head stud torque...why so low?

[bpu699]

Got my Supertech head studs for the 930...

Instructions say torque is 32-37ft lbs???

Geesh... that sounds low...

Most Japanese cars are 50-80lbs. Diesels 150 lbs...

Is the torque setting really that low?!

I figured fancier head studs would have higher torque settings?

[Black_Hat]

different metals have different expansion rates (cylinders & head studs)
once heated up the clamping force is high enough

[Tippy]

Yes, it seems incredibly low for my 993TT studs (which I regretfully torqued mine to causing sealing surface galling), but the older Dilavar's have a small cross section, so the torque is on par.

[chris_seven]

Head stud torque is always a confusing issue.

The principal job is clearly to stop the head to cylinder joint from leaking and I have always believed that the joint must be able to seal against peak cylinder pressure even when cold.

We need to remember that the original head torque on 2.0 litre engines was only 24 lbsft so 37 lbsft is a very significant increase.

For a normally aspirated engine the change in force being applied to the studs by a 3.6 litre engine has increased by 56% - assuming no real change in BHP/Litre.

The ruling element of the design is not really the stud as the tensile strength of virtually all the materials being used is much greater than the forces being generated.

A stock stud has a shank diameter of around 7.8mm so we need to consider the point at which the stud will not be able to withstand load and will start to 'neck'.

I would imagine that a stud with a tensile strength of around 160ksi will withstand around 12000lbs of force.

If we torque this stud to 24 lbsft I believe we will see a preload in the region of 4-4.500lbs which is well within its limit.

If we increase the torque to 37lbsft the preload will increase to around 6500lbs again relatively comfortable.

The impact of expansion is also significant and will depend on the cylinder material.

If we look at Biral Cylinders the increase in force due to expansion is quite low and depending on the stud material the preload could reduce slightly.

Once we move to an Aluminium cylinder then the changes in expansion become significant.

Alusil Cylinders will expand slightly more than Nikasil as the base materials differ but the difference is only around 2.5%.

If we were to make cylinders from alloys such as 6061 however, then the increase could be as high as 30% which could be very significant.

The increase in preload caused by a Nikasil cylinder is around 2000lbs when the shank diameter of the stud is 7.8mm.

This would give a preload of around 8500lbs for a stud torqued to 37ftlbs.

If we increase the shank diameter of the stud then the 'as installed' preload doesn't change but the preload due to expansion will increase in proportion to the square of the diameter.

This means a stud with a shank diameter of 9.5mm will give an increase of around 3000lbs.

If this stud is torqued to 37 ftlbs the total preload when hot could be as high as 9500 lbs.

The stress in the stud, however, would fall due to the increased area.

The real issue is the strength of the thread in the engine case.

If we consider a Mag Case then looking at the shear area of the engaged thread and considering the shear yield strength of the Magnesium I believe a preload of 10 000lbs is the absolute maximum that should be applied and I would like to see some 'headroom' to ensure medium term durability.

As the material in the case deteriorates due to the long term effects of stress relaxation even this figure is likely to be excessive as the material could lose up to 30% of its property.

Case savers of Timeserts would help but I would generally not torque studs in mag cases to 37 lbsft.

I need to re-run some calculations to establish the impact of using a Timesert but it will certainly increase the shear area of the thread.

The Aluminium Cases area little stronger than the Mag Case but only by around 10% but they will retain this property for much longer at typical case temperatures.

I would still consider that large shank diameter studs torque to 37 lbsft are at a realistic limit for the case and again I would think installing inserts may be helpful.

I believe that studs are really a lot closer to practical limits than we imagine.

We have recently been modelling cylinders and forces as we have just finished machining some prototype 98mm cylinders using a Eutectic Alloy and they are currently being plated with Nikasil.

The next stage of trials will be to bolt the cylinders into fixtures and thermally cycle them to around 250 degC and evaluate their dimensional stability.

If this work out well we will then hold them at temperature for an extended time period to try to look at long terms effects.

[Mike Billings]

It's because they are fine pitch.

[Neil Harvey]

Good assembly practices are equally as important as the stretch on the stud.

Head nuts should be tightened by an angular method and not torqued.

