help on Best practices for "large assemblies" 2

[lordrailie]

Hi all

we are getting now into a point where we don't know what to do with the design of our products.

Is it a good practice to disign in "context", i mean to to parts inside the assembly directly?. Up to now what I've seen is that if I do that, everythink works fine up to:

-the moment you take that part out of its context and try to midify it, or link it to other parts
-the moment you want to do a drafting, as wen you explote the assembly, the parts becomes red, and I cannot make an exploded scene out of it as it "not constrained totally" which means that if I explode it in the scene it wont come back
!!

Or what to do....

I tried to instead of working in the context to open the part in a separate window, then copy and paste with link all the references i need for the creation of my new part and then start working. but...

-then when I modify anything, to update the part I have to load al the parts where i took references from, which i might not know, and catia does not tell me, all the references surfaces and solids come with an ?, and in the porperties or link does not say anything about where is pointing.

-Its tedious to have to copy all the references from one part to the other etcetc..

-and when i put it in the product, is placed in the space in its correct place but is not constrained.. so should constrain it...

I actually don't know which way to go.... When the think goes to big assemblies with linked parts it gets complicated...

 

[ferdo]

Hi,

See this link

http://www.eng-tips.com/viewthread.cfm?qid=115631&page=2

(Designing in car position or not in Catia V5).

Regards
Fernando

 

[catiajim]

Designing using Contextual Links can be very powerful when done correctly. But you need to understand exactly what is happening when you do this. You also need to understand the consequences. There are times when it is better to use a Copy/Paste Link over a Contextual Link. There are other times that it is best not to link at all.

You want to be very careful when using Contextual Links. you wan to plan ahead to make sure that you are not creating a complete rats nest of links. We have found it best to create several "Control" documents that have published elements in them (Lines, Planes, Surfaces, etc) and have the user link to these, and only to these. Creating links between this part and that part, and then from that part to another, can result in a very tangled mess that will eventually have cyclic links. This can be extremely messy to straighten out. Plan what you are going to do, and draw out a simplified map of the links. This will show you whether what you are doing is reasonable or not.

 

[Configurator]

^^

I do the same when I'm laying out conceptual designs. I first set up a part with wireframe geometry, and I build parts in the context of an assembly linking only to that wireframe part.

 

[lordrailie]

Hi Catijim and configurator

more or less that is what I got to... having a part in the assembly which I call "reference surfaces" and then linking only to this part. I still don't know how to publish, and what advantages does it have.

And when i open a part out of its context, then how do i know to which assembly belongs? and which parts is linked to?

What i am getting to, is that first of all I have to create a part, that contains in surfaces, my model "raughly", and when i have it all, then i start doing parts which contain individual solids, and that is referenced to that references surfaces. Is that a common way of creating models when the whole process from design to manufacturing has to be done?

 

[Akesson]

Hi lordrailie

The way of working that you describe with a "reference part" is in my point of view a good way to work. I use it in all larger products that I create.

"Publishing" all geometry that you want to use as a reference is important to make it easier to replace a reference. I even publish all planes etc that I use as references when creating constraints between two parts. By doing this I can replace a part and the constraints will not fail.

If you open a part out of its context you can use Edit/Links... to see to witch document the link points and the name of the publication.

One benefit with contextual links is that name of the linked feature in the graph tree gets the pointed feature name added within parentheses.

And last I think it can’t be said to many times (catiajim already said it) only create links from the reference part to the parts not links between to parts. By doing like this you’ll never end up with problems like not knowing where a link points etc. And the update process performance increases.

/Akesson

 

[catiajim]

Publishing your feature is as subtle, but essential step. You can Publish elements by selecting TOOLS, PUBLISH and select the element to Publish. When you do this, you provide a Public Name for the element that can (and probably should) be different than the element itself. This operation puts a special identifier on the element, so that when you go to replace or update your links, CATIA will be more stable.

When you change your source document, you can delete your published elements and simply re-publish your new elements with the same name. Then updating your target document becomes an easier replace option than with un-published elements.

One thing to watch out for when publishing elements is to be aware of whether you selected a FACE or SURFACE. These many times appear to be the same thing, but can trip you up later as they are not interchangable. As FACE is the sub-element of a surface, plane, or solid that represents the functional surface selecte. A SURFACE is the entire element. Subtle, but the difference can make or break you.