I think that more than a RISA question it is a "technical" question. So the cable will have some intent and more or less defined setup, and a someway defined envelope of loading hypotheses. So how you model such something will depend upon all these things. In some cases the design of cable structures involves iterative steps; this also can be used within RISA even for large deformations (geometrical nonlinearities) that of course are not the intent of a program like RISA, but you can approach it by such iterative procedures. Respect a single straight cable you have a tension-only element in RISA-3D; this will help you to not get some problems in analysis when using cables and a number of hypothesis. Yet for a special hypothesis that ensures tension you can model the cable, use P-Delta and have a nice approximation (for problems that not require the iterative process referred above) of the standing stress for the hypothesis. That you may want to compare it to a fraction of Fu or to Fy or whatever will depend of your technical intent. If you throw your cable-like element to RISA under a code, RISA will try to check it as per the code.