nodal mass vs. distributed mass

[mgs]

Dear Frank,

I am modelling a bridge in OpenSees using the linear ShellMITC4 element for bridge deck. I performed eigenvalue analysis of the bridge modeled by elastic beam and elastic shell element. As a result the modal period values and mode shapes would change significantly.

But comparing to the elastic beam, the bridge deck with the elastic shell has more nodes (10 times more) and it wont be computationally cost effective for time-history analysis.

I am wondering to know if there would be any difference, based on your experience, in the seismic responses with treating the mass as distributed with when it is assigned to the shell nodes.

Thanks,
Mimo

[fmk]

i have no experience here as i get very few questions on use of the shell elements. if the meshes are fine enough there should of course be no difference, jst not sure how fine you would need to go. you could always do a small parameter study of some square or rectangular mesh using a simple script that could study the effect AND POST the results or script when done. thanks