For electrical engineering there is KiCad, which is pretty good overall. Only reason I’m still using proprietary software is because I’d have to recreate my libraries and it will be a huge pita.
For mechanical design there is FreeCad, which is usable for simple geometries, but if you come from a proprietary CAD software you may find it lacking.
I got into the 3D printing hobby a few months ago and FreeCAD is pretty much useless. I can be more productive by writing JavaScript code with Three.js library, lol.
Not really. Blender is NOT a CAD. It doesn’t ensure that your bodies are solid, it doesn’t provide any analysis tools, it doesn’t support working with blueprints/sketches, it’s not parametric, etc. Basically, it doesn’t do anything CAD at all.
I had this idea for a while to build a Frankenstein monster of a 3D software that uses real time graphics and has a multi step build process covering CAD, wireframe manipulation and voxel workflows. If I ever actually make it, your concerns will be heard despite being probably not the best softwsre to do your work in :)
CAD system must be reliable. It is simply unacceptable to have math issues which cause unpredictable geometries.
CAD system should have a good UI. This is a big issue for open source software in general as UI and UX is usually an afterthought.
CAD system should be fast and use hardware acceleration. Running single threaded python scripts on CPU to do complex computations kills the productivity. Designing real life objects is already a mentally taxing task, the whole purpose of CAD is to remove the computational bottleneck of a human.
CAD should be object aware. If I draw two gears and put them next to each other, I should be able to rotate one and see the other moving accordingly.
This is a bare minimum, I’m not even talking about computational modelling, stress testing, etc.
That is a question too hard to answer in a comment and one that depends on the use case of the software. Few users need the power and features of CATIA or NX, but those who need it can’t accept anything lesser. SolidWorks is a good spot in terms of flexibility and features if it could be easier for the average person to use. You need proper accurate parametric modeling (e.g. a NURBS kernel) for solid models and surfacing. Hearing things like wireframe and voxel indicates it isn’t suitable to me.
There are no good open source CAD systems at all.
For electrical engineering there is KiCad, which is pretty good overall. Only reason I’m still using proprietary software is because I’d have to recreate my libraries and it will be a huge pita.
For mechanical design there is FreeCad, which is usable for simple geometries, but if you come from a proprietary CAD software you may find it lacking.
I got into the 3D printing hobby a few months ago and FreeCAD is pretty much useless. I can be more productive by writing JavaScript code with Three.js library, lol.
For 3D printing, did you try OpenSCAD? If you’re already a programmer it’s much easier to get into than it is to get into any classic CAD software.
Yeah, it’s quite bad as well. I’m using Fusion360 now.
I’ve made some great and somewhat complex designs using freecad, it’s certainly capable.
I eventually switched to fusion 360 because of the UI and it’s more easy to find help. And less need to find help
Yep, that’s my experience as well. It works, but man… You’re just wasting time fighting the app instead of designing your models.
For that you can also use Blender
Not really. Blender is NOT a CAD. It doesn’t ensure that your bodies are solid, it doesn’t provide any analysis tools, it doesn’t support working with blueprints/sketches, it’s not parametric, etc. Basically, it doesn’t do anything CAD at all.
Yeah, it’s a great tool for the job. Not as good as Zbrush, but I used it for print prep several times and it just has all the tools you might need.
I wonder, what makes a good CAD system?
I had this idea for a while to build a Frankenstein monster of a 3D software that uses real time graphics and has a multi step build process covering CAD, wireframe manipulation and voxel workflows. If I ever actually make it, your concerns will be heard despite being probably not the best softwsre to do your work in :)
CAD system must be reliable. It is simply unacceptable to have math issues which cause unpredictable geometries.
CAD system should have a good UI. This is a big issue for open source software in general as UI and UX is usually an afterthought.
CAD system should be fast and use hardware acceleration. Running single threaded python scripts on CPU to do complex computations kills the productivity. Designing real life objects is already a mentally taxing task, the whole purpose of CAD is to remove the computational bottleneck of a human.
CAD should be object aware. If I draw two gears and put them next to each other, I should be able to rotate one and see the other moving accordingly.
This is a bare minimum, I’m not even talking about computational modelling, stress testing, etc.
Proper math and an intuitive interface, the opensource alternatives really struggle with some basic functions
Modern day, proper parametric modeling with robust and intuitive constraints.
That is a question too hard to answer in a comment and one that depends on the use case of the software. Few users need the power and features of CATIA or NX, but those who need it can’t accept anything lesser. SolidWorks is a good spot in terms of flexibility and features if it could be easier for the average person to use. You need proper accurate parametric modeling (e.g. a NURBS kernel) for solid models and surfacing. Hearing things like wireframe and voxel indicates it isn’t suitable to me.
I got a maker sub to solidworks. I couldn’t keep up with 360’s oddities and feature changes.
Maybe some day.
Once you get the basics down it’s pretty much all transferable. There are some minor workflow changes, but the functionality isn’t all too different.
Ok, good to know!
Blender ;)
Blender is not CAD.