Shell not working as expected...

[sykosys]

This may be a bug, but more likely user error.

I'm trying to build a relatively complex surface. I need to shell it, because it is a plastic part that will be eventually turned into a mold, but I can't get the pwr_shell to work as I need it.

To start, I've simplified it (example in attached max file). I have two surfaces that are offset from one another, that are created using rail sweeps. I create a pwr_sketch to act as trim lines, and trim them so there's a gap in between the way I want. I then create another sketch to define the surface in between, and create it with another sweep using the edges of the surface cuts. I have to increase the rail synch to make the surface conform in some cases. (First model on the left at this point.)

Now if I shell one piece alone (next to right), it works fine. Then (next, to the right), I join them. The NURBS Query tells me that I have the number of open edges as I'd expect. If I shell the join, I get a few different options; 1) I get a box, which I assume is the "I've failed miserably" default, or 2) I get all kinds of open edges, notably at the lower corners of the join sweep.

If I play with settings within the stack of the part, different things happen. Usually it goes to a box. Sometimes, the spanning surface will change; the sketch points drop, move, it connects itself to other edges of a sufrace, and so on.

This is driving me nuts, and I have a deadline...

Thanks for the help...

Kris.

[peterhsieh]

It looks like there may be an issue with our sewing. I the third object on the right (the Pwr_Join of the two trims and the rail sweep), and instead of choosing "Sew Edges Only (Fast)" I selected "Do Nothing". I was able to then apply the Pwr_Shell and it worked without the fail box.

I do not know if this produces a desirable result for you.

Another approach would be to shell the surfaces separately and then boolean union them afterwards.

I don't know if this has been helpful, but hopefully it was.