I have a pipe that run on a structure and on a rack.

i have to analyze earthquake load case.

For pipe the earthquake is 0.35g along the direction x and 0.35g along the direction y.

Client confirmed that during the earthquake:

The metal rack will move:

Direction X: 50mm

Direction Y: 70mm

The concrete structure will move:

Direction X: 80mm

Direction Y: 100mm


On the pipe i insert a uniform load U1=0.35g (X direction) and U2=0.35g (Y direction).

Th pipe rest on structure/rack with shoes having friction 0.35 and there are also some guides with gap=0.


QUESTION:

how to model the structure and rack seismic displacement under the pipe?

how do i consider the deformation of structure and rack under the pipe?


can i, for example, for each support insert a Cnode to a small rigid and for each small rigid associate a displacement equal to structure or rack displacement below the pipe?

If yes, in quake load case I will have

W+T1+P1+U1+D1 for X direction

W+T1+P1+U2+D2 for Y direction

Where

U1,U2 are 0.35g on pipe

D1,D2 are structure/rack seismic displacement


in this way, pipe support shoes will allow sliding by friction even if below it the rigid (structure or rack) moves. do you think it?s correct or do you have any other suggestions?


If yes: have I to consider structure and rack seismic displacements having same direction or opposite directions? Or both cases?


If no: how can I model this case?

My opinion is to run model 1 and ignore all seismic displacements - do accelerations alone.

Models 2 through X will be sections of that model where you have relative displacements that are a problem, and chances are you don't actually need supports in that model, you just have pipe going from point A to B (support 1 and support 2) that are just displacements. You can assume displacements will be thermal plus relative support displacement (again, not total displacement).

Support 1 moving 100 mm and support 2 moving 50 mm is the same as saying support 1 moves 50 mm and support 2 moving 0 mm.

xxxxx

if i well understood.

if i have a pipe running from a structure (on structore there are 5 pipe supports) to a piperack (on piperack there are 6 pipe supports)

i will analyze with different caesar models:

Model_1) i analyze only seismic (0.35g x direction and 0-35g for y direction) and structure don't move.

model_2) i analyze only relative displacement between structure and piperack. if structure moves 100mm and piperack moves 60mm, i will consider both case displacements have same directions both case diplacements have opposite directions.
So:
model_2-subcase_A): 100+60=160mm for structure and 0mm for piperack
model_2-subcase_B): 100-60=40mm for structure and 0mm for piperack

BUT!

(continues below)

in this way i'm analyzing seismic acceleration on pipe (0.35g) separately from structure/piperack displacements.

Structure/piperack displacements are consequence of seismic acceleration on structure/piperack: civil department made calculation to calculate those displacement starting from acceleratio on structure/piperack. So i think (here my starting question) i have to consider seismic acceleration (on pipe) and displacements (on structure/piperack, consequence of seisimc acceleration on structure/piperack) ACTING AT THE SAME TIME. Here is my question.

so: how can i model all this to check structure&piperack displacements (due to seismic on structure&piperack) acting together with seismic acceleration on pipe don't crush the pipe? (overstress,...) so to check if, for example, i need to insert loops or expansion joints to separate pipe on structure from pipe on piperack?

See attached zip folder. Contained within is an mp4 movie file. Of course, I didn't include all combinations, but you get the idea where this could lead.

You have a choice: 1. attempt to simulate pipes sliding around on top of these structures, or 2. make them stick to the structure and ensuring what happens between the structures works for the pipe.

What's truly secured to the structure doesn't care about structure displacements.

The secondary models won't be used to simulate temperature, friction, accelerations, or displacements.

You will INSTRUCT CAESAR what those displacements are to be.

see attached file for two caesar tests (structure/rack displacements according to first post). See txt file inside .zip for the two caesar model explanation.

what do you think about?

here the description of support resting on structure and on piperack

The results make the technique used seem viable. You're going to want to address any relative movements, however. E.G. node XYZ moves in the +X and node 10 moves in the -X, you will have shoes slip off supports - assuming you're using shoes.

it's true. i didn't think about the "shoe slipping off support" problem. mmmm.. so also if pipe don't go overstress, i have to check also real physical shape of support to be sure is "large" enough not to let shoe fall down... more and more complicated :P