Phantom reaction forces?

[Anonymous]

Hi.

I've got this steel structure with wind braces and simply supported column bases all around. This configuration mostly gives the expected distribution of reaction forces: with even vertical reactions and horixpntal reactions concentrated in the bracing. However, after moving some of the braces, their former base supports still get significant horixontal forces. Does anyone have an explanation to this? All beam rotational DOFs are released from the columns, and all supports are pinned.

See attached picture for comparison of two similar supports, but with very different horiszontal reactions.

Thanks!

[Artur.Kosakowski]

Could you attach the missing pictures as well as the corresponding rtd files please? I assume that you did not ignore any instability warnings? In case you did address them first.

[Anonymous]

Certainly, image attached. The .rtd however, is 56 MB, so it's perhaps a bit on the heavy side? Is there a way to "purge" anything but the essential geometry and load information?

No instability warnings.

[Artur.Kosakowski]

Select the rtd without results file type in the Save as dialog and save it with a new name. Ten compress it.

[Anonymous]

Nifty! Much smaller model file attached.

[Artur.Kosakowski · Accepted solution]

Please open the file in the repair mode, save it under a new name and then calculate again. Could you describe me step by step the way you created and edited this model? I assume that you created it in Revit and then exported to Robot. What kind of modifications were done afterwards? In particular what was the way you moved the bracings (I assume that prior to this operation the results were as you expected them to be).

[Anonymous]

Great! Opening in repair mode seems to have done the trick, and all reactions are as expected. Thanks!

In the interest of finding the cause: The model was indeed created in Revit, and I added loads and supportsin Revit as well. After exporting to Robot, I created load combinations and modal/seismic cases before running the analysis. I don't remember the initial result any longer, but I am pretty sure all was well until I moved the braces.

The move was done partly by deleting and creating new ones in Revit, partly by using the move/copy function in Robot. It seems phantom reactions occurred in both cases.

If you have an explanation or tips on how to avoid this in the future, I'd like to hear about it. Anyways, thanks a lot!

[Artur.Kosakowski]

The only dangerous situation I can think of is when you make the changes in both files (rvt and rtd) at the same time rather than in sequences. What I mean is:

Model in Revit > export to Robot > modification in Robot > update (export to) in Revit > modification in Revit and in the same time modification in Robot > update from Revit to Robot

rather than

Model in Revit > export to Robot > modification in Robot > update (export to) in Revit > modification in Revit  > update from Revit to Robot >modification in Robot