Date:         Mon, 8 Dec 1997 10:26:22 -0400
Reply-To:     mark keller <kelphoto@BRIGHT.NET>
Sender:       Vanagon mailing list <Vanagon@Gerry.SDSC.EDU>
From:         mark keller <kelphoto@BRIGHT.NET>
Subject:      Re: heads: corrosion and leak prevention discussion.
Content-Type: text/plain; charset="us-ascii"

Greetings Tim,

Thanks for the response.

Your explanation of crevice corresion seems to be  saying that when a
properly prepared coolant, one that still has good anti-corrsive qualities,
is put in an isolated enviroment, here sandwiched between aluminum and
rubber, that it quickly becomes "exhausted" and then the exhausted liquid
is very corresive.  I'm assuming this is a residue, since the liquid may be
steamed out once the engine is hot. Ok, so much for how the damage occurs.
What do we consider to keep the coolant from getting under the gasket?

The cylinder head I was shown,  had been removed from a NON LEAKING engine.
It revealed corresion, across the entire surface, suggesting to me that the
corresion was not eating it's way to the outside world, so that it would
look like a very depleted area from the water jacket to the point of
failure.

This corresion was VERY consistent across the entire flat surface. I
confess I didn't not check for a line where the metal gasket surrounding
the cylinders or the O-ring line, so I cannot say unequivocally that the
corresion extended into those areas.  The part that stunned me was that
there was not a area of excessive depletion by corresion on the edge of the
head closest to the water jacket, but that the corresion pattern suggested
that coolant was actually getting beneath the seal, and then setting up
crevice corresion as you pointed out.

So, we are still left with preventing the coolant from getting beneath that
seal.  I would hold to the idea that the seal is much more easily
compromised when the engines is cold, AND, operated at High RPMS which
would generate maximum water pressure, and thus force the coolant beneath
seal. As it was pointed out, those flexiable seals are not just relying on
clamping force but on an adhesive to keep coolant from intruding. I would
concluded that the seal is vunerable in low engine tempatures ranges, that
is before the aluminum water jackets expand,and clamping force is at an
operational value.

I would still say that the unique conditions, of the vanagon,and the
engine's operational RPM excerabate the compromised situation of a cold and
thus unexpanded cylinder water jacket/ head area to the flexibale seal.
Plainly stated, it's easy to get in the van, crank it up, and be turning
3000 + RPMs before complete expansion of the water jackets takes place.
Couple that to an aggressive water pump, which has a big job to do, and a
cold viscous liquid, and you get hydralic pressure, which penentrates
beneath the cold flexibale seals.

The design of an aluminum water jacket, and it's expansion related to the
steel cylinder liner would seem to create a sealing and warpage  problem
for the head. Depending on the relationship of the steel stud to the steel
cylinder liners and the edges of the aluminum water jacket surfaces. I
imagine that some type of give has to occur when the engine's temperature
changes. My thinking it that the outside seal would get this responsibility
and is less resilient at low temperatures and thus more prone to invasion
by hydralic pressure.

So we are back to: How do we keep this head area surrounding the water
jacket un-molested by coolant? My reasoning is that be gentle with the
motor until it is warm enough for the aluminum to expand and bring the
clamping force up to an operational value, then coolant will not be force
beneath the seals.

I realize that my suggest is that we as drivers practice an active type of
management,ie wait until the engine temp is say at least showing some
movement on it's guage, which is relatively uncommon in typical cars. Ie,
usually we just get in, fire up, and take off. As a helicopter pilot, in
both piston and turbine powered aircraft, it seemed I was always practicing
some active management of a temperature or rpm range. This practice is very
common in all of aviation, because light alloys are used and the limits are
pushed, much like in the vanagon.  Perhaps the same type of active
management would reduce cylinder head corrosion in the wasserbox.






>Hi Mark,
>
>the head problem is due to 'crevice corrosion', where a small region of
>coolant is trapped, in our case under the gaskets.  this sets up
>localized corrosion pockets, electrolytic action is way higher than in
>the main coolant.  Using soft gaskets in any corrosive environement is
>an absolute 'no-no' in the corrosion engineering field.  I teach this
>stuff.  VW adapted an aircooled, they didn't design a wasser engine.
>They didn't have a corrosion engineer around at the time I'd guess.  The
>head is clamped tightly onto steel barrels, unusual to leak there, but
>the much larer expansion of the aluminium water jackets needs somewhere
>to grow.  Heads the fat soft gaskets. Bad move. Subaru designed a wasser
>engine from scratch, head bolts to 'block' with usual stiff head gaskets
>that seal fully under high preload.  Subaru engines go 300K kms before
>overhaul, and never leak to my knowledge. bye, Tim

Respectfully,
Mark Keller
91" Carat WE
"Lazarus"