Too long, didn't read version of this page:

The main bit of professional 3D modelling I've done is modelling the interiror of the KC-135 cockpit for usage in a VR orientation. This wasn't a flight simulator by any means; its role was to get pilots used to the interior. I did this as part of a four person team for a senior group project. I've known how to 3D model since 2011, but this was really my first major project; everything prior was just little bits and bobs of things that I wanted to 3D print. The process itself started by us taking pictures of the interior of the cockpit. Initially (and somewhat naively), our assumption was that these measurements would be enough to base the cockpit model on.

Naturally it is worth stating that the pictures I have taken aren't such low resolution, but it seems to me to be a better idea to just have it be a lower resolution image just to prevent claims of shenanigans. As mentioned earlier, one of the things that I had to seriously manage with this project was the tradeoff between accuracy and performance. A really fancy model may have been nice, but given that the goal of this project was for mass deployment, mind needed to be paid towards it running on perhaps not the most powerful of computers. With the measurements, before the first of two semesters was done on this project, I was able to get the flight yoke, the chair, and a very rudimentary interior of the cockpit done (I didn't keep this last model but honestly it wasn't really worth keeping anyways).

We had a realization pretty soon into the modelling that there'd simply be too much modelling work for one person, and that's when the person managing all of the senior design projects mentioned 3d scanning. Essentially what this does is that it uses lasers to scan out distances in reference to a reference point, which for aircraft is always one bolt in particular on the left side of the window. These distances are saved as points in 3D space, and what comes out of it is a pretty accurate model that while it may not be complete, is more than good enough for basing some more in-depth modelling off of. The problem with just using the model that the 3D scanner gives out is that it's quite a bulky file (quite easily 200MB+), which we could not use directly as it likely wouldn't work well on underpowered machines. We encountered one very large problem though: as you'll see from the picture on the right, it did not work out well. At all. Even the person doing the scan was shocked how poorly it came out.

This was quite a blow to the expectations of the project, as we not only expected to do the cockpit area, we also planned on doing the boom arm control area as well, since that's the most critical part of the craft (the KC-135 is an aerial refueler after all). What also didn't help is that the boom arm area is quite a cramped area as well, and even the handheld unit we used (the Creaform GoScan 50) wasn't particularly effective. As such, we had to scrap that aspect, but if we were able to get a better scan of the area, I feel like I could've gotten the modelling done in a couple of weeks at most. In total, going from the scan to the finished base model took about 10 hours of modelling that I did in one sitting because I'm really good at locking down and finishing stuff that I've started.

Finally, I ended up with the completed cockpit base you saw when you clicked on the panel on the main page. I was incredibly thankful that I had the pictures that I had taken (over 100!). Flipping back and forth between Blender and the collection of pictures I had, I was eventually able to come up with the final base model you see right now.

Since we 86ed the boom arm area, my other big source of work needing done was the gearbox. This thing was an absolute pain to model, as the scan came out really poorly for it. As well, we only had a couple of good pictures of the thing, which certainly did not help matters. We decided as a group that it'd be easier to deal with the interactivity of the gearbox as one whole unit, so I modelled it as one whole unit rather than a bunch of separate parts. The picture in the center is the actual gearbox.

My role in the project was exclusively 3D modelling. Naturally, the other important part of the equation is texturing. However, i'll leave that portion of the story out as it's quite a frustrating story of a certain member who didn't want to pull their weight. Let's just say that these photos were directly useful for them. When we brought everything into Unity, my final jobs were to take the texture "panels" and place them in the craft where they should be, as well as place the simple buttons I made. The buttons only took about five minutes a piece, and aren't worthy of their own pictures. Instead, here's the completed picture.

Pictures may be worth 1,000 words, but a video is worth a million. Here's the final recording we submitted for the project.

Outside of this, I've also done some 3D modelling for 3D printing here and there. I've worked with a Z Corp Zprinter and a Creality Ender 3 Pro. I'm more than comfortable with working on one as I understand the theory of them well enough.