Creating a test print object

Andrew Mazzotta approached me with the possibility of creating a test object similar to, well, lots of test objects out there already, with a little indirect marketing towards his 3DHacker website. The thing about creating a test object, particularly a comprehensive one, is you have to decide who your audience is, what do they want to test and what can the tests teach them. For instance testing wall thickness is fine... if you're writing code for a slicer to handle that sort of thing. Otherwise to and end user all they'll get is frustration when they realize there's no setting they can adjust to fix it.

So I created something for the 3D designer that will teach them how to model well for 3D printing as well as little adjustments they can make to improve prints with fine details.

1. The organic area just looks pretty and shows how well your printer can print details like this.
2. The arms of the 3 each extend at different angles to test overhang. The top one extends at 60 degrees which may be a challenge for some to print but a properly calibrated machine should be able to handle it. The middle is at 45 degrees, even a badly calibrated machine should be able to do this, the bottom is 70 degrees, bonus points for making that work but it is expected to fail.
3. The D tests arches. With an arch this big it is likely to fail, but it's a question of how far it gets before failing.
4. The H tests bridging.
5. Each letter in "3DHacker.com" is the minimum safe size for text and 0.1mm higher than the one next to it testing relief and raised printing detail. If printing any less than 0.1mm layer heights it is possible there will not be a difference between too letters. Gaps that develop around the sunken letter can be fixed by reducing the number of shells at the cost of... well, shells.
6. The 3 spikes behind the H should be as tall as the H, however this test is designed to fail. Instead it will show how fine the details can be before your slicer will stop putting them in.
7. The main body and secondary block are joined by a filament diameter tester. If your filament diameter is too low you will not be able to mate the two parts. If it's too high they won't stay mated. 
8. When mated relief lettering on the bottom of the print will demonstrate how well this can work, and the trouble that shells can cause for this.
9. On the top side of the second half the top wall demonstrates vertical holes with triangles, squares, on to circle.
10. In the middle is a shape designed to demonstrate how thin, flat organic shapes can turn out sometimes when designed horizontally like the legs of an octopus or a tail. This part may not be fixable simply by dropping shells, it's just one of those quirks to be aware of.
11. The bottom wall is designed as a fail test where the wall is getting increasingly thinner until it's too thin to print, but at what point will your 3D printer stop printing.

You can download the model on thingiverse right now and try it out, but I;d like some feedback. It's a bit bigger than other tests, is it too big? Does it waste too much space? Did I miss out anything important that a designer would like to know? Is putting test in that are known to fail a good idea? Should I change it in any way?

One last thing, did anyone notice that I forgot to plug the book yesterday?

Creating 3D objects... with your brain

So the project is pretty limited and hasn't actually gotten very far yet, but its very promising: an evolutionary algorithm that uses a brain scanner to determine fitness. I suppose that's the best we can expect since brain interfaces aren't suited for running 123DCreature yet, not without a surgical implant anyways. The hard part is going to be developing a robust enough EA, but that is always the challenge.

MakerWare vs RepG

For my book (yes, got it in early) I had to explore MakerWare for the first time, for the sake of completeness. Before I've never really had any incentive to switch over and I still don't. I do like like that MakerWare allows you to plate multiple objects easily and I'm currently trying to see if it's slicer does a better job of making high quality prints or is faster with which so far has been promising.

Overall I think I can say that MakerWare is it's not better or worse than ReplicatorG, which is kind of disappointing. Yes, there are things that MakerWare does better, but then there are still things that ReplicatorG does better, and that perplexes me. And the new version still doesn't address these things. Yes, improved rafts and supports are good, but I still need to edit a json file to change filament diameter, the silver bullet to good prints. I do not understand why this option wasn't put on the gui, nevermind that there isn't a more appealing way to change custom settings than editing a text file. Yes, I can do it, yes, json is pretty close to human readable, but for a piece of software that is obsessively trying to make 3D printing easy for the masses this seems like a pretty egregious oversight.

