New 3D print tools in Blender 3D

[Dolf Veenvliet]

Hi everyone.

I am currently working on a tutorial DVD about modelling for 3D print in Blender 3D with the Blender Foundation. As part of the development of the DVD we are also working on some tools that will help. Some very basic, like volume estimate (for commercial print cost estimate) to some more complex ones like wall thickness estimations... and particularly useful for makerbot users... Overhang indication (so you can see if you need support).

If you want to try the tools out.. there's are some testing builds of Blender available here: http://builder.blender.org/download/

You have to enable the 3D print tools addon in file/preferences/addons then you can find the new tool set at the bottom of the tool shelf.

Right now I'm in the middle of recording, and I was wondering if you have any feature requests or subject matters you'd like to see covered? This is the time to ask!

Remember though... this is not about printing itself, or slicing models... just the modelling part.

Thanks for your input!

Dolf (macouno)

[Joe Larson (aka Cymon)]

I saw this one and were I not in the middle of a similar project myself I would have jumped on the experimenting with it bandwagon.

The only features I've ever wanted is something that will automatically clean up bad geometry ala cloud.netfabb.com. I mean I appreciate Blender telling me where the problem is and letting me fix it myself, but there's a reason I run everything through netfabb all the time.

On Friday, April 5, 2013 8:57:34 AM UTC-6, Damian Gto wrote:

Well I do not use Blender, but one thing I know people do like is an auto support generator.

What I mean with that is that the program will add a minimum support where it is needed.

This would reduce the material you need to use.

One other thing is a watertight check to see your object is right. I have seen many STL files that is bad.

Sure they are easy to fix for me in a 3D program, but many do not have that skill. So an easy tool for this in Blender would be great.

Also a guide how to build an object and what to look out for when you printing. Like you should try not make an object that need support if you can.

[Dolf Veenvliet]

Hi Damian & Joe,

Thanks for the feedback.

Damian

The support generation is something we're considering, though we're on the fence on still. I kind of feel that perhaps that step should be done when slicing models for your specific printer... perhaps not in the modelling software... but we're definitely thinking about it, and I'll certainly show a few ways to manually create supports.

A watertight check already exists in Blender 3d and it is called "Select non manifold" in edit mode... an easier access to that function has already been added to the new toolset as well.

Of course I'll show several ways to model with 3D print in mind.

Joe

What cleanups does netfabb do for you currently? We have a few automated things in there already... But yeah... What we're doing is more about helping you design, than fixing up models after (for instance fixing 3d scan results, though I'll look into multiple fixing methods on the dvd).

On Friday, April 5, 2013 3:29:44 PM UTC+2, Dolf Veenvliet wrote:

[Joe Larson (aka Cymon)]

Honestly I don't know what all netfabb does. It's a little black box. But here's the changes I've observed:

Closed holes.

Normal Re-calc.

Removed internal geometry.

(Apparently) Boolean-ed all loose shells so intersecting objects become only 1 continuous shell.

There have been times when Blender couldn't correctly Boolean objects so I just exported them and sent them to through netfabb and lo-and-behold they're together as if they're Booleaned. I would be very happy if the Boolean would have an autofix option, too. Granted the Boolean is better than it was in the past I still have trouble combining complex geometries.

[KM Design]

Whichever piece of software takes on the challenge of creating an "auto support generator" where the user has direct and flexible control over the end result (or can add/delete individual support pieces after the fact) will win the 3D printing race. After four months of 3D printing, there is no question this is the biggest unanswered need for Makerbot (and other accessible FDM) printers.

If you'd like Blender to become a tool in a 3D printing workflow, it really is a must.

Kevin

[Eighty]

Why not just model the supports yourself? You have all the control you need. Lots of people do it, including me on occasion.

[KM Design]

I model supports all the time. But the question was what would make their software more 3D printing friendly....

K

[Steve]

Dolf:

In terms of videos I would suggestfour videos showing how to start from scratch and design simple projects. Each project should be described in advance and each one should showcase the use of different tools. Let people start out simple and work up. Too often the demos of how to use a tool are written by and for people who already understand it. Make some demos that start simple and then build on the concept showing how to use new tools available in the program. Also do a video on why people should be using your tool instead of the others out there.

Joe is right on with the need for tools to 'fix' 3d models like Netfab does. The design tool should output STL files that can be directly 3d printed without needing to be cleaned up. The extra step of going through Netfab is very important for designers using Sketchup since that program easily outputs STL files with problems for printing. If you have direct output of printable STL files you have an advantage and another reason for people to switch to your tool.

[Dolf Veenvliet]

Thanks for the input everyone!

