Bill Babcock wrote:
>
> I think everyone is on the right track. The complicating factor is how
> much everything moves. These are very strange engines. We have a cast iron
> block, steel sleeves, cast iron head, some very long and relatively short
> bolts, none of which are through-bolts. All these things are heated
> differently and cooled differently. Then we have a wide array of
> gaskets--composition, copper clad, steel clad, steel shim, steel shim with
> copper wire, and solid copper.
>
> One answer for all? I doubt it.
Couldn't weigh in on this at the time... burned yet another keyboard and
had to get another.
All the above is true, perhaps broadly, or in some cases, just in
detail. But, with regard to the wet sleeve and cast-iron head Triumph
engines, I think that the issue of different head bolt lengths is of
minimal consideration, since the clamping done by those studs is more or
less uniform in length between the points captured in the block and
secured by a nut, and in those instances when it is not, uniform torque
produces uniform preload, and that preload is dependent upon
elongation--and elongation is a ratio, not an absolute length; more
precisely, is measured in inches per inch of length.
The rate of expansion of steel (in the case of the sleeves) is very
close to that of cast iron, and since expansion due to heat coefficient
is linear in all directions, torquing hot or cold will produce little
difference in clamping force. If aluminum heads are involved, all bets
are off in that regard--torquing to spec at running temperature is
effectively a lower clamping force than torque done when cold, since
linear expansion of aluminum for equivalent temperature is roughly three
times that of cast iron or steel.
> I always retorque the heads--why wouldn't you? I'm not that committed a
> spectator, so I tend to fuss around my car in the pits. If I set the
> valves I usually torque the heads first. Every so often you find a bolt
> that has loosened a little somehow. If you use either composite or shim
> steel with a copper ring, you'll see a substantial change after a heat
> cycle. Not just once--you see it six months later.
I have no doubt that this is true. Experience with same counts.
As for the use of composite or composite sandwich gaskets, I've
developed a procedure over the years that seems to minimize problems if
the head isn't immediately retorqued, and it seems to work, since I
haven't had any head gaskets fail. On initial torque, I take plenty of
time between cycles of increasing torque, usually fifteen minutes or
more. If torque would be normally split in four cycles, I let perhaps
ten-fifteen minutes go by between each. That allows for progressive
compression of composite gaskets. On the last go-round, I wait maybe
thirty minutes and then do a last check to specified torque. This also
helps greatly with used studs and nuts--if one stud or nut is failing
because it's yielded, that shows up as a loose nut on final torque
check. All this extra time gives composite gaskets more time to relax.
One of the things I have found, over the years, is that it is good
practice to back off the nuts very slightly on re-torque (not enough to
break any sealing accomplished at the gasket surfaces), simply because
after they've been run up to temperature, the nut has a tendency to
stick on the stud threads. I think this probably has to do with minor
thread perturbations or galling combined with peak cylinder pressure
squeezing the stud and nut surfaces together.
If this isn't done, the torque wrench is reading the applied torque plus
the static friction, but the clamping force is dependent upon the
applied torque only.
> A bigger question is are the bolts dry or lubricated. That's a much bigger
> variable than the heat cycle. If you torque a bolt to it's published spec
> dry the chances of pulling out the threads or breaking the bolt are very
> low. If you use a very effective (for sliding friction) lubricant then the
> same torque will rip out the threads or break the bolt. I got obsessive
> about this issue after I broke a long head bolt at the base of the thread
> way down inside the block in the pits before the Monterey Historics.
>
> Most torque specs assume a motor oil lubricated bolt. If you want to
> really do it right, you need to find the stretch specification for each
> bolt you are using, lubricate it with the lube you intend to use, put the
> bolt in a fixture that will allow you to measure stretch, torque it to the
> stretch spec and read the torque value. Then always use that lube with
> that bolt.
I think this is good advice, but one must know the stud/bolt grade (and
its linear elongation) rather precisely for those dimensional
calculations to be of use. A grade 8 bolt stretched to a point which is
75% of proof for a grade 5 or 6 has likely already yielded. The standard
rule of thumb for such fasteners, to provide an adequate margin of
safety, is that the bolt/stud should have adequate clamping force to
resist the maximum load at 75% of proof load. Further, a 100% proof load
is 75% of yield strength. It's also important not to extrapolate the
elongation of one bolt grade to another--the curves aren't linear, so
the manufacturer's or SAE load to failure data is necessary. A grade 5
bolt may have an elongation under load of 16-18% before yield, while a
grade 8 bolt is more like 4%.
As for "most torque specs assume a motor oil lubricated bolt," I will
disagree. This depends entirely upon the manufacturer's specifications.
If the manual text specifically says "lubricate threads before
assembly," then one can be reasonably sure that the manufacturer's specs
for torque of that fastener are for oiled threads. Otherwise, if not
stated specifically, standard SAE torque figures are for dry threads.
Our manuals, for example, derive from GM specs (and, man, you never saw
a bigger bunch of wackos regarding standards than GMC), and all those
are very specific to conform to SAE standards, unless they were
independently derived by GMC.
> With bigger or longer bolts, motor oil or dry torquing will give highly
> variable numbers when you torque to a stretch spec. The torque jumps as
> the bolt sticks and slips. The super slick stuff supplied with high
> quality bolts is too scary to use--on a cheap bolt you'll exceed stretch
> specs way before you see any significant torque.
Very true, but you can test this empirically, if the pertinent test data
is known.
> Gear oil works okay, but I've settled on a cheap, readily available
> anti-seize compound that I use for everything. I've also gotten allergic
> to split lockwashers. Spending any time investigating bolt torque and
> stretch will convince you they are a tool of Satan. Instead I have big
> bottles of every type of locktite. I'm gradually eliminating lockwashers
> from everything.
Important to remind everyone though, that the purpose of lockwashers is
to minimize the loss of a fastener due to vibration, not lack of torque.
There's lots of empirical evidence and studies which confirm this. Shock
loads from vibration can unwind a fastener faster than your wife can
divorce you for overpaying for a rusty 246 Dino in pieces, and
lockwashers help with regard to vibration.
This is not at all to suggest that the right grade of Loctite isn't a
suitable substitute, or that a flatwasher with a prevailing torque nut
isn't good, either.
My favorite story about vibration and washers was when I was working as
a night watchman at a dude ranch. I did that and worked on their
equipment overnight. I came into work one evening and found the hay
baler waiting for me to repair. It had thirty or forty 24" tines with
eyes on one end allowing them to be bolted to the baler arm--those tines
guided the hay in roughly equal amounts toward the baling mechanism. My
boss had decided this was a job for him, the previous day, and had
attached all the new tines with bolts and flatwashers. After about one
hour's running time, half the tines were missing and the rest were
loose--the ends of the tines dragged on the ground, and the vibration
loads were fierce. Before the days of general availability of Loctite,
lockwashers were the answer.
> For simplicity sake I torque the long and short head bolts the same, even
> though they really do require different torques to reach spec stretch.
As mentioned above, I don't think this is precisely true. Elongation is
a ratio, not an absolute measurement.
Cheers.
--
Michael D. Porter
Roswell, NM (yes, _that_ Roswell)
[mailto:mporter@zianet.com]
Don't let people drive you crazy when you know it's within walking
distance.
|