This past week, Jasmine and I finished off the last few parts of our Rhino model. Using a combination of curve networks, loft, sweep, revolve, rail revolve, edge srf, cap, and boolean commands, we managed to complete Bambi’s eyes, hair, ears, tail,  arms, and legs.

For the hair, we extracted iso curves from the main massing of the head and traced the general shape using photos of the front, side, and back views.  We then used curve networks, and then extracted iso curves again. We then adjusted the new iso curves until we were fairly satisfied with the shape and used the curve network command again. Jasmine and I found out that different combinations of the iso curves yielded slightly different shapes, some smoother than others. This process took many tries until we each achieved a shape that resembled the smooth surfaces of Bambi’s hair. After that, we offset the hair and lofted between the two surfaces.

In our last blog post, Jasmine mentioned that we were trying to make the heads of our models turn. We decided to simplify the original “knob”-like turning mechanism in the toy to a basic cylinder. Using an offset of 0.5 mm, we created a cylindrical void inside the head which would allow the neck to turn. After discussing with some classmates, Jasmine and I were careful to leave more empty space above the rod in case we wouldn’t be able to remove all the support material after 3D printing.

For the ears, we traced one section to get the profile curve, rail revolved, and lofted. Similar to the process for creating the hair, we extracted iso curves from this new shape, manually adjusted them, and then lofted again to generate a more precise form of the ear.

After making sure that our model was properly joined/booleaned and  free of nasty naked edges, we were ready to prepare our file for 3D printing.

Our first printing job took about 2 minutes… Due to scaling problems, we ended up printing our model 10 times too small. We soon realized that Tinkerine had taken our model (drawn in centimetres) and changed the units to millimetres. We made the adjustments and proceeded to print again.

We had hoped to print our model in white, but due to temperature issues, and the spool tangling within itself (which happens when the spool is too large to fit on the back of the printer), our second print was a failure to say the least.

For 10 hours, we watched our print job while manually untangling the filament as best as we could.

An open window and other temperature issues caused breaks and gaping holes in our model.  Another lesson we learned was not to leave a print job unattended. After a frustrating 10 hours, we went home for the last 5 hours of our print jobs for some much needed sleep. Unfortunately, when we arrived at school the next morning,  just when we thought things couldn’t get any worse, we discovered the following scene.

Without anybody to untangle and help move the spool along, the filament seemed to have stopped feeding into the machine, leaving our model incomplete.

Our third attempt was much more successful. This time, we decided to split our files onto two separate printers and opted for the clear filament. This time, there were no temperature issues and the print job went smoothly and fairly quickly, taking just under 3 hours. 

Overall, Jasmine and I are happy with the results. The neck connection worked out well, and the heads were able to turn! It was indeed a happy ending to what seemed like a dreadful beginning.

This week I put the finishing touches on my rhino model. My greatest challenge was learning some of the ways in which I can sweep over a surface and maintain a solid polysurface, as I did a lot of sweeping for the arms and legs. For the arms, I used sweep1, sweeping into the body to insure a strong connection. I closed the shoulder side using patch and the hand side using a mesh that was sculpted into the shape of a hand. For the fingers, I turned on control points on the mesh and sculpted the fingers using the physical model as a reference for estimating form.

For the legs, I used sweep2, as the rails were sightly more coplex. I used several section lines to create the bulge of the sock before the shoe. A patch seals the connection between leg and body, and a simple edge extrusion insures a solid connection between the sock and the shoe.

While modelling the shoe I learned that it is much easier to create objects with vertices orientated along the X and Y axis, then rotate after the modelling is complete. This way, I was able to perform all of the snap commands necessary to form my shoe. I used rail revolve, extrude line, join and boolean union to make the leg into a solid polysurface. I also used sweep to create the backpack straps around the arms, and rail revolve and reflect to form the backpack.

Finally my model was a closed, complete polysurface ready to export to the 3d printer. In tinkering, I experimented with infill and resolution settings until I finally got a perfect print. I ended up printing the head separate from the body, and attaching after the fact. This reduced both time and structure material required.

In a new Rhino file, I exploded my model and re-organized the layers based on the colour of my model. In keyshot, I exported the file and got a nice, solid render with accurate colouring.

Unfortunately, instead of modeling our bus as a solid, Shizuka and I were modeling the shape using planes. We only realized that this was the wrong way of modeling once the Boolean command was not performing.

And so, we had to join, cap and extrude our planes to turn them into solids.

Once all the objects were converted into solids, it was like magic! All the functions started working flawlessly!

Finishing up the body – continuing to use my strategy of breaking up surfaces and segments then joining things together later as Hulk is not symmetrical at all points.

Lots of things continue to need refinement…

Successfully using Blend Surface after Tam’s recommendation!

I definitely put off the fists as my last task as it was a hard one to conquer. Edward suggested sharing pieces among group members so the blue was created by Christina (below). Given the style of Hulk however, I thought it didn’t quite match which is the difficult part with having different techniques in Rhino I learned.

After an enlightening power-shower, I realized that given the scope of the project and seeing the 3D models already printed by peers… I realized many details are lost in print. I used this to my advantage and any harsh curves created from using the Pipe + round curve tool would actually work for me.

Last to take on – the hair. This was definitely the most frustrating part of the whole project because I was so close to finishing but the technique that my group members/partner used… just wasn’t working for me. They projected the curves traced of his hair in front and right view onto their heads, split, offset then finally capped it.

Mine was not splitting unfortunately so I had to seek other strategies…

Finally getting close to something… this was the best I could get (although I wasn’t happy with the result). Unfortunately, after my first print – the hair caused a lot of issues.

The past 2 weeks were a struggle as my workflow was put to a rather large halt when “refreshing” or re-PictureFraming the photos I was creating the Hulk from no longer worked.

I continued to see this everytime I restarted, opened, deleted my file. It was frustrating.

My prayers were finally answered however, after much research and even seeking out help from my computer-savvy boyfriend. After already almost concluding I needed a new computer (we had even gone as far as looking into prices of PCs I should seriously consider)…. we finally came upon a miraculous post that was at the back of my boyfriend’s mind… something with the display settings.

All it took was: Rhino Options > View > OpenGL > clicking off “Use accelerated hardware modes” and voila!!!

I was finally back on track and able to continue progress. Next up: The head. I initially left the head for after the body since I thought the body would be harder to do with all the parts… I was surprisingly wrong. I tried several techniques/tools/ideas but for some reason, I was having a really hard time.

It took several frustrated attempts to try and get the best and most accurate head. In hindsight, simplicity is key but information helps. By  information, I mean the information you provide the tools so they don’t average things out but to not overcomplicate it (I made the mistake of putting his head at an angle from the beginning – rookie mistake).