eccentricity in steel columns 1

[bjb]

A statement was mad to me that when designing a steel column with steel beams framing into it with simple shear connections, it is not necessary to account for the bending moments caused by the eccentricity of the beam reactions. He said that because the clip angle connections actually resist some moment, the effective length of the column is really less than 1, and that this counteracts the effect of the bending moment.
I disagree because the clip angles are not designed for moment transfer, so you shouldn't count on them lowering the effective length of the column, and that you should consider the moment from the eccentricity of the beam reactions.

I would like to know what others think about this, please give me your comments. Thanks.

 

[SlideRuleEra]

I consider this type situation to be a matter of priorities. The moments that you are considering exist, but their magnitude is probably lost in the approximations needed for column design, such as selection of the "K" value. Since most structural failures occur at connections, rather than in members, the effort looking at this eccentric column loading may be more wisely invested in designing the connections and other higher priority factors.

 

[bjb]

I see your point SlideRuleEra, thanks for the response. Also, I never select a column that is "maxed out" for its allowable axial load, so this extra capacity would also help deal with these moments. My friend gave me this quote about structural engineering that is relevant to this discussion:
"the art of molding materials we do not really understand
into shapes we cannot really analyze
so as to withstand forces we cannot really measure."

 

[sc]

I usually assume a moment is generated at the connection by the eccentricity of the load transfer point (ie the bolts). The clip angle will have to deal with a small amount of moment due to this load, lever arms, etc. The column will have to deal with a small moment as generated by the connection.

Having said this I treat the beam as pin ended, the column as under combined axial load and moments and the connections as mainly shear but with a small moment in the connection plate.

In reviewing my as constructed designs I have not had a failure yet with this approach. Also as a rule of thumb as I deal mainly with housing, I always specify a minimum of 2/M16 8.8/s bolts and 10thick plates for these types of connections. In a lot of cases this is over designed but the extra cost of the plate and bolt size increases is less than the cost of me doing the calculations when I know from experience that these sizes will work for 95% of the housing work I do.

regards
sc

 

[jheidt2543]

I believe that spring of 2002, OSHA instituted a new rule that requires that columns support a certain amount of moment for erection purposes. How does that requirement stack up against bjb's question. Is the OSHA requirement, in general, larger or smaller than the load expected in bjb's question?