Hello,
I am trying to do a generative design on this part but I haven't been able to be successful any help?
And secondly now that I can't do that I have been extruding straight up from the pipe to the gearbox mount. But when I put it in the overall assembly it does seem to be attached.
Any help Greatly Appreciated.
s
Hi @Anonymous,
If I understand correctly, you're trying to generate the material from the pipe to the cradle mount of the gearbox. If that's the intent, it appears that you've joined the two bodies together which may be preventing the material to be generated as expected.
You can use the Previewer command (that does a quick pre-check of the setup automatically) to quickly get an idea of the trend by visualizing the first set of outcome iterations.
Also, the shape is highly dependent upon how the load case(s) is defined. A quick setup could be:
1. Add a Fixed Constraint to a face of the pipe
2. Add a force load to a face on the cradle, in the direction between the bodies. The load can be be the self-weight of the gearbox.
3. At this point, you can use the Previewer to view the outcome shape that satisfies the design conditions for the load case
You can create additional load cases to define other scenarios such as the startup / shutdown torque, etc. The load cases are combined to create a single outcome shape that satisfies all of them.
If you're interested in experimenting with different shapes, you can create 'dummy solids as preserve geometry seeds' to force shapes that would otherwise lie outside of the critical load paths.
Optionally, you can create a body for use as a Starting Shape. There are awesome examples of its use in the Ready to Explore folder in the GD samples project in the Data Panel.
Hope this helps! Please let us know if you have further questions, comments, or suggestions. If possible, File->Export the F3d and attach it here to allow us to take a closer look.
The images were extremely helpful.
Best regards,
Hi @Anonymous
I'm not sure why, but there was just a blob created when I followed the general steps I suggested for the model.
Since I didn't have the gearbox to use in the model, I made a dummy solid for and made it an obstacle.
Then the pipe and cradle I ended up chopping up to create 'seeds'. I used the Connector Obstacle tool in Edit Model to create more obstacle geometry.
I sent it off to generate, to refine the outcome shape. The volume should continue reduction as subsequent iterations are created. I'll share the final outcome shape if it's anything interesting.
Best regards,
Thank you for the kudo @Anonymous!
I admit, it wasn't that straightforward for me but I do have something interesting to share With the loading in a single direction, there were a total of 42 iterations generated to satisfy the defined conditions. I chose the shape of iteration 30 since it seemed the most appealing, even though Min factor of Safety was above 20.
As you add more load cases, with different loading conditions, the generated shape will have different load paths to consider which could dramatically change the shape accordingly.
What added to my coming up to speed on this design is that we're taking a hybrid approach to GD, since the cradle body is already defined with bult-in gussets and fillets for added stiffness. This is why I had to hack away, because I felt like the existing geometry was being too restrictive for the solver.
If I were to try this again, I'd approach it similar to how the ALCOA bracket sample is set up. We include just enough geometry for the bolted connection points (the material around the countersunk holes) for the preserves. This way, it will allow more freedom for the solver.
However, we need to create and use obstacle bodies that are necessary to block material from being generated that would interfere with the space needed for fastener insertion, tool clearance, and any maintenance considerations ( e.g. I suspect that's a drain plug access hole), etc. It's the same as required for the conventional design approach, which we need to have defined for the solver to consider as well.
Hope this helps! As always, please follow up with any further questions and final result image, if possible 🙂
Hi Sean, I left the entire cradle as a preserve. It should be Study1 in the exported F3D.
I didn't realize you'd already replied before I edited, and added more outcome shapes to my previous reply. Out of curiosity, I added a torque load case to the chopped up cradle in Gen Model 2 😉
This is what I ve done so far.
Constraint the end of the cyclinder, moment on each hole, preserved the mount and cyclinder and made the gearbox and the bolts obstacles..
