Adventures in Resin Printing

I have a 3D printer that I’ve been building for several years now – For a while I was working on it quite regularly and making all kinds of progress, but at a certain point family obligations started to demand more of my time and so for a period of about 3 years I didn’t work on the machine at all. I was meaning to get back to it for a while but then the COVID-19 pandemic hit, and for a while I couldn’t go to the makerspace where I store and work on the project at all. Between the frustrations I faced in the project, my inability to go work on it, and a general frustration in the modeling hobby and a desire to become a productive modeler again, I decided to buy a 3D printer that I could use to print models now, and which could complement the other printer later.
After a fair bit of deliberation I decided on a low-cost resin printer. The great thing about resin printers (even low-cost ones) is that they can produce very smooth and finely-detailed prints. I figured that once my other printer (the typical plastic-extruded-through-a-hot-nozzle FDM type) was done a resin printer would still be useful for common needs like hands, as most mecha kits’ included hands are garbage. I figured I could take the hands from my Zaku Kai model, which I created in Blender back in 2007 or so, and have a good starting point for a first print.
Unfortunately, I made those hands back in 2007 or so. They were terrible. I had to start over. It took me a while but I finally got the new hands to a state where I felt ready to print them.


I really underestimated at first the work that would be needed to make the hands in the Blender model look good. The existing hands really were among the earliest things I made in Blender, and frankly, they were terrible.


The new effort worked directly from the reference art as a starting point and achieved a much more natural-looking expression early on as a result. The work went through a number of iterations as I tried to fit the reference art better, make it fit the Zaku better, and make it function better when posed using the armature. Finally I had something I felt was ready for printing – it really felt great to finally set up the machine and get it going. Though the first few attempts to print the hand completely failed because the supports weren’t strong enough. Over a few iterations I got better at setting up the print with strong enough supports to get successful prints.


Then, having gotten a few successful prints of the hand, I thought, why not print more parts? So I went as far as printing the whole left arm of the Zaku in 1:60 scale.

Ultimately my goal isn’t to give up on traditional model-making methods – but achieving the kind of results I want has always been very time-consuming and it’s harder for me to dedicate that time to model-making these days. I’d like 3D printing to help me achieve those same results more quickly. Both the Zaku Kai Blender model and the new hand model are available in my Git repository. The models in the “CG_Models” repository are provided under the Creative Commons “Attribution-ShareAlike” license. This means you can use the models pretty much however you like, as long as you properly acknowledge my role in creating the model, and if you modify my work, you should share the modifications under the same terms. Both models are provided in the Blender format which was used to author them – while this isn’t the most convenient for 3D printing, it does make it easy to do things like change the pose before printing.

6 Comment(s)

  1. That arm looks so nice, excellent work… is it articulated too?

    attemborough | 2021-12-04 | Reply

  2. This print was not articulated, however I have been working to update the design to include working joints. For a little while I had thought maybe I’d just press on with the static-pose version and crank out the whole Zaku in 1:60 scale, but after a while I decided I’d really rather build in the joints. I’ve since bought a whole bunch of 5mm polycaps to use for this build.

    One thing that came up while adding polycaps to the design is that my printer was actually a little bit miscalibrated: So the shoulder joint, for instance, it’s basically spherical, but the poseable joint is cut into three slices, and the middle one turns independently of the front and back parts of it. But because of the miscalibration, those circular parts would wind up a little bit wider than they were tall, so when I printed a poseable version of that, the three parts of the hinge would match when it was in one position, but be visibly off from each other when the joint was rotated. I went through a dozen or so calibration prints trying to dial in the calibration a bit better, hopefully the next time I print the shoulder joint it will line up better.

    I’ve also worked out a method to make the hands poseable as well: the knuckles are spherical, and they’re about the right size for one of the smaller versions of the Revoltech joints (or something similar, I forget). So I’ll probably build a poseable version of the hand at some point to try it out. (That said, IMO poseable hands on a model are mostly a wasted effort IMO, you wind up with a model that you have to carefully re-pose every time you put it on display somewhere, just too much trouble. But I do think it’ll be fun to make the poseable hands.)

    I’m getting a bit stuck on some of the particulars of the design: Like the shoulder pauldron is really big: As it should be IMO, no question. But it does mean there’s a lot of empty space in there, and I feel like I should do something about that. I just haven’t figured out what, yet…

    tetsujin | 2021-12-20 | Reply

  3. I’m only here to comment on that wonderful Timex software video you made a few years ago.. (almost a decade now, my God how time flies).

    So glad to see you still manage your blog.

    Best wishes,
    Ari P.
    San Diego, CA

    Ari | 2021-12-20 | Reply

  4. Ha, “almost a decade” is also about how long it is between updates to the site.. XD

    I’m glad you enjoyed my video on the Timex software. I still use the software all the time. The theory was that the countdown software would help me stay aware of long-term goals and deadlines, and get more things done… In practice it really hasn’t done that at all, but I still find it very useful.

    I’d like to make more videos, probably model content, but as it is I’m not even getting any model work done so there’s nothing to shoot.

    tetsujin | 2021-12-20 | Reply

  5. There might be nothing to add there, the lineart doesn’t seem to indicate anything being present. If I had to guess, maybe you could add something to do with the holes the pauldron has next to its front and back spikes.
    The red coloring seems to indicate they’re verniers but it’s a bit odd that they’re only present on one shoulder. If I were allowed to speculate… perhaps they’re supposed to assist shoulder bashing, like how the Rick Dom II uses its forearm thrusters in its Yuji Kaida boxart. It seems kinda bizarre that Zeon would have faith in those sorts of tactics so late into the OYW, but that’s the best explanation I could think of.

    attemborough | 2022-01-12 | Reply

  6. True, I don’t think the lineart indicates anything filling space in the shoulder pauldron. Still, when dealing with that part – I feel like there ought to be something in there.

    Also I see areas like this as opportunities to embellish the design. Anime robot designs tend to be kind of simple, and generally I think that’s a good thing, but I feel like when you look closer at a model, ideally there should be little touches to suggest there’s a little more to it than the blocky structure of a cel-animated robot.

    As for the thrusters… Yeah the asymmetry doesn’t make a lot of sense. Generally I think it’s not really worth trying to rationalize things like that, it just kind of is what it is. It can be worth coming up with an explanation if that’s something that leads to ideas about how to make the model better.

    I guess if I were to rationalize the thrusters on the shoulder pauldron, I’d say they’re mostly used for rotating the body (yaw) – and with the pauldron being one of the more distant points (laterally) from the machine’s center of mass, thrusters there can exert more torque even if they’re unbalanced by thrusters on the other side. But then the thrusters should be oriented straight forward and backward – which doesn’t look as good IMO as having them perpendicular to the surface of the pauldron. So I’d leave them as is. :)

    tetsujin | 2022-01-26 | Reply

Post a Comment