Autodesk Inventor

[#1]

Good Day,  As CAD administrator at my company, which is in the process of implementing Inventor/Vault, I am quite sold on the Layout Modeling/Multi-body part modeling functionality that Inventor supports.  Among other things I believe there is a large payoff in assembly stability and simplicity.  However, this workflow is very different from the conventional methods of modeling individual parts and manually constraining assemblies.  What I would like is to hear some verification from some of you experienced professionals (don't let that scare you off).  Is this without question the general modeling process we should adopt or are there any negatives or bad experiences out there?  Bring on the comments little or much!  Thanks in advance.

http://www.cadalyst.com/cad/inventor/layout-modeling-autodesk-inventor-15040

http://blogs.rand.com/files/ma126-1r-thats-solid-mb.pdf

[#2]

I'm with you in loving, what I call relational modeling (i.e. skeletal, derived components, multi-body, etc.). However, I have had comments from some of my designers that they have trouble finding all of the components that need a drawing revision after making a change. For example, if an assembly contains four parts, each with a hole that needs to line up and a change is made to the location of the holes, what flags that there was a rev to the parts. Now imagine hundreds of possible affected parts. With the conventional method, a change to one hole would make the holes not line up and designers could see that. We don't use vault so I'm looking for a method right now if anyone has some ideas.

[#3]

Hello.  In my experience, Layout Modeling is not 'without question' the general process of choice.  That being said, I'm in R&D, not production.  I like the approach but never found it necessary to make it part of my normal workflow for these reasons: lack of  thorough documentation (few examples of the real benefit), limited utility (good for 2D kinematics), and what seems like added complexity to constraints --particularly for kinematic models.  These are just my impressions that may well get picked apart; I don't really use this approach. 

We make a lot of one-off and oddball parts, usually with a lot of changes.  I prefer to use a linked spreadsheet for all component and assembly parameters.  I prefer to keep all constraints in the assembly, particularly when I want to be able to move parts around (make them 'flexible' constraints).  I've been shy of using blocks to do this.  As long as everybody constrains parts appropriately to the coordinate system, this way works fine. 

I expect that Layout Modeling can be a better way for some, probably not all, perhaps not even most designs.  I would love to learn more about it and have my impressions proven incorrect.

[#4]

I would echo much of what Jeff noted.  I use a Layout/Skeletal workflow more and more lately and it generally works well.  BUT - I have learned that I need to put all my Shop Drawings into a single .IDW file to make sure I can update everything at once and not send out partially-updated drawings.  It makes it a little trickier to keep track of revisions that way also, but it works okay for me.

I don't use Vault, however since I am a one-man band as far as Inventor is concerned.

[#5]

f_B,

If I'm understanding what you wrote correctly, I think you're misunderstanding the multibody solids approach.  What I think you're talking about are sketch blocks, which are useful for 2D kinematics, but which I haven't used very much.

Mulitbody solids/layout modeling consists of creating a number of solid bodies in a single part file, which can be easily related to one another by sharing the same features (a through hole, for example), or projecting geometry from one body or sketch to another.  The individual bodies are then derived into individual parts, each with their own materials, appearances and iProperties.  They can then be put together in an assembly, either grounded to the origin or constrained in the conventional way.  When that hole (for example) needs to be moved, you only need to edit the layout, then update the assembly and all is properly updated.  Much, much more robust than adaptivity, and easier to edit in most cases than skeletal modeling.  I've found lot's of uses for them, although doing an entire 100-part assembly as one layout is not one of them-- too many eggs in one basket, in my opinion.

[#6]

I don't think that one single method works for everyone. We do a lot of commercial vehicles (trailers & truck equipment). I find that either a skeletal approach using a single IPT that only contains a sketch for the placement of the components works well as does just using labeled work planes to control component placement.

I haven't really that the chance to try the solid body and push out the components.

[#7]

No, what you describe in your second paragraph is what I'm talking about.  Typically the workflow I would follow would be to start a new assembly, click Make Layout which places a grounded part in the assembly and opens it for editing then using the Make Components command to populate the assembly with parts created from the layout part solids and grounded on the origin.

I would like to thank you and everyone for the comments so far.  Definitely the reviews are probably a little more mixed than I would have expected.  I don't think we would be anticipating using the Layout Modeling workflow completely across the board.  Virtually all of our equipment is custom configured, many times to a high degree, but we have many standard manufactured and purchased components.  We would be thinking in terms of using layout modeling for the basic custom componentry and then manually adding standard components from libraries to the resulting assembly.

Any more thoughts?  I would love to hear a lot more.  Thanks again.

[#8]

Sorry, Sam B, you can disregard my first paragraph as I realized you probably weren't replying directly to my original post.

[#9]

Sam, yes, you're exactly right.  I thought the OP (and the linked article) referred to sketch blocks as layout modeling.  Thanks for clarifying.  I think your description summs it up nicely, too.  Multibody parts are great up to a point (complexity), and WAY better than adaptivity.  Sorry if I confused things.