SubmURSA’s mechanical systems were largely successful at the 2011 RoboSub Competition, but many opportunities for improvement were identified. The central rectangular pressure hull that houses the electronic components was very effective, but the frame to which it was attached was quite pliable and not very adaptable. By having the pressure hull slide into the frame on rails, it was intended that static balancing could be achieved easily, but the hasty addition of two vertical thrusters made this impossible. The hand-made Lexan pressure hull lid sealed poorly and deflected downwards at high pressures. These issues framed the mechanical team’s design philosophy approaching the 2012 Competition.
A major portion of the new platform design which needed significant overhauling, was the pressure hull. In the pressure hull design of SubmURSA, the areas of visibility and accessibility were the top priority, with secondary goals of modularity also being achieved. The hull is essentially an aluminum box, with a polycarbonate lid. To reduce the weight of the hull, the wall thickness of the aluminum sidewalls is 4mm, with a slightly thicker base, measuring in at 5mm. Since the connectors have 20 threads per inch, this wall thickness only provides approximately 3 threads, which was deemed unsatisfactory. To overcome this issue, special extrudes were designed with a thickness of 8mm, which are then adhered to the hull. This allows the connectors to sit securely in the hull, without significantly increasing the mass of the hull. Also adhered to the walls of the hull are ten mounting points for the draw latches used in the new sealing mechanism. By manufacturing the mounting points separately, this allows for ideal placement of the latches to optimize the clamping force, and also allows for some mass reduction, since the wall thickness was not increased.
The visibility goal was achieved by the design of the clear polycarbonate lid. As previously mentioned, this allows for quick observation of the electrovisual signals designed on the electronics boards. The lid is created by heating a 1/16″ sheet of polycarbonate, and then blowing it through an MDF frame, which allows the polycarbonate to be blown into a dome shape. This allows for a simple and quick forming method, which allows for the manufacturing of many different lids, with different heights or colours, within a short time frame. It also allows for the optimization of the lid shape.
Complementing the visibility of the hull, is the accessibility of the inside of the hull. Due to the use of draw latches, the unsealing time of the hull is under a minute, which allows for quick repairs of electronics to occur, while not consuming amounts of time that could be detrimental to the internal components. Also adding to the accessibility of the hull are two underside rails and a rail system on the main frame. This allows for the hull to easily slide in and out of the frame so that the internal components can be accessed uninhibited by the rest of the frame. To aid in the removal of the hull, an easy-access handle is mounted to the back of the hull, which is pulled when required. To secure the hull while in the water, one locating clip, held by one cap screw is sufficient.
The modularity of the hull was achieved by the selection of the micro contact series of Subconn connectors. By limiting ourselves to these connectors, and to any pin count between 2 and 8 pins, the threading of all out connectors is 7/16″-20. As such, the external connectors can be moved around as required, and, in future years, the connector can be changed to another fitting connector to use different external components.
Also of note regarding the pressure hull, is that the inertial measurement unit has a separate case to prevent electromagnetic disturbances from other components. A small rapid prototype case was designed and mounted onto the frame. The case is designed to allow the IMU to fit snugly into it, while minimizing the volume in such a way as to aid in our overall buoyancy.
To address the concerns identified at the previous Competition, the mechanical team elected to retain the pressure hull, but redesign and re-build the frame, with the
following key requirements in mind:
- Maintain high adaptability. The frame must accept various parts in various configurations as competition requirements change.
- Design a carry handle for safe, easy transportation and hoisting.
- Maintain a minimum distance of 18” between the inertial measurement unit (IMU) housing and any source of magnetic interference.
- Minimize weight while retaining sufficient rigidity to withstand thrust forces.
- Use readily available materials, and uncomplicated design, to expedite construction.
Based on these requirements, the frame was designed to be constructed from 3/4” square aluminium tubing with 1/8” thick walls, with numerous threaded holes for attachment of components such as thrusters, hydrophones, and future additions like torpedo launchers and grabber arms. Previous designs had successfully used 6061 aluminium, but when tested by immersion in Edmonton city water for several days, this material experienced significant corrosion. The 2012 frame design uses 6063 aluminium, which trades some
mechanical strength for lower cost, improved weldability, and excellent corrosion
resistance. Wherever a tube was required to make a 90° bend, it was divided into two 45° bends to reduce the stress concentration at individual welds. The geometry is simple to construct, and is as light as possible while maintaining an absolute worst-case safety factor of 2 before any part of the frame will fail. The frame typically mounts six Seabotix BTD-150 thrusters: two in each axial direction. Their arrangement allows for five degrees of freedom (surge, sway, heave, yaw, and roll).
SubmURSA’s pressure hull requires a lid to protect its contents from the aquatic
environment, which is known to be somewhat hostile to electronics. It was originally assumed that an adequate lid could be cheaply hand-made by forming a clear Lexan sheet over the hull with a heat gun, but this solution lacked the mechanical strength and watertight seal required to protect the hull contents. A much-improved lid has been designed for use in the 2012 Competition, made from 1/2” clear polycarbonate. Ten jaw clamps on the pressure hull latch onto notches in the lid to pull it onto two rubber O-rings lining the hull rim, and a groove in the lid fits snugly over the hull lip, with an allowance for the lid’s expansion when moved from Edmonton’s subarctic climate to the balmy temperatures of San Diego. Assuming a worst-case depth of 10 m, the lid will experience a maximum deflection of 2.6 mm, clear of the 10 mm distance between lid and hull contents. In this situation, the lid’s deflection and stress are well within the limits of the material, giving a safety factor of 3 before failure.
The pressure hull contents can generally be freely arranged within the available space, but some design attention was given to simple trays capable of organizing and retaining the electronics within the hull. The electronics racks were not as exhaustively simulated as the other mechanical components, due to the trivial loads they will be required to support. Special attention was paid to the trays’ modularity, to allow them to be moved based on wire length considerations or for static balancing. The trays are generically sized to accept various future boards, and, like the lid, are clear to maximize visibility into the hull.
Cette année ARVP introduit un noveau robot autonome, SubmUrsa, donc on introduit aussi une nouvelle plate-forme méchanique. Cette plate-forme a été conçu pour être aussi modulaire que possible. Pour cette raison, la même taille de filetage était utilizé pour tous les connections externes, permettant le lieu de chaque connection externe d’être changé si c’est necéssaire. Pour observé les signaux électro-visuel qui indique la vie de la pile et s’il y a des cartes de circuit qui ne fonctionnent pas rapidement, un coquille claire était choisi. Une autre amélioration est l’utilisation d’une combination de pompes et the propulseurs. Le robot est plus facile à conduire à cause de les pompes, et plus vite à cause de les propulseurs. Un trait de Bearacuda qui a était améliorer est le temps qu’il prend pour sceller le coque pressurisé. Il prenait à peu près 10 minutes pour scelléBearacuda, mais SubmURSA peut être sceller en à peu près 30 seconds à cause des nouveaux loquets.