Hi Bob, by 3D printing standards that's unfortuantely an extremely difficult part that you're trying to send to Shapeways.

One problem is that the 3D printer ideally wants to have just one single solid in the file to print, but instead you've got 113 different objects all in the same file there, with some of them not being closed solids and even the parts that are solids are partially overlapping on top of each other. It's rather doubtful that the 3D printing prep software is going to be able to make sense of such a complex multi-object model like you've got. The printing software is going to try to calculate the outlines of planar slices through the object and it wants to form a uniform closed boundary line for each slice. When you've got non solids and parts running into each other it makes it very difficult for it to arrive at a good slice through it.

So that's one big problem - your entire structure there is very complex and difficult to process from the standpoint of what 3D printing software expects to receive, you've got to try to get things into just one solid rather than so many separate and overlapping pieces.

The other problem is the tiny size of some of your details. The shapeways materials chart here shows the minimum detail size for different materials:
http://www.shapeways.com/materials/material-options

You can see that the minimum detail size there ranges from something like 0.1mm to 2mm depending on the particular material. But in your model you have many features that are quite a lot smaller than that for example this little area here has a thickness of only 0.03 mm:






So those are some of the problematic things, I think you'd need to probably significantly simplify and unify your model here before it's going to be very feasible for it to get printed...

- Michael

That's what I've started doing. The potential buyers for the part are so far okay with a simplified part without railings and whistles and such, so the entire part is being detailed down, and the funnel base and the funnel itself will likely be two different parts for the modeler to simply glue together. The big thing is the gussets. Any idea on how to accomplish that?

Bob

Hi Bob, re: gussets - I'm not familiar with what a gusset is so it's difficult for me to help you on that.

Is it the 3 sided area that you're trying to fill in? I'm unsure if that's what you are talking about because it looks like you have that in your model you posted. Are you trying to make it different than what you currently have?

- Michael

Its that roughly triangular area on each side of the forward funnel, left and right of the piping. I'm trying to get it to look like the picture. Sweeping the various edges hasn't worked - some REALLY funky results. Lofting works best, but then it seems to sag, or lack of a better word. The result isn't smooth or organic, with bots and pieces all raggedy.

Hi Bob, I'd think you would want to do a sweep there, but currently the edges don't really match up directly with one another, like for example you'd want this edge to span all the way across:



Currently that edge does not meet up with the other ones, there is a significant gap between their ends:



If you could post a 3DM file with just the curves you are sweeping and the all raggedy result from it, it would be easier to give you some advice on what's going wrong with that - it's difficult for me to pick out these things from your full model file that has everything in it.

Are you maybe selecting multiple edges all across the same station for the profiles? That will make for bad results because Sweep thinks that each profile curve that you give it should be one station along the rails - if you have multiple segments selected it will try to use each one of those as a station independently. You've got to join things like that up into a single curve so that you have just one profile curve per station along the rails.

Here's an example (also see attached 3DM file) where the curves are ready to sweep - select the profile, then run Construct > Sweep and then pick the other 2 curves as the rails:



The more narrowed down you can make your example the easier it is to give you advice on it, so for example if you have a question about why a particular sweep is weird, if you extract just the geometry that's directly involved in producing the sweep and the weird result and then post only that, I'd be able to figure out what you are running into a lot easier.

- Michael

Thanks, that helped a lot. The forward gussets now look fine and the after ones are a little bit problematic, but they're coming along. Seriously, I hadn't noticed the curves not matching up. I suspect I'm still thinking parametrically in terms of solids than in terms of MOI curves. In other words, operator error!

Bob