Cl 3.4 of B31E allows for Design by Analysis. Longitudinal stresses are not to exceed the lesser of:

  • 2.4 S
  • 1.5 Sy
  • 408MPa

With a typical carbon steel, the 2.4 S value governs – so applying B31E to a B31.1 system using A106B @200ºC, 2.4S = 2.4(117.9) = 283.0 MPa.

For B31.1 longitudinal stress is not to exceed kSh, which is 1.2Sh (events acting for no more than 1 hr or 80hrs per year).

So the allowable stress value for B31E is twice that which is allowable under B31.1.

The definition of the moment to be used in calculating the stresses is subtly different:

In B31E the moments are defined as:
  • Mseismic = elastically calculated resultant moment amplitude due to seismic load, including inertia and relative anchor motion
  • Msustained = elastically calculated resultant moment amplitude due to sustained loads concurrent with the seismic load

This contrasts with B31.1 where:
  • MA = resultant moment loading on cross section due to weight and other sustained loads
  • MB = resultant moment loading on the cross section due to occasional loads, such as.....earthquake

The key difference here is between “elastically calculated resultant moment amplitude” and “resultant moment loading”.

The term “amplitude” in B31E is not defined. Amplitude is variously described as the height of the half wave (zero to peak), or as the full wave (peak to peak). It isn’t clear to me which of these is intended in B31E.

Is it:
  1. the peak-to-peak amplitude (in which case B31E doesn’t change the allowable stress from that allowed in B31.1), or
  2. the zero-to-peak amplitude (in which case B31E allows the occasional stress to effectively be twice that allowed in B31.1).

Any inputs welcomed.
_________________________
Ross Sinclair