The swerve modules, placed into the chassis
So to clarify, this frame is purely a test bed for the swerve units? Because it doesn’t look like it would be very useful for any other use, due to sheet metal across the middle.
Is this something that you’re actually going to build, or just a CAD exercise?
Well… yes and no. If we decide to use swerve next season this could conceivably work as our drive train. Could you clarify what you mean when you say it wouldn’t be useful because of sheet metal across the middle? Any feedback would be welcomed, so I can tweak and improve my design.
It looks like you’re relying purely on material stiffness to prevent bending, and not section. This is a very heavy approach.
A better way is to design in flanges and box/geometric sections in some way. In sheet metal, this is done with flanges, in a tube frame, this is done with box, C, and angle channel. You will handle the loads much better in this way.
Basically everything in the chassis design is unsupported, and there will be extremely high bending loads on virtually every component. I recommend re-evaluating your construction methods and load paths.
What all of this means:
-Think about where you would place a mechanism on this. Where is it loading the frame? Say it weights 60 pounds. Where are those 60 pounds going? If we assume the mechanism is a point in the middle (worst case), the only thing to support it is the plate aluminum, which will likely bend. The only thing preventing this are two braces in tension on the sides, which are attached to the red boxes, which become a large lever arm, which is likely to bend the connection to the top plate (over time or rapidly, depending on material thickness and connection methods).
Consider an alternative design:
-If using sheet metal (I <3 sheet metal), consider forming a chassis panel on all four sides around the pods from top to near the ground, flanged inwards on the top (and inwards or outwards on the bottom, your choice), and attach the pods to these flanges. Then run two chassis panels down the middle to pick up the inner sides of the pods, with four additional smaller panels to pick up the fourth side of each pod (does this make no sense)? Since this is a swerve, torsional stiffness isn’t as important as in a drop center drive. With this, you could use an HDPE disc near the axle of the pod to transfer side loads into the sheet chassis better, reducing loads on your upper bearings (allowing you to design them differently/less beefily).
-If using box or channel, I recomment a similar layout but out of 1x1 box tubing directly under the support plate for the pod. This will provide enough stiffness in 1/8" (I’m not good at the mecanical analysis, I’ll leave that out), and would probably work in 1/16" with additional stiffness from the mechanisms. From there you could bring back the red risers (1x1 box) and run another box around the perimeter out of 1x0.5" L,C,or box. Make sure you don’t cut the flange to retain stiffness, that is important. You can also triangulate the risers to the upper frame, you will want this if your bumpers are attached to the lower frame.
Try to design around stiffness by sections. This starts with using panels/tubes with sections (folded sheet metal vs unflanged, box/C/L tube), then making larger geometric shapes with the structure (triangles are especially strong, boxes are strong but possible to turn into parallelograms, etc.). Then think of the load paths, where the large forces are coming from, where impacts (bumper loads) are going, where gravity is supporting the robot, etc.
What I’m saying is that your swerve modules look pretty good, and would be a good design to use on a competition robot, but the frame you’ve designed here is pretty impractical for a competition robot.
The sheet metal I’m talking about is your “X” shaped brace right in the middle of the bot. Having that brace there means that you’re limiting where you can place other robot components. What if you wanted a floor pickup mechanism of some sort? You’ve “filled” that space with framing.
If you study successful FRC frame designs, you’ll find that the area in the middle of the frame is generally left open, and instead, support for things like swerve modules is created through structural members around the sides of the bot (Hint: That’s also where the bumpers go, and it’s a good bet you want strong framing behind your bumpers 
)
If you’re designing a drivetrain before you know what the game is (And therefore what mechanisms you need, and where those mechanisms will fit on the robot) you want to leave your options open, you want a frame that could be adapted to any number of mechanisms.
Wherever you have framing, you can’t have mechanisms (at least in general) I can’t think of many mechanisms that make use of the edges of the robot, but I can think of a thousand that make use of the very center of the robot. 