Sam,

So far, I haven't work with PIPSYS, but I assume that similarly to Caesar II, it is a "linear-elastic" piping stress software.

Therefore, for multianchor piping systems or two-anchor systems with intermediate restraints, where additional nonlinear effects may accur (not necessarily local yielding, but friction, gaps/clearances in guides or axial-stops), the relaxation effect cannot be quantified.

Therefore, in my opinion, it is meaningless to talk about maximum reaction forces&moments for the original/installation conditions on the linear-elastic analysis basis. In fact, under linear-elastic analysis circumstances, if no initial anchor displacements are imposed at installation moment, and considering weight, pressure and thermal expansion as cyclic loads, the final "cold" reaction forces&moments developed on restraints after a full loading-unloading cycle should be ZERO.

That's why the common practice is to communicate to the structural dept. the OPERATING REACTION/LOADS developed by the simultaneous action of the external loads (weight, pressure, thermal expansion, possible imposed displacements etc.). Meanwhile, the "Cold" situation corresponding to the weight action exclusively is taken into account.

In fact, considering the self-springing coefficient C1 = 1, and estimating Ra = C1*R, you obtained the same OPERATING LOADS too.

However, since the dead-weight action is not a cyclic load, we may evaluate the cyclic component of the reaction loads (on linear-elastic basis), considering the following loading cases (as are defined for Caesar II programme):

L1 WNC (...+H...) (SUS)
L2 W + P1 + T1 (...+ D1 ....) (OPE)
L3 L2-L1 (OPE)

The reaction forces&moments identified for L3 case represent the maximum cyclic components of the support/anchor reactions and may be useful for structural fatigue evaluations (if required).

Now, regarding the reaction sign, it's true that in many situations, the reaction forces and moments are sumbitted to structural dept. in their absolute values.

Personally, I use to attach the sign to the support reaction values forwarded to the structural dept. The reason is that I do not have practice in structural calculations and I believe that more accurate evaluations may require the loads sign if necessarily. When the reaction (force or moment) corresponds to an actual movement restraint (vertical rest restriction, axial-stop or lateral-guide), I attach the sign as it has been established by the programme (meaning the action of the pipe upon the structure). When the reaction corresponds to the friction force (such as the typical case of the horizontal friction forces induced in the simple rest points), I attach the both "+" and "-" signs, so that structural calculations are to identify the most unfavourable load combination.

We may identify the origin of the reaction watching the displacements of each supporting point.

Regarding friction, as many other Forum members pointed out before, due to the stochastic character of this phenomenon, I use to perform both the friction and non-friction static analysis for the same system and to retain the most unfavourable results.


I hope these details clarified my personal point of view related to this subject.


All the best,