I have met a case to analyize the stress of GRP piping, anyone can tell me what kind of code should be select to analyize?

I have considered BS7159 or UKOOA, but the manufacture's product does not comply with the two codes, how can I to do this?

Thanks!

FPSO

BS7159 is a broad church Code of Practice it gives reference laminates but you can apply it to bespoke data. Never used UKOOA.
In Caesar's "Configuration set-up" are you using any of the manufacturer data AMRN2020 or WAVIN55, WAVIN63 or Wavin73? Or would you say none of these. This is important because your selection here will fix the elastic constants for the orthotropic modeller. SO MORE INFO PLEASE.
Starting a Caesar model you have to specify material 21 to get the orthotropic modeller.
At this point it is entirely at your option in clicking the allowable stresses box to specify which post processor Code to follow. You could specify B31.3 but BS7159 is best since you can then specify the ratio of hoop to axial modulus.
If you used a manufacturer in "Configuration set-up" you also get an allowable stress to the output.
Now I recommend BS7159 because you get a comprehensive stress output. I do not think anybody can reasonably argue with the output for straight pipe.I used this with Future Pipe Industries (FPI)WAVIN55. FPI state fittings are to be thicker than pipe according to the ratio of so-called "Hydrostatic Design Stress". Well for the pipe this is 63MPa and for elbows 40MPa. Yet we can also infer from this a fixed bend SIF 63/40 = 1.575. For the 362.4 o/dia x5.7mm pipe and 9.0mm thickened bends the BS7159 SIFs were 1.960, 1.444, 4.096 and 4.044 for in/out plane and axial/hoop co-directions. Obviously BS7159 was not suitable here.
To cut the story short FPI state the laminate is limited by the equivalent stress "calculated with the use of the continuum theory". So by looking at Caesar's so-called "Von Mises Stress" I was able to ascertain the stresses were OK. [Caesar's "Von Mises Stress" does not equate to an equivalent tensile stress but rather to the "octahedral shear stress" the same value multiplied by [sqrt(2)]/3. You see a sudden drop in values if you change configuration from Tresca to so-called "Von Mises" using B31.3. This is why.]