Mass Relays are harder to model than you'd think

 When modelling the mass relays, I learned an unpleasant lesson: The number of polygons in a cylinder matters. I began the top and bottom halves with a simple cylinder, which I converted to an editable poly and extruded into the correct shape. Unfortunately, I was unable to extrude parts of the model - this cost me a lot of time, as it wound up being not entirely circular, and I probably lost at least an hour of my time by trying to fix this - I wound up having to add in a whole other cube, instead of extruding a section for the bottom. The result doesn't quite look perfect, especially towards the loop end of the relay, where it doesn't match up to the rest of the part very well, but it would take far too long to re-do the entire shape of the relay.

Learning from my mistakes, I ensured the number of polys in the second cylinder matched up nicely - this side of the relay progressed far more quickly. Lesson definitely learned: Plan ahead the number of polygons in a primitive before converting it to editable poly.

Some amount of modelling later, I was left with this:

Clearly something is not quite right here. The very top and the very bottom of the curve is smooth, and the rest is not.

http://docs.autodesk.com/3DSMAX/16/ENU/3ds-Max-Help/index.html?url=files/GUID-FF7D7633-03AD-4427-821A-65F8AC484CDD.htm,topicNumber=d30e177574,hash=WS73099CC142F48755-1257E12111BF108800E348C

Some research gave me this link, leading me to the conclusion that the smoothing groups had the model smoothing over the shallow curves there, but not the rest of the model. The top half of the model was easy to do (Auto Smooth worked perfectly), but of course the tricky bottom half of the model wasn't as easy.

The shape looked a bit squashy here, so I did some editing, and when Auto Smooth didn't work, I had planned to turn on smoothing groups at random, and try to work out a pattern until I got it right. Luckily for me, the first two I turned on sorted the problem out entirely, leaving me with no need to investigate further:

The model is still not quite finished - there are some small parts left to model, but they won't take long.

After this, I did some research on the spinning rings in the centre, since they'll inevitably need animating - and since it's a simple rotation, it shouldn't be hard to figure out. I used the following link to analyse the mass relay:

https://www.youtube.com/watch?v=bGr9T1k8QOY

This provides a good close-up of the spinning rings - the inner one spins on the Y-axis in the screenshot. Meanwhile the whole lot spins on the Z-axiz, as shown in the following screenshot:

This will require a bit of planning, and two separate animations - I know how to do it in Unity, unfortunately that's not much help here. I'm sure I can work out how to do it in 3DS Max, though - a simple matter of going over last semester's notes on animation. Thankfully, the lovely slow-motion close-up of the Alpha Relay in the Arrival DLC for Mass Effect 2 gives us a lovely image of the two rings spinning. 0:50 in the video is the best example of this.

The model is currently missing only a few small spindly parts, then I'll call it complete. These shouldn't take long at all to model - a few cylinders ought to do the job. After that, I'll provide a full analysis of what I've done right and what I've done wrong (plenty) in creating this model, along with the previous two. My progress so far: