Seismosoft | Earthquake Engineering Solutions

Dear Sir,
I was wondering if in future versions it could be possible for the user to choose in which end of the Plastic Hinge Elements plasticity is expected to occur. Sometimes users may want to subdivide a beam or a column into more elements thus requiring a single plastic hinge on a specific end.
Thank you. Best regards,

Hi Gonçalo,
I believe you would want such feature to be able to subdivide the 'elastic' part of the DBPH element, isn't that so?
In that case, if the central part is assumed to behave linearly, you can play around with the plastic hinge length in order to get close to your desired level of discretization, without relevant loss of accuracy in the analysis, I suppose.
What do you think?
Cheers,
Ricardo

Hi Ricardo,
I'm thinking for instance in a beam which has different reiforcement distribution along its length and therefore subdivided into 3 elements (e.g.) with diferent reinforcement ratios. In that case I would want to model the central element as elastic and the others with the PH, each with a hinge at only one end.
Other examples are cantilevers, and beams that are free to rotate in one end such as beams supported by another beam,...
So, the answer to your first question is yes.
I dind't quite understand your point though. Do you mean that I can asign a zero length on the hinge? Even in that case, length is equal for both sides.
Thank you,
Best regards,

Hi Gonçalo.
I was not thinking about the situation of different reinforcement. For the case of different reinforcement patterns along the beam element I agree with you, it becomes impossible, unless you assume some sort of equivalent simplification criterion to use one element only. For the case of cantilever beams, though, I don't believe you would need to worry about assigning a plastic hinge to the free end, given that no significant forces would come up therein (actually, the same could be applied to the previous situation, based on the definition of plastic hinge behaviour itself).
In any case, I suppose the best option, for the time being, is to use a hybrid model, where the particular cases you pointed out (such as the 'beam-supported' beams, for instance) would be modelled by means of complete infrmDB/FB elements.
Let us wait, therefore, for the inclusion of your suggestion in upcoming SStruct versions.
Cheers,
Ricardo