Specific Question - ARP Head Studs on 2.7 Magnesium case - Insane?
 

I searched and read.... Wondering what the experts say - I have the new studs already but have scared myself by reading. I plan to load to ARP spec. Case already has timecerts and front oil cooler in cold climate (New England) for entire life. 1974 so not a case subjected to the thermal reactors. It has been suggested I go with the generic Canyon studs sold by the host here instead... I'm not sure I see the benefit over ARP though because the expansion rates are about the same with only lower loaded torque. if I'm going to have an issue with a cert pulling then I expect I would whether it's with ARP or Canon. Anyone?

As I discussed in a recent post I find the use of Aero Engine grade material a bit over the top.

ARP in the UK tell me that their 911 Head Studs are made from Inconel 718.

Its room temperature CoE is about 12.5 ppm/degC compared to 8720 Steel which is about 11.6ppm/degC.

Not a great deal different and not enough to worry about.

The difference in the forces due to expansion will be about 7% higher and not something I would worry about.

I am not sure there is any benefit to using the Canyon Studs compared to the ARP except cost and if you already have ARP why buy more?

It is interesting to note that ARP say on their Technical page that Inconel 718 is 'completely immuse' to Hydrogen Embrittlement.

Whilst I don't think that HE is an issue for 911 head studs - so please don't worry - I seem to remember that Inconel 718 does have some issues in this area.

Vol.35 No.10 pp.689-694

Thanks Chris - I appreciate the feedback. I've been harassing the ME's that I know for their insight and one who works for Lockheed is going to run the numbers this week. Your line of thinking jives with what I had originally researched. I'll likely stick with the ARP and thoroughly clean the threads of the case before install.

As far as HE with the inconel 718 - I have SharkHide that I planned to coat the aluminum of the engine with. This is also compatible with steels and other metals... probably no harm in coating the shaft of the ARP as well once they're installed and torqued. Any feedback there?

I just had some Timeserts pull on my 2.7 that was rebuilt about 10 years ago. I had used all Dilivar studs at that time, and the engine has not been abused or overheated since the rebuild. It is now at the machine shop having case savers installed, so even the Timeserts are not bulletproof.

Re: Specific Question - ARP Head Studs on 2.7 Magnesium case - Insane?
 

Timesert and Dilavar and still pulled... yikes.

-Andrew

I just can't see a source or externally applied hydrogen being an issue with a 911 engine as there is nothing associated with the environment to produce nascent hydrogen.

Inconel 718 is very unlikely to corrode even if you don't apply a coating.

I think that a standard Time sert uses the same size tap as a Helicoil Insert and this could mean that you still have space to fit a 'Big Sert'

These have been designed to use when helicoil or other inserts pull out and if there is enough case material they may provide a second shot at a solution.

http://www.timesert.com/html/bigsert.html

I think case Savers are a similar idea.

Re: Specific Question - ARP Head Studs on 2.7 Magnesium case - Insane?
 

Would it make sense to go incrementally higher with the expectation that they all pull eventually and the longest life for the case would be refitting the next largest saver each repair cycle? It doesn't sound like there is a bullet proof solution for these.

CaseSavers with Supertec head studs. Hundreds of sets sold and personally installed to date, not one complaint about studs pulling.

Re: Specific Question - ARP Head Studs on 2.7 Magnesium case - Insane?
 

Henry, what material are your studs made of?

Aged 17-4PH

The design of the stud is equally if not more important than the material.

I see there are cast iron cylinder options which would maybe negate the materials expansion issue that are present with aluminum cylinders. Would installing iron jugs be part of a reliable fix?

Here are some shots of a mag case engine with 20K easy miles.
Timecerts and Dilivar. Only a few studs were loose but every head was leaking.

http://forums.pelicanparts.com/uploa...1363209048.jpg

3.5 liter race engine with 60hrs. Every head was leaking. 993TT Dilivar stud. Anyway you slice it, they are just plain crap.





http://forums.pelicanparts.com/uploa...1363209388.jpg

http://forums.pelicanparts.com/uploa...1363209415.jpg

What about the QSC iron cylinders? That would be less stress on the case than aluminum jugs I imagine. Are they junk?

How about JB Racing Aluminium Fin Steel Cylinders

JB Racing - Porsche Engine Components

I didn't like to cut and paste their photograph due to Copyright Issues.

Great link - The billet finned steel cylinders are a work of art. I would love to have those in my budget but I'm not sure I could reason $3000 into this area. I'm not sure quite the benefit of the cast irons... if they are not within spec 95% of the time due to cheap manufacturing then I wouldn't want to waste the time on experimenting.

