Should Friction Be Turned Off for Seismic (Occasional) Stress Computation?
Should Friction Be Turned Off for Seismic (Occasional) Stress Computation?
(OP)
Hi guys. I have searched the forum and did not find an answer so please forgive me for posting if the answer is already out there.
I cross-posted this from the Coade forum to hopefully reach a wider knowledgeable audience.
Question: Can credit be taken for friction when computing seismic stresses (occasional, aka OCC in CAESAR-speak)? In other words, do I turn off friction, or leave it on, when evaluating seismic stresses in piping?
I am talking here about the computed stress level in the pipe, not loads on restraints.
B31.3 is silent on this issue, as far as I can tell. ASCE 7-05 says "Friction resulting from gravity loads shall not be considered to provide resistance to seismic forces". Note that that section of ASCE 7 in context is referring to the design loads to be used for the design of the restraints, not the evaluation of stresses in the piping. That tells me that, yes, I do turn off friction to calculate restraint loading, and I use those loadings to design my restraints. But that has nothing to do with the stress level in the piping itself.
In my opinion it is an out-of-context extension of ASCE 7 to assume this section tells me I have to turn off friction when evaluating OCC seismic pipe stresses. To do that I believe will cause excessive conservatism in the OCC stresses. It is not known a priori whether a seismic event will lift the entire system off the restraints. That would be the most conservative approach however so perhaps that is why folks do it. But that approach also leads to overdesigned restraints and anchors.
I can't be the first guy to have asked this. What say ye? Thank you!!!! Pete
I cross-posted this from the Coade forum to hopefully reach a wider knowledgeable audience.
Question: Can credit be taken for friction when computing seismic stresses (occasional, aka OCC in CAESAR-speak)? In other words, do I turn off friction, or leave it on, when evaluating seismic stresses in piping?
I am talking here about the computed stress level in the pipe, not loads on restraints.
B31.3 is silent on this issue, as far as I can tell. ASCE 7-05 says "Friction resulting from gravity loads shall not be considered to provide resistance to seismic forces". Note that that section of ASCE 7 in context is referring to the design loads to be used for the design of the restraints, not the evaluation of stresses in the piping. That tells me that, yes, I do turn off friction to calculate restraint loading, and I use those loadings to design my restraints. But that has nothing to do with the stress level in the piping itself.
In my opinion it is an out-of-context extension of ASCE 7 to assume this section tells me I have to turn off friction when evaluating OCC seismic pipe stresses. To do that I believe will cause excessive conservatism in the OCC stresses. It is not known a priori whether a seismic event will lift the entire system off the restraints. That would be the most conservative approach however so perhaps that is why folks do it. But that approach also leads to overdesigned restraints and anchors.
I can't be the first guy to have asked this. What say ye? Thank you!!!! Pete





RE: Should Friction Be Turned Off for Seismic (Occasional) Stress Computation?
IMHO friction should be ignored for seismic analysis. The reason is that you never know whether or what degree of friction is applicable at each support due to the effects of the seaimic vertical loading. The weight distribution , and hence friction, (or even horizontal loads and horizontal friction) are all based on the pipe reaction load which varies during a seismic event. Also seismic analysis in Caesar is linear whilst friction is non-linear so you cannot have non-linear supports in the seismic analysis (unless you are doing a static G approach).
RE: Should Friction Be Turned Off for Seismic (Occasional) Stress Computation?
I appreciate your thoughts and methodology. As I said I believe it leads to conservatism but I understand the rationale behind it. I'll wait and see what others have to say. Thanks!
RE: Should Friction Be Turned Off for Seismic (Occasional) Stress Computation?
I had a similar query and I guess the only response i got was from DSB123.
But in my view not considering friction is (overly ?) conservative.
Take for example the case of analyzing water hammer on a mainly horizontal laid out pipeline (with dia. > 24" usually).
Given the large magnitude of frictional forces at the supports I am given to believe that this should contribute to reduce pipe stresses during an event.
I just cannot find support for this though.
I know that dynamic analysis considers linear restraints only but the Caear II manual also provides a method for including for support friction.
I was really looking for feedback from analysts who have used this method.
RE: Should Friction Be Turned Off for Seismic (Occasional) Stress Computation?
So, if friction is 100% (a slide fails or is rusted up) that increases certain reactions (forces extra movement) in some restraints and some support steel, and decreases others.
Certainly a frozen (rusted or painted over) Teflon-covered slide is going to stop motion. But is a 100% rusted restraint realistic every time? More likely, the water hammer impulse is going to hold for a few milli-seconds, then break loose, force allow movement of the pipe that will reduce the final impact of the water at the end of the line. the energy lost while breaking the rust will (realistically) "help" reduce final impulse, but quantifying how much that is is very hard to do.
RE: Should Friction Be Turned Off for Seismic (Occasional) Stress Computation?
The OP was talking about seismic loading as were you in your previous post you refer to so what I responded was correct. You are now bringing into the equation another form of loading (Water hammer) so my response would have been different. My response to Seismic loading is valid!!! For water hammer we have a different scenario. I tend to answer the question posed at the time as most people do. Do not "interpret" an answer on Seismic loading to all other forms of dynamic loading!!!!!