[Catorce]

The part you are missing is that on a 911 engine, when you clamp down on the heads you also squeeze the cylinders. Excessive torque in this area would start to deform the cylinders and heads....remember, they are light alloy.

On a diesel or even small block, the head to block joint is a giant plane of flat metal, like 2 slabs of pure metal. They get high torque because it is almost impossible to deform them by over torquing.

A 911 needs just enough torque to hold the heads in place, not a drop more.

[chris_seven]

One of the things we are planning to do is measure cylinder bore deformation when the cylinder is clamped.

It will be difficult to make the measurement with a hot cylinder but we can measure with the cylinder clamped to simulate static preload and then look at increasing the preload to simulate expansion.

We are planning to bore our cylinders using torque plates so the decision about how tightly to clamp them seems worthwhile.

[Mark Henry]

My VW beetle with Porsche 914 Type 4 2600cc, 2.0 heads with ARP studs and 102mm LN nickies (94mm is stock size), 180hp dynoed.

24lbs head torque.

[pampadori]

@chris_seven could you somehow test different clamping torque by using a sealed fluid under pressure to increase the clamp force instead of a metal fastener? I'm just thinking out loud but something like a hydraulic press applying force on the heads where the washers typically reside. Then you could adjust the pressure in real-time and document your measurements, adjust pressure and measure again, etc.
To be fair, me have no real edumacation but thinky real hard like sometimes. could this work?

[Tippy]

Hydraulic forces are easy to calculate, and based on the size of a washer, I do t think you'd have near enough size. But I could totally wrong too lazy to dig up the formula.

Now, a cylinder the size of the piston inside the test mule might do it. Push on the cylinder itself, no cylinder head attached.

I've read the problem isn't getting enough torque, it's distorting the barrels when going higher.

[Mark Henry]

Originally Posted by Tippy

I've read the problem isn't getting enough torque, it's distorting the barrels when going higher.

I've seen this real world with head sealing on 911. I've seen more sealing issues with ARP studs which call for an big increase in torque than with stock torque value studs.
I've also seen this on VW heads where the builder over torqued the heads, including VW type 4 with Nickies.
Aircooled VW is very similar to AC Porsche engines on head sealing issues. Too much head torque is a bad thing in most apps.

[Tippy]

Mark, to me, its such an Achilles heel. Not for the NA engines, but for us boosted guys. We need lots of torque to prevent lifting the heads.

[chris_seven]

Originally Posted by Tippy

Mark, to me, its such an Achilles heal. Not for the NA engines, but for us boosted guys. We need lots of torque to prevent lifting the heads.

EB Weld them on maybe ?

Not too difficult to achieve and if you buy Billet machined heads and cylinders the alloys maybe more suited to welding.

[Mike Billings]

Now I see the pitch is irrelevant

T=KDP

T = Torque (in-lb)
K = Constant to account for friction (0.15 - 0.2 for these units)
D = Bolt diameter (inches)
P = Clamping Force (lb)

Screw diameter matters so the Chevy 7/16" bolt would need 11% more torque to get the same clamp load as the Porsche M10.

[Tippy]

Originally Posted by chris_seven

EB Weld them on maybe ?

Not too difficult to achieve and if you buy Billet machined heads and cylinders the alloys maybe more suited to welding.

Hmmm. Sounds very pricey!!

[Catorce]

Originally Posted by Tippy

Hmmm. Sounds very pricey!!

Yeah, pricy like only used on 959s and 962s pricy......

[chris_seven]

Originally Posted by Catorce

Yeah, pricy like only used on 959s and 962s pricy......

May be not so bad now.

Modern EB systems can use much lower vacuum levels than first generation systems which needed around 10 x ^-6 Torr and the time required to pump down the chamber containing the work was significant.

The welding time for a head to barrel would be less than 60 seconds so the costs would be to cost to develop a jig, to clean and dry the parts so they don't outgas (The guys doing the welding would have a suitable plant/process) and load them in the machine.

Pump down time with new Turbo pumps would be short so it could be manageable.

We made a batch of gears for our rally car project a few years ago and we EB welded the dog rings in place.

The cost for 30 gears was about $10.00 per gear but the machine did use a 'soft' vacuum.

A colleague of mine used to make Titanium Propshafts for the Subaru WRC Rally Cars and they used EB and costs were not insane as the complete prop only cost around $1750.