Sender: <marv@lancaironline.net> (Marvin Kaye)
To: lml
Date: Sat, 01 Mar 2003 19:49:27 -0500
Message-ID: <redirect-2041070@logan.com>
Disposition-Notification-To: Brent Regan <Brent@regandesigns.com>
From: Brent Regan <Brent@regandesigns.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.0.1) Gecko/20020823 Netscape/7.0
MIME-Version: 1.0
Subject: Re: proper toque values/calibration
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit

As ubiquitous as threaded fasteners are, I have always found it ironic 
that most people  (including most engineers) hold significant 
misconceptions about their use and installation.

Some fastener facts:

Maximum fastener performance is achieved when the fastener is installed 
with it's tensile preload is into it's yield range.

Accurate fastener preload can only be achieved through fastidious 
attention to installed conditions OR by measuring overall stretch.

A properly preloaded fastener gets no benefit from retaining devices 
(split washers, star washers and even (gasp) safety wire). These devices 
are only used to give the pilot a little extra time to find the "never 
properly tightened" or "over stressed in use" fastener before it fails. 
This is why you NEED to do nut and bolt tightness checks periodically.

Because preload and torque are related by the coefficient of friction 
between the mating threads and the fastener and the part land AND the 
coefficient of friction varies widely with surface conditions and 
lubrication IMHO, for most situations, calibrating a torque wrench is a 
like calibrating a hammer.

Most (99%) of all fasteners out there are crap. Their thread form, 
material properties, surface coatings, residual stresses, fillet radii, 
overall dimensions, surface lubrication, surface corrosion (you get the 
point) vary so much that using a calibrated torque wrench on them is 
comparable to using a micrometer on Jell-O (brand flavored gelatin). 
Surface lubrication alone can change to torque to preload ration by more 
than two to one. Besides, why would you need to torque a bolt to an 
exact, calibrated point when the point was selected from a table of 
general recommendations?

Using a calibrated torque wrench IS a good idea when you are installing 
$25, individually packaged  bolts into a dimensionally controlled, high 
performance structure.

If you want to be consistent (and you do want to be consistent) then the 
method I would recommend for fasteners where you don't have a 
manufacturers recommendation (as compared to the engine or prop 
fasteners) is to select the nut - bolt - washer combination you plan to 
use, assemble them in the part and then torque to failure. Now grab 
another set, install it and torque to 75% of the failed torque. 
Remember, everything has to be the same. You are, in effect, calibrating 
the torque wrench to the fastener performance. Your torque wrench only 
needs to be repeatable, not accurate. It seems that everything I know 
that is useful I learned by breaking something.

If a threaded joint is creaking then it IS moving, and movement at a 
threaded interface is always bad. Remove the bolt, add shims to the aft 
spar - gear box interface until the gap is gone and ream the spar bolt 
hole to the next bolt size. Be sure to chamfer the hole before 
installing the new, oversize bolt.

Remember, if you install a fastener in your airplane, always tighten it 
even if you are SURE you are going to remove it later. This way you can 
never install a loose fastener.

For further reading see Carroll Smith's book "Engineer to Win" chapter 8.

On a final note, I asked a builder why he wasn't using Locktite instead 
of safety wire on a non critical fastener. "I don't trust that glue" was 
his reply. Pointing to his airplane I said "Friend, now is NOT the time 
to loose faith in adhesives".

Regards
Brent Regan