Extracting Simple Geometry from a Complex Model

[Anonymous]

I have been tasked with building a new part that is similar to an old part that we have. The new part is different enough that it will be a ground up re-build, but there are a few features from the old part that will be reused. Additionally, the old part was created in such a way that it has become unstable and the file is very difficult to work with. The simplest features will simply be measured from the old part and re-built on the new part. I am looking for the best way to re-use the complex features from the old part. 

The complex features are the result of maybe 100 operations. Sketches, extrusions, thickens, etc, etc, etc. The operations to create these features are scattered throughout the operation tree and it would likely take me a few days just to isolate all of the operations to attempt to extract them and re-use them. I'm looking for a simple way to grab the current geometry of this feature while ignoring the operations that created it. 

My current plan is to:

1. Use cut extrusions on the old part to isolate the feature that I am interested in 

2. Extrude a new body that covers the feature.

3. Use the combine operation to cut the old feature out of this new body. 

4. Create a second new body that fills the void in the first new body then use the combine operation again to fill that will fill that void. 

I've basically made a digital mold of the feature, then digitally cast a new part in that mold. 

This seems to work well. 

My questions are: Is there a down side to this method of part creation? Is there a better way to accomplish my goals?  Can I now sever all of the relationships between this new "cast" feature and the old, complex part? I'd like to simply save the geometry as-is without the links/relationships to ensure that my new part is not linked to the old, unstable part. 

I've attached a picture showing what I'm trying to do. 

THANKS!

John

[swalton]

Why reuse features from a model known to be unstable and hard to edit? 

Use what you learned about what doesn't work to plan your modeling approach to the new-from-scratch model.

[Anonymous]

Maybe I didn't explain it well,

The features work well. We've printed them many times, after several iterations we have the mechanism working well. I do not want to go through that process again. It works, it's good, I'm going to use it. 

The rest of the part will be redesigned. 

But I want to be able to import the current geometry of this feature so that I do not have to go through all of the iterations again to get the feel of the mechanism just right. 

Does that makes sense?

[johnsonshiue · Accepted solution]

Hi! There are multiple ways to do that along with your way. You can derive the complex part to a new part. Then use Thicken/Offset to create a zero offset on the faces of interest. You can turn the surface body into a solid body (or keep as a surface body) and move it around to wherever you want. Then join it to the main body.

Or, you can place the complex part in an assembly. Then insert an empty part at the origin. Then use Copy Object command to link the face geometry from complex part to the new part.

Many thanks!

[Curtis_Waguespack · Accepted solution]

Hi john,

Can you share the subject part? I suspect the features such as the slide can be modeled more simply than you imagine, and by doing it as such will make reusing it much easier.

I used to do a lot of what you describe of working out the details and refining features based off of unknowns, previous models, and other components that were being designed concurrently. The results were not great, often with all kinds of features scattered all over the model tree based on weeks of back and forth revisions.

The solution I arrived at, was to look for the point in the design process that allowed for it, and re-model the entire thing.

It's surprising how it can take weeks to arrive at a kludgy model that is unstable when figuring out what the design needs to be. But we can re-do it using best practices and all the "knowns" and have it done in a couple of hours,. And the model feature tree is lean and mean, and the model lends itself to making changes and revisions with just a few clicks. Some of that comes from the simple sketches approach:

http://inventortrenches.blogspot.com/2011/03/inventor-101-simple-fully-constrained.html

I suspect you might want to know about the Rule Fillet tool also. You can often keep your features simple, and add the fillets using the rule fillet at the end of the model, and that will save a great deal of time and effort:

http://inventortrenches.blogspot.com/2013/02/rule-fillet-incident-edges-for-machined.html

I hope this helps.
Best of luck to you in all of your Inventor pursuits,
Curtis
http://inventortrenches.blogspot.com

[kelly.young]

@Anonymous I see that you are visiting as a new member to the Inventor Forum.
Welcome to the Autodesk Community!

First off, thanks for the detailed initial post you will get much farther providing concise thorough information from the beginning.

The previous replies are excellent and I urge you to follow up with their responses regarding modeling practices. I was wondering have you ever tried implementing the Vault into your projects? Using the Copy Design and part revisions could help you sort out some of the referencing struggles. You can keep all the different iterations of the design process, go back to previous versions, and document along the way to notate what works well.

The reason I bring it up is you state that you have some parts you like and some you don't. Without a revision manager this could become troublesome. Once you get the modeling behavior down this might be something worth looking into if you're not already utilizing it.

[Anonymous]

Johnson,

Thanks for your input. I like the idea of zero-offset surface. I'm going to play with that. Thanks!

[Anonymous]

Curtis,

I agree with everything you've said. I think I could recreate it. I've been told that I am wrong. It's a point of friction. I'm trying to follow orders...

I read both of the blog posts that you shared. They're great. Thanks for sharing them. 

[Curtis_Waguespack]

---

@Anonymous wrote:

....I've been told that I am wrong. It's a point of friction. I'm trying to follow orders... 

---

Well that's part of the design process as well.   In that case johnsonshiue's workflow is probably the way to go.