I am trying to find further information or explanation regarding the requirements for Code Compliance Cases to be used for GRP piping using both BS 7159 and ISO 14692.

1. BS 7159

BS 7159 does not provide any explicit description of the load cases to be considered. It does however state that the maximum combined stress should not exceed the maximum design stress for the laminate at any location (Cl. 7.3.3.2).

Since fibreglass doesn’t go through shakedown like steel pipe the code doesn’t have or require a check of the thermal range - EXP stresses (although the EXP stresses are used to evaluate the influence of fatigue).

So it is my understanding that we should evaluate any possible loading combinations and check that the combined stress for each of these cases does not exceed the maximum design stress.

So for example if we had a model which included the following loads:

W, P1, T1, U1, U2

then the required load cases for code compliance would be:

W
W + P1
W + P1 + T1
W + P1 + T1 + U1
W + P1 + T1 + U2
W + P1 + T1 - U1
W + P1 + T1 - U1

2. ISO 14692

ISO 14692 takes a different approach. Part 3, Annex D.1 states that:



Further detail is provided in Part 3, Section 7.6, summarised as follows:

The part factor “f2” is used to define an acceptable margin of safety between the material strength and operating stresses for three load cases.

1) Sustained excluding thermal loads (long term), where f2 = 0.67
2) Sustained including thermal loads (long term), where f2 = 0.83
3) Occasional (wind or earthquake or water hammer or blast loading, considered in combination with sustained loads excluding thermal effects (short term), where f2 = 0.89

So using the previous example, I would envisage that the required load cases for code compliance would be:

W (f2 = 0.67)
W + P1 (f2 = 0.67)
W + P1 + T1 (f2 = 0.83)
W + P1 + U1 (f2 = 0.89)
W + P1 + U2 (f2 = 0.89)
W + P1 – U1 (f2 = 0.89)
W + P1 – U2 (f2 = 0.89)

So in my view this takes some of the B31.3 approach, but modifies it to include thermal as a part of the sustained loadings (somewhat like the Appendix P method in B31.3), but it doesn’t look at total stresses in the way that BS 7159 does.

Does anyone know if there are any reference papers or texts that discuss these issues and/or provide further guidance on the implementation of code compliance checks for these two design codes?

Ross Sinclair

In my opinion, you are correct in how you read the load case requirements for the two codes. BS 7159 simply looks at FRP from the point of view that all stress/strain is (equally) bad, while ISO 14692 uses a more "modern" approach, where stresses/strains may have different impacts, depending on such factors as the length of time that they are present, or whether the loads are self-limiting or not.

BS 7159 simply says that the stress/strain shall not exceed a specified value. Therefore you should build your load cases in such a way as to maximize your absolute stresses/strains. (Plus, as you mention, an EXP case if you wish to evaluate fatigue.)

As you mention, ISO 14692 increases the allowable when the loads are fleeting. This increase of 33% for SUS + OCC is somewhat typical when doing structural or piping analysis. The increased allowable for thermal loads might be confusing, especially in the light that FRP does not experience shakedown, but the consideration here is the fact that displacement-driven loads such as thermal expansion or imposed external displacements are self-limiting: as the pipe begins to give, the load dissipates (rather than continues on to catastrophic failure) since the constraint against displacement is reduced.

You ask, are there any documents which specifically name how load cases are meant to be combined? Not that I am aware of. One document that you maight want to get your hands on is the Shell DEP 31.40.10.19-Gen, which recommended modifications to the UKOOA method. This is the document which first recommended the allowable stress increases for Occasional and Operating loads, as you see in ISO 14692.

By the way, if you are trying to use CAESAR II to do ISO 14692, you can get most of the way there through UKOOA. The major differences between ISO 14692 and UKOOA are:

1. SIFS are different for bends for piping diameters below 20”.
2. Failure envelopes for ISO-14692 can be slightly more detailed (and therefore less conservative) often being defined by one more data point (having pentagon shape rather than trapezoid shape).
3. Allowable stress levels may be increased for Operating and Occasional load cases, by 24% and 33% respectively, as compared to Sustained load cases.
4. Failure envelopes are defined for different types of joints in ISO 14692.
5. The “f2” term is excluded from the bending stress in ISO-14692, as opposed to being included in UKOOA.

So what we suggest to people who wish to “simulate” ISO 14692 is:

1. Model and analyze the system in CAESAR II using the UKOOA code.
2. Make sure that the checkbox for “Exclude f2 from UKOOA bending stresses” is checked in the Configure/Setup.
3. If you have ISO 14692 failure envelopes, construct a trapezoid that will fully fit within those envelopes, and use those to get the values you need for UKOOA (Sa(2:1) and R).
4. Use a system factor f2 of 0.67.
5. Calculate the bend SIFs as per ISO-14692 and manually enter any of those which significantly differ from those of BS-7159 (which is what UKOOA references).
6. After analyzing your model, compare your Sustained stresses to 100% of the allowable, your Operating stresses to 124% of the allowable, and your Occasional stresses against 133% of allowable.
7. Where joint strengths may govern, you will need to check them manually against the forces from the CAESAR II local force report.

If ISO 14692 is the code to be complied with the following points have to be noted:

1. ISO 14692 requires qualification procedures from the supplier. This will set some of the design parameters that will be used in stress analysis.
It would be good to confirm if ISO 14692 is to be used on a project.

2.It is also important to finalise the supplier at a very early stage in project and see if they agree to ISO14692 requirements. The allowable stress envelopes may be requested from the supplier for a check of the "operating" stresses.

3.Anticipate some FE( Finite element) studies if Large Dia piping is involved (Support saddles and some fittings may require FE study).

4. The jointing methods (Laminate or Mechanical coupler) also influence the design.