Expeditions in 3D

We have taken the plunge into maker thinking with a 3D printer in the school library, available for anyone in the school. We have started out as an extracurricular club, with two of us staff-advising (thank you Mr. Le!) with the hope that as the students learn and refine the process they will be able to share tips with others. The students' first project has been to learn to design in a 3D digital program (originally SketchUp, but we may also to try TinkerCad, 123D, 3Dtin, and perhaps others) to create architectural scale models of the library furnishings in order to use them with a blueprint as we redesign the space.

It is a steep learning curve, it all still seems very "beta": the software (either for designing or printing) is not wrinkle-free yet, and not intelligent enough to discover (and announce or fix) gaps or geometric consistency errors in our designs that will goof up in the printer, but we have plowed on with new successes bit by bit.

It is super easy to print from a ready-made and tested design (ex. from Thingiverse, and some of our prints on the left),

but the real challenge is to design something of your own creation and have it successfully transfer and print.

While it is easy to joke about a pricey machine just to make 25 cent plastic doohickeys, that is of course not the point. The students are learning to think in 3D, learning new technologies, to design a 3D object in a 2D digital platform, to transform life size objects into proper architectural scale, to be at the forefront working in a medium that will be normal in a couple of years, to be a maker not just a consumer, to solve problems as they come up, to investigate, research, and test solutions, to be patient, to explore, that failure is part of learning, and success is all the sweeter when you've worked hard for it.

We are using a Makerbot Replicator 2X.

Here are a few of our discoveries:

1. Prints take a LOOOOOOONG time. We're not talking Star Trek replicators (yet!) A small object the size of a quarter may take 12 minutes, but anything larger, like this, can take almost two hours, and more solid objects can take many more hours.
2. You really have to be there and watch over the printing process. Don't just start it and walk away. Prints can fail at any stage and leave you with a big mess (the "blob" being one). Our first blob (which was supposed to be my house) fortunately only wrapped around the extruder metal heated parts so was easy to scrape off, but one poor fellow had to replace his extruder after an overnight blob crept up into the wiring. Blob vs. house: 
   
3. Not everything sticks to the build plate (even with a raft) so you have to stop the build. We will play around with acetone on the kapton surface, heating the build plate a bit more, slowing the print. It would be nice if we could slow the beginning of the print then speed it up once established. The software is not sophisticated enough to allow that yet.
   
4. We had originally purchased the bot with the intention of using PLA plastic instead of ABS plastic for environmental reasons (PLA being a kind of sugar-based plastic, with a lower melt temperature, and fewer fumes). But after trial and error we discovered the 2X really isn't optimized for PLA after all. We were lucky enough to have only opened 2 spools of filament and could return and replace our other spools for ABS.
5. Our initial default prints (pre-designed and loaded in the Makerbot SD card) went fine, though we were disappointed to see little flaws in the final products (some threads of ABS don't seem as tight as they should be, some little blobs at the end of the M Token print, some rough surfaces). Such is the current state of the technology (at our price point).
6. The first student-designed item, a wrench, went quite well, though testing it at different profile resolutions (standard, high, low) gave confusing counter-intuitive results, we'll have to play with that more.
7. Sadly several of the next student-designed prints failed. They were simple objects, bookshelves, but the prints came out badly, with missing shelves, or shelves only partially completed and more. We have no idea why.
   
8. We were recommended (by Makerbot tech support) to run any of our .stl files through cloud.netfabb.com, and that has really helped "clean up" the files to successfully print. But we don't know why the translation from Sketchup to Makerware created such a poor prints in the first place.
     
   
9. We still don't quite get how to work with "shells" and what to do about them. Netfabb consistently reduces the shells in our files, and we can alter the number of shells in Makerware, but haven't learned enough yet why or when to do so nor what effect they will have. We're hoping it cleans up and tightens the prints.
10. I had a terrific Google+ Hangout session last night with two (new to me) twitter colleagues who are also integrating 3D printing with their programs, thanks @TeacherHann and @hdurnin!), it's a great way to pool our experience. We'll share a sample .stl file to print on each of our machines for comparison. Here are a few more tips to check that I learned from them: 
1. be sure the build plate and nozzles are properly aligned and calibrated 
2. check that "preheat the build plate" is in fact checked off in Makerware > make > advanced > temperature)
3. design everything as "components" in SketchUp to avoid gaps and errors
4. we might try cleaning the kapton (with acetone?), and slowing the prints for ones that don't seem to stick well.

We'd love to hear your experiences too, please add to the collective knowledge as we learn to integrate these new technologies!

(images updated April 4 2014)