MakerGear Forum

I've noticed that complex models are a lot less prone to non-manifold edge errors if subcomponents are kept separate and held together using the Group command rather than welding them together using Boolean Union. S3D does not seem to care if the models are really just nested parts instead of one continuous piece of geometry-- it prints the same either way. Making changes to the original model is also a lot easier when subcomponents are grouped instead of joined together with a Boolean union. It's very frustrating to complete a complex model only to have it generate non-manifold edge errors in S3D, then have to go back and almost rebuild the entire thing because unioned parts are preventing a face from being able to move or some other minor change like that. I had gotten into the habit of keeping one version of the model with grouped parts for editing and then a copy with the unioned parts for final printing, but it appears you don't even need the final model to be a one-piece deal. Anyone else notice this or have an opinion?

Yes indeed, I've noticed this, and I've had to use it a couple of times when files I was designing in Rhino refused to weld through Boolean Union for whatever reason. (Likely because I jogged something out of alignment a hair when moving things around and resizing.)

But a non-unioned slice job from S3D is not the same as one that has been through Boolean Union. It keeps those separate parts separate, so each part will have it's own perimeter walls, and infill. The walls might be touching, but you are going to sacrifice strength where the two parts touch, and it will be prone to break there.

On the other hand, if you like to mess with changing out colors - that's a good way to make it easier to see where to drop in your code.

Recheck: Come to think of it - not sure if grouping makes a difference in that individual perimeter thing or not. I might not have grouped the the touching parts - I might have just forgotten to BU them.

Never mind!

I usually make a finished part, then copy it, union the copy and move the unioned copy to a new layer, usually called "completed parts" or whatever is pertinent to the project. I always keep a layer called "layout lines" that I use in case I need to go back and rework something. It holds the profile curves used to create solids. At times, I have made "surface layout" and "solid layout" layers to further help me work backwards and revise if I need too. These layers are usually kept turned off. Layers are free, so I use them a lot!

The reason I take it to the union stage is that when you union and follow up with "merge all faces", you get an immediate check on how well your model is put together. Did all the extra planes disappear? Great! They didn't? That means that something out of line. You may not be able to see it, but Rhino does.

If your plane is off even the tiniest bit, Your result can be a print with one area that prints on layer one and the rest of the print base starts at layer two, resulting in poor adhesion. Sometimes, your group may have small separations that look good on the monitor, but not in the print.

For me union is like a final check of how well I paid attention to details while designing. It keeps surfaces aligned and allows you to check for naked edges and bad objects. You can certainly group and print, but union and edge analysis confirm good modeling practices.

x2 on will and jules comment. i do the same thing. before union i will make a copy. i always print fully joined models. s3d will print the two differently in many cases. i have never exported something that was fully joined and had rhino produce a non manifold edge. i will have parts not want to join however but that always my fault in how well i built the model. usually there is a small gap somewhere. you can easily run a test before export to check for naked and non manifold edges. i would assume that tool is available in rhino4. if not then thats another good reason to update.

You fellas are right on the money regarding making and keeping the extra Rhino files with the parts unjoined though. I learned the hard way that if you don't want to rebuild the blasted thing from scratch several times - it's a great idea to keep an unjoined file in addition to the final one.

i use the layers a lot as well, but I've got a bad habit of forgetting that I've got the extra stuff in there (hidden) and S3D will pick it up and try to slice it.

Makes for some very interesting results!

use export selected. this way you pick the model you want exported and its not the whole file

jimc wrote:use export selected. this way you pick the model you want exported and its not the whole file

Picture me smacking my forehead!

I've seen the layers feature but never used it. Sounds like a good way to organize stuff though. It does get cluttered having multiple model clones laying around in the same view space and having to move them out of the way so they don't block the model your actually working on.

I've never been able to find a feature for detecting non-manifold edges in Rhino 4.

http://docs.mcneel.com/rhino/5/help/en- ... wedges.htm
This might help.

Just make sure all of your normals are facing the right direction or some of your sub-parts might make rather interesting holes in your part!