KM Design Well... I don't think it is our goal to make Blender "The" go to software for 3D modelling for 3D print. Blender is already a very good program, and we can help people who want to learn how to use it for 3D print get their head around the tools in the program, and to help those who already know Blender to model for 3D print. And... we have the opportunity now to add a couple of helpful tools. And yes... support is one we're thinking about... though I will definitely do a video tutorial showing some ways to do it manually.

Steve Yes... I try to keep the less experienced user in mind.

Automagically fixing models is hard... Some we can do easily... but a lot of it is quite complicated... like the booleans issue. A lot of work has been done on the Boolean operators in Blender in recent times, and they improved a LOT, but they're not quite perfect yet. The best way is still to make a clean model. Not make a messy model and depend on software to fix it for you... because it will have to make some arbitrary decisions for you... which you may not agree with at all. And "learning to model clean from the start" is where I can help with some tutorials ;)

On Friday, April 5, 2013 3:29:44 PM UTC+2, Dolf Veenvliet wrote:

[David Celento]

Glad to hear of this effort!

Having taught 3D modeling at university level for six years -- with watertight modeling the goal for surface modelers -- the one feature I desire most is "colored backfaces" for NURBS, polygons, and subdivisions.

Rhinoceros (windows version only) does this and it is FANTASTIC. Far better than millions of useless "hairs" sticking out of complex objects.

I'm a Blender novice (tried several times, impressed, but frustrated that norms from Maya, Rhinoceros, and others are SO different) so I'm not aware if Blender already does this in the latest version?

Biggest wish for Blender, in general, would be hotkeys from other programs that actually work (Maya hotkeys were largely a fail when I tried them) and auto-complete command line (like Rhino).

I'm newly under contract with Wiley to write a book on digital modeling with the goal being CNC and 3D printing outcomes, so I could use some insights about Blender's capabilities in this area. Would love to feature it in my book, but initial explorations have not been encouraging -- perhaps due to being (ab)used to other common methods and a lack of Blender knowledge.

Feel free to email me off-list if you'd like to discuss how Blender might fit in to this book project.

[Joe Larson (aka Cymon)]

Now that I think about it there are a few quirks that I'd love to see cleared up somehow.

First of all Blender doesn't like the scale things have to be made for 3D printing. If I ever want to use the rendered to show off my models generally I have to scale it down to 1/10 it's original size. I think there may be a a way to fix that but the point is by default Blender's render depth is fairly low at 1unit = 1mm scale. It would be great if there were some default that would set Blender up to be happier with scenes like that.

Secondly Blender's measuring tools are... confusing. You can turn on edge length but what you get is a jumble of useless information. It would be nice if the line I was moused over or selected was the only one that showed the length.

Also, snap to grid or snap to units isn't really useful. This was one thing that TinkerCad did exceptionally that allowed simple movement to generate confidently precise results. When snapping to the grid (completely configurable at what level it would work, 1mm, 0.5, 0.1, 5mm, 10mm etc) the bounding box would line up with the grid points in the directions being moved. If something were 1.2mm wide and it were being moved along the X it would snap to the left side lining up with the nearest grid line, then the right side.

[Shawn]

The biggest issue I ran into when working on my first model for 3D
printing was the units (mm/cm/inches/etc.). I had a mental block on how
to set the correct units so things would be scaled to the right size.
After some digging, and realization that STL simply has points and
doesn't make use of units per se, and realizing that one "unit" in
Blender is equivalent to 1mm in print, then life got easier on that
note. Such a simple thing, but it blocked me for a day or two.

If your video(s) can address that somewhere, it may help other users
avoid that pitfall. (Or maybe I'm just a little slow sometimes. :) )

> --
> You received this message because you are subscribed to the Google
> Groups "MakerBot Operators" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to makerbot+u...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>

[Dolf Veenvliet]

David Celento I love the idea of coloured backfaces. I'm sorry the Maya short cuts don't work as you expect, but then... personally I'm not necessarily in favour of making programs behave exactly like the others... But that's a whole sepparate discussion... lets not start that here ;)

Joe Larson (aka Cymon)/Shawn I agree... "scale" in general is tricky in 3D programs... because it's all relative. But that is also it's greatest strength. Because we make scale relative, you can model anything from a molecule up to a galaxy. If we measured both in Meters... it would literally be "hard to compute". I will definitely take some time to talk about scale on the DVD. And... btw... you will be glad to hear that new measurement tools are being worked on and will be available in the next official Blender release.

[David Celento]

Hi Dolf, if you have a windows version of Rhinoceros, check out the colored blackface setting. It's one of Rhino's nicest features, but it's so buried in the prefs that I used Rhino for about 3 years without stumbling on it. Many thanks goes to Marc Fornes of THEVERYMANY for pointing it out to me.

As for Blender's conventions (compared to most others) -- I'm fully aware that this is a loaded point. Blender has every right to do as it sees is best and I've seen amazing results from those who have adopted them to be reflexive. The only thing I'd add is that a LOT of people are interested in Blender (who know a number of other programs quite well, and could/would contribute to Blender development) but they find it very difficult to do the simplest things due to radically different conventions.

