar head studs would be to replace them with stock steel studs and replace all 24 barrel nuts with 12 point nuts.

[Walt Fricke]

Or 6 point flange nuts, which have a reduced socket size? Those have worked fine on one of my race motors. I've not had a head nut of any type come loose.

[chris_seven]

Originally Posted by Henry Schmidt

I'm confused; in an earlier thread you stated that you disagreed with my assertion that increased clamping under warm conditions was good.
What changed?

Nothing changed.

I still disagree that you only need increased clamping under 'warm' conditions.

If the engine develops too much force to maintain correct clamping it will suffer from this problem when cold.

I disagree that using expansion to ensure correct clamping is a sensible design approach.

The best approach is to design the joint correctly and ensure that it survives when hot.


Anything that then reduces the increase in clamping due to expansion must be a good thing.

I can't see how changing the type of steel - as long as the CoE doesn't change- creates this result.

I can agree that reducing the spring rate of the stud will help but the standard steel stud is quite good in this respect.

In this way I just don't understand the dislike of Dilavar - other than the brittle failures - which I do agree make the material suspect. From a clamping point of view I just don't see the problem.

The only real way to ensure that nuts will not 'vibrate off in a controllable manner is to use a prevailing torque device as is done by Casper Labs who use a Carpenter Alloy that is very similar to 17-4PH.

I looked at using a Simmonds Precision 'Silver' plated Kaynar Fastener as these nuts have reduced heads and have a prevailing torque feature but the variation on torque on a nut to nut basis was around 10% of the 24lbsft needed on a stock stud and I felt that this was an excessive error for this application.

I think Casper Labs may use a Binx Type Nut which may have less variation buts is 14mm AF

I think the 12 point nut with the serrated flange may go some way to achieving this but I am not 100% sure that this works with a hardened and ground washer.

This issue would be less of an problem with later engines and higher basic torque settings.

Originally Posted by tom1394racing

I have never seen a head stud pull from an Aluminum case on a stock motor.

From all of this discussion, it seems to me that the best solution for a stock, Al case, '78-'89 motor with the lower Dilivar head studs would be to replace them with stock steel studs and replace all 24 barrel nuts with 12 point nuts.

This is exactly the point I have been trying to make for quite some time

I will continue with the development of the Titanium Studs as this idea does have some merit and it is interesting.

Originally Posted by Walt Fricke

Or 6 point flange nuts, which have a reduced socket size? Those have worked fine on one of my race motors. I've not had a head nut of any type come loose.

We have seen a few SC motors with loose upper studs but we are not sure if this have been due the the differential expansion issues or the failure of the CE ring - Chicken and Egg situation.

I do really think mixed studs is just bad news.