Re: Specific Question - ARP Head Studs on 2.7 Magnesium case - Insane?
 

I looked closer at JB and I see he previously sleeved the cylinders which meant the shoulder was still aluminum that tensioned on the gasket surface of the case. The new design with billet sink appears to be a larger diameter sleeve that is turned so the steel sleeve is now tensioned on the gasket surface of the case. Working on the sloppy math still for CoE of the iron replacing the aluminum under the head nut. 35mm of the material is aluminum in the head, the rest is cylinder. Should have something to share later tonight once the day job let's me out. If there is much gain reducing pressures on the tensioner/case then I may have to save for another month to have less aluminum under the head nut.

Has anyone measured the temperature of the cylinder while on a dyno yet? I've only been able to find some examples of cylinder head temperature (CHT) but not cylinder temp. This would be useful for determining maximum expansion of the jugs.

We're looking at the impact of conduction/convection of the replacement materials. Fun exercise to explore anyway.

Okay so did some rough math/calcs for CTE based upon certain assumptions but also hard measurements. In a nutshell, you'll have far less tension on the studs with steel cylinders under the fastener. Here's how we arrived at that conclusion:

Coefficients
ARP Fastener (Inconel 718) - 7.2Min/inF
Cast iron - 5.9 to 6.2 Min/inF
Aluminum - 12.8-13.1 Min/inF
(we're working on dilavar numbers to compare now as well).

Here we're using one of my Mahle aluminum cylinders from my RS spec setup for measurements. The fastener is clamping 1.322" of aluminum in the head and 3.367" of aluminum in the cylinder at room temperature of 70F.
http://i458.photobucket.com/albums/q...psebea1ee9.jpg

http://i458.photobucket.com/albums/q...psa07c4201.jpg

http://i458.photobucket.com/albums/q...psb75b4d02.jpg

http://i458.photobucket.com/albums/q...ps68b69b6a.jpg

http://i458.photobucket.com/albums/q...psb020fc18.jpg

Assumption 1 - 400F material temperature of the cylinder during heavy operation giving a delta for our calculation of 330F (400F minus the room temperature of 70F for an increase of 330F).

First combination is what I have in hand now, all aluminum under the head nut for the fastener to hold:

3.367" (cylinder) + 1.322" (head) = 4.689"
13.1 x 4.689 x 330F = 20270.547 Min/inF or .020270547" of growth once the cylinder reaches 400F from room temperature.

Second combination is with cast iron cylinders in place of the aluminum.
13.1 x 1.322 x 330 = 5715.006 or .005715006"
6.2 x 3.367 x 330 = 6888.82 or .00688882"

Combined stack expansion .00688882 + .005715006 = .012603826"

Third item for consideration is the ARP Inconel 718 material to be used:
7.2 x 4.689 x 300F (assuming this runs slightly cooler than the cylinder because there isn't flame/friction inside it) = 10128.24 or .01012824"

At this point it's a fair conclusion that cast iron cylinders with the ARP fastener would be a wise choice to relieve pressure on the case threads due to thermal expansion of the materials above.

Growth of fastener: .01 inches (engineer says he would bet closer to .008 in his experience)

Growth of all aluminum stack: .02 inches
Growth of iron/aluminum stack: .012 inches
Growth of fastener: .010 inches

With the cast iron cylinders we would be reducing the growth of the stack under operation at 400F by roughly 60%. We need a text book to go further with this hypothetical for determining the pressure of the expansion of the stack relative to the sheer strength of the magnesium threads in the case. It would also be nice if someone with an infrared temperature gun could confirm temperatures of the items during or immediately following heavy operation. I believe we'll be able to dive further and review these basic calculations above in the next couple days. I won't be putting my engine back together with aluminum cylinders and ARP fasteners however. Feedback is welcome - we understand the assumptions leave some wiggle in the conclusion.

Andrew,

I am not sure that I can agree with all of the assumptions that you have made otherwise any Nikasil Barrel used with a Magnesium case would fail quite quickly.

The main area that I disagree is the Temperatures you have considered.

I believe that 400 degF as a running temperature is too high and most Elektron Alloys would suffer serious problems at this level.

It is fair to say that it is now possible to find Elektron alloys such as WE43 which are rated for use at up to 300degC but this is a relatively modern alloy.

It also loses about 35% of its yield and 50% tensile strength at this temperature so would need to be down rated

I believe that operating temperature of the engine case is more likely to be around 125 degC and this should be the cylinder bulk temperature as the numbers all become more meaningful. It would be great to obtain good temperature information.

Assuming a good high strength Mag Alloy was used the pull out strength of an individual stud should be about 12000 lbs. Yield would start to occur at around 9500 lbs so load would be lost at around this level.

