Dear Claudio,
I would recommend to model that using shell elements (see http://www.cervenka.cz/products/atena/f … -elements1 ; maybe with small regions of volume elements near the supports or connections). Regarding reinforcements, usually most of the bars can be modelled by smeared reinforcement and a few (+ any prestressing cables) as embedded discrete reinforcement. The analysis running time is hard to estimate as it depends not only on model size (number and type of elements), but also many other factors, nevertheless, I think for such model you can expect analysis running times in the order of hours. However, sometimes the calculation can get very lengthy if you are interested in postpeak behavior or when there are heavily damaged regions.

I think possibly even a 2D model (resp. two, as it is not neccesarily clear which of the direction is worse) may be sufficient - of course it depends on the setup and what results are needed. Even if it is not enough, it would probably make sense as a good starting point before one starts with a complicated 3D model...

You can try our software online
http://server.cervenka.cz/vtls/
or download the demo version (limited number of elements)
http://www.cervenka.cz/downloads

I assume you have already seen our bridge poster
http://www.cervenka.cz/products/atena/f … of-bridges

Please contact us when you have further questions.