Stiffness at a welded joint

[CaneAE]

This question may seem a bit silly when you first read it but I'm struggling to convince myself either way:

For a welded structural element which is designed based on deflection (controlled by stiffness versus stress) does a welded joint in the system need to develop the full moment of interia of the connected element?

Example: Vertical to chord connection of members in a verrendel truss where the selection of the member sizes was based on the gross stiffness. Should the weld of the fully restrained member develop the full stiffness of the vertical or only the required strength for the moment? If so, can you point me toward a design example or text?

 

[MikeHalloran]

Isn't that why Vierendeel trusses typically have big-ass fillets at every corner?

I claim zero expertise, but thought you might find this helpful:

http://www.eng-tips.com/viewthread.cfm?qid=85066

Mike Halloran
Pembroke Pines, FL, USA

 

[CaneAE]

Different example to illustrate my question: do you consider the actual moment of inertia for the welded joint when you deisgn the connection of a cantilevered beam whose size was selected based on deflection limit states (ie, the net moment of inertia is decreased at the connection)

 

[ATSE]

If you have a reduced section at the welded connection for a "restrained" moment connection - not a good idea, whether the connection is technically controlled by stress or deflection - you need to add an incremental rotation at the joint.

I assume you are trying to use a PJP or fillet welds instead of CJP for cost savings.

If you are modeling in a program such as RISA, then you would add a very short beam element between the chord and the end of the vertical strut that has a reduced Ix.

A vierendeel truss vertical deflection is particularly sensitive to joint stiffness.

 

[CaneAE]

ATSE:

I agree. I worked a forensic project (not the one I'm working on now) where the EOR didn't properly detail and it caused problems.

On this job, I'm trying to evaluate the required weld for a vertical to horizontal joint in HSS sections. The EOR has shown a 7/8 fillet weld and we're trying to make a more cost effective joint using PJP.

However, this study this is bringing up a bigger question about FR connections. Historically we have specified CJP in our connections but more and more fabricators are asking to 'optimize' the connections for the actual design moment. Based on your response I assume you would design the top and bottom flange plates on a flange-plated FR moment connection to 'replace' the moment of inertia of the section?

 

[ATSE]

Yes, I would design the welds at the flanges to both "replace" the area of the flange and develop the strength of the flange using matching electrodes. PJP with reinforcing fillet can accomplish this and eliminate the backgouging requirement and be cost-effective, but probably not permitted for cyclic structures.