Shape Optimization Odd Results

Altair Forum User

Hello,

I am using shape optimization to find the best shape for a structure. My structure is a simple 2d plate. It is fixed at both ends and there is a uniform load on top. I select the bottom nodes to be my design space.

My problem is that with this loading, I should expect the results to show simple arch that one would see in a typical bridge. However, the results I get start like an arch, but then the material bulges out in the middle again.

If one draws a moment diagram for a fixed-fixed beam (essentially what my problem is), they would see that optistruct is putting material where where there is a high moment (either positive or negative).

However, this does not give the optimium solution, which should be an arch.

I am minimizing mass while constraining deflection on my top surface.

Also, if anyone tells me how to attach an image, I can include an image to show you my 'bulge' in the center of my arch.

Does anyone know why I wont get arch for this problem? Also, if someone can point me to a verification or closed-form solution to this, it would be greatly appreciated.

Thanks,

Ed Mueller

Altair Forum User

You are correct, to better understand what you are trying to do, a picture would be most helpful. You need to upload a picture to a free picture site like postimage.org or similar. Then you past the url in the image code provided by the insert images button on the forum tool.

The best I can say from the infomation you provided, is that I am not sure how you are specifying a design space by selecting nodes unless you are using free-shape? To get the arch the you are looking for, it seems a topology optimization would be more straightforward. Again, a pic would help.

Topology optimization will show you the textbook Michell truss design of a bridge if you modeled it correctly.

Altair Forum User

I am unable to access any sort of file sharing site from my work, so I am unable to link to pictures.

I am using free shape, and I get the same results when I do topological optimization.

Thinking about my results, I think the bulge is actually correct. An arch is the best shape for a point load moving across the structure with 2 fixed ends, but with a distributed load on the top, there is a high moment in the middle of structure (draw a moment diagram), so I think it makes sense to have material there.