Reply

Design intent in assemblies

I ran into a problem with changing assemblies via variable tables.

I have an assembly with the first part containing overall size control variables and a second part whose size is controlled by an inter-part copy of the first part.

I use peer variables in the assembly to adjust the first part size.

Everything works fine unless I have left some design intent settings off when editing another part. Even a totally unrelated part because design intent settings seem to persist within a SE session.

I think that is bad design. The DI settings should stay with the part, not the SE session.

So when changing a variable if, say, I don't remember leaving symmetry or some other needed design rule off, even on some part from 2 hours ago, it can mess up the part and assembly. And I don't have a clue unless I happen to see it because the design intent box does not appear while changing the variable table.

It would be useful if design intent appeared while changing variables.

This is a big problem with synchronous to me in general. Especially with more complex parts where it's very difficult to notice changes or the changes are so small you don't notice until a lot of little changes have built up to an extent that your parts have become skewed enough by the last several changes that you can't help noticing. By that time the mess can be so deep it's hard to fix.

 

Bruce Shand
ST9 MP6 - Insight - Win10 - K4200
11 REPLIES

Re: Design intent in assemblies

Hmm, crickets. Dang!

Bruce Shand
ST9 MP6 - Insight - Win10 - K4200

Re: Design intent in assemblies

@bshand

This is one of the (many) reasons why we don't use sync - unpredictable results.

HP Z420 16GB RAM
Quadro K4000
ST7 MP10 on Windows 7 Pro 64.

Re: Design intent in assemblies

[ Edited ]

@beachcomber, Yeah, I get it. But if just one or both of the changes I mentioned were implemented I think it would make it a little better, a little more predictable.

1) Stick DI conditions to files, not globally in the current SE session.

2) Have the DI menu callable in assemblies by user prompting or by the mere fact that you are changing a part via variables. Right now the only way I know to see what the DI conditions are is by clicking on a part feature in an open part.

Bruce Shand
ST9 MP6 - Insight - Win10 - K4200

Re: Design intent in assemblies

I wish I had an answer off the top of my head for you, but I don't. I'm getting people more knowledgeable than me involved, so we can find a real answer.

 

As far as your statement about " changes are so small you don't notice", you can really control stuff by just turning off most or all of the Design Intent settings so that nothing is going to change unless it is explicitly selected by you.

 

I think if you also get conversant with the Solution Manager, you can quickly see what's going on with a particular change.

 

When you say " The DI settings should stay with the part, not the SE session", I get a little confused. I think the Design Intent is more restricted than that. It should only have an effect during a change.

 

See, here's the problem. Sync is meant to make changes more controllable. That's the "direct" in direct edit. But you're mixing that with associative (interpart) and automated (peer variables) tools.

 

I would not be surprised if you have a circular relation going on somewhere here. Synch is meant to avoid this kind of thing, but it's the non-sync tools that are doing this to you.

 

You have part 2 driven by part 1 by the inter-part copy. And then is it possible that part 2 is also driving part 1 via the peer variables? That would constitute a circular relation, a logical loop. Sync techniques on their own do not do this.

 

Anyway, that's just something to think of. Again, I'm trying to get people who will be able to give you a real answer.

Re: Design intent in assemblies

[ Edited ]

@MLombard, No, it's really very simple. There is no loop. In fact I can leave out the driven part altogether just to eliminate doubt. If I peer variable change the single part in the assembly it will change according to the last DI conditions that were last set by editing any part. So if I turned off symmetric, for example, in a totally unrelated part, then symmetric, will not apply to my peer variable edit even though I need it to. Solution manager isn't even relevant because it's not even there. If I'm fortunate to see that it didn't change as expected, I can open any random part and click on symmetric again, go back to the assembly, peer edit the part which will then change as expected. Or unless I exit SE and come back in. Because the DI conditions apply globally in the session. And you can't see what they are in an assembly.

So, it seems to me if they can't carry with a file rather than a session, then make them visible in an assembly for times when you edit variables.

Edit: The single part is in sync BTW.

Bruce Shand
ST9 MP6 - Insight - Win10 - K4200

Re: Design intent in assemblies

Perhaps this was one of the issues that led me to abandon sync completely. My largest problem was figuring out what conditions were added automatically when I want to add every single relation/geometric assumption manually. The little bit I did with sync I learned to do everything asymmetrical so that symmetry would not be implied until finished with a design. I hate the idea of figuring out what was added automatically to free up a design so that it can be completed. Once a part is complete, I want to add the relations and geometric assumptions manually, not automatically. In that way, I know how the part will respond to changes.

Re: Design intent in assemblies

[ Edited ]

I'm starting to agree. I can see how sync is good for relatively simple single part editing or editing in simple assemblies. Like the ones featured in demos and training. But with more complicated parts with many tiny features, without sketches controlling features in a stepwise fashion then all features (not controlled by DI, and that's where global DI settings get you, where DI settings are off when you need them on or vice versa) have to be dimensioned at the same level and it gets very confusing trying to see through all the PMI. And sometimes a part will not modify just because it's missing a dimension. But there is no clue where that lack of definition is. So I end up throwing more dimensions on guessing where the problem area might be until it does change as expected, if it ever does. It's nuts-driving.

Bruce Shand
ST9 MP6 - Insight - Win10 - K4200

Re: Design intent in assemblies

This is why I still prefer ordered. I can control so much of my assemblies size, shape, etc just from the peer variables, and do it in a controlled manner.

ST9
Windows 7 Professional
Nvidia K2200 & Intel Xeon
"You're doing it wrong"

Re: Design intent in assemblies

[ Edited ]

I would actually like to use sync. There are just a few things that prevent me from using it.

 

#1 capability to add (and delete) live rules one by one. NOT automatically.

#2 capability to control the size of a sheet metal flat pattern directly. IE a tab or flange(s) floats to satisfy the flat size.

#3 ability to back out of Sync to Ordered if the part was origionally created ordered.

#4 More flexable flange design relative to resultant flat patterns. IE Flange side angles and normal cuts work from the K plane and can be adjusted to and from the bend line. Including the side angle cutting material from the tab it is attached too. (same desire for ordered)