problems with dmass element

[lukanar]

Dear Seismosoft,
I'm developing a quite large rc building model using infrmFB elements and dmass elements to simulate distributed loads. I found some problems from the beginning with permament loads application obtaining such warning message in the Analysis log "Unable to apply the entire permanent load". I've already developed the same model with concentrate loads (lmass elements) and I had no problems.
Removing dmass elements the warning message does not appear.
I've also tried the comparison between dmass and lmass elements with a smaller size model with only gravity loads and static analysis: I've noted some differences in vertical reactions even if their sum is the same (thus I've applied the same total weight).
Have you ever found similar problems with dmass elements?
Other problems also arise when I have to subdivide infrmFB elements (i.e. a rc beams with varying longitudinal rebars scheme) and corresponding dmass elements.
If possible I would use dmass elements in order to take properly into account the effects of gravity loads on chord rotations and bending moments at element level.
Thank you very much for your kind attention and help.
Luca

[seismosoft]

Dear Luca,

Can you please send us (support@seismosoft.com) smaller size versions of your two equivalent dmass and lmass models, so that we may try to understand why the former is not managing to balance the same level of permanent loading as the latter?

Many thanks,

Seismosoft Support

[lukanar]

Dear Seismosoft support,
according to your request, I've just sent you the files of 3d frame models. Some of the problems I've encontered with dmass elements seem to be solved using the last version of your software (v.5.0.4 build 5).
Thank you very much for your kind help.
Best regards,
Luca

[seismosoft]

Dear Luca,

Thank you for your message. We are happy to hear that you are no longer experiencing the numerical problems you had previously described.

With regards to the differences between the use of 'dmass' and 'lmass' elements, these are expected since the transformation of distributed loads in point loads requires the introduction of not only end transverse forces but also of end moments (check a textbook on structural mechanics/analysis for further details on this). Hence, the use of 'lmass' as opposed to 'dmass' is bound to give different results, with the differences obviously diminishing as you increase the number of 'lmass' elements.

The differences between the use of non-subdivided and subdivided 'dmass' elements are also expected (the more discretised the mesh is the more precise are the results, see Help System page for this element), but they are also very small (considering the relative small displacements/rotations in question), hence feel free to use the simpler mesh, which leads to smaller mass matrices.

Thanks again for the detailed email.

Seismosoft Support