There is some guesswork involved in this as I don't have accurate material properties.

The 24lbsft of torque will produce a preload of about 3600lbs (mean value) in the stud.

If we assume a Delta T of 100 degC the differential expansion between the stud and the cylinder will result in a compressive strain in the cylinder of around 8.3x10^-4 and using the simple relationship for Young's Modulus this will result in additional force of about 7500 lbs per stud.

The Total force on the stud would then be around 11000lbs which suggests a problem.

I think that the numbers I have used a conservative as many engines survive and my rough calcs indicate that they should fail - but only just.

If you substitue a good quality AluSil material the Pull Out force increases to around 16000lbs with yield not occuring until 12000lbs and as Aluminium Cases don't seem to suffer I can again assume my rough calcs are conservative.

If you increase Delta T significanlty then the pull out forces generated by expansion are directly proportional to temperature and failure is almost assured for both types of case.

A typical Grey Cast iron has a Coefficient of Expansion of about 9.5ppm per degK and this says that the stud should be losing preload with increasing temeprature.

I think the one thing that we can see is that Nikasil Barrels and Steel Studs of any Grade or type of material that is not Austenitic (hence high expansion) used with Mag cases runs close to pull out and the variations in torque/preload may be an issue.

Clearly using a Timesert or Casesaver increases the Pullout force needed and improves the situation.

Great feedback Chris - thank you for sharing. The percentage of aluminum in the head (28% of total compressed by fastener) that would remain and expand would potentially offset the low expansion of the cast iron cylinder relative to the ARP fastener. We are shooting from the hip without real data which will mean we're only working in the general neighborhood. I promise to put the infrared thermometer on my engine once assembled to get a better idea of the true temperatures. To get a better idea of the case strength it was suggested that I contact AMEC for a materials analysis/test and then to consult matweb.com in order to get decent ultimate, yield, and CTE values. I imagine someone has done this already and maybe I just haven't been able to find it on the vast internet yet.

On another note, the ARP fastener calls for a torque of 36 ft-lbs. Would this in itself exceed the case thread capability? Another question for us to answer once we know for certain what the case is made of... we could assume weakest case scenario for a certain margin of safety. How much torque is transmitted through the nut/thread and washer to the root/case with the ARP lube? If the fastener wasn't loaded properly then a host of issues may present.

Although it's fun to watch the metal masturbation, I think you guys are investing way too much in this exercise.
We've been dealing with this issue (stud pull out) since the mid seventies (even earlier in type I VWs) and there really is nothing new to learn.

Too much heat kills mag cases.
Timecerts have a thread pitch that is too fine for head stud repair.
Dilivar studs offer too little clamping force.
Supertec studs with professionally installed CaseSavers have never had a reported pull out.
We have built in excess of 400 air cooled Porsche engines (mostly mag case, race and street) and find this combination to exceed our warranty requirements.

:)Andrew,

I agree that the head expansion will produce a small amount of expansion but this won't do a great deal of damage as the pull out force increase will be quite low.

I would never increase the preload on a stud to a higher level than that recommended by the manufacturer without significant measurement and testing.

The increased static preload can only cause probems and be more likely to cause both stress relaxation and an increased tendancy to pull the stud out of the case.

It would be excellent if Henry would share his design calculations and we could then all give up worrying and be totally confident and stop all the 'tossing off'.

I do think, however, that these studs must create their own special Elastic Singularity and the math is probably beyond my capability to understand . :)

To be serious I think that there is a difference between finding a solution and understanding the solution.

Henry, with all due respect I prefer to do my own homework rather than taking an answer that someone else found that works. There is thinking to be done outside of the fastener.

Chris, you know I'm not an engineer nor have I ever made such a claim so that is why I hired a company that specializes in fastener design to pencil out the specifications of our stud. There's no magic, just a reasonable design.

I wonder if you truly respect real world experience?
Building a 911 engine has a plethora of challenges and investing time in or trying to reinvent the wheel seems counter-productive in the rebuild process.
If you find something that works better than the solutions that have proven results, please let me know. I'm always looking for a proven "better way".
Cheers

Re: Specific Question - ARP Head Studs on 2.7 Magnesium case - Insane?
 

Thanks Henry. Yes, I do respect experience. That said I'm not here for a bay of pigs. There's hard data to collect and analyze and I have some of the best engineers with years of experience at Honeywell, Pratt-Whitney, Lockheed Martin etc helping me along and guiding me. You should be able to respect that they will have the ability to arrive at a recommendation for nearly any solution a 60yr old design requires. The largest root into the case as possible, as you suggest, may be a viable long term solution... but only time will tell as these units cycle under extreme pressure of mismatched materials pounding away. In 30yrs will these savers still be holding? A long term solution would be to eliminate the forces pulling on that root. Stay tuned, you may learn something new like you said.

