trevorshiloh
Civil/Environmental
Hi all,
I'm working on a job where I'm reinforcing an existing WF shape with a channel on 1 side, centered on the centroid, with plug welds at 16" o.c. also centered on the centroid. The channel does not extend to the supports on either side, but rather stops short. The PM on this job is choosing to put plug welds on the centroid since shear flow is 0, and is arguing that there is no shear transfer, and therefore any weld size, spacing, etc would work and the 16" o.c. is just chosen based on convenience.
I disagree that the shear that the welds see is 0. Yes, the shear flow is 0, but doesn't that just describe horizontal shear from the compression and tension above and below the neutral axis? There is still some vertical load that gets distributed to the channel, and the welds have to transfer that load, or else the channel would not carry any moment nor deflect with the WF. PM said we don't do this check for WT reinforcement underneath the main beam, but this is because at that interface there is compression not shear, i.e. the WT is centered on the WF and is pushing up on it because of the tension in the top face of the WT. You wouldn't transfer compression through the welds.
I have resolved to perform a quick check to see if it is even worth arguing over, by taking the maximum moment in the beam (mix of distributed and point loading), backing out the equivalent distributed load, and checking the welds for that amount of load transfer. This would seem to be conservative because it assumes that all the load is transferred to the channel, when it should depend on the stiffness ratio (?). We were way under capacity, which I figured would be the case.
But this made me think of another scenario, where the loading is applied on the channel instead of the WF. That means the welds have to transfer all of the load, since it eventually has to make its way to the supports. But how would this load distribution look? Would the shear still distribute based on stiffness throughout the length of the channel and jump back into the WF at the last weld? Or would the WF take all the shear throughout? Or a mix of these?
I modeled this (albeit crudely) in RISA, with 2 identical shapes, and found that the more concentric the loading is to the y-axis centroid of the 2 pieces, the more evenly they share the load, but moving the shapes further apart makes one take more, and in the channel's case, dumps it into the WF at the last weld.
To sum up, my questions are as follows:
1. Is there vertical load in the channel that needs to be accounted for, even though shear flow is 0?
2. Is the way I checked this vertical load reasonable?
3. Does the load distribute based on stiffness?
4. How much load goes into the channel if the loading is applied to the WF vs applied to the channel?
I'm working on a job where I'm reinforcing an existing WF shape with a channel on 1 side, centered on the centroid, with plug welds at 16" o.c. also centered on the centroid. The channel does not extend to the supports on either side, but rather stops short. The PM on this job is choosing to put plug welds on the centroid since shear flow is 0, and is arguing that there is no shear transfer, and therefore any weld size, spacing, etc would work and the 16" o.c. is just chosen based on convenience.
I disagree that the shear that the welds see is 0. Yes, the shear flow is 0, but doesn't that just describe horizontal shear from the compression and tension above and below the neutral axis? There is still some vertical load that gets distributed to the channel, and the welds have to transfer that load, or else the channel would not carry any moment nor deflect with the WF. PM said we don't do this check for WT reinforcement underneath the main beam, but this is because at that interface there is compression not shear, i.e. the WT is centered on the WF and is pushing up on it because of the tension in the top face of the WT. You wouldn't transfer compression through the welds.
I have resolved to perform a quick check to see if it is even worth arguing over, by taking the maximum moment in the beam (mix of distributed and point loading), backing out the equivalent distributed load, and checking the welds for that amount of load transfer. This would seem to be conservative because it assumes that all the load is transferred to the channel, when it should depend on the stiffness ratio (?). We were way under capacity, which I figured would be the case.
But this made me think of another scenario, where the loading is applied on the channel instead of the WF. That means the welds have to transfer all of the load, since it eventually has to make its way to the supports. But how would this load distribution look? Would the shear still distribute based on stiffness throughout the length of the channel and jump back into the WF at the last weld? Or would the WF take all the shear throughout? Or a mix of these?
I modeled this (albeit crudely) in RISA, with 2 identical shapes, and found that the more concentric the loading is to the y-axis centroid of the 2 pieces, the more evenly they share the load, but moving the shapes further apart makes one take more, and in the channel's case, dumps it into the WF at the last weld.
To sum up, my questions are as follows:
1. Is there vertical load in the channel that needs to be accounted for, even though shear flow is 0?
2. Is the way I checked this vertical load reasonable?
3. Does the load distribute based on stiffness?
4. How much load goes into the channel if the loading is applied to the WF vs applied to the channel?
