Slashdot Mirror
FOSS CAD and 3D Modeling Software?
Paul server guy writes "I work at a privately funded, open source, manned, return to the moon mission — Yes really, and Yes, we really are going to put man (and woman) back on the moon. Since we are open source, we want all of our tools to be, too. What we are looking for is CAD software that we can feed into Blender (or the like) to do 3D modeling with. Many of the engineers have tried working with Blender and Art of Illusion, but have not been pleased. They want to just draw the parts, then feed them to the art people who will run them through the 3D modelers for videos, illustrations and such. What is your preference?"
It looks like that campaign was supposed to end last year, on Dec 31st. Why should we waste time answering your questions now, given the seemingly unrealistic goal, when you can't even format a donation box? Or is the a scheme to get money out of stupid geeks by driving traffic to your website?
BRL-CAD is probably the only full fledged package. Link:
http://brlcad.org/
Nerd rage is the funniest rage.
FreeCAD https://sourceforge.net/apps/mediawiki/free-cad/index.php?title=Main_Page has made huge progress recently.
There are several CAD apps out there. It will eventually come down to trying them and personal preference.
Check them out at:
http://sourceforge.net/search/?type_of_search=soft&words=cad
http://freshmeat.net/search?q=cad&submit=Search
I actually do a lot of work with Blender, and it can import a wide variety of formats. I would be very surprised it the CAD programs you are using don't export to at least ONE of the various formats Blender can accept. If you are using AutoCAD I think you have a good shot of an export to a ,dxf or .dwg are probably your best bets.
HeeksCad is making progress. I don't know about feeding your parts into Blender, though. You may be able to shape the project some if you get involved, though. Somebody else mentioned FreeCAD. I've not yet tried to use it.
There's also gcad3d. I found that one to be tough to use, though. For 2D, I don't think that you have many options but qcad.
A Couple of others have already mentioned it, but take a look at BRL-CAD.
It's pretty much the standard. It originated as a US Government backed project and was later open sourced. This is a VERY mature piece of software, unfortunately with a steep learning curve.
Red
In addition to the comments here, you might find useful suggestions in this 2005 and this 2003 Slashdot discussion.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
Doesn't work, and I speak from experience. I have done work for the CSA (Canadian Space Agency) doing similar things and what you are looking for doesn't exist on all sorts of levels.
First, engineering software is a very specialized beast in exactly the wrong way to exist as a FOSS project. For FOSS projects to exist you first need someone who is capable of doing the programing. Then they have to have a need that they want to fulfill. And they can't need it urgently enough that simply going out any buying a working package makes sense. None of this describes the type of people who are trying to design next-generation parts of anything.
It comes down to this: if you have the funding to actually make anything that you plan on designing you have the funding that paying for a high quality industry standard package is peanuts. And if you don't have the funding then it doesn't matter, does it?
It's the same reason that film and television production has always been happy to pick up FOSS solutions that already work but have never particularly cared about developing them. If you are operating at the professional level where you need these tools the cost of them is almost meaningless. It something that always confuses GIMP and Blender supporters who view it as personal software. For them shelling out $5000 a pop for software is such a big deal and they can never understand how the pros don't seem to care.
If you are seriously attempting to design aerospace hardware then you have moved into the realm where these types of software costs are basically meaningless. Suck it up and act like it. If, however, you are actually trying to become a proof-of-concept for FOSS in engineering work then I wish you the best of luck. However, those are two different goals and likely not compatible.
However, beyond the FOSS issue what you are trying to do will not work. Period. These types of software packages are very specialized for specific types of work and beyond a basic level are no good beyond that. 3D modeling software such as Blender or AoI (or Maya or Lightwave or 3DS Max...) are not CAD software. They are not even remotely CAD software. Yes, they appear superficially similar but they are NOT. 3D modeling software is intended to fake the appearance of large numbers of real objects. CAD software is intended to do what is basically visual math. 3D modeling packages have margins of error built in. Many of them will auto-round any equations or numbers entered. As such they are not suitable for real-world design of any complexity.
The types of data that CAD and modeling software generate are also not particularly similar. If you try and just toss engineering blueprints into animation software your artists will not thank you are the end result will look like ass. CAD tends to have too much and the wrong type of detail where animation software is looking for simplification and tends to simplify areas that need detail to look proper once animated. It takes almost more work to clean up a CAD model for animation that it takes to create one from scratch.
You can't really even send a CAD design right to a 3D printer without a significant amount of clean-up unless it was designed with that in mind.
So, to summarize, decide what you want each section of your operation to do and shell out the cash for whatever it takes to let them do it properly. Let everyone worry about their own needs and don't try and meddle by forcing the internal needs of other departments on them. If you were seriously planning on saving costs by not buying professional software for an AEROSPACE project then you are already fucked. You may as well blow all the investor's money on a massive party because it's lost anyways.
Actually, that's probably the easiest bit. The European Space Agency funded a couple of open source CPU designs, so you can get rad-hardened open source SPARC32 CPUs quite easily. They will quite happily run *BSD or older versions of Solaris. Maybe even Linux if the painfully broken MMU handling on SPARC32 has been fixed.
I am TheRaven on Soylent News
I wouldn't exactly call the N-1 "ready to go". Its first flight (Feb. 1969) exploded 69 seconds after liftoff; its second flight (the one in July) blew up 23 seconds after liftoff. Even if it had flown successfully in July, it wouldn't have had anyone on it--not even the Soviets were daft enough to put a crew on a rocket that had only flown once before.* And they certainly wouldn't have been doing anything more than an earth-orbit checkout. It would have taken really good luck on the Soviets' part, plus another Apollo 1-level disaster to NASA, to give the Soviets even a slim chance of putting someone on the moon first. And that's being generous. The N-1 never did work right; something about having 30 engines in the first stage just left too many things to go wrong. All four flights ended in explosions.
*Of course, the US did exactly that 12 years later. Actually, they had a crew on the very first space shuttle launch--no step-by-step or unmanned testing with that one.
The meek may inherit the earth, but the strong shall take the stars.
Supply, meet Demand.
Moon rocks are $2000 a gram because they are astonishly rare, something you'll happily be taking care of for us.
Your income isn't the price now, it's the area under the curve of the price as your 200 kilograms of rock drives the price down.
If I were riding a spacecraft to the moon (or riding any vehicle that could easily kill me), I'd want it designed with the best tools for getting the job done. If that's a closed source tool, buy the closed source tool.
Actually, we have a large number more than that. That page is just the team leaders. Stop wining and come help.
Your Moon, Your Mission, Get involved! http://www.openluna.org
Have you seen the list of Open Engineering Tools over at DevelopSpace (http://wiki.developspace.net/Open_Source_Engineering_Tools)? Not sure if it helps with the immediate need, but it is the most comprehensive list of open source tools for all things engineering I have found.
"To hasten the advancement of humanity into a spacefaring civilization..." -- http://www.mach30.org
Don't get me wrong, Inventor is a pretty decent software package. We used it for quite a while in our SAE Baja team and I believe they are still using it, mostly due to a reluctance to move to UGS NX. However it IS a lowend package compared to some more capable software like UGS NX or CATIA.
One of the largest benefits of working in a software like NX for example is being able to move ideas from the Design to Engineering to Manufacturing phases all in one software package.