The first modification replaces their pentagonal spring rotor base with a circular rotor base (Figure 1).

This was so that we could countersink flat-head bolts into the rotor structure, compared to having raised bolts such as pan-head bolts (Figure 2).

When we recreated Aachen 2015’s pump, we discovered that no matter how carefully we mounted the rotor onto the shaft of the stepper motor, one screw head would always drag along the pump plates, implying that the rotor was tilted at an angle.This caused the pump to have insufficient torque to move fluid through the silicone tubes.

The screw attaching the rotor to the stepper motor shaft is a flat end screw, and the rotor should not be tilted at an angle. Troubleshooting further would yield diminishing marginal returns, as other possible sources of error such as inaccuracies in 3D-printing would not be within our power to resolve. Moreover, the angle was very slight and otherwise undetectable by visual inspection. An obvious solution would be to decrease the height of the central rotor column, and then shift the entire rotor assembly upwards, but this would not be ideal. As the load moves further along the shaft, the angular deflection of the shaft increases. While the increase in deflection would definitely be negligible in this case, such a solution would be bad design practice. Hence, we elected to countersink the bolts because this approach allows us to easily visually inspect the rotor to ensure clearance between it and the faceplate of the stepper motor.

However, a flat-head bolt of equivalent size could not be countersunk directly into the Aachen design as there was too little material remaining to hold the bolt in place. We then simplified the rotor base, and thus produced what you see in Figure 2.

Our second modification was to add material for the fastening screw. A drawback of the 2015 team’s design was the weakness of the bolt and nut combination fastening the rotor to the stepper motor shaft (Figure 3).

We thickened the wall separating the fastening nut and the shaft by 2.75 mm, placing the nut 6mm away from the centre of the shaft (Figure 4). We also thickened the wall between the bolt head and the nut. This design can thus better withstand mechanical stresses, and is more durable.

The third modification we made was to reposition the inlet and outlet for the silicone tubing.

Aachen 2015’s design stacked several Plexiglass layers on the stepper motor’s faceplate, forming a housing to hold a silicone tube inside a circular path. We noticed that Aachen 2015’s silicone tubes had a smaller outer and inner diameter than the tubes we planned to use, as the working volume of their bioreactor is less than ours. After analyzing their design, we conjectured that we only needed to modify the thickness of the pumping layer (Figure 5) to accommodate our own silicone tubing.

However, our silicone tube exerted forces on the walls of the pumping layer and caused the walls to deflect (Figure 6).

Because of this, the silicone tube could not be pinched closed (occluded), and the pump was unable to force the fluid to move through the tube.

We thus repositioned the inlet and outlet, combining them into one opening (Figure 7).

Now, when the silicone tube attempts to regain its original, non-deformed shape, the lines of action of the forces it exerts at the opening will be much closer to the metal fasteners (Figure 8), causing a much smaller bending moment. Additionally, the force exerted by the silicone tube is now shared by 4 fasteners instead of 2 as the shape is now continuous.

After this modification, the pumping layer no longer deflected visibly, and operation was smooth.