Welded Moment Connection End Plate Design 2

[waytsh]

I have bolted moment connection where due to a detailing error the end plate and bolts are undersized for the connection. I am looking for options to help salavage the parts and am considering welding the end plate all around to the supporting flange in addition to the original bolted connection bolts. I can get the weld to work for all the forces in the connection (even without considering the bolts) but am now concerned about bending in the end plate. Is there a prescribed method for checking an end plate welded to the support in this manner? I have checked various sources with no success including the AISC design guides. All the connection types I can find weld the flanges and web directly to the supporting member with no plate in between.

Thank you!

 

[nutte]

There is an AISC design guide that addresses end plate moment connections. Perhaps your configuration (which I haven't fully grasped) is in there.

 

[waytsh]

Thank you for the quick response nutte. I did check that design guide already. All the end plates in the design guide are bolted to the supporting member. Mine is going to be welded.

 

[hokie66]

If it is the thickness of the plate that is undersized, I can't see how welding around the edges is going to help. You won't find design guides to help with connections that don't work.

 

[waytsh]

Well, I guess my thinking was that a plate would no longer be bending just about the bolts and would now have continous support along it's edges to resist bending. Just like a simple span beam may not work for a certain load when it ends are pinned but will when its ends are fixed. Maybe I am way off base here but it seems like a reasonable approach to me and would save the detailing firm and the customer a lot of headaches.

 

[Lion06]

Two things. First, I hope it does work without the bolts because AISC will not let you combine the capacity of both.

Second,I would try doing this. The thickness of the plate is being dictated by the bending in the plate. I would analyze it as a baseplate, but make it shorter. If it is currently 18" long, try making it 14" or 15" long and analyze it as a baseplate with moment only (equal to the moment in the beam, obviously). This reduced plate length will mean that your required welds to the supporting member will be larger, but it will take away some of that plate bending that is causing you problems by shortening the cantilever length of the plate.

 

[Hanif08]

Hi,

I agree with 'waytsh' that after welding the end plate all around, there will be no bending in plate as the welds alone will take all the forces.

This type of fixing will result in the following:

- The T/C couple due to moment will be in the flanges of connecting member.
- Welding will have to resisit both the existing shear force and the tension(T) simultaneously and this requires a more more careful analysis of stresses in the weld.
- A simple way is to perhaps find the centroid of weld group. Draw a free body diagram of all external forces forces and determine distribution of forces in weld (horizontal or vertical legs). Size the weld according to magnitude of forces. For additional safety slightly oversize the weld and if the weld size is 'large' then ask for ultrasonic testing.

On the other hand if this is a standard beam to column moment connection then it is also possible to directly weld the beam to the column flange. In this case size the welds in flange area to resist force couple (due to moment) and weld in web area to resist the vertical shear. For shear flow add stiffners in the column.

 

[hokie66]

Hanif08,

You are on the wrong track. Waytsh has a beam with an end plate that is too thin. Welding that plate to the column is not going to solve his problem. With the plate between the end of the beam and the column flange, he can't directly connect the two by welding. Bending strength of the end plate is the issue.

 

[waytsh]

Let me give some more details to help clarify the situation. The moment connection is an interior column to roof beam connection. The plate that is now currently on the top of the column is for a pinned connection and is only slightly larger than the column. Enough to allow for the welds. It does not have the exensions on it like it would have had for the thicker moment connection plate. If this end (cap) plate is welded to the roof beam flange the welds across the ends are going to be very close to the flanges of the column. I think the plate is going to see very little bending because of this. I have checked the weld for the combined moment shear and axial loads and a 1/4" weld will work. I would plan to use a 5/16" to build a little extra strength into the connection. I was interested to know if there is a prescribed method for checking the plate when welded in this manner. For instance if I did have the extensions and was welding the connection how would that plate be anaylzed?

 

[hokie66]

Well, that is not what you said in your original post. My comments are applicable to your original, 2nd, and 3rd posts.

 

[waytsh]

Thank you all for taking the time to respond.

hokie66, even according to my original post how would you know that the plate is going to be to thin for the welded connection? Are you just going to assume that it will behave like a bolted connection and ignore the influence of the welds along the sides of the plate?

I still assert that an end plate is going to see less bending stress when welded all around than when it is being restrained solely by bolts at the ends of its extensions. The trick becomes trying to quantify the bending stresses when restrained in this manner and hence the reason for the post.

 

[hokie66]

waytsh,

Sorry if I was curt with the last reply. Your question is legitimate, but the solution is so complex that I wouldn't venture an answer. The welds on the side would have an effect. You could model the connection with edge supports and the beam flange and web forces imposed. You might get some reasonable answers by taking the top flange as a tensile line load, with the edges and the bottom flange as simple supports, and using some equations from Roark. Or do FE analysis if it is worth it.

 

[waytsh]

hokie66, funny you should mention Roark's I'm flipping through it right now to see what I can find relative to this situation. I'm just suprised there is not more literature addressing this splice configuration. No problems with your previous reply, as I said I am grateful for you guys taking the time to reply.

 

[patswfc]

From how i read the situation you think that the connection detail wont cause much bending of the plate but dont have an example or reference to justify your engineering judgement.
Can you add some stiffeners on the underside of the cap plate to the column and analyse this assuming that the cap plate will bend, which will be conservative.

 

[dik]

Can you weld the end plate to the beams and cut off portions of the end plate to minimise moment in the end plates, ie. cut off the plate nearly flush with the steel beam and then do the final welding?

Dik

 

[civilperson]

Adding stiffeners to each side of the plate between the bolt holes can make the plate suitable for moment connection using either welds or correct size bolts. 3/8" thick is standard for stiffener plate.