U-bolt bending stress

[cornemo]

Dear,

a 16" pipe (heavy wall thickness) is supported on a beam and held in place using a U-bolt.
Normally I design the U-bolt based on tensile stress due to possible upward force and shear stress based on lateral force. For easy calculation I divide the pipe force by 2 to get the force for each cross-section of the U-bolt legs.

Now I got the question to also design the U-bolt for bending stress. I found this old thread:

http://www.eng-tips.com/viewthread.cfm?qid=190914

giving some formulas in the last post, but there is no reference with those formulas so I can't use them.
In my opinion the bending of the U-bolt will be very limited because bending is prohibited by the pipe shape due to which it will create very large tensile stresses before large bending stresses can occur.

Is bending to be considered in U-bolt design and how to determine the bending stresses (preferably with reference to design codes/books)?

Thanks in advance for your reply.

 

[racookpe1978]

What bending do you anticipate on the U-bolt, and in what direction do you think it will bend?

 

[cornemo]

I think it will not bend, because the wall of the pipe will prevent the U-bolt from bending. The client thinks the U-bolt will bend due to the lateral (sideways) force from the piping such that the legs of the U-bolt will not be vertical anymore. The mechanical model client is proposing is a U-bolt with a (point)force at the height of the centerline of the pipe. This of course makes the U-bolt bend but doesn't take into account the stiffness of the pipe wall. In my opinion clients idea is very very conservative and it can be assumed that the stiffness of the pipe wall will prevent almost any bending.

I would like to hear some opinions about this.
Thanks in advance for your reply.

 

[desertfox]

Hi

So if I understand this correctly, the u bolt clamps a pipe against a steel beam and at the centre height of the pipe a transverse force will be applied which is trying to move the pipe side wards? So if that's the case the first thing that needs to happen is the force needs to overcome the friction generated by the u bolt and then one leg of the u bolt would see all this force but the other leg would still resist because it's trying to be ripped out of the nut.
I would then treat that as the u bolt trying to shear as opposed to bend although in practice what actually happening is probably very complicated.

I agree with you it is being very conservative and probably unrealistic.

 

[Ron]

The pipe will only prevent the U-bolt from bending in the direction perpendicular to the pipe. Bending can still occur parallel to the pipe and will be exacerbated by friction between the pipe and inside curve of the U-bolt.

Analysis of this is done by considering the U-bolt to be a cantilevered beam with a section modulus of 2x the shaft section modulus.

 

[sdz]

For an outwards load causing tension in the U bolt legs I would only design for the tensile capacity as you have done. A square U bolt might not behave exactly the same.

For loads longitudinally or laterally to the pipe there is probably limited rotational restraint at the bolt hole so the U bolt will tend to act like a cable.

I've been looking for this information myself and considering the wide use of U bolts I'm surprised there is no information available.

 

[cornemo]

Just for information to all of you.
I made a finite element model with a U-bolt with both ends fixed on the steel structure. In this situation the bending moment for my layout with a 16" pipe is approx 0.38*F*R, where F is the lateral force and R is the pipe radius.