Pipe Support evaluation - Friction for dynamic load cases?

[pettervu]

In our pipe support evaluations we've always considered friction for all load cases, both static and dynamic (including waterhammer, global vibrations etc).

We do feel that this is quite conservative, but haven't found a convincing argument to reduce this assumption.

Now we've found that secition 6.10 in the AREVA document "U.S. EPR Piping Analysis and Pipe Support Design" (link below) assumes that only frictional loads from deadweight and thermal expansion need be considered. The argument is quite short "Since friction is due to the gradual movement of the pipe..." and there are no references for the basis of this assumption.

https://www.nrc.gov/docs/ML1214/ML12146A080.pdf

As we've seen it the pipe may move quite a bit while vibrating which could cause additional frictional forces from dynamic loads.

Could someone please elaborate on the reasoning behind ignoring frictional forces caused by vibrations or direct us to a more comprehensive explanation to this?

 

[LittleInch]

I would say if the pipe is vibrating or jumping due to surge then the vertical force on the support reduces.

Most pipe systems don't see significant vibration or surge.

More important in my opinion for support loading is what friction factor you use. The seemingly standard 0.3 looks low to me for a rusty metal to metal support.... I would do a check at 0.7, especially for the high movement locations.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

Not sure how you can have friction caused by vibration. Vibration is a movement not usually associated with support locations, but a transverse movement occurring somewhere within the free span, normally with a max amplitude at center span. Sustained vibration movements are usually equal in both directions. If friction was to be created by vibration, it would be first in one direction, then the other. Since friction is caused by sliding, that implies vibratory sliding at a support. Stress in the pipe would first be shear and bending caused by movement in the initial direction, then followed by equal and opposite stresses from movement to the opposite direction. Net average stress would appear to be zero.

 

[HTURKAK]

for all load cases,[/b] both static and dynamic]

Although i read the thread a few times , still am not sure that i understand...

Let me explain my opinion, if the source of global vibrations are seismic loading;

- The vertical accelerations can reduce and increase the effective friction force,
- Friction resistance is not be considered for seismic restraint for process piping and for short runs,
- Friction loads is not combined with wind or earthquake loads..

 

[KevinNZ]

Vibration will reduce/release friction load. Movement like thermal growth drags the pipe across the support and causes force in the direction of movement. Friction restrains movement so it pays to check movement with no frication in case the pipe falls off the support. From experience of failed supports/foundations we use 0.6 friction for design of supports. You can get high frication loads when pipe shoes are new and have coating to scrape off or when the are warn, corroded or damaged.

For seismic loads we would ignore friction as a pipe moving freely over the support gives higher loads on the anchor points and will load up the sides of guide supports. You can also check EQ loads with friction.

 

[1503-44]

KNZ, You mean vibratory sliding, like a back and forth movement while maintaining contact between pipe and support, reduces friction? How? Can you explain? Would not friction be the same 0.6W, first in one direction than in the other?

 

[KevinNZ]

1503

I am thinking of vibration with a vertical component and piping with anchors. Yes there will be cases of longitudinal or transverse vibration loads that do put frication on supports. When there is thermal expansion to design for, then these loads on the support will all ready to allowed for.

 

[1503-44]

I think that would be unusual, since friction should dampen all but the worst vibration quite a lot. It wouldnt even move until higher than static friction load (and thats really high with 0,6) and sliding friction would be far less, 1/2 or lower, since movement has broken any "corrosion tackwelding" and polished it up a bit by then. But maybe. Static friction condition will be worse in any case, esp with that ole 0.6