Helping them over the initial hump would be most welcomed! As a side note, and to illustrate the point, there was a very good thread here about what software would be best since Tinkercad closed up shop. Blender was suggested, but had very few advocates not because it lacked capabilities, but because many found the conventions too difficult for people who knew 3D modeling well to even help their children.

I spent three (fantastic!) months in The Netherlands (creating digitally inspired ceramics at EKWC), so I know that you Dutch appreciate a good argument! :-)

[pashakun]

Wow! Very happy to see this thread~

I got started with Blender about 3 month ago with little prior experience in 3D design and have already had massive returns on time invested.
The community has a wealth of resources for beginner blenderers, but unfortunately there is not a lot of 3D printing discussion and most of that is centered on modeling for Shapeways and the like.
 
From my admittedly brief experience:

1) Handling non-manifolds 

- easy enough to find in Blender, difficult to fix
- Netfabb works more often than not, but it doesn't seem to give minute control unless you have a paid version

2) Hollowing out

- Sometimes n00by acrobatics can leave you with a model that looks good on the outside, but is jumble of subsurface geometry

- I suppose you can do smth like select visible only, duplicate, thicken etc, but there must be a better way

3) Boolean / Shrinkwrap / Soft body / Fluid / Cloth modifiers
- Love them when they work, stumped when they don't

- Would love a 3d-priting-specific tutorial on methods, pitfalls etc

4) Custom supports

- Slicers tend to be overzealous when it comes to adding supports

- A tutorial on modelling supports would be great, a supports modifier would be simply awesome!

5) Custom infill
- Customizing the structure of infill to vary rigidity & flexibility of the printed mesh

And I would totally buy that tutorial DVD :)

[Shawn]

One other thought... Perhaps it's just me, but I find that almost all
the literature out there on modelling covers either before or after you
have your mesh. i.e. starting from a blank slate, here are the basic
tools for basic mesh creation, or after you have the mesh, here is how
you color/shade/texture it. What seems to be missing is the tips and
tricks for manipulating a mesh.

I started by creating a cube, put it in edit mode, and deleting the
vertexes. THEN I would add points as needed to outline my object in 2D.
It was only after watching an expert working for a bit did I see the
power of loop-cut, scaling, and moving the resulting points. Since then
I've found the different selection tools (the B and C keys), figured out
that pushing the +/- keys selects more/less points around the current
selection, etc. These techniques (and more) seem to be considered as
common knowledge and thus glossed over after a quick description of the
tool. How and why the tool is important to mesh editing is kind of skipped.

I've found that for 3D Printing, I really don't care about the textures,
materials, shading, rendering, animation, etc. All that seems to matter
is building the mesh and then saving it to STL. This is the guide I'd
like to see, or hope to finally stumble upon on the interwebs...

I hope the changes/videos you are talking about help to address this
somewhat. And I do appreciate any effort to make things better/easier.

Disclaimer - I am most definitely NOT a 3D designer (though I can fake
the basics now), and understand that my experiences (or lack of) may not
be representative of the larger population. :)

Shawn

[Mike Payson]

On Fri, Apr 5, 2013 at 11:09 AM, Dolf Veenvliet <mac...@gmail.com> wrote:

The support generation is something we're considering, though we're on the fence on still. I kind of feel that perhaps that step should be done when slicing models for your specific printer... perhaps not in the modelling software... but we're definitely thinking about it, and I'll certainly show a few ways to manually create supports.

The correct answer to this question is "there is no correct answer". There is no reason this should be considered an either/or question, there is a place for support in both tools. The slicing software should definitely be able to automatically create support for most simple cases, but there are always special cases where the standard support doesn't work for a variety of reasons. Being able to do it automatically generate basic support and then giving the end user the ability to edit the result would definitely be useful. 

[Dolf Veenvliet]

pashakun Most of your suggestions will be on the dvd. The one that won't is "custom infill" because that is something that's done by the slicing software... and you don't do that inside Blender 3D.

Shawn I have created an hour long video of me doing a simple model start to finsh... and that will go on the DVD. That should address the in between bits.

Mike I completely agree! We'll have to see whether we can get around to it... I will certainly show some ways to do it manually.

On Friday, April 5, 2013 3:29:44 PM UTC+2, Dolf Veenvliet wrote:

[James McCracken]

Support in blender is a great idea - our slicing software does currently generate support but we don't have much control over it, and the amount of support material created tends to be overkill for the application.

Overhang detection is also a very nice idea - it'd be great if Blender could draw shaded polys from the model down to z=0 for overhangs over a certain angle.