Oh, and be careful where you install MakerWare. It wiped out my ReplicatorG when I upgraded MakerWare because I accidentally put them in the same directory a while back. Silly me, I thought they'd use the same convention as everyone else and they don't. MakerWare installs by default in a directory called %programfiles%\makerbot, not %programfiles%\makerbot\makerware like everyone else. And their uninstall doesn't just remove the files it installed, it nukes the whole directory.

What do we get with the new version of MakerWare? Improved rafts, improved supports, and better infill. The hex fill doesn't bug out when set to 100% and that's good. But do we really need cat infill? The feature that would have convinced me to start using Makerware more would be if I could make those fancy support come out of the other extruder so I could start playing with PVA.

As for their new firmware I still do not know why they don't just license Sailfish. What they've added to the new firmware is only a percentage of what sailfish has been doing for a while already. I'm leaving my firmware right where it is.

EDIT: Just discovered something else strange: MakerWare generated files heat up both extruder heads the whole time which causes the unused extruder to leak. This is a big problem so I think I'll be avoiding MakerWare a little longer.

EDIT2: Apparently the uninstall problem and the dual heater problems came from upgrade. The old installer was the naughty one, not the current one. The old custom profiles are causing a conflict when upgraded. But delete those custom settings and that problem goes away. If MakerWare adds filament diameter to their main menu and if they add supports with the other material I will defiantly consider making the switch.

Challenge Accepte..ugh.

DesignMakeTeach issued the challenge and I rose to answer it... sort of.

To be fair, I already knew I was going to have problems with the overhangs. Overhang issue aside this, I think, is the best way to print this object. All overhangs become gradual slopes with few exceptions, and the majority doesn't even need the supports. You don't need to cut the model in half and glue anything together. Just a little supports and your done. And some PLA. Use PLA when using supports.

Am I going to plug my book with every post I do from now on? You bet I will.

I love it when people make cool pictures from my designs

On Thingiverse.

Also, did you know I'm writing a book?

Andrew Mazotta's interview with Shapeways

Andrew Mazotta recently visited Shapeways an had an interview with Bart Veldhuizen who, incidentally, was one of the people who bought Blender and released it free and open source. (By the way, I'm totally writing a book that's all about Blender.) So an all around good guy and awesome to have Andrew interview him.

I like Shapeways. I'm on Shapeways. I really appreciate what they're trying to do. But unfortunately I'm not thrilled with how much Shapeways costs. That's the reason I don't have more things on Shapeways. To bring cost down your things either need to be whispy thin or you need to plate 20 or 100 of them and then sell them yourself on etsy.  (It seems Shapeways has an initial base cost plus material, so to bring the overall cost down spread the initial base cost across more materials.) And then if something is the wrong size or anything there is no return policy because it's not like they can resell the item to someone else. Plus that initial cost deters first time designers from buying their own things so they don't discover that their models are too fragile to work.

Obviously Shapeways has found it's niche with some people willing to spend $15 to make their transforms able to gesture with their fingers. (And I've already lauded modibots.) But since shapeways is too expensive for really solid things it seems it's relegated to the realm of novelty toys or jewelry, which doesn't do much to advance the field of alternative manufacturing.

But like I say, I love what these guys are doing and I wish them well. I just wish they were cheaper too.

Makerbot Project Blueprints on Preorder Now!

It is with great excitement that I can finally announce a project that I've been working on for months!

I've written a book! Makerbot Project Blueprints will pass on my masterful knowledge of making things for 3D printing to you. 9 awesome projects give you the knowledge you need in an exciting and fun way at your pace, as opposed to the dry technical discussion by an expert going too fast you'd get elsewhere. Don't go get help from them, get Makerbot Project Blueprints available on pre-order now at Packt Publishing and later on Amazon, Barnes and Noble, and Safari.

As a special incentive if you order pre-order now you'll get the raw chapters before they're even finished as we work on them so you can start learning now!

Go to http://www.packtpub.com/makerbot-project-blueprints/book to find out more.

Music in the video by Kevin McLeod http://incompetech.com/