Pushing the sheet metal envelope....

kent_henderson

https://cad.onshape.com/documents/3e61b87ec35d8ef44954444d/w/929672606d096e2d3d4d1b76/e/d8c1a10906f2f13c918eecaf

Can anyone advise a way to convert this to a sheet metal part so I can unfold it. It's a custom fire vase for a public space. We will be fabricating from 14ga stainless steel. It will have an open top and a closed bottome (with holes). I know I'm pushing the envelope here but feel compelled to ask.

NeilCooke

Hi @kent_henderson - sheet metal can only unfold flat and cylindrical surfaces, so your twisted loft shape cannot be converted. How is this intended to be manufactured? That might give you an idea of how best to model it.

kent_henderson

It will be laser cut from 14ga stainless sheet in 4 sections and a bottom, then the 4 sides will be rolled and welded in each corner.

NeilCooke

Getting that sort of flat shape is difficult and subjective. Each manufacturzing process will differ. Try ExactFlat from the AppStore- that will get you the closest. 

kent_henderson

I would love to be able to do things like this. We're always pushing sheet metal unfold to the limits, and then some. Inventor will not do this either.

kent_henderson

Thanks Neil. I'll give ExactShape a try. 

brucebartlett

I really think there is room for un-uniform unfolding tools. However, it may not be easy to get consistent unfolds. Similar problems here.
https://forum.onshape.com/discussion/9759/ideas-for-curved-sheet-metal-skins#latest

I'd love to hear how you get on with ExactFlat. My experience wasn't great when I tried but that was a few years ago to it may have improved since then. 

This IR sort of relates. https://forum.onshape.com/discussion/9764/flatten-unwrap-surface#latest

NikNYC

ExactFlat has an online browser version outside of Onshape, to test things on. For a brute force 90 degree twisted metal strip, built by blending the edges in 3D with curves then using those to make a surface twist, Rhino will "unroll" it with a few percent area error being reported but it wonderfully results in a thinner twist area that in real life is much easier to achieve than without the indents that reduce the width from 3" to 2".

Interestingly, in real life for 1/8" stainless steel, 3" wide, the Rhino "pure geometry" result is a whopping 0.7" shorter due to the twist contracting the metal length, but in real life due to actual metal stretching, it becomes only 0.1" shorter.

However, ExactFlat does something quite bizarre. It concentrates all the stretching in the center point of the twist so it magically restores the 2" width of the twisted part to exactly 3" to recreate a simple rectangle flat pattern. That's not going to happen in real life.

Ah, but when I use their multi-minute long Optimization for steel it fixes this back to the Rhino result: