Heat and head expansion

[cgarr]

Just an FYI

Wanted to measure a cylinder head at different temps while I had to install some seats anyhow. From the cam carrier surface to the sealing surface factory spec is 84.5mm or 3.3267 Heated to 250 degrees it grew to 3.3327 and at 300 expanded to 3.3352 that's 8.5 thousands larger than cold. I didn't check any cylinders or studs but seems like some serious stress on these parts, Must be why most go with good head studs and inserts in mag cases.

[itsme]

Would be interesting to do the same test to different head studs to see what the expansion rates of the different materials is. ie cold length vs hot length and compare to your head figures.

[Lapkritis]

Don't forget the cylinder too... I had worked these numbers and forces up last year and shared in my top end thread. Head with aluminum cylinder measurements against head with iron cylinder measurments. Simply put, the old timers don't want to hear it. Just listen to them and do as they say.

[Walt Fricke]

Craig - tricky part here is distribution of the heat. Got to be hottest up at the top. Not at all that hot farther down. So actual expansion in a running motor would be less, wouldn't it? If you knew the heat distribution, you could recalculate (somebody could - not sure I have the math skills myself - maybe not too hard if the distribution is linear).

[Lapkritis]

Hottest at the top depends on thermal conduction of the materials. Aluminum is roughly 4 times more thermally conductive than cast iron. You're going to get even greater growth with aluminum because the heat from the spark plug end will more easily carry down the cylinder body of aluminum relative to cast iron. Aluminum also has a higher coefficient of thermal expansion. If you heat a 1" bar of aluminum at the same rate as a 1" bar of cast iron, the aluminum will grow more and thus exert greater force when constricted beneath a fastener.

[Bill Verburg]

Originally Posted by itsme

Would be interesting to do the same test to different head studs to see what the expansion rates of the different materials is. ie cold length vs hot length and compare to your head figures.

Bruce Anderson published that years ago

[Henry Schmidt]

Good information Craig and Bill

Originally Posted by cgarr

I didn't check any cylinders or studs but seems like some serious stress on these parts, Must be why most go with good head studs and inserts in mag cases.

There are stresses but to characterize these forces as serious may be an overstatement.

I think I would characterize them as calculated and understood in the engineering process.

The design engineers at Porsche fully understood the forces created by thermal expansion and considered all of these dynamics when designing the 911 engine. Neither the advent of mag cases or Nikasil cylinders in and of themselves was the problem.

What happened with the mag case was the addition of ancillary equipment (75-77 2.7) that caused the engine to run temperatures outside the design parameters. Every 2.7 car delivered in California (perhaps 50% of the world wide sales) during these years, regularly (by design) saw increased head temperatures, oil temperatures in excess of 245 degrees while running and cylinder temps as high as 900 degrees immediately after shut down.
This excess heat damaged the mag case requiring the addition of an insert to repair the threads.

Magnesium cases that were not exposed to these crazy temperatures rarely saw issues with head stud pull -out, NO matter what stud was used. Cases exposed to high temperatures not unreasonable stresses are the issue. With properly installed CaseSavers and temperature control, stud pull-out is no longer an issue.

The reason to install the best stud possible is to create a stable platform in which all the components function as designed.

[chris_seven]

There is, of course, still the issue with the change in mechanical properties that occurs in many Magnesium and Aluminium Alloys subjected to even relatively moderate temperatures for relatively long period of time.

The effects of stress relaxation shouldn't be underestimated in engines that are now around 50 years old - a factor which was unlikely to have been considered back in the early Sixties.

We should also consider that the differential expansion of the stud and the engine components is only one part of the equation which governs the forces developed when the engine heats up.

The elastic design of the stud/bolt has an equally relevant impact on the forces produced.

The more elastic the bolt the less the force increase due to expansion. Clearly if we make the bolt too elastic it may yield as it could have insufficient strength.

It relatively easy to define the expansion of a cylinder and the forces that will be produced.

If we assume that the cylinder and the head are infinitely rigid to simplify the calculation then we can easily find the means to model the increase in pull out force and rank individual studs with respect to this force.

If we begin with a standard steel stud with a shank diameter of 7.7mm and assume an expansion due to an average temperature increase of 150degC the increase in pull out force will be approximately 1500 lbs per stud.

A Dilavar Stud with the same shank diameter (7.70mm) will give an increase in pull out force of just less than 500lbs.

If we take an Inconel or other Steel Stud with a shank diameter of 9.5mm then the pull out force increases massively to more than 4000lbs.

This could be a real pain for an old Mag case which has already lost some of its properties.

The real stresses produced would be a bit less than these estimates as some deformation of the head and cylinder will occur but they give some idea of the importance of bolt elasticity.

