Extrude In Absolute Direction

[travis.true08]

I cut a solid body into two pieces using a construction plane. I want to create an inner lip to all the top and bottom to kind of snap together by extruding the inner part of the baseplate's wall upward 1mm, and cut into the faceplate's inner part of its wall by that same amount. I created a sketch for each part on their cut surfaces, and added an outward offset from the inner boundary of the wall. When I extrude/cut, the operation is always performed in the direction of the arbitrary plane that split the objects instead of in the absolute up/down direction. Is there a way I can create the sketch on the absolute XZ plane, project the sketch onto the part's wall and perform the operation?

To illustration here's the full body with the slanted split going through it:

Here's what the faceplate looks like with the plane going through it:

Here's what the sketch should look like whether I sketch directly to the surface or not:

Here's the result:

You can see that things look good on the right side up until it gets halfway up the slope on the side where it forms an envelope. It maintains the envelope until it gets to the other side.

Potential solution:

I sketch onto the absolute XZ plane, and extrude the geometry downward to make the inner lip. To get the cut extrusion to match the curvature of the surface, I can create another construction plane with the same direction construction plane I used to split bodies except it's 1mm deep into the surface.

I'm new to CAD in general, so I'm wondering how this will that affect my workflow... Are there better ways to perform this?

[TrippyLighting · Accepted solution]

If you share you model I can create a screencast and show you how this can be accomplished. the key word here is sweeping a profile along a path. That path can be the edge of a body or a line in a sketch.

[travis.true08]

That'd be great. I didn't think to use a sweep. I'm still developing the mind set for sweeps, lofts (lofting?), and revolve. I don't think my project has any cases where their use would be obvious-enough for me yet. I'm still trying to get sketches down.

A screen cast would be great too. I've attached an archive of my project. The archive implements my solution, so you'd have to revert some of the timeline a little bit.

[TheCADWhisperer]

How are you going to manufacture this design?

I see several issues from a Design for Manufacturability/Assembly viewpoint.

[TrippyLighting]

Just now looking at this. I'd have the same question as @TheCADWhisperer. The question would be how would you ever pull this from an injection mold.

Not only does your design not have any drafts - that would probably too much to expect from a person new to CAD - but the bosses also are extruded in a direction that would lock the part into the mold:

the way you have created the lip is perfectly valid and creates the same result I'd have suggested in another way.

[travis.true08]

Hey guys!

To answer your question: I'm not sure yet. I've only got about 2 weeks of experience in CAD (all in Fusion 360), and no experience in CAM. I'd be surprised if there weren't plenty of problems with the design from that perspective. I also wouldn't doubt it if I'd have to redo my design from scratch once I've completed it, and accept that. CAM appears to be too broad of a topic, so I'm just focusing on CAD right now. From what I've gathered, making a CAM-compliant design is an iterative (and expensive) process to get a design that will yield a suitable mold. Right now, I'm happy with the results as long as I can 3D print (very) rough prototypes. The few tests I've printed have exceeded my expectations, but I'm easily amused haha.

I also have friends who work with Fusion 360 professionally, so I can pay one of them to develop a CAM-compliant design, and give them my designs as a reference if it comes down to it.

I'm glad my solution to the issue I ran into also works. Gives me confidence that I'm learning something. I played around with sweeps and lofts yesterday to get a better feel of it (haven't used the tools much yet), and it's given me some other ideas one where to use them.

[travis.true08]

@TrippyLighting When you say "bosses" are you referring to the binding shafts? I'm still new to most of the terminology, but I'm learning. The image of project you posted did get me thinking, and I began to wonder the same thing. The inside of the original SNES controller has slightly tapered shafts for the screw holes as well due to the draft applied on the faceplate. So, the faceplate's portion of the binding shaft is straight while the baseplate's is slightly slanted to account for the slant of the faceplate. I could fill in the under-hang area on the baseplate portion creating negative space in the mold for plastic to flow into. The downside is that more material's required, but now I can pull the part of the mold that fills up the interior straight out of the part without causing stress on the binding shafts. The SNES controller doesn't do this, so I'm wondering if they were made from another process.

I've also wondered how it will work with side-holes on single parts for injection molding. The wall thickness is only 2mm and ABS is somewhat flexible. Maybe I'm able to squeeze the side to remove them from whatever part of the mold still has it captured. That's how I've seen it done with phone cases anyway.

Also, when you said my design doesn't contain any drafts, were you referring to the draft feature in Fusion 360's modeling workspace, or is it a feature in the CAM workspace?

[TrippyLighting]

---

@travis.true08 wrote:

@TrippyLighting When you say "bosses" are you referring to the binding shafts? I'm still new to most of the terminology, but I'm learning. The image of project you posted did get me thinking, and I began to wonder the same thing. The inside of the original SNES controller has slightly tapered shafts for the screw holes as well due to the draft applied on the faceplate. So, the faceplate's portion of the binding shaft is straight while the baseplate's is slightly slanted to account for the slant of the faceplate. I could fill in the under-hang area on the baseplate portion creating negative space in the mold for plastic to flow into. The downside is that more material's required, but now I can pull the part of the mold that fills up the interior straight out of the part without causing stress on the binding shafts. The SNES controller doesn't do this, so I'm wondering if they were made from another process.

I've also wondered how it will work with side-holes on single parts for injection molding. The wall thickness is only 2mm and ABS is somewhat flexible. Maybe I'm able to squeeze the side to remove them from whatever part of the mold still has it captured. That's how I've seen it done with phone cases anyway.

Also, when you said my design doesn't contain any drafts, were you referring to the draft feature in Fusion 360's modeling workspace, or is it a feature in the CAM workspace?

---

Yes, I was referring to the Binding Posts/Shafts.

My native language is German, so my plastics manufacturing terminology on occasion might not be correct either 😉 

With drafts I am referring to what can be done with the draft tool. Its' use would be fine for this model but in more complex geometry one might have to resort to surfacing methods in the patch environment). The amount of draft depends on the manufacturing method and materials used and the dimensions. A shallow mold might not need the same amount of draft as a deep-draw part.

The methods you are describing for releasing not so well drafted phone cases might also apply here, but consider that a simple 2-part mold with proper draft is a lot less costly than mold tooling with a number of other moving parts, all sorts of ejector pins etc. For high volume production that might be OK, but for this controller it's not necessary and easy to avoid.

[travis.true08]

Sounds good! But yeah, I will get into those details when I've got more experience with using Fusion 360. I'm still getting used to a lot of it, and I haven't really done much with t-spline modeling. I've researched it here and there, but that's about it.