plate design question 2

[delagina]

I have an existing rectangular manifold. Think of it as a pipe but rectangular and very big 8'x3' cross section (see sketch).
It's 50' in length and connected by flanges (splices).
The support of this manifold needs to be removed temporarily and will be put back in later.
I need to temporarily support/shore this manifold during that time and the only location I can put the support is right below the flanges.

I plan on finite modelling the flange with plate elements and removing some plates where the bolts are.
The load is just self weight. I can get the stresses on the plates and compare it to Fy.

My question is should I be checking other things like any local or global failure than will not be shown in finite element analysis?

 

[delagina]

forgot to attach the sketch

 

[WARose]

My question is should I be checking other things like any local or global failure than will not be shown in finite element analysis?

Definitely check (localized) plate buckling. Never leave that to a FEA program.

 

[Agent666]

How's it currently supported? Are you supporting it temporarily at the same location/centers as the original supports, is the manifold spanning further than it was previously for example that might require a closer look?

 

[delagina]

It has big diameter pipe bearing on concrete foundation. It will be cut by 3'.
The only way to temporarily support this is by putting a beam under the flange (long story).

My question is about checking the flange against all modes of failure.

 

[delagina]

there are also flanges at the ends of manifold so it's not going to be on cantilever

 

[delagina]

Warose,

I saw this online and will verify the values.
My question is how do I account for the bolt/bolt holes in this buckling check?

https://www.engineersedge.com/calculators/plate-axial-load.htm

 

[Agent666]

Just use the net section area for any checks with the full tributary load. Effectively saying all the load goes between the bolts.

This is how you would check a column that had holes drilled in it for example at mid height (usually codes have a clause that might state you can use the gross area if the area deduction for holes lies under a certain percentage limit based on the ratio of f_y to f_u).

What exactly are you checking when you refer to checking the 'plate buckling', are you checking the flange that is sitting on the new support as a small sway column (pinned at the bottom with some restraint from the 1/4" plates at the top)? The flanges are pretty stocky looking so there would presumably be plenty of capacity, I'm more concerned that the 1/4" plate doesn't really restrain too much in the middle of the 8' width if you shove the whole thing sideways (flange plates simply twist and fail in torsion), you don't want it to fold over sideways if you know what I mean. You can always temporarily weld it to the support to give it some stability (the sides also help I guess to prevent twist of the bottom return, and its probably predominantly going to sit at the corners).

Seems like a lot of effort to go to when you could just unbolt each section one/two at a time, lift them off and cut the supporting pipe and then rejoin them at the lower final level (guessing you can't do this for one of the reasons you state and have been through and discounted it already!).

 

[delagina]

Lots of existing pipes above that will prevent this manifold to be removed from above. The 1/4" plate manifold has angle stiffeners around it. But I'll check it if the supports are at the flanges.

The bottom portion will be removed and slided towards the direction parallel to manifold. It wont be removed sideways. Not sure exactly what you meant flange will fail in torsion.

 

[delagina]

They will cut bottom 3' and weld a new one but thats not part of my scope. Elevation will remain the same. I think the bottom of big pipe is rusting and they want to remove amd replace it.

 

[SAIL3]

I would say it is ok by inspection...but, to confirm that, maybe the OP can give the total load @ the temp support, fla size, pl thickeness...the load/reaction should be reacted by the vert fla on both sides of duct ...the horiz fla really don't do much...if the bottom 1 or 2ft of the vert fl can deliver the load to the side pl of the duct then I would assume it's ok....

 

[WARose]

[blue](delagina)[/blue]

My question is how do I account for the bolt/bolt holes in this buckling check?

If they are small holes......I'd ignore them. By my comment was more about the box itself than the flange.

[blue](Agent666)[/blue]

What exactly are you checking when you refer to checking the 'plate buckling',...

It's the same concept as checking aspect ratios for flanges for local buckling in I-Beams and web crippling. With the former, he is likely changing the flexural stresses with this temporary support so that is worth a look. And with the latter, the support condition could induce buckling from shear so that is worth a quick look as well.

As far as the flange goes, I would think that would be pretty straight forward based on the moment at that cross section. (I'd account for some temperature stresses as well....if applicable.)

 

[delagina]

Sail3, I agree about vertical flange as critical. But not sure how to check it. The plate of manifold is welded to it which will prevent buckling on weak axis. Maybe I should model this in FEA

 

[SAIL3]

not knowing the load involved, the following suggestions may change if the load is substantial...

1..check vert fla @ corner for bearing..using also a portion of the horiz fla and also 16t of side pl as bearing section.

2. if the load is not substantial, it can be sheared into the side pl over a small distance so that overall
out-of-plane buckling of the vert fla would not be a concern...check allow shear in side pl locally to get that distance( kind of looking at this as a local support bracket)...or check the vert fla as a col that is loaded by a continous load over it's length...others on the forum may want to weigh in on this.

3. once the load is sheared into the side pl the OP may want to check the side pl as a deep girder from support to support..

4. one concern of the set-up I have is that the fla may have a tendency to "walk-off" the WF if it is not ligned up with the web of the support WF and held there...I would, at least, consider putting a web stiffener on the WF especially @ the corners or use a box bm instead of a WF..

5. I personally would not expect much usefull info from FEA in this case...

 

[delagina]

I’ll just model the whole flange in FEA including the plates. I’ll put joints in the flange every 3” so i can connect the plates for proper load transfer.

I’m not sure yet the correct unbraced length I should use for the vertical flange because it has the plate welded to it. I’ll go conservative first and check it using using whole length as unbraced length.

 

[charliealphabravo]

I was thinking you would need to consider rotation of the flange from vertical due to deflection of the manifold. When I worked it out it seems that the combination of stiffness and self weight results in a very small deflection.