A Titanium Stud with a 7.7mm shank will produce an increase in pull out of around 800lbs due to its lower modulus. Not quite as good as Dilavar but nothing like as expensive to produce and may be a good compromise for early mag cases.

[Lapkritis]

Age, thermal expansion stress... anyone think about harmonics of engine operation? The acoustics of a running engine can do funny things. I've seen ears snap off alternators...combined heat and weight/vibration break exhaust manifolds etc. Surely, even if at a very minor relative level that would play a part on these older engines combined with the other considerations.

[KTL]

The Liberty Bell was a casting issue/brittleness, no? Or did the Great Chicago Fire cause it to expand too much and crack?

[Lapkritis]

Correct-o-mundo ... so called brittle metals and harmonics don't mix.

The Americans were throwing a rager for Washington's birthday and just couldn't lay off the tunes from the bell.

[KTL]

Was producer Bruce Dickinson there and demanded MOAR COWBELL from Blue Oyster Cult?

[Henry Schmidt]

Originally Posted by Lapkritis

Age, thermal expansion stress... anyone think about harmonics of engine operation? The acoustics of a running engine can do funny things. I've seen ears snap off alternators...combined heat and weight/vibration break exhaust manifolds etc. Surely, even if at a very minor relative level that would play a part on these older engines combined with the other considerations.

Good question. Yes we have considered harmonic stresses. The fact is that the 2.4-2.7 crank is poorly designed and actually flexes excessively creating an imbalance in these engines at higher RPMs. The harmonics created by this imbalance is well documented to spit flywheels off. The issue is that the webs/ flyweights are too thin allowing an usual amount of crank flex. The 2.8/ 3.0 RSR cranks remedied the issue until Porsche built the 9 bolt crank which had wider flyweights and narrower rods.

These harmonics are also responsible for cracked main webs (usually # 7) in 2.7 engines used for racing.

[Mitch1]

Originally Posted by KTL

Was producer Bruce Dickinson there and demanded MOAR COWBELL from Blue Oyster Cult?

"I gotta a fever...", haha, too funny Kevin. So classic.

[chris_seven]

I am not sure harmonics will have much of an impact on head studs.

The maximum excitation frequency the head stud will see is around 60Hz which is well below their natural frequency so resonance won't occur.

60Hz cyclic stresses could cause fatigue damage to occur in the threads of the case particularly if it has already lost some of its mechanical strength due to stress relaxation.

There really isn't much that can be done about these stresses other than the installation of case savers or Timeserts to reduce local stresses.

Studs with improved elasticity will also help as again they will reduce local stress.

I am surprised that 911 crank/engine cases suffer from lateral vibrations and if a crank were to reach its first order critical speed (lateral) I could imagine it destroying the entire crankcase.

Small 4 cylinder Coventry Climax engine with 3 Bearing cranks suffer from this problem.

The normal failure of these engines at high rpm is that the crank deflection causes the centre main to rub, leading to seizure which in turn destroys the engine block.

I feel that the issues with the 2.4/2.7 Cranks is due to gas torques creating high order torsional vibrations.

These vibrations lead to the type of behaviour that will throw off flywheels and cause fatigue failure of crankshafts.

They are the reason that all in line six cylinder engines are fitted with crankshaft dampers or they tend to fail in minutes rather than hours.

By improving the counterweights on the crank the small amplitude high order torsional vibrations will be more easily damped and the problems resolved without too great an impact on crank inertia.

There is another issue that could be due to torsional vibration and a harmonic resonance.

As the crank twists due to torsional vibration it will also to shorten and this will create an axial thrust.

The engine through bolts will also be seeing an axial vibration and if this vibration approaches their natural frequency they will lose clamping.

The axial forces being generated by a torsionally vibrating crank will cause the cases to shuffle if and when the through bolts lose clamping.

The RSR through bolt with the 'lumps' in the centre will not have an eigen resonance so are less likely to lose clamping.

Bells, of course always operate at their natural frequency

[aquazulu2]

My limited experience with 911 head bolt failing is recent, where two head bolts/studs broke while the engine was sitting on a stand. Although I have reads several post where individuals have found broken studs when valve covers were removed, I assumed that the bolts had failed due to over expansion (heat). Reading the comments on this thread seem to attribute a variety of reasons that may cause the delvar stud to fail. My question; could stresses of the engine sitting on the stand cause a head stud to fail. The two studs that failed show that a section of stud had failed earlier, seem to have been stretched and cracked causing that area to show signs of rust approximately 1.5 MM in an area that measured 7.5 mm just below the lug. I plan to replace both bottom row of head bolts with steel studs, should I also replace the top row of studs?..

Thanks for any advice.

