Need a book? Engineering books recommendations...

Return to index: [Subject] [Thread] [Date] [Author]

RE: Shear-off bolted connection

[Subject Prev][Subject Next][Thread Prev][Thread Next]
Looks like there are strong opinions on this subject. It seems to me you have 
more control over variables with the twist-off vs. the LIW. With the LIW, 
you've got to deal with highly variable surface conditions in the base metal 
surface resulting from paint, mill scale, rust, surface roughness from the 
rolling process, etc. These are all in addition to the variables associated 
with the bolt - which are essentially the same between LIW systems and 
twist-off systems.

For our military construction projects in years past, we always asked for 
mock-ups of the LIW bolt assemblies to be tested in the field (The military 
used to require LIWs). The results were always highly variable to the point 
of being unreliable. Then we had problems bolts breaking during installation. 
Mill cert's were required for all of these bolts. I believe reasonable field 
controls were in place to adequately ensure proper installation so that this 
experience was a fair indication of "real life" performance of LIW 
assemblies.

It's that experience and feedback from erectors and fabricators that have 
moved us to specify twist-offs vs. LIWs. No system is perfect, or ever will 
be. In my opinion (par value .02), twist-offs are the way to go.

In response to Mark, I agree that you are probably OK tensioning the bolt to 
110% of proof in most cases, but what happens if the LIW is not compressed at 
110% and you end up at 140% of proof? That's almost ultimate for an A325.

Bill Keen




-----Original Message-----
From: Majid Sarraf On Behalf Of Majid Sarraf
Sent: Monday, January 18, 1999 6:24 PM
To: mail@ih {seaint(--nospam--at)seaint.org}
Subject: RE: Shear-off bolted connection


You have made interesting points, Bill. Let me add what I think about these
two systems. I would appreciate you enlightening me, or correct me if I am
wrong.

Although I like to see twist-off bolts work, I am a bit concerned about it.
The way I understand is that the magnitude of the torque applied is supposed
to correspond to the right pretensioning force in the bolt. However, the
torque is not just overcoming the force for turning the bolt, but it has to
also overcome frictional torque between bolt head and the plates. So the
surface condition of the plates is also important and it may vary too. In
fact, in a bit exagerated senario, if the bolt head gets stuck for some
reason (high friction, bumps or dents on the surface, etc..) the bolt end
will twist off way before any significant tension is developed, and you may
think you have a good pretensioning force in the bolt, because the end is
twisted-off!.
Probably it would be also hard to calibrate the torque for every condition,
and come up with different size of twist-off bolt end for a given bolt dia.

As for DTIs. First, I thought this would be the best, but this did not last
for long, since I thought about the senario that you just mentioned too
(over-pretensioning).
I have an explanation though, and I hope this can be used in a way to
improve they way the washers are made or used. Due to the size and form (
arch type) of the typical bumps on these washers, froce-deformation after
initial yielding becomes highly non-linear, as opposed to a nice yield
plateau. Non-linearity in geometry and property of the material, can both
make their resistance highly sensetive to the deformation imposed. Perhaps
some test results would confirm this, and can indicate how much overstrength
would be developed in washers.
Just as an example, if at the begining (after snug tight condition) every
turn would increase the force in the bolt by 10%, after initial yielding of
the bumps this may suddenly change to 30%. So we may easily loose control
over force because of the sensitivity in behavior of the washer under higher
load, and risk failure of the bolt. Even if it does not fail we have no easy
way of knowing how far it is from failure!

So, I suggest, If I may, either we change the way washers are made, or find
out an easy way to avoid excessive squishing the washers.

I hope I can learn something here.


Majid Sarraf
www.uottawa.ca/~msarraf


>I think we need to balance the theoretical aspects of this issue with the
>practical aspects. Erectors and fabricators are telling us they like the
>twist-off bolts because they are easier to install and they can maintain
>higher quality. Usually, if it's hard to do something in today's construction
>industry, there's very little likelihood that it will be done well without:
>a) a LOT of involvement by the SER,
>b) more field inspection,
>c) a contractor who cares about quality and
>d) someone with enough chutzpa to tell the contractor it's not right and do
>it over again
>
>I agree with Charlie - they all CAN perform well if installed properly. The
>question is, how hard is it for all parties involved to make sure that
>happens.
>
>I like the twist-off bolts. I've experienced problems with DTIs myself. The
>biggest problem I've seen is that they "squish" too late in the torque
>process and bolts break off. Over-tensioning worries me more than
>under-tensioning. I'd much rather have a bolt tensioned to 50% of proof when
>90% was required vs. 110%.
>
>Bill Keen
>
>
>
>
>

              ~      ~           ~ ___
                       [_________]-|__)
              __________OO______O__O_O___________________________
                  |==========================================|
                  |                                          |
                   \               Majid Sarraf             /
                    \             Ph.D Candidate           /
                     \   Department of Civil Engineering  /
                      |        University of Ottawa      |
                      |           Ottawa, Ontario        |
                      |            Canada K1N 6N5        |
                      |   Tel:(613) 562-5800 Ext. 6159   |
                      |         Fax:(613) 562-5173       |
|                                  |
                       ~  ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~