[Vwdiesel] head gasket woes(way too long)

Tyler "Casioqv" Backman casioqv at usermail.com
Wed Feb 26 05:11:11 EST 2003

I was considering the ARP head studs for my motor, but have decided not
to get them based on what I have learned recently. I am glad, because I
wasn't looking forward to spending $200+ on bolts. At some point I would
like to set up a web site with some sort of FAQ containing some of the
good advice that we see all the time on this, and other lists. Stuff
like Hagar's "Dieselfuel 101."
Anyway, this is what Tom Bryant (a guy that really seems to know what he
is talking about because he as rebuilt a lot of these motors) told me
about the head bolts on VW Diesels:
"No.  Head bolts don't break on these engines, properly torqued or not.
  As is
generally true of most fasteners, if it doesn't break during
installation, it's
not going to break.  I re-use head bolts routinely.  But then, I know
how to do
it without getting into trouble.   (Of hundreds of head bolts re-used and
torqued per my procedure, one has broken, and I attribute that to the
overheating and meltdown my son subjected the engine to.  The engine didn't
fail because of the head bolt; the headbolt failed because of a broken
shutoff valve and the resultant engine meltdown.)"

This is his torqueing procedure:

"Now, for an educational (I hope) tour of headbolt torquing.

In the old days, headbolts were torqued to a specified torque value,
below yield, and that was that.  Re-torquing was generally required after a
thousand miles or so, but once again to the same torque.

The main problem with this procedure is that the head gasket requires a
amount of compressive load, not a certain amount of torque.  That is
equivalent to saying that the head bolts require a certain amount of axial
load, rather than a certain amount of torque.  In an ideal world, one can
compute, through a rather simple formula, the axial load on the bolt that
results from a given torque.  So, engineers use that formula to
calculate the
amount of torque required to provide the bolt stretch needed to properly
the head gasket.

The difficulty is that one of the most important parameters in the
equation is
the friction, which occurs in the threads and under the head of the
screw (or
under the nut).  Friction is hard to quantify or control, and it depends
on the
roughness of the threads and the applied lubricant, as well as the phase
of the
moon and what the mechanic had for supper last night.  :-))

The result is that the actual torque required to produce the necessary bolt
loading is quite variable and unpredictable.  In the old days, they just
bigger-than-necessary bolts, and bigger-than-necessary torques to ensure
sufficient loading on the bolts (or the gasket.)  Nowadays, however,
things are
not that simple.  We can no longer afford to over-design engines.  So
what did
engineers do?  They came up with angle-tightening, which is designed to
any wrench monkey to get the proper load on the headbolts, consistently,

Here's the theory:  You torque all the bolts to a specified torque,
below yield, then you turn each of them through a specified angle.  That
results in passing through the yield point, above which the load on the
bolt is
pretty much independent of how much you turn the wrench.  The bolts are
designed to be able to stretch a fairly large amount after the yield
point is
reached, so it's quite forgiving as to the exact amount of rotation put
on the
wrench, and the load on the bolt is quite uniform.  There's also plenty of
margin in the bolts, in case someone should turn them too much.

Bear in mind, also, that bolts (at least well-lubricated, smooth-threaded
bolts) don't fail due to the "twisting" motion if over-tightened,
rather, they
fail due to stretching.  That is, the axial stresses in the bolt are above
yield, but the torsional stresses are not.  Also, if the bolt is going
to fail,
it will almost invariably fail during the torquing procedure, not later.

So, that's basically the "official" book procedure for torquing D24 head
as well as any other torque-to-yield, or "angle-tighten" headbolt.

Here's how you avoid buying new head bolts:

1. Select your used head bolts carefully to ensure smooth threads, no rust,
corrosion, pits, etc.

2. Compare the used bolts with a new one to see how much stretch they
have.  Reject any that are stretched by more than the width of 1-1/2
over their length.  That's an arbitrary criterion that I've established,
has proven to work in practice with hundreds of head bolts.  I believe
it to be
conservative because the bolts should be able to tolerate a lot more
than that.

3. Oil the threads liberally with (Get this...) Mobil 1 15W-50.  :-))
(Actually I chose that lubricant because of its high viscosity, and because
it's available on my workbench because that's what I use in SU carburetor
That's also what I use to pre-lube most every bearing surface when
engines.  See, I'm not wholly against high-viscosity oil; it has its uses.

4. Now, torque those bolts in sequence, around and around the head you
go until
you reach 80 ft-lbs or so.  Now go around the head again, torquing each
until the torque no longer increases (at least not by much) as you
continue to
turn the torque wrench.  When that happens, you have reached the yield
and further tightening is pointless.

What you will find is that the maximum torque is quite possibly in the
range of
80 ft-lbs to 140 ft-lbs.  I've seen both extremes, and both are perfectly
acceptable torque readings.  The reason for the variation is that the
forces on the threads and under the heads are so variable.  It's also that
variation in torque that drove engineers to specify angle-tightening, which
virtually eliminates the variability.

Now you know all my secrets, and any one of you can properly torque a
used D24
headbolt.  :-))


LBaird119 at aol.com wrote:
>   I'd either stick with the non stretch bolts or get Raceware head studs.
> BTW has anybody done the ARP studs yet?  I'd suppose they should
> be similar.  There's sure a difference in price though.  I've been happy
> with the Raceware.  You can feel the difference when torquing them.
>      Loren

More information about the Vwdiesel mailing list