Hi Forum. I have the attached model of starboard spacer ( high-density polyethylene ) that historically I've crudely milled on my own that I'd like to have milled for me. Towards that end I modeled in Moi really unaware of what shops that do this kind of thing are looking for as file. I'm aware that with 3D printing there are programs that care of filling voids, add structure, etc. like UltiMaker Cura, but I'm at a complete loss for what a milling/plastic shop that does this work would like from me ? Any helpful resources on this ?

As for the specific model, the spacer has a groove for a couple wires that will be machined out of the bottom of it. To perform the boolean diff of wire groove on spacer I'd like to close off the triangle sections highlighted in magenta that are open ( see grouping "how to close off triangle section ?" ) before performing the boolean diff. I remember that planar on sections like this are problematic. But how else to realize and then create a solid of this grouping which can be diffed from the spacer ?

Sven

p.s. I appreciate when people create posts with animated gifs showing what they are looking for or before and after. Is there a post that explains how to do this ?

Hi Sven, so typically the main thing that shop will likely be looking for is to have the model be a closed solid object.

Usually to fill in the triangular area you show you would use Construct > Planar. But this case here is having some problem with that so instead you can use Draw solid > Plane to put in a rectangular plane and then use the Edit > Trim command to cut the plane.

- Michael

Hi Sven, is the attached version what you're looking for?

- Michael

Hi Sven, so for how to do it - booleans can have difficulty when there is a lot of overlapping coincident surface areas like you have here.

So I didn't use booleans in this case, I used Edit > Separate to break things into individual surfaces, deleted interior pieces and used Extrude and Trim to cut several pieces along their own seam curves, like this:



- Michael

Hi Sven,

The most important thing to give a shop is a good 2D drawing with dimensions, tolerances, material, and surface finish all spelled out. A machinist can't tell from a 3D model what your intent actually is. It's not possible for them to machine the ideal "platonic" version of your model, so they need to know how much they're allowed to deviate on each feature.

That said, some shops will accept just a 3D model, an overall tolerance, and an overall surface roughness. But if there's no drawing and you don't like what they deliver it will be difficult to get them to fix it.

Re:
> high-density polyethylene

As you might have found, that's a difficult material to mill. In some ways worse than the hardest steels because it requires razor sharp tools and gentle clamping since it's both soft and tough.

I would try to find a very small (1 man) shop that's close enough to you that you can go there and talk it over with them in person. That way they can give you feedback on the design and you can fully describe what the important areas of the part are.

Finding a good machine shop can be pretty difficult.

Re:
> I appreciate when people create posts with animated gifs
> showing what they are looking for or before and after.

LICEcap is a good program for Mac. Not sure about what's out there for Windows.

- Peer

Yes, exactly Michael. Thank you for explanation and video which makes sense.

Thank you Peer. Good sounding practical advice. I'll try LICEcap.

Sven

As a follow-up to this technique I've attached a simplification because although I could create planar on one side after following your technique Michael as I was unable to reproduce on the other side. A mirror does work, but planar or the following technique does not. I extruded a line vertically from a magenta curve on one side and a cyan curve on the other. The magenta curve can be boolean diffed or trimmed with the "curiosity" object without issue while cyan curve cannot do either. I'm unable to see any difference between the two sides and reproduced in a symmetrical manner so I'm stumped ?
Furthermore when viewing from left side I should see cyan curve and from right side magenta curve, but I see magenta from both sides.

Sven

Hi Sven,

re:
> The magenta curve can be boolean diffed or trimmed with the "curiosity" object
> without issue while cyan curve cannot do either.

I'm sorry I'm not understanding this part.

Is this the curiosity object here? :



And the magenta curve is this one?



I'm not understanding how you would boolean those together, if a plane is extruded from the magenta curve upwards, it just barely grazes the end of the curiosity object and does not divide it into pieces:



- Michael

Hi Michael, I assumed too much...



so in this image I can successfully boolean diff the blue object from the magenta base object ( trim works equally well ) resulting in...



but when I try the same thing with cyan curve ( either boolean or trim ) it doesn't work. I've tried recreating ensuring symmetry so I'm not sure what I'm doing wrong...



lastly, I notice that viewing from the right results in same image as viewing from the left, but should I see cyan and not magenta line from the right ?

right view


left view

I hope the next model help you

Hi Sven, ok thanks I think I understand better now.

Surfaces that barely graze each other and are also tangent can tend to be more difficult in surface/surface intersection calculations.

It will be easier for Trim to work on your cyan plane if you select the edges as the cutting objects rather than the full object, like this:



Then it should be cut up ok:



Also another way you can do it is to draw in 2 connecting lines like this:


Then select the lines and edges like this:



And then Construct > Planar can make a trimmed plane there:



- Michael

Thanks for explanation Michael. I realize that there are other solutions, just figured I had to be doing something wrong given the symmetry of model and it working one side, but not the other.

Thanks for the models vic.

Sven