I play in the higher end cad systems, but I have downloaded a while back Autodesk's 123D. Last I checked it was still free, and it had reasonable capabilities. There is one open source cad that I remember looking at a while back, but it was the worst cad experience I have ever had (CATIA has 100X better user interface, and that is saying a lot). There is also Sketch-up, but I do not have experience with that one. I think that there are some mesh based, open source modelers, that while they will not make a lot of friends in the MCAD world, they will do some very organic shapes. I hear that the DesignSpark group will be coming out with a 3D modeler here soon.
In the professional world, some of the mid end systems ($2-8k) would be Rhino, SolidWorks, and Solid Edge. There are a few others as well such as Keycreator, but all of these are rather expensive.
What types of shapes are you trying to design, and what is your end desire with the products? These will also help you to narrow down your selections.
@aeroengineer: What types of shapes are you trying to design, and what is your end desire with the products?
I can think of all sorts of smallish things that it would have been great to build for hobby projects in the past.
For the moment, did you see my blog Are We Losing the Secrets of the Masters? In thsi blog I mentioned an amazing book "Five Hundred and Seven Mechanical Movements" -- I'm thinking of starting off by replicating some of the gears used to implement these movements to play with them.
Sorry, I did not post any links, I just has a small break between projects at work. I did see your blog post that you mentioned, and for what it is worth, I also agree with you that many have lost some of the basic skills. I am not sure, though, if it is that there are less people with those skills, or if it is just that there are more people that are button pushers, and hence by percentage there are fewer that have those skills compared to the whole.
I had the opportunity to get my start doing aircraft restoration when I was 12. This helped pay my way through engineering school. As of about 5 years ago, I branched out and started to learn electronics, so I commend you for branching out to the mechanical side of things. There are a lot of fun things that can be done. I might end up starting a personal blog, I am not sure, but these are some of the things that I want to write about. The integration of electronics with mechanical hardware. I think that most consider it a bit of voodoo magic, but in reality it is no more challenging than designing the electronics. The biggest issue with the mechanical side of things is that tooling for production is rather expensive. It is a whole different though process, and hence the greatness of the expansion of 3D printers.
I think, though that Autodesk 123D will meet the immediate needs that you have. I found it rather intuitive for simple stuff, but I may not be a good benchmark for this.
@aeroengineer: I think, though that Autodesk 123D will meet the immediate needs that you have. I found it rather intuitive for simple stuff, but I may not be a good benchmark for this.
This does look interesting -- I'm going to be busy for the next few days, so I'll wait to see if anyone else has any input here, then I'll decide which package to play with and download it and install it and start preparing for the arrival of my 3D printer (I smile whenever I say that :-)
@aeroengineer: The book that you mentioned, is very similar to a set of books that I loved to check out from my high school library. They were called Ingenious Mechanisms for Designers and Inventors. I would spend hours going through those books.
It looks liek they have folks who love them and folks who hate them :-) Plus they are mega expensive ... but maybe after I've finished reading my "507 Mechanical Movements" I'll be earger to read more...
While blender is free and impressive in its ability to pull off some of the same features as bigger software by bigger companies, it isn't great for designing for 3d printing. CAD software would be better for this.
You can do it in blender, but measurement of items after their initial construction is done by eye. This is less than optimal.
Yes, blender is, for lack of a better term, a competitor to 3ds max, maya, softimage and the like. These are all effects packages and used for movies/gaming. While this can be nice (it is where I learned and what I'm comfortable with), it is lacking in precision for drafting.
What you probably want for 3d printing is a drafting package for precision positioning and alignment of assets... like gear teath.
Sometimes you want to model something organic or fun though and that can be incredibly difficult in CAD. Both tools are nice to have.
I just had my first 3D model printed. It went very well. I found a free 3D CAD package I love, OpenSCAD.
It is not what you are probably expecting. It has no point and click interface to draw models. What it does offer is a consice language to describe your model in code. You develop and modify the model with a text editor. It also has a 3D model viewer so you can see the results and spin it around to your heart's content.
If you code the model properly you can use variables and change a dimension of many features by changing a single value.
With 100 lines of code my model was ready. No fussing with a mouse to line everything up. It generates STL for the printer.
It's definately not for everyone, but if you're a code / script developer it's very natural. I think it is much like coding VHDL instead of drawing schematics.
One of the things that a lot of people do not know is that most mid level and above MCAD packages offer scripting. I know that both SolidWorks and CATIA offer VB scripting. I also know that both can be made extensible with custom programed interfaces that use their API's.
The reason most people do not do this is that for most shapes, it is much easier to use hot keys and a mouse. The other reason is that most mechanical parts will change in their lifetime. The GUI allows for fast modification of these parts.
This becomes readily apparent when you start to work with complex curvature. Most of these shapes are defined by higher order math, and it would be a nightmare to try and program that into a script. Even the programs do not always get it correct. I was helping a friend last night attempt to fix the leading edge of a wing model done in a MCAD program. Despite the inputs being correct, that particular MCAD package could not handle that shape.
Hence many companies will have two or three cad packages to be able to take care of issues like this. Each place that I have worked at has run no less than three MCAD packages. This is excluding modelers that are used for finite element programs.
As has been said below, each MCAD package will have its strengths and weaknesses. It is best to try a few and see which feels right. It will take a while to learn its limitations. At that point you might then start looking for a secondary package to back up the first one.
@Aeroengineer: At that point you might then start looking for a secondary package to back up the first one.
Oh joy! :-) All I want is a simple life ... initially I need a package that will allow me to capture things like gears (not just round ones -- ones with unusual shapes) ... maybe later I will move to a package that supports more "organic" shapes.
