Pipe Hanger Seismic Analysis - Moment in Rods?

[AThor]

I am trying to understand and analyze a detail for a hanger support with seismic brace for a run of multiple pipes. The pipes are sitting on beams that are held up by threaded rods, coupled to anchors embedded in the concrete above. So far in my career I have not managed to fully wrap my head around the topic of seismic support for pipe systems, and it seems the literature out there is sparse, lacking detail, and usually manufacturer specific.

Looking at this detail, my first thought is that the pipe seismic forces from ASCE 7 chapter 13 are not applicable, because the pipes are not coupled to the supports, they just rest on the beams. Seismic motion in the concrete structure would cause the hanger itself to move, and the pipes would just bang around inside the beams and rods, no?

The other aspect to this is whether or not there will be any bending induced in the hanger rods. Let's say the pipes were all strapped onto the beams so they moved with the hanger under seismic motions. In that case, wouldn't the seismic force from the pipelines create bending in the rods, transferred through the beams, that would clearly fail the rods? With the pipes not being coupled to the beams, is this no longer an issue? The rods have such minimal bending capacity, I'm worried that even the impact from the pipes banging around and hitting the rods might cause too much moment, but I don't even know how I would go about checking that.

Your thoughts and ideas are appreciated.

 

[bones206]

Bracing the hanger makes little sense if the pipe is not fastened to the hanger.

If the hanger is properly braced, there should not be significant bending in the rods. I believe Unistrut and maybe Tolco have seismic design manuals with worked examples.

 

[Tomfh]

The seismic bracing is to the stop the hangers and pipes swinging around. The pipes have to be fixed to the hangers/bracing for it to work.

 

[bones206]

It's all about restraining degrees of freedom. The goal is to provide the minimum amount of restraints in each direction required to keep the piping system stable.

The gravity DOF restraint is pretty straightforward. The pipe just needs to rest on a hanger every so often. No clamp needed.

In the lateral DOF, lateral braces are required at some spacing depending on the magnitude of the lateral seismic force. The pipe needs to be restrained to the hanger laterally so the hanger can deliver that force into the brace. If you have braces on each side of the hanger they can be designed as tension braces. If it's just on the one side, the brace has to be designed for tension and compression.

Similarly in the longitudinal (parallel to pipe) direction, a restraint is required. Usually you don't need as many of these, since the axial capacity of the pipe is relatively high. Same principles apply to brace design and restraining the pipe to the hanger such that the longitudinal force gets delivered from the pipe to the hanger to the brace.

That's a really simplified overview. Read this document front to back and you will be an expert:

https://www.unistrut.us/resources/literature/seismic-bracing-systems-opa

 

[AThor]

Thanks for the replies. That's a good summary bones.

I have just read through the Unistrut examples, and I also found this reference from Kinetics Noise Control that I'm going to post here for any future users who come across this thread, because it gives a good explanation also.

https://kineticsnoise.com/seismic/seismic_restraint_pipe_duct.html

I still have one issue. Let's say the pipes are attached laterally so the seismic braces are required. All the examples I've seen have a single trapeze beam or clevis hanger attached to the bottom of the hanger rods only. I haven't seen examples or discussion for a multi level trapeze hanger like this. Wouldn't there be moment in the rods since the higher trapeze beams tie into the middle of the rods?

 

[bones206]

In a multi tier hanger, each tier should have a brace.

 

[strucbells]

Yes, for a multi-tier support frame if you are only bracing the bottom beam, the lateral load on the middle tier(s) will put a point load on the hanger rod that is spanning between the attachment point to structure above and the brace below. Typically this doesn't calc out for most hanger rods unless loads and spans are very small.

A couple of ways I've seen this resolved is by bracing every tier (gets a bit messy esp for >2 tiers) or by swapping out the hanger rods for 1 5/8" strut vertical members that have higher flexural capacity. You may have to dive into the strut catalogs a bit more to figure out what fittings/attachment options the second solution requires.

I provided this comment recently reviewing a pipe support deferred submittal where the contractor submitted a design with only the lower tier braced. The engineer they hired briefly tried to argue about some sort of frame action due to partial fixity at the connections, but then when they returned the submittal, they had provided braces at every level. Think it would have been difficult to quantify/prove.

 

[bones206]

I worked on a nuclear plant project back in the day where my colleague did a really elaborate nonlinear analysis to show unbraced hangers could survive the sway action in an earthquake. He used plastic hinges and checked ductility and plastic displacements, etc. It was a lot of work for a special circumstance. Most projects wouldn’t take that approach.

Another approach that can be used in certain scenarios is swivel connections where the rods connect to the supporting structure. This lets the hanger sway while minimizing rod bending. Check the applicable code for where those can be used.

 

[bones206]

ASHRAE Practical Guide to Seismic Restraint is another really good reference if you can get your hands on it.

 

[AThor]

Thanks for all the input, this discussion is helpful. Strucbells, that sounds very similar to my situation. The detail has unfastened pipes, and only lower tier braced trapeze beams. It could be a two wrongs making a right situation, but I'm skeptical. If we don't have the pipe weight as part of the swaying action, then we don't have bending in the rods, but we also don't really need the braces. As soon as/if we fasten the pipes to the hanger, then we certainly need the braces, and a solution to stiffen and/or brace the rest of the rod. Swivel connections would make sure no moment develops at that end of the rods, but if the brace is present as shown in my sketch, it wouldn't necessarily prevent moment in the middle of the rod, because the rod is still like a pin-pin beam loaded in the middle with the upper tier trapezes carrying the pipe seismic force.

I work with a few mechanical engineers, I will check if they have that ASHRAE standard laying around. Thanks for the suggestion.

 

[bones206]

I think you are looking at it a little backwards. First you determine if the pipe needs to be braced. If it does, then you detail the hangers to be braced every so often. You also detail the pipe-to-hanger connections to ensure the pipe force transmits into the brace and is not sliding around freely.

You shouldn't look at the detailing and decide from there if the pipe needs to be braced or not.