moment from simple shear connection

[smvk3]

If you are connecting to the flange of a WF shape with a shear tab or double angle connection, do you have to design the column for the moment M = P*e where "P" is the end reaction and "e" is half of the WF column depth? What about when you are connecting to the wall of an HSS tube with shear tab connections. Do you have to design the column for the moment where "e" is half of the HSS column depth?

 

[Trenno]

Centroid of bolt group to supporting member centreline or 100mm - whichever is greater.

Also with HSS's, you need to check local buckling of the plate element the shear tab is connecting to.

 

[DaveAtkins]

I design tube columns for the moment, but not wide flange columns (although, to be theoretically correct, both should be designed for the moment).

The way I look at it, a tube column will have a single plate or WT shear connection, which will cause a moment which should not be ignored for a tube column (because a tube column will have a small section modulus compared to a wide flange column, and because the single plate or WT connection will introduce more moment than a double angle connection). However, the moment induced by a flexible, double angle connection to the strong axis of a wide flange column will be fairly insignificant.

DaveAtkins

 

[bookowski]

Theoretically yes, in practice it is often ignored (or was in the past - software is making it more typical to include automatically).

 

[JAE]

You should include the Pe moment on the column.

For HSS columns with single plate connections, this is from the AISC website:
[blue]QUESTION: Are through-plates always required for single-plate connections to HSS columns?

No. Sherman and Ales (1991) demonstrated that local yielding of the support was not a concern due to the self-limiting nature of simple-shear connection end rotation and that the compressive strength of the HSS column was unaffected by the associated local deformations. However, this same research indicated that punching shear may be of concern for relatively thin supporting material thicknesses. Punching shear can be prevented by selecting an HSS with a wall thickness tw that meets the following criteria:

tw is greater than or equal to (Fy pl )(tpl) / Fuw

where

Fy pl = the yield strength of the single plate

tpl = the thickness of the single plate

Fuw = the tensile strength of the HSS wall

Note that this equation differs slightly from that given in Sherman and Ales (1991). Here, the expression is derived at the design strength level (omega factors included) whereas it was previously derived at the nominal strength level (no omega factors)

If the actual maximum stress is known, it can be substituted for Fy pl in the above equation for a less conservative result.

The above minimum thicknesses would also be applicable to a welded plate tension connection (uniform stress distribution assumed). However, for cantilevered bracket connections, which do not have self-limiting rotations; yielding must also be checked.[/blue]

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[smvk3]

What about AISC's response in a previous issue of the modern steel construction magazine?:

http://msc.aisc.org/globalassets/modern-steel/steel-interchange/2002/2002v11_si_web.pdf

They mention research that shows members framing into such columns provide rotational restraint and therefore the column would not see this moment. It seems that as long as you provide a "simple shear connection" per AISC guidlines, the historic practice has been to ignore these moments.

 

[WillisV]

What smvk3 said. I do not design for the P*e moment in typical building conditions with "simple" connections. There is nothing wrong with it, but it is unnecessary. It was not accounted for in the past and the only reason people are accounting for it now is that it is a little check box for it in their computer program.

 

[haynewp]

I know software I have used automatically takes this eccentricity into account, but Charlie Carter's response is more in line with what I would have reasoned and always saw done in practice by engineers since retired (and many times myself, not retired) of going to the axial load tables in the Manual and picking an HSS column from the axial load basis only. I would look to include it if I thought there was high flexibility in the connection for some reason, or horizontal slots present at the bolts.

 

[haynewp]

I got distracted and Willis posted before I hit send. Essentially the same opinion.

 

[Thanh123]

Sometime if the web of column is thin, it could be fail by buckling regarding the additional eccentricity moment. Therefore, a stiffener is needed without check.