Need a book? Engineering books recommendations...

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

Re: Steel: Moment Connections

[Subject Prev][Subject Next][Thread Prev][Thread Next]
Majid Sarraf wrote:
> Interesting question by Jan and more interesting comments by Stan.
> I have a few points which may make you ponder even more, and question the
> whole idea of using slip critical bolts for moment connections.
> In a moment connection there is two type of bolt slip which may cause
> problem, 1) Horizontal 2) Vertical
> Horizontal: ...web contribution will not exist until connection fails (most
> likely due to non-ductile weld failure). If we simply ignore the
> contribution of web, then we have to rely on overstrength and ductility of
> top ,or bottom plate (depending on which one is in tension) with their
> typically brittle welds!

The typical assumption used in
design up until the fallout from
Northridge has been that the 
flange welds take all the moment
and the shear tab connection takes
all the shear.  The problems with
the welds or areas around
the welds fracturing is, of course,
what all (almost) the hoopla is about
out here in California.

> Vertical: ...
> If we want to totally eliminate slip of bolts for a moment connection in a
> eartquake resistant frame, we have to design for a resultant factored force,
> horizontal factored force to provide moment resistance for the web, and
> vertical component, factored shear due to gravity and factored shear due to
> end moments  combined (contarary to routine practice of using service loads
> to design SC bolts). Lets for simplicity forget a bout horizontal component.
> Now,
> As for loads, I will give it, at least, 30% extra shear compared to shear
> due to specified loads. As for bolts shear resistance (considering no
> special surface preperation) you have roughly half shear resitance of bolts
> in bearing if they are to be used as slip critical. That means you need 2.6
> (1.3 x 2= 2.6)times no. of bolts you would normally design for shear to
> control slip! Increasing no. bolts means, more holes. This causes serious
> reductiuon in the web cross section, as well as tab plates. Tab plate
> thickness easy to increase, but beam web not as easy!.
Since our allowable loads have been factored down
from expected real values, it may be that, in
order to prevent a great deal of slippage in the 
connection, one would need to back out this factor
for the design of the slip critical bolts.

> >For bolted web (slip critical or not) and full
> >pen. flanges:
> >First weld flanges.  Tension the bolts later.
> But even doing this does not help much. Because most of the shear due to
> gravity is applied after placement of the beam (Dead loads, live loads and
> earthquake). unless we almost complete the construction and let the live
> load on and then tension the bolts! Or, still a high percentage of shear can
> be taken by top and bottom plates.

I should have clarified myself better here.  The 
primary reason for welding the flanges first is 
not to adjust how much dead or live load is 
imposed on the bolts, but instead to avoid 
building in a lot of unnecessary stress in the
joint as the weld cools.  If you restrain 
relative movement of the beam and column with 
tensioned bolts before you do your complete
penetration welds, you may load the bolts and
increase residual stresses in the weld as the
weld cools.

Stan Johnson, PE, I could use heat treatment in a
right now to get rid of some residual stress!  :)