In addition to the many valuable recommendations already made, I'll add these things to check for (I don't know if you've already done some/all of these, just wanted to double-check):
- Make a 2D sketch (or 3D model, if you feel like it) of the chains/belts to ensure none of them hit your standoffs.
+ It looks like a couple of your standoffs may be in the way, but it's hard to tell, obviously
- Design for chain tensioning and/or use a chain length calculator
+ Could be a floating wheel, tensioned by a screw or turnbuckle
+ Could be a larger sprocket placed in the chain run (like 118)
+ Could be a small sprocket (or piece of delrin) pushed against the chain until tight
- I'm having a hard time judging the scale, but if you're pressing bearings into your sidewalls, ensure your holes are toleranced to get a good press fit (depends a lot on machining method used... I'm unfamiliar with the plasma cutter)
- Standoffs can be a lightweight way to make a more rigid structure between two sheet metal plates... I would recommend at least using some pretty large OD standoffs and potentially putting a couple plates across that have two holes in each sidewall. This should add significant rigidity
- Also, you do want to ensure that your drive modules mount to the rest of your frame rigidly and without creating stress-concentrated areas...
What thickness of aluminum do you have modeled so far? With a better-optimized lightening pattern and good structural design I would say you don't need anything thicker than .125; however, I don't have any experience with flangeless sheet metal drivetrains.
Best of luck in your design!