Proper way of doing assembly design 2

[BostjanPerdan]

Dear All,

I would like to do an assembly from several subassemblies and parts and was wondering what is the most proper way of doing this?

Should I first create subassemblies as individual assemblies in separate files and then insert them in the final assembly or should I start with the final assembly and then crate subassemblies in this single file?

Thanks & Regards,
Bostjan

 

[jackk]

Either way is OK. Whatever way makes the most sense to you, as long as you end up with an Assembly with multiple sub-assemblies.

 

[vsawye2]

As Jackk stated, either way is fine. i prefer to create subassemblies first and then drop them later into a final assembly. I do this to keep the contextual links from accidentally forming when i dont want them too. beware the brown gear, its not bad, its just not always good.

 

[MarkAF]

Both methods you mentioned have correct uses, and the situation of your model will determine which method is better. I'll try to give some example so you can understand the different effect of each method.

Think of a model of a car. The door panel is a part that always goes in only one location. Therefore it may be best to model the door in context of the whole car assembly. But the wheels go in several locations, so those should be modeled separately and then placed into the car assembly.

Specifically, I work with aircraft tooling. In that world, we follow a method that the subassemblies (and parts) are created within the top assembly for anything that is unique to a particular part of the aircraft. This way, when you insert both the tooling model and the aircraft part into an assembly, all the models line up in the correct location and orientation and do not need to be positioned. This method of positioning is extremely important for my type of work.

For other assemblies (and parts) that are more generic, used in many places, it is better to make up the subassemblies on their own and position them in the top assembly later.

Cheers,
Mark

 

[loki3000]

is it true that flexible (sub)asms consume a lot of cpu power when you move them (like in solidworks allegedly)?

 

[KevinDeSmet]

Flexible sub-assemblies basically solve all of the constraints for the sub-assembly in question, together with the currently active assembly. Assemblies' default rigid behavior allows it to ignore the constraints in a sub-assembly. All constraints need to be solves concurrently and for every constraint you add, it needs to re-check every other constraint for validity.

It's the case of: do you check your bumper, when you go to adjust your mirror? if it's not in separate sub-assemblies, Catia will. (SolidWorks is identical in this regard)

Certified SolidWorks Professional

 

[ferdo]

Hi,

In my opinion you have to follow company's rules and methods...and most of companies are saying that you have to create first the parts , then sub assemblies and then big assemblies (or integrate them in a SAM - Space Allocation Model).

This is normal if you think that we live in a world concentrated on short terms deliveries and profit , which means reusing parts and assemblies or variants of them is something currently and normal... that's why most of companies prefer to create everything with their own axis and not with a specific axis (of course, you can find exceptions)...

Also, companies are defining the methods how to model parts, in a specific way and many times as you don't like...and this is done for different reasons (luck of license, agreements with different manufacturers, a.s.o.)

Regards
Fernando