Designing a Safe Beam Support for an Exhaust Pipe

[hughfrancis89]

Hi there! I have recently taken on an engineering job and have been tasked with designing a simple beam support for an exhaust pipe.

Because I'm still studying and I want to make sure I do this properly, I have decided to consult Physics Forums, as reading others posts have helped me so much in the past.

I have attached a picture of the design to this post. Basically I've just got to make sure there is a significant safety factor against yielding with the setup.

I don't have any numbers to work with currently, I'm just getting the theory down pat first.

I assume I need to find the yield stress of the material, and compare that to the bending stress currently caused by the load from the pipe. Also, how would I calculate the load caused by the pipe on each support? Would it just be the weight of the pipe divided by two (due to the two supports) or would it be more complicated than that, because the supports arent evenly spaced?

Also I assume to model the top beam I would just use a fixed end beam as in FIG 2.

Thanks for the advice, apologies for the simple question!

Hugh

 

[nvn]

hughfrancis89: We don't really know the fixity of your top beam. Therefore, it might be safer for designing the midspan of the top beam if you assume a simply-supported beam for the top beam, instead of a fixed-end beam.

Draw a free-body diagram of the pipe and hanger supports. Use static equilibrium equations (summation of forces, and summation of moments) to solve for the tensile force in each hanger.

If you want a safety factor against yielding of, e.g., 1.50, then the applied bending stress on your beam should not exceed the beam material tensile yield strength divided by 1.50.

 

[minger]

Important thing to remember:

Yield strength is typically a single cycle load application. In your exhaust system, do you anticipate any vibration or cyclic loadings? In so, then you need to also look at fatigue.

 

[hughfrancis89]

Thanks for the replies! I should model the pipe as a distributed load, right?

 

[minger]

Looks like two point loads to me. Why would you think of it as a distributed load?

 

[nvn]

I should model the pipe as a distributed load, right?

hughfrancis89: Yes, that is correct, if you are referring to paragraph 2 of post 2, to solve for the hanger point loads.