From: jgd (John De Armond)
X-Source: The Hotrod Mailing list
Date: Jul 1992
Subject: Turbos and heat soak

>In the chapter on lubrication, however, he makes mention of a problem
>that I've seen mentioned in many other places, but never quite accepted.
>He describes what happens after a turbo vehicle has had a hard run, and
>then the driver suddenly reduces speed, stops, and shuts the engine off
>right away. According to MacInnes (and others?) the sudden loss of oil
>pressure leaves the bearings in a condition where they will burn up as
>the turbo shaft continues to merrily spin on at 100,000 rpm. I don't
>understand how this is possible.

It's not.  I've run a turbo engine on a dyno with no tail stack on
the turbine exit (for no other reason than I'd not fabricated one yet)
and I could watch the turbine wheel.  The turbine stops almost immediately
when the engine does.  Coastdown is a bit slower when the engine is
running for what might be a not so obvious reason.  If you've ever
capped the intake of a squirrel cage fan and noted how the fan speeds
up, this is the condition that exists during coastdown assuming an
upstream throttle.  The blower has a high vacuum on it and thus little
drag.  The turbine partially evacuates the exhaust manifold so it is
also running in a partial vacuum.  When the engine is stopped, a different
situation exists.  The turbine is still in a partial vacuum but the
blower is now at atmospheric.  It stops very fast.

As you surmised, the real problem is cookoff.  I'm not really sure
a short idle interval really helps.  The temperature decreases asymtotically
with time so once the first flash of visible heat is gone, much more
time is needed to cool to a safe zone.  Water and/or oil cooling are
the real answer.  My solution is simply not to push the car hard for
several minutes before shutdown.  The flow of relatively cool exhaust
flow helps speed the cooling process.

>Anyway, maybe I'm wrong on all this, but I noticed something else that
>bothered me even more. Nowhere in the chapter on lubrication does
>MacInnes mention the problem of heat-soak from the exhaust manifold,
>and the resultant problems of oil coking and varnishing on shafts,
>bearings, and oil feed lines. This, to me, seems like a much more
>important consideration than the scant few seconds that a turbo
>*might* have to endure moderate RPM's with no oil pressure.

Heat soak from the manifold is not a particular problem.  The gasket is
a relatively good insulator so not much heat gets transfered.  The
problem is the turbine housing and the wheel itself which is at least as
hot as the manifold.

John