topic Re: Best practice of relative positioning of components in parametric design in Fusion Design, Validate & Document Forum

Best practice of relative positioning of components in parametric design

denis.itskovich — Sun, 08 Dec 2019 18:50:23 GMT

What is the proper way of positioning one object relatively to others in parametric design?

Let's take an example when I need to model a mount which should connect between 2 components, whose dimensions and positions depend on parameters. Usually I position 2 components first, and then create sketch for the new component, using projections of existing components as a reference.

The simplest way of positioning - is to use align + move + capture position (align to required faces and then use move in order to shift to make some space for the mount itself). This is really easy, but the problem is that in this case the history stores just a captured absolute position, which will not be correct when parameter values are changed.

Another option would be to use joint tool, this solves the parameters problem, but this seems unnatural: let's say I have 2 component A and B and I'd like to model component C which will be mount between A and B. I need to position B relatively to A before C is designed. So if I use joint between A and B directly I will not be able to make a joint later between A <-> C and C <-> B which is the final goal.

What is the proper way of doing such things in fully-parametric design?

Re: Best practice of relative positioning of components in parametric design

davebYYPCU — Sun, 08 Dec 2019 19:36:35 GMT

Based on your theoretical, A being jointed to B, should work with parameter changes.

C is positioned with projected references from A and B.

Why does that not work?

Re: Best practice of relative positioning of components in parametric design

denis.itskovich — Sun, 08 Dec 2019 19:41:33 GMT

The problem with this approach that if make a joint between A and B directly, then later, when C is modeled I will not be able to make proper a joint between A<->C<->B, because existing joint between A<->B will already constraint the movement. This is a problem when A<->C or C<->B should be joints with some degree of freedom, but I already made a joint A<->B for design purpose only, which is rigid

Re: Best practice of relative positioning of components in parametric design

davebYYPCU — Sun, 08 Dec 2019 19:45:30 GMT

Will need to see the example,

or are you asking for something that can’t be done?

Re: Best practice of relative positioning of components in parametric design

denis.itskovich — Sun, 08 Dec 2019 19:54:46 GMT

I'm asking a general question about best practice for such cases. I believe that modeling some kind of connector/adapter/mount between 2 imported components is a very common use case, so I'd like to know what is the recommended approach for modeling such mount components in parametric design? When I just started working with Fusion 360 I was usually using alignment tool - but as expected it caused a lot of problems each time when parameter values that affected positioning had changed. So now I'm considering indeed to go for joints between components that should not be connected directly, but this feels to me like a walk around, because this causes me to create joints that should not exist in reality (in reality A and B are not connected directly, but through component C). May be I should create a joint A<->B for design time, temporary, until C is modeled and then to remove this joint and replace it with 2 new joints instead, but this also feels me strange. So concluding all above, I'm wondering what is the recommended approach in this situation

Re: Best practice of relative positioning of components in parametric design

davebYYPCU — Sun, 08 Dec 2019 20:07:36 GMT

You say A is joined to B, these two are now a single group call it AB.

AB has one component grounded, so now both can’t move now by accident.

what parameters in this group will be changing - considering they are imported, are they Linked Components?

Component C can be joined to either, and or both, now that is Design Dependant, and

I don’t think any of us can give a hard process / rule to help - without some example (doesn’t have to be your actual design).

Without some detail, you get best guess that probably does not fit the detail.

Re: Best practice of relative positioning of components in parametric design

g-andresen — Sun, 08 Dec 2019 21:12:46 GMT

Hi,

Pictures say more than words even there are many of them.

günther

Re: Best practice of relative positioning of components in parametric design

chrisplyler — Mon, 09 Dec 2019 19:49:34 GMT

@denis.itskovich wrote:
The problem with this approach that if make a joint between A and B directly, then later, when C is modeled I will not be able to make proper a joint between A<->C<->B, because existing joint between A<->B will already constraint the movement. This is a problem when A<->C or C<->B should be joints with some degree of freedom, but I already made a joint A<->B for design purpose only, which is rigid

Joint A<->C and/or C<->B as desired, then Delete (or at least Suppress) the original Joint A<->B.

Re: Best practice of relative positioning of components in parametric design

laughingcreek — Mon, 09 Dec 2019 20:11:34 GMT

As already indicated, it's hard to give good answers based on just your description of work flow. I can imagine what you've described branching off in several directions that would need to be handled differently.

BUT

you seem to already know what the retaliative position of components A and B should be. So create component C and make a sketch indicating those positions. then joint A to it's location on the sketch in component C, and do the same with B.

create a skeleton sketch at the top level and joint all the components to it.

either approach COULD negate the need to later delete a joint and establish new ones.

Re: Best practice of relative positioning of components in parametric design

denis.itskovich — Mon, 09 Dec 2019 20:23:48 GMT

I think deleting a joint is not an option - because in this case the joint will be deleted in the timeline as well, making all the dependent design to be broken. Suppressing I didn't try, but I'm afraid the result will be the same, since it will be suppressed in the timeline at the moment when the joint was created and not when I'm going to replace it with 2 another joints

Re: Best practice of relative positioning of components in parametric design

denis.itskovich — Mon, 09 Dec 2019 20:28:02 GMT

I agree, but I don't have any specific design to show. In the past I did a lot of mistakes in previous designs, when used aligning / capture position along with parametric design. Later I realized that it was a mistake. So before starting a new design I was wondering what would be the best practice in this case. I think @laughingcreek's answer provides a rather convenient way to go

Re: Best practice of relative positioning of components in parametric design

chrisplyler — Wed, 11 Dec 2019 14:27:01 GMT

@denis.itskovich wrote:
I think deleting a joint is not an option - because in this case the joint will be deleted in the timeline as well, making all the dependent design to be broken. Suppressing I didn't try, but I'm afraid the result will be the same, since it will be suppressed in the timeline at the moment when the joint was created and not when I'm going to replace it with 2 another joints

If you make the two new Joints with the regular Joint tool (not the As-Built Joint tool) the relative positioning will be maintained regardless, because the regular Joint tool brings the two location coins together no matter their prior positions.

But yes, if you created Component C with design dependencies to the locations of A and B, then that's might cause problems. I'm wondering why you didn't model A and B "in-place" to begin with.