Pipe support design 2

[Bounty001]

I have the load on restraint output from the stress analysis of the pipes using caesar II. Do I still need to calculate and apply friction force in my pipe support design or has this been incorporated into the caesar output?

 

[dik]

The Caesar output should provide forces at the pipe supports and they should be designed for these forces, incremented as needed.

 

[dik]

Should have added, if there is a guide, then the support for the guide should be designed for the vertical load x coef of friction.

 

[TLHS]

Personally, I would design any vertical support for a lateral load equal to the lower of:

- The vertical load times the friction factor
- If resisting thermal loads, the load necessary to create a deflection in the structure equal to the largest deflection the pipe could see

If the Caesar analysis provides a greater load than the above, I will use it.

Generally, the stress analyses I see are fairly rough. You aren't going to fully take into account things like structure stiffness, install tolerances, and other things that can shift the loads around. Generally a failure of the structure is a worse situation than a failure in the pipe, so I'd prefer to design to a limit state that I'm a little more confident in. I know the pipe will slip when it exceeds friction and I know the pipe can only expand a certain amount.

Make sure you're being realistic though. Make sure you aren't assuming forces that would obviously fail the pipe out of plane and things like that. Have a conversation with your pipe guys.

dik, maybe I'm not quite getting what you're saying. If there's a guide then the guide should be designed for the calculated force on the guide. Friction won't be a limit on the load in this case, as there's a mechanical connection to the pipe. If it's near a bend, that can be a calculated value due to expansion, transients, or other forces. If it's outside on a long stretch it may be calculated based on wind loads. If it's just making sure the line doesn't snake it might be a percentage of the axial capacity of the pipe based on standard company or client practice. Guides, stops and anchors are the spots where you really have to count on the stress analysis because the forces can exceed frictional loads. There is a mechanical connection at those locations.

 

[BAretired]

Provide more details and a sketch.


BA

 

[Bounty001]

Good day all,

Thanks for your response to my earlier post. The pipes are heavily rated with huge valves on the line.
An extract from the caesar output at one of the nodes which I'll be providing support for is shown below (the lines are straight run with no bends).
Node Fx N. Fy N. Fz N. MX N.m MY N.m MZ N.m DX mm DY mm DZ mm
0001 Rigid X w/gap; Rigid +Y
2 (OPE) 0 -51179 0 0 0 0 -0.290 -0.000 0.238
3 (OPE) 0 -27603 0 0 0 0 -0.166 -0.000 1.099
4 (SUS) 0 -45998 0 0 0 0 -0.014 -0.000 -0.157
5 (SUS) 0 -45998 0 0 0 0 -0.014 -0.000 -0.157
MAX 0/ 2 51179/ 2 0/ 2 0/ 2 0/ 2 0/ 2 0.290/ 2 0.000/ 2 1.099/ 3

Apologies, for the disorganized data. My interpretation of this loading is that vertical load is Fy based on pipe stress orientation and no lateral forces due to friction or thermal forces.

My concern is based on the displacement and the gap in the Rigid X. Would there be friction force generated in real life and is it proper to assume 0.4 (steel on steel) friction coefficient?
This assumption would throw outrageous moment into the foundation of the proposed T-type support.

Thanks.

 

[TLHS]

Talk to your piping guys. It looks like they didn't include frictional restraints in their model. That's fine, but it means you definitely need to account for something. You'll get *some* sort of force on your structure just from a pipe expanding under thermal loads between a pair of supports. If these are the only support structures in the system, I'm also kind of curious where all of their forces are going. Have they got a directional stop, anchor or something else in their system that connects to structure you should be checking? Are they taking it entirely at the equipment nozzle? Maybe they're just giving you the output from the gravity case?

I'd probably design for 30%-40% friction in the axial direction. Depending on the geometry, I'd either match that or use a lower value in the direction perpendicular to the pipe axis.

 

[JLNJ]

I agree with TLHS about talking to the piping guys. I have found that they generally do not model the supports in the same manner a structural engineer would. Too often they apply restraint where there isn't any (or enough) and combine restraints, springs, and releases in un-buildable supports (e.g. "I want to fix X and and a spring on Y but have a slider on Z with fixity for Mx and Mz but make it free in the My and do all this 50' above the ground where there is no structure to tie off to. And all the fixed points need to have infinite stiffness 'cause that's how it's modeled by the program.")

You might need to bargain with them for certain kinds of supports where you can get them. A good piping guy knows how and where to place the supports to keep the reactions few and reasonable and keep the stresses in the pipe below the allowables.

As for the friction, you need them to tell you how it's modeled and what you need to provide. It all plays in to the behavior of the pipe and the stresses therein.

 

[MiketheEngineer]

OR how about hiring someone who has been there and knows what he/she is doing??

 

[Bounty001]

My appreciation goes to everyone.
TLHS; the line is buried at one end,about 15m in length and has huge valves + fittings along it's length. I'll discuss with the pipe stress guys as advised.
Thanks.

 

[ukengineer58]

Mike surely learning new skills is what it is all about. if we all hired someone else all the time pretty soon no one would be left with experience