Designed and printed 2 iterations of this short stroke linear actuator that uses an M4 screw, a T slot nut and a stepper motor, but the stroke on this is currently only 2mm, which is not enough to pinch the silicone tubing, and it needs some chassis that contains the tubing as well.
Printed molds and casted silicone gaskets for the water reservoir nipples. 
Wrote test code to sweep servo motion. See video.
Mark 2 of the pinch valve uses the entire 180 rotation of the cheap servo but will hopefully have enough torque. I could shrink the length of the larger component slightly. It's unclear to me how this would work in a water line since this in effect would need to create 2 bends to keep the line running the same direction, but maybe if there is enough slack it would be OK either way. Need to hook this up to a different controller to get the full rotation, since the one I soldered previously had resistors selected for 45deg. 
Soldered 03012-001 servo control board but with 20k and 24k resistors for R1 and R2 to reduce the range of motion to just 45 degrees. Printed and assembled corrected pinch valve geometry mark 1. Unfortunately there is not enough torque to compress even this soft tubing, so we will try something different. I don't want to pivot (no pun intended) to a beefier servo than the lowest-level hobby grade. See video of the fail. 
Printed out models and realized I need slightly less distance between the stepper mounting surface and the recess depth in the rotor. Also realized I need to improve bed adhesion before the next print.
Received 8mm OD 6mm ID silicone tubing and did a quick fit test on the 03015-001/2 pinch valve. It does successfully block air when close and allow air through when open, but I noticed a few things that needed to change, so I updated the model to a 3-piece 03015-001A/2A/3A. Started printing this new model overnight. 
Designed housing assembly to accept 8mm OD silicone tubing and printed both components of the pinch-valve. Installed everything on a small servo. The clearances on the interlocking divots are too conservative; more engagement is needed. If the silicone elasticity/friction is not enough to lock the servo in the closed position, these divots will be required, but for now, the lack of engagement is OK. Ordered the softest silicone tubing I could find for a preliminary test. 
Designed test cam for test fit with servo arms. 