No announcement yet.

Structural Analysis Interoperability

  • Filter
  • Time
  • Show
Clear All
new posts

    Structural Analysis Interoperability

    In the past I have trialled E-Tabs, and found that the interoperability wasn't great, whilst we could import an E-Tabs model into Revit, the member sizes did not come through correct, and vice verca.

    I have had some very good feedback regarding the use of Robot, although I understand that Robot does not do certain analysis that is required on some projects. Obviously being diverse with your choice of analysis software is critical.

    I have used CSC Fastrak successfully to create a very 'rough and ready' set of documents for value engineering.

    The model was constructed in Fastrak, and exported into Revit, all members came through into Revit correct, and it meant we were able to value engineer a 6 storey steel frame with a turn around of maybe 1 days work for the engineer and half a day for the Revit modeller.

    One problem I can see with using analysis packages as an interoperable solution is the requirements for each model.

    A structural engineer needs to produce a 'rough' model for structural analysis, maybe with an accuracy of 100mm... or more? The analysis model doesn't need to be as accurate as the Revit model.

    Another problem is the Revit modeller will need to have an understanding of how an analytical model needs to be constructed, and be aware that by doing things such as running a concrete beam past another concrete beam, then using an opening by face to tidy the beam join up, could have an impact on the analytical model, if this process isn't done correctly.

    I would propose that if the time constraints permit, the analytical model is produced first in design stages and an export to Revit is used for preliminary documents, until the Architects design is set.

    Once there is a requirement for Construction documents, the Revit model should be constructed in parallel with the analytical model, given the tolerance required for each model, I don't think it is practical to use an 'interoperable / bi-directional' model for analysis and documentation.

    I would be interested to hear other peoples thoughts on this subject.


    Check out my Revit blog -

    Hi Glenn

    In my experience it doesn't take much to strip and reduce a model back to it's bare bones for initial use of passing along an analtyical model. However keeping it updated becomes harder especially when it comes to connections, loads etc...

    I have great experiences with ETABS but as I have said we have customised routines that make life easier for the engineers to validate and use the model. I also have reasonably good success with RAM concept since Bentley releases ISM.

    SpaceGass for the essential bare frame works quite well and the passing back and updating of members works quite well as long as you keep a standard database setup.

    I also spent some time with the Inducta tools again both slabs and building have smooth tools to get the models into the packages, but there are issues like with buildings you still have to setup seperate dummy elements to simulate roof dead loads and things like that. Also sometimes the beams sitting on the outside of a slab did not play well at all.

    We trialled Robot and have a copy, while by far one of the most comprehensive pieces of engineering software around and great visualisation of results, the lack of codes and lack of stressing analysis really leaves this one in the dark for us. To train up a whole new team get everyone up and running and abandon all our other software just isn't realistic.

    The other issue is once a model is setup between Revit and "insert package here" There is still a lot of holes of what can and can't be transferred back and forward and still requires testing. I've used Navis to help solve issues and differences between analytical models and Revit models which can work quite well on larger projects. Same with AutoCAD. Not too mention any sort of curvature in a model it just plays hell and in most cases I use our dummy analytical elements and turn off the true item in order to bridge a connection and make things smoother.

    Microstran also works well as long as you setup the dxf you export correctly and provide a little bit of cleanup, it would be nice for direct import/export but this still saves us alot of time compared to traditional methods.


    Related Topics