We are carrying out design of an above/below ground FRP piping system and have some queries re the seismic design (which we are looking at for the above ground piping only).

We have the option of designing the piping to either BS7159 or B31.3. When we use B31.3 the seismic design to this code is fine.

However when we design to BS7159 the seismic design is proving a little more difficult. BS7159 mentions wind loads but does not appear to mention seismic loads, so we decided to use the same occasional load combinations as we would use in B31.3 as a starting point,ie sus+U1, sus+u2 etc. However doing this we have run into some problems.

In some areas of the FRP piping the code allowable stress is 30MPa and in others it is 60MPa (due to the properties of the FRP pipe). We have a node point with a code allowable stress of 60MPa and a sustained stress of 45MPa. However when we define the occasional seismic loadcase as sus+U1, we are getting an occasional code allowable stress of 39MPa (30+30% occasional load factor).

We then do not comply with the code at this point as the sustained load is 45MPa whereas the sus+U1 allowable is only 39 MPa. We are already exceeding the allowable stress with just the sustained component, before we even begin to consider the seismic loads.

Any body with any ideas? Any suggestions/advice would be greatly appreciated.

The allowable is based upon which of three stress equations govern. So one load case could have 60 while another has 30, etc. (especially if the first is governed by hoop stress).

However, to answer with any sort of detail, we'll need to look at the input and the load cases. Can you e-mail us the ._A and ._J files?

Itchy

I would recommend using ISO-14692 as a design standard. This is the latest and most up to date standard available. The philosophies in ASME B31.3 for metalic pipe should be avoided for displacement and occasional loading.

I am currently revising my load case files for occasional loads on the recommendation from our GRE pipe manufacturer. The load case I am using is effectively W+P1+T1+(occ load), where P1 and T1 are the operating conditions not design. This is the total load the pipe can be subjected, and must be considered. No thermal shakedown can occur in GRE piping.

The points made by Nigel Marsh are very true. FRP and GRE materials are not ductile, so the philosophies in ASME B31.3 for metallic pipe are not as applicable for the failure mechanisms of FRP. Negligible plastic yielding occurs, so "No thermal shakedown can occur in GRE piping" and it would be recommended to change the configuration to use the B31.3 "Conservative" allowable equation 1a ,instead of the "Liberal" allowable equation 1b.
This does not change the SUS allowable for your problem, though. My approach to solving the SUS over-stress problems would be typically to add more supports. Does the FRP piping seismic restraints include extra guide restraints and line stops (at least 2 guides and 1 stop per run between changes of direction?)

Thanks for the above suggestions. We'll look further into these ideas.

Since under the normal sustained loadcase we do not have a code failure we don't need to add more pipe supports - it is purely a matter of finding the correct way to define the occasional loadcases.