-Andrew

There are no perfect solutions but there are solutions with proven track records.
Time will tell......oh wait. It seems I have Porsche engines that have won championships and run for 150,000 miles and more. I can name one belonging to a JPL engineer the had 190K on one of our rebuilds. I will bet none of your "experts" will be able to claim that.

I always get a kick out of the new guy who thinks he has some insight beyond that of the professionals who have built these engines for decades.

I wish you all the luck in the world on your first build.

Re: Specific Question - ARP Head Studs on 2.7 Magnesium case - Insane?
 

Your input here is no longer welcome. Good day sir.

You are an interesting twist.....
This is an open forum and as long as proper decorum is exercised, we don't allow individuals to censor other contributors.

I seem to recall you specifically asking for my expert opinion and when I suggested you go a different direction than your "as cheap as possible" solution you no long desire my advise.
Interesting to say the least.

Re: Specific Question - ARP Head Studs on 2.7 Magnesium case - Insane?
 

Good day sir.

Ok...I have to ask a question.
Harley Davidson...for decades...has bolted the base of the cylinders to the block...and then the heads to the top of the cylinders.
This now seems like a probable solution for Porsches.
I know it would require some thinking...principally in the oil return tube...and timing chain areas...but is it possible?
It would have a slight effect on compression when the cylinder expands (in length) as the engines gets hot...but that in itself may be a good thing.
Brain working overtime...sorry 'bout that.
Bob

Re: Specific Question - ARP Head Studs on 2.7 Magnesium case - Insane?
 

Interesting notion. I think the issue might be weight hanging that far out. Are the old BMW motorcycle engines the same with base cylinder mounting?

Bob - I get your drift: If everything expands uniformly, no extral thermal expansion stress added to the usual combustion forces.

Do HDs use aluminum cylinders? Somehow hawg riding/wrenching has passed me by.

I'm thinking that the 911 aluminum cylinder walls are quite thin in between the fins, which might make the cylinder marginal for holding the forces of combustion (assuming a design could be worked out so a flange could extend out from the tip of the spigot over where the head studs now emerge, and feed the forces equitably).

Then there is the question of cycle length. As I understand it (from Carroll Smith), if you keep the tensile forces on steel below its elastic limit, you can get almost unlimited cycles from that material (maybe it has to be some percentage lower - but somewhere). Whereas, with aluminum, each flex ticks off one of the allowable lifetime cycles. I think of this when I look out the window of an airplane and watch the wing flex up and down - comforting myself with knowing the engineers have all that figured out long ago.

VW hot rodders - drag racers, actually - add a 5th stud to help keep the heads on their super stressed motors with Nitrous and Superchargers and whatnot.

Ultimately, all these forces, no matter what, have to be held in by threads in the case. I've not heard anyone say that the makers of inserts are wrong in saying that the wider the threaded hole, the more stress it can handle. Maybe that's a simple function of more area with the same strength of the case material?

Almost all old bikes had iron cylinders...some of the newer designs have aluminum with steel (or iron) liners in them.
I believe the newest design (by Porshe I think) has aluminum/iron barrels...and a very good HP/weight ratio .
Bob

This debate is a bit like Torque and Horsepower - it will continue to run and run.

I have tried a couple of times to summarise the basic issues and I will have onle last try:

There is no real solution to pulling studs out of aged and overheated mag cases.
they were borderline even when fresh.

The use of a Big Cert or case Saver is the only thing that can help if you use any Ferritic/Martensitic Stud material.

This is the area where I disagree with Henry and have done for some time.

The dominant property that affects studs pulling out will just be due to the initial preload + the forces caused by expansion.

The difference in the two coefficients of expansion is the only significant driver.

Having a couple of extra threads in the case may give a 5-10% improvement and this may help move the edge of the borderline.

Using Dilavar does reduce pull out forces significantly and I am interested to understand why it doesn't give enough preload?

I can understand that this may be an issue with High Pressure Turbo Motors but normally aspirated engines won't develop enough peak cylinder pressure to lift heads.

If the engine needed the extra clamping force generated by expansion they would all tend to give problems when cold.

I agree with Henry that practical solutions are good but again where our opinions diverge is about 'proof'.

