There is a problem of stress analisys that represent an important impact in the evaluation of shells of vessels by WRC 107, this is called the thrust load or thrust force, that for me is an conservative way of analisys of shells, but however always use it. I have had cases when the only application of the thrust force is not withstand for the sell of vessel (why? is bad design of vessel?), but up to now is very difficult explain the "Why Not" use it, although every piping system (arrangement)is different and must has cases where not apply....Do you have any explication?


Adan Rangel Santiago
Coatzacoalcos, Veracruz. Mexico

This is difficult to answer without a few sketches, but ...

The pressure is applied at first closure, the thrust load (PA) is then resoled at the closest, stiffest boundary condituion to that closure. Hopefully, a few examples will illustrate this.

• Assume leaving the nozzle, you have a straight run of pipe to an elbow, with no intermediate restraints. The PA load acting on the elbow will be resolved at the nozzle. Therefore, pressure thrust load should be included in the nozzle evaluation.
• Assume leaving the nozzle, you have a straight run of pipe to a pump. The PA load will primarily be resolved at the pump anchorage. Therefore, pressure thrust load should not be included in the nozzle evaluation.
• Assume leaving the nozzle, you have a straight run of pipe to an elbow, with an intermediate axial restraint. The PA load will be resolved at the axial restaint. Therefore, pressure thrust load should not be included in the nozzle evaluation.
• Assume leaving the nozzle, you have a straight run of pipe to an elbow, with an intermediate untied expansion joint. The expansion joint is too flexible to transmit load, so the PA load will be resolved at restraints on the far side of the elbow. Therefore, pressure thrust load should not be included in the nozzle evaluation. <em>If the expansion joint was tied, you would include the pressure thrust.</em>
• Assume leaving the nozzle, you have a straight run of pipe to an elbow, followed by additional runs, elbows, and restraints. This is a <em>maybe</em> situation. It depends on how far the restraint is from the first elbow.

I hope this helps.

------------------
Regards,
Richard Ay (COADE, Inc.)


[This message has been edited by rich_ay (edited May 04, 2001).]

Richard

I can follow your reasoning to evaluating the pressure thrust load (which always exists) and opposing load from restraint, however I would suggest that this is purely a theoretical argument and the pressure thrust can never be completely balanced. I put the following two reasoning forward for discussion:
1. For these loads to balance the restraint would have to be very rigid. The deflection produced by pressure only at the restraint point if the restraint was released would be very small. The restraint would have to counteract the pressure thrust with a deflection much less than the unrestrained deflection. Even a &#8220;real&#8221; rigid support will be flexible enough to transfer at least part of the pressure thrust back to the nozzle.
2. Even if the restraint were rigid enough to counteract the pressure thrust, any thermal expansion of the pipe would be a far greater than the expansion due to pressure. Therefore practically it would be imposable to balance the pressure thrust against a restraint.

Another case worth considering is case of a manhole in a vessel. Should this WRC calc include the pressure thrust, as there is no restraint to balance? Most low pressure vessel manholes designed in accordance with ASME VIII area replacement method would fail with the pressure thrust alone. Clearly this cannot be correct.

Considering your case where you have a straight run of pipe to an elbow, with an intermediate untied expansion joint. I agree with your point that the shell to nozzle intersection would not see the full pressure thrust (the junction will see the pressure thrust from the difference in area of the pipe ID and exp joint ID). However care must be taken with the loading calculated by Caesar, if the pressure thrust from the expansion joint is included in the piping analysis then the pressure thrust should be included and will balance the WRC pressure thrust.

I would suggest that pressure thrust is not an external loading to be used in the WRC calc, but an internal loading. If this is taken to be the case the allowable stress range (expansion case) should be limited to 2.0 (3.0Sa allowed by ASME for local stresses less 1.0Sa for primary stresses including pressure thrust). Similarly for sustained external loadings the allowable stress should be limited to 0.5Sa (1.5Sa allowed by ASME for local stresses less 1.0Sa).

Nigel, you have some interesting points there. There is no clear-cut procedure for calculating pressure thrust. Here are some aspects to consider:

There are practical issues to modeling the pressure effects in the piping system; for typical piping system the deflection due to internal pressure will be small. It may even be smaller than the geometrical tolerance of a restraint. Therefore, we do not know if that restraint is "active" or not, for the sake of internal pressure. So, that is why pressure thrust has to be (conservatively) estimated and accounted for. The points that Rich outlined in his post are good guidelines to estimate the pressure thrust. Engineers also need to rely on experience when dealing with it.

If one cannot compute the pressure thrust, then let's find a better method to model the thrust load. The typical approach is, applying the thrust load as a radial load on the nozzle, using the WRC 107 module. That approach can produce very high stresses especially the membrane + bending stresses. This can be seen when the nozzle fails with only internal pressure and the thrust loading.

One reason is that, WRC 107 only addresses external loads and not the pressure loading (internal pressure and thrust load). Moreover, if the curves used in a WRC 107 analysis are exceeded, then inaccuracies will be introduced. (In these circumstances, the program uses the last point on the curve.)

An article about pressure thrust is due to appear in our July 2001 newsletter (I will put the link as soon as the newsletter is available). It also contains a comparative study between FEA, WRC 107 and WRC 368 (which models the effect of pressure loading and thrust load) and recommendations on how to account for the thrust load.

Another point, ASME Section VIII area replacement method accounts for the general primary membrane stress in the vicinity of the nozzle, it does not account for the local primary membrane or secondary stresses due to pressure. You should analyze that manway using either WRC 368 or FEA.

I am not sure about your suggestion of modifying the allowable stresses to account for the thrust loading, can you point me to the reference for it.

Feel free share your comments/suggestions.

Regards,
Mandeep