Mechanical Blender

Wow, this would be one of the most awaited features in Blender for me!!! A way to import a non-mesh file format, like STEP or IGES or anything industry standard. Good work!

There was an old GSOC project for NURBS integration in Blender, and it was almost working, but then it was abandoned by the developer. I hope yours is here to stay. Maybe you could package just the importer as a standard Addon, so that it would work with any Blender version.

Is there any Windows binary available for testing?

Nice things!

We are interested on a more generic tools :stuck_out_tongue: and more focused to mechanical/parts development, because this is our area :wink:

At the moment, the importer is translating data to mesh data, so there is no support for nurbs / curves. Itā€™s done as an addon and avaliable at

Itā€™s not finished, needs testing and best way to solve some issues


What you refer about cad curve tools ?

Hello,

I have offten problems with importing step-files from my CAD-System to Blender. I have to Export step, convert it to obj and then Import to blender. The resulting mesh is terrible and the polycount is offten to high for any pc. So an Assembling with many parts is not easy to handle.

A better remesh while importing files would be helpfull, the function Alt+F wonā€™t work for every part of the mesh.

greetings
Kai

Attachments



in blender we are missing some of tools for curves on CAD
not certain which one but been said before several time in threads

would be nice to add these tool in SVN ?

and may be even including the new T spline surface but that one might be difficult unless it already exist in Freecad !

Happy bl

THats a neat little animation, done in blender using line effects?

Thanks. Yep, bog stad Blender renderer with ā€˜lineā€™ option turned on. No cartoon shaders, just fiddled about with some of the specular options.

Some one mentioned spreadsheet output?
That would have been pretty handy. I had to hand build a parts list for this project.
Would have been great if I could have had all the ā€˜nailā€™ and ā€˜screwā€™ instances counted and listed.
Also nice to have all the timber lengths catalogued.

@mauge
mauge123/mechanical-blender

 [Geometrical solver based on SolveSpace](https://github.com/mauge123/mechanical-blender)   

KmolYuan/python-solvespace

Hi, I received an e-mail a bit ago, I saw that you are integrating a parametric geometric constraints solver, cool. I hope to see something as soon as possible.

Has been some time without development on this project because lack of resources and timeā€¦ I hope this new year 2018 get time for it and get it working over blender 2.8 :slight_smile:

Best wishes for this new year!

Hi Mauge,
Best wishes for 2018 too. Itā€™s always hard to program on free time. Blender 2.8 will take some time to stabilize, I would recommend to stay on 2.79x until they have a more stable 2.8x branch, it will offer a much better user experience for your testers.

can you tell us about that UCS thing
I thought there was already some project at BL org for that !

hope to see some new features added and new build for wind 10 soon

good luck
happy bl

thanks! I have to say that part of the work was done during work time, but this is not possible at this moment, and yes itā€™s hard to program on free time, specially if there are kids on the family!

Another part I would like to work with is the step importer, which is in python and need implementations. Also an exporter is required to communicate with other systems and applications.

http://www.mechanicalblender.org/static/reference.html#ucs

Any comments on it will be appreciated. What was done is adding and Origin to transform orientations, and grid and Standard Views (Top, Leftā€¦) and also grid snap should be according to UCS if selected

Maybe need more and better explanation on the reference.

Do not know if the BF project is working, and what was done there. Maybe a patch could be submittedā€¦

Good luck finding time to continue this year! I really love what you are doing with Blender.
Yes the added tools are more CAD like but I honestly believe they are simply better than current default behaviours.
Is there (or has ever been) any talk between you and the BF that at least some of what you are doing could be part of master Blender?
Your kinda snapping, pivot alignments and UCS (<3) - at least - could totally work there.
But I also can be very wrong and I guess this was already discussed before but could you please explain me how things work with useful branches like yours?

Some years ago i got a subproject on blender 2.48 but was discontinued, and lost definitively on 2.5

I 've got lot of experience since this first time, so the project was started from beginning, with a completely different focus. But this time didnā€™t get any space on BF blender, i starting with snapā€™s and submitted a patch and although imho was getting a good feedback, it was postponed. Do not know current status.

Anyone knows that blender is considered by heads as a tool for artists. There are some opensource projects focused as cad. But i like blender, and the idea of have a suite for all: i do not like exports, to modify or add info to data. because you end up duplicating data, and in CAD this data could be requested to be modified and reused in different projects or during product lifetime.

So the solution to get faster development was forking and trying to get always the project sync with BF trunk.

Thanks for your answer!
As I said I believe that many of the features you implement could fall in the ā€œtools for artistā€ box.
I can say that because Iā€™m an artist myself and Iā€™d enjoy working with them. You should talk with Dalai Felinto if they see any of your tools as a good extension to the 2.8 project longer term goals. At least you could ask them about snapping and UCS :slight_smile:
But sorry if Iā€™m pushing too much.

Iā€™ll try! As i heared from Ton some time ago, you must insist and insist and maybe it will be includedā€¦

Like IĀ“ve said before, Ton is allergic to the word CAD in Blender (Blender is not CAD software! Better use a proper CAD software for that!). Sell it as improved modeling tools and snapping. Marketing, marketing, marketingā€¦

I agree with theApe, if you are serious about CAD design, you need a history based parametric modeler. However, for the home user who just wants to make a custom swaybar down link for his autocross car or some wood furniture, Blender could be helpful. The biggest limit right now is the inability to make dimensionable line drawings. To make line drawings, you need to draw the silhouette and the edges of all the features. The basic routines for these already exist in Blender. When you turn on Outline Selected in Object mode, the silhouette is shown in yellow. For the edges of features, the Edge Split modifier doubles the vertices at all edges equal to or greater than a specified angle. Using routines from these two, one should be able to make a new 3D view mode, solid with visible edges shown (or just the silhouette and visible edges). If you set this view mode in quad view, it would have all the information for a printed drawing. Even if you had to add the dimensions by hand, you would have something you could take into your shop to reference when making the part. If there was a simple way to aid dimensioning that would be even better.

Blender as is is not productive for CAD enviroments.

The history based model is not as important as the availability to modify data fast and with precision dimensionally.

Printed technical drawings are also important, not only for manufacturing, although I can see a transition there, being less necessary in most areas.