I decided to design and build a VMC. I don't have the space for a larger router setup and I mostly find myself working on smaller parts that would benefit from the added rigidity in the Z axis. Capable benchtop CNC's on the market utilize epoxy/granite or cast frames for rigidity and damping. Although I'd love to go down that rabbit hole, I don't think it could be feasibly done within my time frame. Using RHS makes the build a little more straightforward and damping can be added by filling the tubes with E/G. 

For a VMC, the base is stout and fixed to a more massive table. The X and Y axes move close to the base, keeping the torsion on the base relatively small in comparison to the Z axis. The center of the spindle has to sit off of the face of the column by ~250mm. For design and simulation purposes, I'm going to assume that the spindle is completely rigid, passing all forces on to the linear bearings and column of the mill. The force on the spindle would result in significant torsion and/or bending in the column. 


I'm using HGR20 linear rails and blocks. The spacing of the rails is fairly constrained due to the RHS I was able to cheaply acquire.  The spacing of the bearings was decided on by some basic hand calc on the forces going into the bearings and the material cost of the plates that the bearings will be bolted to. I'll include those calculations shortly.


I'm using Nema 23 closed loop stepper motors. 



The controls are built up centered around a Mesa 7i96s FPGA card. This allows me to operate the mill through LinuxCNC, a free and well supported software that opens doors to many other robotics applications.