The trick is to find something cheap (free) that is easy to learn, reasonably sophisticated, and can drive my 3D printer (I still smile when I say/type "my 3D printer")
Do not worry, your primary MCAD package should be able to handle 95% of what you want to do. Once you get into the more nuanced issues, is where you start wanting to have a secondary MCAD system. Most of my experience comes from designing wind tunnel models and composite rotor blades. These are areas where the small stuff counts. Geometric deviations of ±.002" on a wind tunnel mode was considered a normal profile tolerance. We once lost many thousands of dollars because a MCAD program interpreted the geometry improperly on import and caused things to be misshapen by ±.020.
I have not "used" blender, per se, just checked/tested to see that it *would* work. It is pretty much like any other 3D software ... in that it requires literally a whole new dimension of thought. How good of an artist are you? How good of a draftsman? If you can draw standard 2d projections of a 3D object, then you should be capable. That said, Blender looked to be more capable than I had skills to do. So, IM(NotSo)HO, it seemed decently straight-forward, but with plenty of upside potential. As a certified (AARP member) geeser, I can relate to feeling the challenge, but engineers have always had to learn new tools, new software, new approaches, so this would be just another chapter and verse. AutoCAD (and really, ANY other 3d software) is about the same level of capability and challenge, as learning Verilog, "C", FPGA design, Chip design, etc. It is mostly a matter of putting in the effort to learn.
Blender's big advantage, is that one can import the main 3D formats, so just about everything "out there" that is available, can be *yours*. With the total investment cost being "merely" time and energy, it beats everything else available. I strongly believe that using free software, when available, is a benefit. The community of users and developers tends to be a LOT more helpful, even considering other types of professional level --- versus --- free software I have used in the past.
As for the actual machine to 3D print, it seems nothing magical past the heater element. Everything else amounts to a well-controlled 3D CNC, of which there are many examples and designs. Remove the heater element, mount a spiral saw to this platform, and 3D machining (to a degree) can be accomplished. Additionally, a double-sided PCB could be "milled" for the routing, holes could be drilled, solder paste added, (should one push the platform), all with little modification to anything but the "head" assembly. I could see buying the heater element, but the rest should be a DIY design capable project.
IIRC, there are some 3D printer designs supported by a hobbiest community, where existing owners/users will print the parts needed to build another unit (short of the heater element, and motors).
Nothing but money ... (stands between me and great wealth)!
This is just a short list of free ones. Each has its pros and cons, and each has a different approach to creation. There are a few more that I didn't list because I can't quite remember their names right now.
I could actually go in depth about the style differences in each one and who would prefer to use them. In one of my past lives I tought 3d modeling and animation.
FreeCAD (From Germany) is also not too bad -- I have a friend from college that has done some really impressive stuff with Blender. The AutoCAD 123D online will allow one to have parts made, and shipped to you (Both Metal via CNC, and 3D Printer - Resin)
I dabble with them all at work, but at home I prefer Sketchup with the third party .stl exporter added in. Sketchup can NOT import .stl files though (or at least well ?) - for that I use NetFabb to slice, and/or rotate and/or scale donated .stl designs. This is more art than engineering though using Sketchup and NetFabb (compared with Solidworks (3D), Autocad (2D) and Inventor(3D)). Makerware (from Makerbot) is a great way to import multiple .stl's, and independently scale, rotate or move each, then export back out as a single .stl (or two different files if doing dualstrusion). This free package also exports to their proprietary editable format (.thing) and machine code .x3g (gcode choice no longer works), but don't let that stop you from trying it out ! (since it also saves out graphicly tweaked .stl's as another .stl) Best, Davy
@Davy.Baker: I dabble with them all at work, but at home I prefer Sketchup with the third party .stl exporter added in.
Thanks for the great advice -- I must admit that I'm leaning toward SketchUp based on other things I've heard -- is there a website for the .STL exporter or do you get that from the SketchUp site also?
I did my first actual 3D parts a couple years ago and I used regular SketchUp version 8. I believe I used the su2stl.rb plugin which USUALLY works OK, there's a webpage about it if you have problems. I actuallly got usable hardware first time using Redeye as the service bureau. I would however give two pieces of advice. One has to do with scaling, you need to bear in mind that SketchUp is only really written to work well for the scaling architects would use, the tools DON'T work particularly well for small things like enclosures so you'll need to scale up the drawing by 100x or even 1000x while you're editing it so at the very LAST step you scale it down before export (which of course works flawlessly). The other is you need to download a REALLY GOOD .stl file viewer and learn how to use it. The reason is if you "drag" one part to be "adjacent to" another, it may LOOK LIKE the two components are "touching" then if you have it made you find they come out as separate parts, not such a great result if you're paying hundreds of $ for a prototype! (I guess you're thinking about having your own machine so failure comes relatively cheap but our time and materials are worth SOMETHING.) It was actually the manager at Redeye who spotted and fixed the problem for me, again showing there IS value in working with folks who know what they're doing! I would also wholeheartedly recommend the same book I bought which appears at the following link to help you understand the various RP processes and services available:
Max, I'm sure the first folks who heard Colonel Powell's lectures about the discovery of the Grand Canyon said the same thing (recall he lost most of the members of his group to various incidents). There's certain penalties that we "explorers" or early adopters have to be able to deal with, then again having conquered them we wear them as badges of honor. In the final analysis I didn't really have too much trouble thinking of inches as mils but yes you do have to take it into account.
I have heard good things about Rhino3D but I also recall hearing that's about $1K, you pay your money and you pick your poison. I do know I got SketchUp to deliver a functional two-piece gauge enclosure with a twist-lock front. Nothing new is ever easy.
Blog Doing Math in FPGAs Tom Burke 2 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...