Tail

Observing from the commercial shark air swimmer, we realised that it might have been unnecessarily heavy as there was a rather large structure, as well as a foam sheet that supported the motor driving the tail. We also deduced that it was running on a servo motor as the flapping of the tail was accurately controlled, and could flap left and right on command. The commercial air swimmer tail comprised of a crescent-shaped membrane connected to two carbon fiber rods, which was then connected to the structure as mentioned above.

To reduce weight, we removed the structure and the foam sheet and attempted to replicate the flapping motion. We also used a piece of styrofoam core sheet that resembled the shape of the original tail.

We also used a 1.2g micro servo that could output 0.08kg/cm worth of torque. However, the servo was poorly connected to the inflated balloon itself and often dropped to the floor. Additionally, the servo could not withstand the load and could not flap the tail to the full extent of the servo, and eventually overheated, causing the outer casing to melt and failing to work entirely after 15 minutes of usage. The propelling force generated by the tail was also insufficient to push the balloon at a satisfactory speed.

From this test run, we learnt that the foam sheet and the structure that we removed were actually essential as components of the propelling force. They acted as a platform for the servo motor and carried the weight of the tail itself. This way, the servo motor only needed to provide the forces to flap the tail left and right. Additionally, the platform helped to distribute forces around the balloon to propel it, as too much force concentrated on one spot of the balloon would cause that spot of the balloon to dent inwards, instead of propelling the entire blimp as intended. The tail membrane was also made of a much thinner material, and had consistent holes (2mm diameter, 1.9cm apart) across it. This reduced the torque required for the servo to exert.

Furthermore, we deduced that the motor driving the flapping tail was actually a DC motor that switched polarities upon toggling the controller, which made the tail flap left and right respectively. When the tail reached the full extent, a spring would prevent the DC motor from pushing the tail any further. However, this required the user to constantly toggle the controller to mimic the flapping motion of the tail, which proved tiresome. Hence we programmed our servo motor to automatically provide left and right movement when prompted.

Additionally, upon further research, we realized that we could utilize a bigger variety of shapes for our tail instead of restricting ourselves to a crescent-shaped tail.

Thus, we 3D printed our own support structure to act as a platform for the servo motor, and also to carry the weight of the entire tail itself. This included ledges and nooks for the carbon fiber rods to be superglued on, as well as the servo motor itself, such that the motor’s axis of rotation would stay the same. The discussion about 3D printing the tail can be found here.