Articulated V1 Figure

After seeing a lot of people create their own posable 3D-printed figures (particularly with Dummy 13, which can be found here), I wondered if it would be possible to design and create my own figure. I thought that I would try to make my own articulated model of V1 from ULTRAKILL, as I'm a big fan of the game and I think having my own figure of the character would be pretty neat :)

For this project, I used:

PLA Filament + Creality K1 3D printer (for prototyping)

Fusion360 (for designing elaborate joints)

Blender (for designing limbs and overall shape)

Super glue (for any immediate repairs)

And was planning on using:

ABS-Like resin and FormLabs SLA printer

Before I actually started designing anything, I wanted to get a feel as to what the joints would look like and how they'd fit together in a way that would be clean and functional, while also maintaining the original design.

I already decided beforehand that I would print out the model fully in PLA at first, just to make sure that everything was functional since resin is pretty expensive.

I also acknowledged that my expertise on Fusion360 as a 3D modelling software instead of a manufacturing software is somewhat limited. I don't know how to make all of the intricate parts for V1 just using Fusion. However, I do know the basics for 3D design in Blender, and I felt confident in at least knowing how to get rough forms for each part that would be to scale to each other. Because of this, I decided that all of the design for the limbs and parts themselves would be made entirely on Blender, and then they would be imported into Fusion where I could make accurate joints for each limb while also refining any parts that were too messy or complex. I'm aware that it's possible for .stl files to be made from Blender models, but the printing of many of these parts would be in my school, which doesn't allow Blender on its computers. I also don't know how to make parts suitable for manufacturing (parts w/ accurate measurements and assemblies) on Blender, so the usage of both programs was the best option for me.

The designing aspect of the project is probably what took me the longest in this project. Although part of it was definitely because I had to test out how to properly insert files from Blender into Fusion360 and how to work with meshes, another part of it was that 3D designing just.. takes a long time. All of this section was done in Blender.

First, I had to get a reliable reference sheet. Thankfully, there's plenty of drawn reference sheets online, and I ended up using the reference sheet that can be found on the ULTRAKILL Wiki page for V1.

From there, it was modelling the parts themselves while also keeping in mind that these parts had to be able to go together somehow. Multiple "layers" of components had to go together so the part would be able to move and also have a form that looked close to the original model. The assembly of the parts themselves and assuring that they'd all fit was an issue that I would address in Fusion; all I really had to do now was create a rough model that could be adjusted.

The layers of the parts themselves were mainly comprised of 2-3 sections: the base or skeleton, the frame, and the outside plating that would cover the former two sections. (ex. Photos 1, 2, 3)

After completing this process for the whole figure, the parts were exported individually as .obj files and imported into one Fusion file as a mesh.

After importing the limbs into Blender, the next part was to find out how to connect them. As mentioned previously, I didn't have a lot of knowledge on how to create accurate measurements for joints on Blender, so this step was done almost exclusively on Fusion360. This meant that all of the files from Blender had to be exported as .obj files, and imported as meshes into Fusion. A lot of them needed repairs in some way, but luckily using the Wrap function fixed them up well enough for the geometry to not be too complicated. From there I converted the mesh into a body and began to work on the joints.

The only joints that I really had experience with or knowledge on that would've worked with this model were swivel and ball joints, so it was joint determining which type would work best for each connection from there. The only prominent ball joints in the model were the head-to-neck (attached), the shoulders, the hips, the wrists, and the waist-to-frame connections. Everything else worked better as a swivel joint, including the elbows, knees (attached), and ankles, since they didn't really need more than two axes of rotation.

Initially, I added caps for the swivel joints to keep them in place, but since the print itself was so small the cap was too fragile to actually do anything. Decreasing the diameter of the rotating limb by around a hundredth of an inch and adding a chamfer proved to be effective in allowing mobility and still being secure enough where having the model fall apart didn't seem like much of an issue.

My original intention was to print out the final project with resin, so the prototyping was done with PLA.

The first part that I printed out was the left arm and it did not turn out that well (Image 3). The ball joint for the elbow wasn't really working, as the movement was limited and the method I used to convert the original mesh into a body was making the finished print a little uneven. The choice I made to separate the shoulder from the bicep made the parts unable to connect. With all of these issues, I had to fully remodel the left arm by replacing the ball joint used for the elbow with a swivel joint, which turned out more successful.

The second part was a more successful print. The parts for the chest were all able to be connected, and only minor edits had to be made to accommodate the arms and the neck. The first test for the leg also went relatively well, but it also had to have minor adjustments to make sure that the limbs would be fixed in place (Image 1 + 2).

I would have to later edit the leg pieces (and many other joints) to still be sturdy after shrinking the model to around 80% of the original design. I made some edits to the front part of the chest so it would just be all one piece attaching to the torso frame (Image 4). I also decided that in order to save some time, I'd just super glue some of the parts to other mobile ones. This only had to occur for the waist cover and the lower part of the neck (Image 5).

What was intended to be my last print for prototyping didn't turn out as well as I hoped it would. I was using a friend's printer at this point (thank you Camden) since the two functional PLA printers in the shop were either in use for other projects or out of commission, and for the first prints, we still used Creality Hyper PLA. Unfortunately, a major print got interrupted and wasted most of the filament that was available. Because of this, the "final" prototype had to be printed with Silk PLA, which is not built for parts that are going to be put under pressure. A majority of the parts broke in some way, including the ankle joints, the neck joint, and some of the parts for the left arm (Image 6). Although a final assembly wasn't possible within this time frame, I'm almost certain that if a different filament type was available I'd have more success. One of the more successful assemblies were the legs, and can be seen in images 7 through 12.

Although some of the components of the prototype were successful and functional, the entire assembly wasn't able to be completed due to material and time constraints. However, considering that most of the parts themselves would have been functional if not for these limitations, most of the steps that I had planned for after this stage wouldn't need much editing.

If I had more time, I would have tried to done a full print of the model with resin. Using the SLA FormLabs printer that we have in the shop, I would have done a small-scale print at first to see what adjustments (if any) I'd have to make to the model in order for it to have the highest quality print. I also would have looked into making the layer heights for the model smaller than what the PLA model had, since resin prints are able to have a higher quality finish compared to FDM printing.

From there, I would have assembled and painted both of the models, using spray paint for the PLA print and acrylic paint for the resin print, followed with a sealer on both to keep the paint from flaking.