Dear all,

I'm surprised the topic of flange supports hasn't been visited here in depth, yet.

This is what I found.

This is what some folks at eng-tips had to say.

Questions I would offer to the group:
1. How would you approach mathematical qualification such a design? Failure of the bolts or support plate is one thing, but qualifying this flange against leak is a separate issue.

2. It's common practice (amongst some end-users) to replace bolts on flanges in service, only one at a time. However, it looks like swapping out the support would be predicated by a shutdown. I see a few designs that attach to roughly half the bolts, and I see designs that attach to the bottom two bolts. Has anyone heard of any practice regarding the ability to disconnect two adjacent bolts while in service?

Thanks for taking the time to read and offering your thoughts.

Never used voluntarily such support. It is quite popular in German companies, especially for sustain the valves.

I don't know how a transverse force affects leakage evaluation because intuitively it seems that such force does not participate directly to the tendency of flange (or a part or parts of flange) to get opened (releasing the seal stress on gasket) as it is happening under bending moments and/or tensile axial forces. As theory, would be a "second order" effect of the transverse force applied on the deformed flange under bending moment, but I've never seen such work.
Of course the support placed there changes the bending moment in that point of system comparing with the case when the support is missing, but this is already captured in the calculation.

Never accepted a procedure to replace bolts on flanges in service.

It would seem to me that any axial displacement of the pipe relative to the base of the support would result in bolts being placed into additional tension. Any lateral or vertical displacement relative to the base would place the bolts in shear. Rotations result in combinations of the above to degrees relative to the distance from the axis of rotation.

Bolt fatigue is an obvious repercussion. By adding tension to the bottom bolts, you compress the bottom half of the flange. The flange may react by itself applying tension to the top bolts, and possibly leaking out the top, and even if not, fatigue the bolts on top, as well.

But there would have to be significant movement in order for a leak to make its presence known early on, and that hinges on the support itself not having any give.

Again, thanks for your thoughts.

I think such support is usually a rest type and the axial force in piping is a result of friction. Such force will create also a (local) bending moment and the additional displacements would create the effects you mentioned (these are "second-order" effects in structural engineering terminology). As practical approach with Caesar, I think one can neglect the second-order effects (mainly because we don't know how to handle them) and estimate roughly the benefit of using such support by the net effect of the support in terms of bending moment, comparing the bending moment in flange when support is missing with the case when the rest support is modeled in place.