Interference Behaviour

[HarrisonClassic]

Hi all,

Fusion 360 is current build (2.0.11186 ), no errors in the model according to Compute All.

I'm seeing some interesting results and behaviors when using the interference feature. All of the bodies are nested under around 8 components

I don't think I actually have a problem. If I select differing sets of bodies I don't get any errors or interferences, I've tried lots of differing "overlapping groups" of bodies and all is well, just seems to be when I include all of them. I think it just could be Fusion spitting it.

When I do an interference selecting all the bodies in picture, without selecting "Include Coincident Faces" I get an errors about a couple of coincident faces and it doesn't show me the result - ie. if there are any interferences.

Is there a way to locate where the issue is, I can't seem to find a way to do that?.

If I check "Include Coincident Faces", it shows me a lengthy table of interferences all of which are zero volume.

Upon inspection, most of these seem to be around reasonably complex splits, or part geometry as a result of combine operations such as below - body splits with sketch line.

Thoughts anyone?

Thanks

David

[TheCADWhisperer]

---

@HarrisonClassic wrote:

Thoughts anyone?

---

In the real world we can't manufacture perfect parts.

Allowances for Manufacturing Tolerances and Design for Assembly should be included at the the design stage.

Are you modeling perfect assembly joints or are you designing in appropriate clearances between mating parts?

Is tolerance "stack-up" an issue in your actual assembly process?

Are the individual components hand-fitted with filing/grinding/chiseling/scraping during assembly?

Those are my initial thoughts.

[jeff_strater]

Hi @HarrisonClassic - nice to see your boats progressing.

Regarding your question here:  "Is there a way to locate where the issue is, I can't seem to find a way to do that?".  If you'd be willing to share your design, we can take a look.  I'm not familiar with that error, but we could maybe find out what it is trying to tell us.

[HarrisonClassic]

Hi,

As for the perfect part issues - I agree.

In design, right at the end of the process I use dog bones for instance to enable intersection fits. In some cases I build in offset faces after extrudes/combines/splits.

When I move things to CAM, I generally add negative offsets of around 0.25 mm to CNC'ing the parts. In my case pretty much all of the machined faces are actually faces that hold epoxy glue so there needs to be room for that and ease of assembly. I wont suffer tolerance stack up simply because most of the joined parts don't change the dimensions of the project in any real measurable way.

David

[HarrisonClassic]

Thanks Jeff, @jeff_strater 

In the main Fusion 360 has been behaving really well since the last few releases, so very pleased.

Have emailed you at your Autodesk address with a link to the file.

thanks

David

[HarrisonClassic]

@jeff_strater 

Jeff,

I worked back to 3 individual bodies that when included in an interference cause the errors.

TR:FR-04 Complete

TR:Body 179

TR:Body 205

I managed to remove the error by doing an offset face which is ok for my purposes.

Hope this helps.

David

[jeff_strater]

Thanks, @HarrisonClassic for taking the time to isolate that problem.  That had to have been a ton of work, which I appreciate.  Glad you found a workaround.  I was able to verify the problem, and have sent it off to our modeling kernel team to investigate.  Will update the thread here if I learn anything of interest.  At the root of an Interference check, we use the same API as is used for Combine.  I know that this API has problems with coincident faces (which is why that "include coincident faces" option is there), so I suspect that is the root of the problem here, but it is good to have the input to let them track it down.

[jeff_strater]

@HarrisonClassic - heard back from the kernel guys.  They think the error might be related to these small edges in the body:

it sure seems possible that this is the case, since those happen to coincide with the two pieces in the "line 3" folder of the "TR" component, that you narrowed down to the ones that caused the problem:

I traced those lines back to the Combine86 feature in FR in the timeline:

If I roll back before that, and do a measure between the bottom edge of those pieces and the bulkhead, I see that there is zero distance (i.e. the edge is exactly coincident with the face of the bulkhead, and the Combine leaves this standalone edge imprinted:

That's as far back as I traced it, though.  I did not look at how those cross-pieces were built, but I suspect that some adjustment could be made to those to prevent this.  Since you are able to move forward, though, I'll leave it there for now.  I'm not sure what to tell you about how to prevent this in the future, (this points to the need for some good body checking tools in Fusion), but I thought I'd pass along what I found out.