[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [New Search]

Re: [T3] LONG response... How-To help: Figuring out torque.


Hello-

> I am hoping someone can tell me how to figure out the torque necessary for
a
> given fastener (bolt, nut, screw, rivet, etc.).

As simple as it sounds, the actual answer is fairly complicated and
confusing.


> I have HP Books "High
> Performance Hardware" (a good beginners guide to fasteners) and the 26th
> edition of "Machinery's Handbook" and I can't figure out how to determine
a
> torque setting for a given fastener.

Many people who _do_ understand how to do it are _extremely_ hesitant to
give out the information.  This isn't due to some proprietary thing, but
rather due to liability.  Tightening torques are one of the messiest
specifications to make without guessing.  There are some parts to the
formula that are quite hazy...  quite frankly, the general "Joe" isn't
trusted to understand what to do :-).


> It's just a general would-be-neat-to-know
> for me.  The MH does have a variety of formulas but I'm not sure which to
use
> and the one I did use (wrench torque) gave me values that I don't seem
right
> so...can anyone give me an idiot's guide to establishing torque values?
;-)

The basic formula is this:

T = k * F * D

Where:
T = tightening torque
F = preloading force
D = major diameter of thread
k = a coefficient

This is an _emperical_ formula - it is not perfectly derived from theories,
but rather created by glossing over the details to get an approximation.
There are several problems why this cannot be exact.  The first is this: we
are trying to relate TIGHTENING TORQUE to PRELOADING FORCE.  The preload
force is axial - it's the actual clamp load of the bolt/stud itself.  The
tightening torque is a resistance or friction torque to turning the nut any
further.  Think about the physics for a second... these are two very
different animals!

The goal of tightening a bolt is to get the correct _preload_.  Well,
there's problem number one: what is the correct preload?  As screwed up as
this one is, this one is actually the simpler of the two major problems to
answer.  It depends on many criteria.

The first criteria is to determine what the weakest point of the stud/bolt
is.  This is done by finding the smallest stressed cross-sectional area.
This is often the threads.  The stressed diameter of threads is:

Ds = D - 0.9382*m
or
Ds = D - 0.9382/p

Where:
Ds = stressed diameter
D = major diameter of thread
m = module of threads (i.e. 1.25mm, 1.50mm, etc. - like metric)
p = pitch of threads (i.e. 13 1/in, 20 1/in, etc. - like SAE)

But, if you have a stepped stud, or an undercut stud, the weakest point may
be the length of the shank itself.  This is the case for every head stud of
every German-boxer-like automotive engine (VW T4, Porsche 356, Porsche
547/587/904, Porsche 911/930/964/993, Porsche 956/962/959, Corvair, etc.),
except the el-cheapo VW T1 :-).

Another important criteria is the yield strength of the material.  Get it,
then determine what stress you actually want to impose on the material for
the preload (i.e. NOT the yield strength, but lower).  Ah - now we're
getting hazy.  You could use the "proof strength," which is often about ~90%
of the yield strength.  This is good for parts that aren't very dynamic and
aren't exposed to high temperatures.  You could use a more conservative
measure, like ~75% or so (a decent 1st guesstimate for dynamic situations
where forces oscillate a lot).  Or, you may have a far more complicated
problem, such as the head studs of a VW engine, where temperature effects
radically change the loading conditions of the stud.

As you can see, this is not a straightforward or simple answer.  More on why
engineers keep it to themselves... experience and training are necessary to
fully understand and appreciate what is going on, then make decisions that
effect the safety and wellbeing of both people and their monetary
investments.

Okay, so great.  We magically figured out the correct preload.  Now, how the
heck to we get the bolt there?  We have a big problem - that "k", the
coefficient, is a big question mark.

What does the "k" represent?  It represents a sort of "conversion factor" to
get from a torsional friction torque to an axial preload force.  As a rough
idea, it is often somewhere around ~0.2.  But, the exact details are far
more complicated once again.  And this time, there are no equations to help
you out - just emperics and experience.  Basically, EVERYTHING that could
effect friction needs to be taken into account here.  More friction means a
higher k.  Less friction means a lower k.

For example, what lubricant is on the threads during tightening?  What is
the surface finish like on the threads?  What is the pitch of the threads?
What is the length of threads on the nut?  What is the age of the pieces?
What is the washer like that's used under the nut/bolt (especially important
with serrated locking washers or the equivalent type of flanged, serrated
bolt)?  What lubricant is on the washer?  What temperature are these
lubricants at?  Do these nuts have self-locking nylon or the equivalent on
them?  How fast is the assembler tightening these things (note the
difference between static and dynamic coefficients of friction...)  It gets
real sticky real fast!  :-).

Then, you have to deal with "factors of safety."  This is the engineering
way of "covering your ass," basically figuring out how much you have to
underball/overball the ideal so the worst case won't result in destruction.
How accurately are the hardware pieces made?  How well were they stored
(dings, deformations, etc.)?  How accurate is the installer's torque wrench?
You have to take _all_ of this into account.

Personally, I've seen coefficients as low as ~0.11 and as high as ~0.24.
That's over a 2x range.

With all the haziness surrounding tightening torques, it's no wonder that
where precision counts (and the geometry of the application allows), bolt
stretch is far preferable to tightening torque to determine preload :-).

Now, with all the engineering stuff behind us, where does that leave the
average Joe who just wants to tighten the friggin' bolt and get on with his
life?  Well, first, find a reference measurement.  If you have a standard
SAE (5, 8, etc.) or metric (8.8, 10.9, 12.9, etc.) grade bolt in a fairly
"run of the mill" application, then consult a standard table.  If the type
of bolt you have is a slight deviation from the types given in the tables,
then remember the above equations and make a variation off of it.  For
example, if you have the specs for a Grade 5 (120,000PSI yield strength)
bolt, but you have a Grade 8 (150,000PSI yield strength), then you can
probably multiply the table value by 150/120 and get a decent number.  If
you have a certain preload you'd like to keep that a bolt used to hold, but
you have a different bolt with a different kind of thread, then you can
adjust accordingly (i.e. VW T1 10mm vs. 8mm head studs... if you want the
same preload, then the 10mm will need more torque).  Or, if your bolts are
made to a higher specification by a specific company (say, ARP), then
consult their tables and adjust from there.

I guess what I'm trying to politely say is that if you're not an engineer, I
recommend doing as little engineering yourself as possible :-).  Start with
what an engineer already specified (noting the context, i.e. "regular"
application versus something like a VW head stud) and intelligently adjust
accordingly.

Take care,
Shad Laws
LN Engineering - Aircooled Precision Performance
http://www.lnengineering.com

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
List info at http://www.vwtype3.org/list | mailto:gregm@vwtype3.org


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [New Search]