Pretty much everything else I brainstorm about, I can figure out a way to do it with what Blender already has... mostly through vertex selection... I would like a much richer extruded text tool - and as others mentioned, boolean operations in Blender often result in weird printing behaviour if the model isn't cleaned up first.

--

[Joe Larson (aka Cymon)]

The tutorials on http://cgcookie.com/blender/ I find are excellent. I thought I knew blender until I went through their Blender Basics. Wow, the things I learned there. And I'm not talking about animation, screw that. I'm just talking about the basic tools.

[tunell]

OK, I've been toying with the idea of jumping from the sketchup boat for awhile. I think this may be the deciding factor.
Although I have no Idea how to use blender, I'll take a look at those videos suggested by Joe and get started.

On Friday, April 5, 2013 6:29:44 AM UTC-7, Dolf Veenvliet wrote:

[Pasha Alpeyev]

+1 for CGCookie.

Also, CGBoorman's tutorials and Andrew Price's channel

There is more than plenty to get started with Blender basics

--
You received this message because you are subscribed to a topic in the Google Groups "MakerBot Operators" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/makerbot/ln3f4UHz5V8/unsubscribe?hl=en-US.
To unsubscribe from this group and all its topics, send an email to makerbot+u...@googlegroups.com.

[Blake Messer]

Can't the blender project take on all the replicatorg people and together they could make one beautiful and userfriendly open source one-stop shop for 3d printing design and implementation.

I'll fork both projects and make it myself. That'll show em.

[Bryon Miller]

You need to know what to look for.  3D modeling and design is complex, it's not a simple task and has always had a very steep learning curve.  Even then, once you've learned the tools, if you have no artistic abilities, it's like learning how to use a paint brush but lacking the talent to make anything compelling.

But, look up modeling techniques for tutorials on how to manipulate a mesh, create a mesh, boolean, etc...

Box Modeling - this will give you insight on how to manipulate simple geometric shapes, convert to editable mesh, manipulate vertex, lines, poly's, quads, etc.

Spline modeling - Not sure if this is still in use, but you can create more organic models without getting into the complexity of NURBS.  This was an old technique that was used years ago.  It is probably still used today.

NURBS - High quality organic models.  This is much more difficult and probably too complex if you are really just wanting to make simple meshes.

Cross Section modeling is like spline modeling but you put up two intersecting cards with reference images on them in your build area that act as blueprints and you trace and pull out the lines.  It is important to understand this process, because the way you trace, the order in which you place lines and the direction you create them in can alter your finished product.

The good thing is once you understand the concepts and the tools, you can use most any design software, it's just a matter of learning the interface and where the tools are located.

I would suggest that you look into box modeling, that would be the easiest way to learn. 

I learned how to use 3d on 3D Studio Max years ago, but this will apply to other packages such as Lightwave 3d, Maya, or even Blender.

[Bryon Miller]

That's what I would do.  I've used 3d to just add pieces to hold down corners and act as supports when I didn't want to use a raft or remove a bunch of support material when I only need to address a small section of the model.

[Bryon Miller]

The problem is 3d design isn't something you can just learn over night.  It is very complex and has a steep learning curve.  You'd be better off going to youtube to learn the basics of 3d design and the tools because making tutorials explaining how to model would take a very long time and there is a lot of information out there.  Some of the terminology might even be confusing and interpreted as "being written only for people that already understand it" when in reality you just need a basic understanding in order to continue.

Look up Box Modeling on youtube and you'll see information on one of the most common forms of 3d design.  There will be a lot of tutorials on them.

This seriously takes people years to master.  You can get the basics down in a short period of time, but expect there to be a steep learning curve.

[John Sondericker]

+1000! I would love to see what needs supports and what doesn't. Thanks. Blender is my tool of choice.

[Dolf Veenvliet]

Thanks for all the input everybody.

I am now in the middle of recording my turotials... and pretty much exactly on the other side of the globe from me... some fancy new features are being written into Blender.

I'll take all your feedback and ideas on board... and I'm sure you'll see a couple reflected in new features and turotials.

Watch this space... I will let you know when I'm done and pre-sales start!

Dolf (macouno)

[Richard]

Re: A watertight check already exists in Blender 3d and it is called "Select non manifold" in edit mode.
Dolf,

One thing that would be helpful in all these programs for people who have not been designing in 3d for years is to include plain English terminology.  For example, how about this:

Change "Select non manifold" to  "Select non manifold/make watertight" and include the term watertight in your search engine so in Help, someone could search on manifold or watertight.

Sometimes the difference between needing support or not is just a fillet or a change in angle of a degree or to.  Blender may already have this but if not, it would be good to have a feature that lets you run an inspector at different times during a design process that will highlight problem areas such as overhangs and offer fixes that you can accept and/or undo.  I think you need to be able to do this during the process so you can fix things as you go along as well as at the end.