WHY? Generic Engine Questions for the ignorant:

[mtelliott]

Valve Adjustment: Why does the connection between the rocker and the valve get out of adjustment? Does it have to do with thermal expansion and contraction over time? Is this something that Porsche has fixed in newer engine design? And why do Porsche’s have this problem as compared to other engine manufacturers?

Torque Settings: As I understand it, torque settings are generally set to approximately 90% of bolt failure. Is this correct? Why then do you torque valve covers to a lower setting than the sump plate even though the bolts are bigger on the valve covers? I’m guessing that is has something to do with gasket design?

Aluminum Engine Design: How much more would an engine weigh if it was made out of steel? I understand that weight saving was a decision criteria however what other reasons for going with aluminum over traditional steel? Does heat dissipation have something to do with it? And while we're at it, why did Porsche go from Aluminum, to Magnesium cases? What made them want to try magnesium?

[sammyg2]

Valves and valve seats and rockers and cams and bearings wear over time.
If the valve seat in the head or the valve itself wears the clearance will get tighter.
If the cam or rocker or rocker bearing or cam bearings wear the clearance will get greater.
adjustments are really only supposed to be valve clearance checks, but occasionally it is necessary to make adjustments.

Modern engines (including porsche) are equipped with hydraulic style lifters (or adjusters) instead of an entirely mechanical set up. That means they adjust themselves. Used to be that hydraulics limmited the rpm but they got it figured out better nowadays.
When they first designed the 911 engine there really wasn't a reasonable way to do hydraulics on an OHV engine, plus it was generally regarded as a higher performance design.

Some bolts are designed to be taken to 90% of yield but that is the exception and not the rule. Most are not supposed to be torqued that much. You are correct on the valve covers, the torque is designed to the gaskets (or somethimes the parent metal), not the actual bolt. If you tried to take those studs to 90% of yield they would rip out of the cam towers before you got there.


They use aluminum to save weight.
The switch to magnesium was also for weight savings. Possibly to save money too, not sure about that.
As the engines got bigger they went back to aluminium for strength after the head studs ripped out of the magnesium.

an all steel 911 engine would weight a *****load. If I had to guess I would say somewhere around 800 or 900 pounds.

[Jubbie]

Thermal conductance has a lot to do with it.

Material = Thermal Conductivity (W/m k)
Steel = 47 - 58
Cast Iron = 58
Cast Aluminum = 209.4
Magnisium Alloy = 70 - 145

Aluminum is great at wicking heat away

Mag is lighter, but harder to cast. The 2.7L were more than the mag cases could bare.

[Wayne 962]

Originally posted by mtelliott
Torque Settings: As I understand it, torque settings are generally set to approximately 90% of bolt failure. Is this correct? Why then do you torque valve covers to a lower setting than the sump plate even though the bolts are bigger on the valve covers? I’m guessing that is has something to do with gasket design?

Definitely not true. Most torque wrenches are only accurate to something like 10-30% anyways (depends on lubrication of the joint, etc.). So with this in mind, most of the bolts torqued would snap.

In most cases, it's much, much, much, much lower than that...

-Wayne

[F1BADUDE]

I've done a few accidental torque to failure tests at work using an extreemly accurate $20K nut runner. Generally after we accidently run the torque using the wrong program ( like running an M8 class 9 nut nut using an M12 class 10 program) we will graph out the failure. Although it varies quite a bit from joint to joint you can generally tell that the stud will start to yeild at over twice the recomended torque. The graph will look like an EKG of a heart attach at just under twice the recommended torque and then will drop off pretty rapidly until ultimate failure.
You'll think, OH S%^T when it starts to creek (that's the heart attach) Worse part is, this is usually about a $250 oops on a prototype mechanism!

[NevenM]

Hi

Dont you just hate that, My advice for anyone tighening a bolt is that the torgue you apply should steadily increase as you turn the bolt, Turning the nut increases strain (elongation) which should increase stress linearly and therefore the torque you apply. If you are turning and getting no 'tighter' stop. This happened to me the other day on a case perimeter nut, thankfully it was the crush washer failing.

My current problem is the opposite, I have just discovered an
exhaust stud broken half way down the hole (I wondered why that stud was longer) Hmmmm

Neven

[Jim Sims]

Fasteners subjected to fluctuating stresses such as connecting rod bolts are often tightened to within 80 to 90% of the proof strength (which is roughly equivalent to the yield stress). This works, as a significant portion of the stress is the torsional shear stress from tightening which disappears when tightening is finished or shortly afterward. Sometimes fasteners are loosened a small bit to ensure the torsional stress is released. Recall the OEM 911 connecting rod bolts are tightened to 100% of the proof strength or yield stress which is a method of offsetting the inaccuracy of torque wrenches. The torque for a fastener is a function of thread friction, fastener size, type (screw, bolt or stud) and grade or class, the strength of the material it threads into and the loading it will be subjected to. Cheers, Jim