From: John De Armond
X-Source: The Hotrod Mailing list
Date: Mar 1992
Subject: Re: Bolt torques
> Last night I was reading the latest issue of Popular
>Hotrodding and ran across an article on engine building. In the
>article they suggested lubricating the main bearing cap bolts
>with assembly lube before torqueing to avoid torque errors from
>binding. I have very limited engine building experience, but I
>believe that this is very wrong. Torque specs are given for
>clean dry threads. Lubricants will cause the bolts to be
>overtorqued. Please correct me if I'm wrong.
Actually both are wrong but oiled threads are more correct. Torque has
only an oblique relationship to the real parameter of interest -
fastener stress. This stress or preload, measured by the enlogation
of the fastener, is the critical value. SAE specifies a stress load
for each kind of fastener in ... (oh hell, I can't find it now) anyway,
the "bolt spec." Turns out that a fastener achieves its ultimate
strength when stressed to about 85% of its yield strength. A good
layman's summary of this subject can be found in Carrol Smith's
"Engineer to Win" book.
The problem with "clean dry threads" is that an unknown proportion of
the twisting effort goes to stiction, dry friction and perhaps
galling. Lubricating the threads and head gets the situation much
closer to the ideal (coiled incline plane). The problem is many
American engines have their torques specified dry. Since there is
no mathematical conversion, other methods must be used.
The best method is to simply measure enlogation directly. This is
possible on rod bolts where both ends are available. A small dial
indicator assembly is readily available for the purpose. Blind holes,
such as cap and head bolts are more trouble. I specialize in one
engine (the Datsun Z) and even though the japanese sensibly give torque
figures as wet, I've gone to the trouble to develop my own.
This is most easily done by making up an assembly of parts such that the
bolt can be torqued in a similar environment but where access is had to
both ends of the bolt. I carefully clean and polish the threads
and the underside of the cap, lube them and then torque the bolt until the
specified stretch is achieved. Averaged over several samples, this becomes
my assembly torque. My assembly lube is STP or STP with moly disulphide
mixed in when I can locate it (Dow Corning Moly 77 powder.)
During assembly, I carefully chase the female threads, polish the threads
and the underside of the head of the fastener using a buffing wheel and
red rouge, use a hardened thrust washer and lube the whole assembly.
Even then, if it does not feel right, the joint comes back apart and
the trouble found. I never use "click" torque wrenches because the
feel of the joint is destroyed just when you need it the most. I also
don't use them because when I worked in a metrology lab, click wrenches
were the most common instrument to arrive grossly out of calibration.
This may seem like an excessivly tedious procedure but it is standard
practice in the aerospace/nuclear fields, places where failed fasteners
have a bit more serious consequences. I can proudly say that I've never
had a bottom end failure on one of my engines attributable to fastener
Interesting trivia: The 6" diameter studs that hold the head on a
nuclear reactor are hydraulicly stretched while heated to operating
temperature with a heater inserted in the hollow shank. the "nut"
is then hand tightened down against the reactor lid and the hydraulic
pressure released. No wrench or torque is used.