During phase II, the team chose the finalized conceptual design after evaluating two different potential propulsion systems for the micro underwater robot. As the team progressed through this portion of the project we developed a working SolidWorks model of each concept so that a stress analysis could be performed in CosmosWorks, allowing the team to see any potential methods of failure associated with each design.
After working with conceptual models and performing a technical analysis on each design, the team was left with two concepts that could be refined into prototypes. The two designs both mimic the waving motion that propels fish and other aquatic life. Before the final design is carried out, the team had to objectively rate the two designs to eliminate one before proceeding with the phase III design. For all the designs, the factors of key importance to the team were the durability of the robot, the time it would take to fabricate a working prototype, and micro-fabrication concerns.
Concept 1: Piezoelectric Truss
The first concept was a tail connected to a head segment that would be constructed of two strips of piezoelectric material separated by an inert material to provide the required separation. The design relies on the piezoelectric strips to expand and contract to create a moment about the tail and deform it, generating motion. This model requires a low voltage actuator because the two strips are separated by the zero force members. However, the strain produces in the tail will be very high due to the distance the actuators need to move to great the required deformation. This raises a concern of durability, since high strain will eventual lead to failure of the part after a short time.
In addition to the durability concerns, this design raises two more concerns that the team had to address. The first was the time it would take to produce a prototype. The design works simply in theory, however, piezoelectric actuators that have a high expansion factor and low force output are difficult to find commercially, which would delay out prototyping and design refinement. Finally, the truss design in the robot is useful for magnifying the moment created by the piezoelectric actuators, but at a micro-scale the work required to implement this design would be beyond the current micro-fabrication capabilities. After reviewing this design, the team has decided to not proceed with it to prototyping.
Concept 2: External Piezoelectic Actuation
The second approach that was generated from out conceptual designs is tail that is driven by piezoelectric actuation but, where the actuators are not the support for the tail. This means that the tail will be constructed of an inert, ductile material that will form the backbone of the tail, and the piezoelectric actuators will mount onto the outside of tail, similar to how the muscles of animals are connected to a supporting bone structure. This design is more durable that the previous concept because it downs not rely on the actuators to maintain the structural rigidity. The "backbone" of the tail will be more resilient because a material that has a high ductility will be chosen to ensure proper elastic deformation of the tail, reducing the cold-working of the tail material and subsequent failure.
This design can also be rapidly prototyped because the actuators are high force, low strain models that are commercially available in a wide range of shapes and sizes. The tail segment will be a high ductility plastic to allow easy mounting of the actuators as well as reducing the weight of the robot and increasing its buoyancy. In addition to its ability to be rapidly tested, the design will also be easy to manufacture at a microscopic scale. The design could be built using simple fabrication techniques that are in use today.
After comparison and debate, we chose this design as our final design based on its feasibility. This design has advantages in three aspects. First, reduced tail height shortens the overall length needed to create propulsion compared to the truss concept. Second, this simple design can be implemented with basic masking and deposition of materials. Third, no micro-machining is required in this design which lowers the level of complexity when fabricating. But there are challenges that follow while pursuing the concept. One such challenge is that the shorter height causes a higher stress in the actuators from the higher force output.
Detailed Design
The final prototype will consist of parts that are constructed by the team as well as several components that will be purchased pre-fabricated. The team will build the tails backbone out of low density polyethylene (LDPE) plastic because of its high range of plastic deformation and low cost. The piezoelectric actuators will be purchased from a third party company, called FACE International. The actuators chosen were two TH-8 thunder actuators because they produce a large force over a high range of deformation. These actuators will cost approximately $250 a piece, and will be the only major cost of the project. The head of the robot will also be constructed of HDPE plastic to ensure buoyancy of the prototype and proper sealing of the control hardware in the head of the robot.
The robot will be assembles using thermoset epoxy to create permanent, water proof bonds between the plastic backbone and mounting plates of the linear actuators. The following drawing shows the layout of the actuator attached to the tail's LDPE backbone. The actuators sit above the backbone due to the mounting plates; this will allow the actuators to create a proper moment around the backbone, allowing it to bend.
Final Assembly Model
The final assembly will be the tail component connected to the units head. The control wires for the actuators will run into the head and the holes sealed to prevent water flow into the robot. The head will be constructed of HDPE plastic to ensure that it creates a stable surface to mount the actuators.
