Grasshopper-ANSYS Interface

Comment

Comment by Riccardo La Magna on June 12, 2013 at 5:16am

The topology and the underlying polygons are the same. What is missing in the model of the video is the plane cutting the triangulated cells (it made the model easier to deal with). That's why it differs from the built prototype. The number of modules is also the same. What changes is just the global geometry.

Comment by djordje on June 11, 2013 at 5:47am

Thank you. So the only thing that was changed in comparison with the final shape were modules?
On your upper video modules are triangular, while the ones on this link are pentagons and hexagons?

Comment by Riccardo La Magna on June 10, 2013 at 5:05pm

Thanks djordje! The one in the video was a study for the final shape of the Pavilion. And the final shape turns out to be the built Pavilion itself ;)

Cheers,

r

Comment by djordje on June 10, 2013 at 10:48am

Hi Riccardo,
This is a fantastic project. I really admire you.
Is this the final shape of the project:
http://www.oliverdavid.de/?page_id=36
Or it's a completely other one?

Grazie.

Comment by Przemyslaw Kolodziej on October 29, 2012 at 12:59pm

Thanks a lot!

Comment by Riccardo La Magna on October 29, 2012 at 6:41am

Huh, I don't seem able to upload the file on the forum, weird.

Anyway Przemek, here you can download the Python code to automatically update the geometry and run the analysis on it.

In a Grasshopper script component you need to add the ANSYS Process object through which you fire the program and run the above Python script. It would look somewhat like this (in C#):

System.Diagnostics.Process ANSYS = new System.Diagnostics.Process();
ANSYS.StartInfo.FileName = "YourPathToANSYS\\runwb2.exe";
ANSYS.StartInfo.Arguments = "-B -R YourPathToPythonScript\\AutomaticUpdate.py";
ANSYS.Start();

..and that's basically it.

Comment by Riccardo La Magna on October 26, 2012 at 7:10am

Hi Przemek

Sure I can share the script. I just need to find it somewhere on my hard drive, as it's been quite a while since I last used it. By the way the GH script just fires ANSYS in batch mode, which in turn executes a Python script (which is the Workbench's scripting language) that starts the analysis. No more, no less. The rough part lies rather on the ANSYS side, as it's not so trivial to find the right setup to make it do the right calculations.

Anyway, as I said I'll send you the script. As soon as I find it..

Cheers,

r

Comment by Przemyslaw Kolodziej on October 18, 2012 at 6:25am

Hi Riccardo, 

Well done! Is it possible to share yours GH ANSYS plugin? It'll help me a lot on my phd research. 

Best, 

Przemek

Comment by Riccardo La Magna on September 6, 2011 at 10:42am

I'm sure it can be done.
I would suggest starting to build your model in ANSYS (I'm giving for granted you're using ANSYS Workbench. I hope that's the case) and setting up the whole simulation process.
Once you've done this you need to write the APDL macro to export all the pressure values you're interested into, either to an Excel table, a *.*txt file or whatever format you feel more comfortable with (in the video I'm exporting to a text file). Most of my time I've spent it finding the right ANSYS command lines to export the displacement values, I bet this is the tricky part.
At this point you can setup your GH + Galapagos model, linking the Fitness Input to the exported values. Remember that you'll need some post-processing of the global values, as Galapagos accepts only single numbers as optimization parameter.
To finish off you link everything with the script component which fires ANSYS and invokes the Python script at each simulation. Moreover you need to pause Galapagos during each simulation, or else it will continue to output solutions while you're still waiting for the old simulation in ANSYS to finish. I did this by (rudimentarily) imposing a Thread.Sleep() until the export data isn't updated. I hope David may have more elegant solutions for this. David, anything?

Well, I think that's about it. If you need any further advice let me know.

Cheers,
r

Comment by Amaury THOMAS on September 6, 2011 at 8:45am

Yes i understand your process, it's very clear and logical. thank you very much for sharing. I would like to make uniform the wind pressure on the skin of a building. Do you think it possible with ANSYS,GH,Galapagos ? I would be very glad to get some advise to achive it.

• ‹ Previous
• 1
• 2
• 3
• Next ›
• Page