[KTL]

Originally Posted by aquazulu2

My limited experience with 911 head bolt failing is recent, where two head bolts/studs broke while the engine was sitting on a stand. Although I have reads several post where individuals have found broken studs when valve covers were removed, I assumed that the bolts had failed due to over expansion (heat). Reading the comments on this thread seem to attribute a variety of reasons that may cause the delvar stud to fail. My question; could stresses of the engine sitting on the stand cause a head stud to fail. The two studs that failed show that a section of stud had failed earlier, seem to have been stretched and cracked causing that area to show signs of rust approximately 1.5 MM in an area that measured 7.5 mm just below the lug. I plan to replace both bottom row of head bolts with steel studs, should I also replace the top row of studs?..

Thanks for any advice.

There have been instances of dilavar studs (older style, non-fully threaded) failing in new installations, not even being subjected to the heat of engine operation. The stress of sitting on the stand isn't necessarily the reason for failure. Its a fault of the stud itself.

I would not bother replacing the top row of steel studs. They are very durable and they never seem to have problems.

[chris_seven]

The studs that you have that have evidence of prior defects must have been at or around their failure stress when you tightened them.

The corroded surface you observed is clear evidence of a fatigue crack and Dilavar is relatively notch sensitive at room temperature.

The main reason for failure of Dilavar must be due to fatigue assisted Stress Corrosion Cracking. (SCC)

Old Dilavar studs that have broken have fracture surfaces that would typically be associated with this type of behaviour.

Precipitation Hardening Austenitic Stainless Steels (Dilavar is this type of alloy) do exhibit a sensitivity to chloride environments.

The failure starts due to corrosion at the grain boundaries of the steel which form small cracks or pits on the surface.

Cyclic stresses than cause these surface defects to grow into the body of the material. Once the crack grows to a critical length the stud will then fail in a brittle-elastic manner with the failure being in the plane of the maximum resolved shear stress.

There is also evidence that brand new studs have failed in engines that have never been run.

The mechanism that would cause studs to fail in this manner would be completely different to an

This type of behaviour is commonly associated with Hydrogen Embrittlement (HE) where atomic hydrogen dissolved or diffused into the steel moves through the grain boundaries until it finds other hydrogen atoms to combine with and form a hydrogen molecule.

This normally occurs where three grain boundaries join and the stresses produced are sufficient to develop internal cracks within the body of the material.

If these cracks are sufficiently large the preload in the stud can cause a catastrophic failure.

Materials such as Dilavar are not known for their sensitivity to HE and the normal 'fix' should it be a problem would be to increase the Molybdenum content of the steel.

Dilavar has a relatively high Moly content making it more surprising that hydrogen would be an issue.

There are rumours that for a time there were some aftermarket studs available that were masquerading as Dilavar and if these studs were made from materials sensitive to HE and produced in a poor quality environment this type of failure could result.

[aquazulu2]

I introduced two types of chemical to the engine, a degreaser call H-7 to degrease the engine before the valve covers were removed, and a high concentration of PB blaster. The second stud that sheared unlike the first which showed a distinct colour difference, this broken end seem to have been impregnated with the PB blaster, it was completely blackened. If I did not know better one would have assumed that it had failed a long time ago and was just sitting in the valve cover rusting.. Some type of chemical reaction seem to have taken place. The broken portion looks quite grainy
and it seem the stud has absorbed the PB blaster.

Note: I did not tighten or attempt to loosen the studs, I only removed the Valve cover, and adjusted the valves. If any force were applied it was compression caused by rotating the engine through the various Valve adjustment points from TDC.

Your point to a possible chemical reaction is quite interesting, since I flooded the stud portals with PB blaster with intent of doing a valve guide replacement, but had changed my mind out of fear of the timing work. The very nature of PB blaster is a penetrator. I may be reaching here, but' I hope this is not the case I would hate to have degraded bolts waiting to shear off when I try to remove them..

Thanks folks for the feed-back I really appreciate the points.. I will start my tear down tomorrow, have a couple cans of propane and waiting for my stud remover to be delivered..

[chris_seven]

Fracture surfaces are often uneven and will pick up light oils quite easily.

PB Blaster is virtually a mixture of Naptha and light oil and not likely to cause damage.

The studs had probably cracked long ago.

I took one out recently, it came undone looked sound until we dropped it on the bench and it fell apart. There was a very small area holding it together and I was amazed that it came undone without breaking.

I think it extremely unlikely that either chemical you have used has caused any degradation.

Chloride attack of grain boundaries in steel is a process that takes place over years unless temperatures are elevated and engine temps are not really that significant to this process.

There is a huge body of work on this subject published by the Oil and Gas industry.

It is a bit odd for Dilavar studs to show rust as they are an Austenitic (non-magnetic) stainless steel and relatively corrosion resistant.