First week of 3D printing with the MakerBot Replicator

Posted on August 13, 2012 in 3D printing

On monday UNK received it’s first MakerBot Replicator, which I couldn’t wait to unbox and calibrate the day it arrived. Within about a half hour I got the Replicator printing a calibration cube, which wasn’t perfect, but not a catastrophic failure either. Here are a few high-res images of the very first thing ever to come off the new Replicator:

Problems, accidents and lessons learned

All week long I’ve been printing out models and trying to figure out exactly what works, what doesn’t and why. Because the MakerBot Replicator is shipped pre-calibrated, there isn’t much fussing to be done, but I think it is always a good idea to take the time to make small steps whenever you are trying something new. If you take too big of a leap early on, you may spend more time than you’d like in frustration and trying to scale back until you’ve found the problem. While I’ve been printing simpler objects, Mark Hartman has been throwing the most complicated models he can find at it just to see what happens. Surprisingly, they have been coming out pretty well! Here are a few of the things I’ve noticed and some possible solutions:

  1. Curling and peeling away from the bed
    Every once in a while, especially with hollow, thin-walled models, prints will tend to literally curl up a tiny amount and peel away from the surface. Usually this results in catastrophic failure as the print head slams into it over and over, eventually picking the whole damn thing up and spitting out useless fibers. Very entertaining, but bad! Apparently, this is a very common beginner problem and can be due to a number of things: the temperature of the nozzle being not quite perfect, the plastic not cooling down fast enough, the extruder moving too fast and more. Near the end of the week this problem didn’t occur as much, so I’m curious if its possible that the Kapton tape or heated build platform needed to be “broken in” in some way. Prints are sticking to the surface much beter now than they were on day one!
  2. Small amount of contraction on the first couple of layers
    Many of the objects I have printed so far appear to have a small amount of contraction on their first few layers (most noticeable with a uniform object like the calibration cube). I don’t have a solution for this just yet, but I have a feeling that if I raised the build platform temperature by a couple of degrees the problem may go away. We’ll see!
  3. Objects sometimes print missing their top and bottom layers
    Starting sometime around Wednesday, the prints I was making were missing their top and bottom layers, resulting in only the walls of the objects being printed! While this was kind of cool in a way, it wasn’t supposed to happen. One day, objects were printing out great (like this “Barrel of Monkeys” monkey), and since then, I have not been able to print a solid calibration cube or whistle, and all of the models Mark has printed are completely hollow! I re-loaded the Replicator defaults in ReplicatorG and even re-installed ReplicatorG a couple of times, but it wasn’t helping. Then a user on the MakerBot Google group mentioned that there is another settings reset function located in ReplicatorG’s File menu (which I assumed was for the basic program settings like file paths and font colors). I gave it a shot and lo-and-behold, prints are coming out great again!
  4. Gaps occurring at the tops and bottoms of curves
    While printing more complex shapes full of curves, we noticed that the tops and bottoms of these curves had some big gaps. I had a hunch that this was due to the default layer height being too large to accommodate some of the more intricate details that the curves need (the default layer height is 0.27mm). After printing these objects the default layer height, we bumped it down to 0.1mm and tried again. While this did help a little, it didn’t solve the issue. Right now, it’s seeming like these gaps are inevitable and may just be an unavoidable artifact of our printer.

Photos of prints from this week

A closer look at models and costs

One of the most common questions that new or prospective 3D printer owners and operators have is, “How much does it cost?” Sometimes the answers are pretty unsatisfying, which I think is partly because without being able to feel, measure and compare things physically, it’s hard to get any idea of what the correct answers are. So I went ahead and weighed some of the models we’ve printed this week, so I can give you an idea of what kinds of costs and investments have been involved so far. I hope this at least helps a little bit, but keep in mind, seeing is believing! While initial investment costs in the equipment may seem large, each actual print is very, very cheap.

Keep in mind, the standard size of all plastic spools is 1kg. We used a 1kg spool of natural ABS purchased from MakerBot Industries for all of these prints, which costs $46.

IMPORTANT NOTE! We discovered that, somehow, infill was completely turned off for these prints, meaning that they are completely hollow. This means that they printed relatively quickly and used little plastic, but some are quite fragile. Because of this, the numbers below are, although accurate, not representative of what a typical print will look like!

Human skull
Weight: 18g
Size: ~3.2″ x 4.4″ x 3.6″
Cost: 83 cents
% of spool: 0.018%
Possible copies from 1kg spool: 50+
Relevant settings: 0.27mm layer height, 1 shell, no fill (accidental)

Klein bottlehttp://www.thingiverse.com/thing:22734
Weight: 13g
Size: 2.3″ x 4.49″
Cost: 60 cents
% of spool: 0.013%
Possible copies from 1kg spool: 70+
Relevant settings: 0.27mm layer height, 1 shell, 10% infill

Allosaurus clawhttp://www.thingiverse.com/thing:18665
Weight: 3g
Size: 1.9″ x 0.66″ x 3.7″
Cost: 14 cents
% of spool: 0.003%
Possible copies from spool: 330+!
Relevant settings: 0.1mm layer height, 1 shell, 10% infill

Plans for next week

This coming week (Aug 13th-17th) is the week before the fall semester starts up, so a lot of professors and administrative peeps are scrambling to get the facilities tidied up and curriculum underway. In fact, I still need to sign up for another class, but I’ll get it figured out one way or another. This week will be a good opportunity to really use the printer to push out some epic models with relatively fewer errors. In fact, Mark just finished an awesome 17+ hour print of a detailed human skull, and I’m typing this as I’m sitting next the bot as it sings me the song of its people and prints me a whistle. We’re planning to print some neat gears, classical art figures and goofy stuff we can find on Thingiverse, at least until we have to actually get some work done 😉 Once we get all this playful goofing around out of our systems, we’ll start working on some original artsy work.

I will continue posting photos of every print on my Flickr account, and will also be posting across other social networks like G+, Facebook and Twitter (all the links can be found at top of the site). Get in touch if you’d like to learn more, or try to print something cool out!