Good engineering finds solutions and then explains how and why they work. This creates confidence and improves the 'State of the Art'

I would also suggest thet it is usually the 'practical engineer' that refutes science and asks that we all believe in the 'Magic Fix'

I would agree with Bob about motorcylce engine construction but the secret os the cast iron cylinder. As soon as you use an Aluminium Cylinder you have the same problem.

the old BSA Victor used a Chromal Cylinder to replace the cast iron and this shortened crankcase life.


Walt,

The Fatigue Endurance Limit' of most steels is around 50% of the materials yield strength but this would only be valid for smooth bar 'fully reversed' loading. Once you introduce notches and loading that has a tensile mean stress then you need to really look hard at Stress Intensity Factors'.

It is true that Aluminium had no theoretical fatigue endurance limit but you could estimate that you could use up to 35% of the elastic limit for around 10^8 cycles.

This isn't great in terms of steel but does work for Civil Airliner Airframes which are relatively unloaded and the loading profiles used to test these materials and structures are based on wind gust dominated forces.

The main fatigue event of an Airframe is the pressure cycle in the Hull and the 'life' used to be around 100 000 cycles.

Fighter Aircraft are altogether different ans are manouver dominated and use entirely different test configurations.

The first time fatigue was considered an issue with Airframes was the Comet, there were some issues with the DC10 so the Science is relatively new.

Fatigue is not the main issue with case failures.

The forces that really cause the damage are the expansions being superimposed on the static load and it is clearly the heating cycle that causes the load to increase.

The material may also be prone to stress relaxation if the temperature is high enough compared to the local stress at the root of the threads. The relativley high mean loads could cause some microstructural damage that would locally weaken the material.

I have been convinced for some time that the mag cases suffer from this problem and there is a simple force/time/temperature realtionship that governs this behaviour.


If we knew the alloy used it is likley that Magnesium Elektron would have good design data.

I was surprised to learn that ARP recommend 36lbsft ot torque for their 911 studs as this will tend to increase this problem.

My conclusion is that there is no real way to redesign the case in this area but if we understand the basic mechanisms that cause the problems we can chose an optimium solution.

At the moment if I used Nikasil Cylinders on an old Mag case I would use Big Serts and Dilavar.

If I used cast iron cylinders I would probably use Big Serts and Steel - the type of steel is relatively unimportant.

On Aluminium cases I would generally use Big Sert and Steel Studs.

I have based these views on trying to understand the engineering issues and the basic material properties.

The solution used on most modern motorcycle enigines seem to be based around different construction methods with wet liner blocks and lots of short head bolts of small diameter.

Re: Specific Question - ARP Head Studs on 2.7 Magnesium case - Insane?
 

My mickey mouse calculations agree...

Non dilavar fastener, steel cylinder.
Dilavar fastener, aluminum cylinder.

My confidence, or lack thereof, in the dilavar fastener has me looking at non-dilavar fastener and steel cylinder. Simple as that for the sake of the case.

I think the diameter percentage increase of the saver is a good portion the strength improvement. The pitch may be the same but more biting with larger diameter.

The fatigue concern I allude to is based upon the evidence of stock thread holding then failing, timecerts holding then failing...now we're en vogue with case savers and not looking at the bigger picture of what is behind all the previous failures (thread pulling only). Everything so far works for awhile until the cycles get to it. If you cycle the engine to operating temperature and cool down often then the miles don't matter nearly as much.

Foul.

Foul.

Re: Specific Question - ARP Head Studs on 2.7 Magnesium case - Insane?
 

Please keep this thread on topic instead of promoting quarrels. Thank you.

CaseSavers have been used "en vogue" for 30 year by those qualified professionals who build Porsche engines in mass.
It's not a fad but a proven solution.
CaseSavers in combination with top quality studs and temperature control are in the real world performing flawlessly.

For those reading this thread who don't want to reinvent the wheel and want your best chance at success, there are professionals building these engine who have offered suggestions that will give you the best chance.

For those scratching on note pads who wish to spend the next 30 years trying to catch up on your own, good luck.

Re: Specific Question - ARP Head Studs on 2.7 Magnesium case - Insane?
 

The negativity and discouragement to innovate here is notable and disappointing. I expected more from reputable members. Do we all grow so proud of our accomplishments that we truly believe we know it all and then demand progress by all others to stop while touting those accomplishments? The feedback you provide is not constructive and does absolutely nothing to encourage development. Imagine if you received the same reception when you were developing your super-duper da ta da fasteners.

Perhaps you should lead by example.

Having spent my entire life in the Aerospace field, I have to respect Henry’s work; Porsche had issues with these studs even with their R&D departments. If the parameters are to push this engine to Mach 3 and the edge of space, of course it can be built better, but for the guy driving around the street and track the problems have been solved.