Sub-Pages

Conclusions and Recomendations

The submersible robot project began as a feasibility study which the team turned into a finished design project.  The project called for designing a submersible robot that was capable of traversing the Wakulla underwater cave system while recording relevant data such as water velocity, water pressure, direction and heading.  The team recommended that a sensor package be purchased such as the MAMMARK sensor package that can provide the required information so that the team could focus on the design of the actual robot which would be required to house the sensor package.   The team evaluated several initial design considerations which were used as guidelines for beginning the design process.  These considerations gave the team constraints to follow to allow the team to develop a design based on realistic requirements which directed the design process.  Through these considerations the team decided whether to tether the robot, whether to have a means of self propulsion, whether to maintain constant communications, where to retrieve the robot, and finally whether to make many inexpensive designs or to use a single more expensive design.  After these design considerations the team came up with several objectives that would define the project.  The objectives were to develop a resurfacing system to recover the recorded data, to develop a static buoyancy system that was based on the housing and component masses to allow the robot to float in the water, and to develop an alternating buoyancy system to prevent the robot from getting trapped in any obstacles.  The team decided that these three objectives were required to complete a successful design based on the conditions the robot was to be operating in and also based on the decisions made in the initial considerations.

This led the team to its first design idea. The first design called for a small dimensioned sensor package to be housed in a spherical design.  The design was to be inexpensive and lightweight, allowing for the release of many robots into many different cave openings.  The driving factor in this design was that it would allow for the maximum collection of data at one time.  After researching the required sensors, the team found that size and weight would be two of the major issues which would constrain the design.  Due to these two issues the team headed into a different direction with the housing design and ultimately with the design for the entire submersible robot.

The team then came up with a design using PVC pipe as the housing due to its rigid structure.  The PVC housing was a semi-spherical design in that the length was equal to the diameter.  This was done to try to prevent the robot from getting stuck in any obstructions while it was floating downstream and also provided the robot with a durable design capable of absorbing impacts.  While the semi-spherical design would not be used it was the basis for the team’s future designs.  The design of the PVC housing would change several times over the next semester as the requirements put on the robot changed. 

The resurfacing system was developed out of the requirement that the data had to be recovered and that there would not be constant communications with the robot.  It was also required because the robot needed water current to move so it would not be able to return to the surface on its own.  The basic idea behind the resurfacing system was to develop a system in which a mechanism could trigger the system to allow it to operate.  The mechanism was to be designed to trigger the resurfacing of the robot after the occurrence of some event recorded by the robot.  The initial plan was to use a pressure transducer to serve as the trigger once the robot rose to a certain depth.  The problem with this design was that there was no way to prevent the robot from triggering prematurely if it went into a cave system that was not open to the surface.  Another problem that was found using this design was how to take the input (a predetermined pressure reading) and get an output (electrical signal) that could be used to operate a valve. After discussing this problem with the team’s teacher and instructor the team went with using a photovoltaic cell as the trigger mechanism.  The solar cell would operate by generating electricity from the sunlight as it lit up the water at a cave opening.  While this design required that the robot be able to reach a depth where sunlight penetrated, it seemed the best approach because it allowed the robot only to resurface when it reached an opening that was accessible from the surface.  The team conducted several tests on the solar cell and was able to verify that the cell generated power in less than ideal conditions and was theoretically capable of generating the power required to operate the resurfacing system. 

The final design for the resurfacing system called for a solar cell for generating electricity, a carbon dioxide canister which served as the supply for the buoyancy system, a solenoid valve which would allow for the carbon dioxide to flow from the canister to a bladder, a bladder which would be used to displace water and create positive buoyancy, a microcontroller to control the opening and closing of the solenoid valve and finally the tubing system which provided a airtight seal for transporting the carbon dioxide.  The premise behind the teams design was that once the robot was carried through the cave system and reached the end of the tunnel the water current would carry the robot to a higher elevation and eventually to a height where sunlight penetrated.  The solar cell would then generate the required electricity (5 volts) as more sunlight is absorbed.  After the solar cell powers up the microcontroller, the controller the opens the solenoid valve and allows the flow of carbon dioxide from the canister to the bladder through a series a copper fittings and tubing.  Once this occurred the water displaced by the bladder would then cause the robot to rise to the surface where it could be recovered. 

Due to the requirements of the solar cell needing to be able to absorb sunlight, the team changed the housing design and went with a clear PVC housing that would allow sunlight to penetrate the housing. 

The next system that was developed by the team was the alternating buoyancy system.  The theory behind the alternating buoyancy was that if the robot could oscillate around a static buoyancy value it would allow the robot to free itself if it got caught up in an obstruction.  This was considered a vital part of the design due to the surroundings that the robot would be operating in.  The team approached this design by considering several key factors; the robot needed to be as small as possible, the fewer moving parts the better, and incorporating the design from the other systems would allow the team to maximize its’ budget.  The team knew that the robot had to be able to increase and decrease its’ buoyancy around a fixed value.  This led the team to calculate the static buoyancy of the entire robot with all of its components.  This will be explained thoroughly in the following section.  With a static value the team then had to consider how to generate the required buoyancy.  There were several methods the team considered but due to the decision to use water current as the means of propulsion the one realistic design based on all of the constraints the team considered was using carbon dioxide to inflate and deflate a bladder like in the resurfacing system.  The original design was to use a separate carbon dioxide system for the buoyancy and resurfacing system but the team integrated the two systems to save time during the construction of the robot and also to save weight.  The one centralized carbon dioxide canister removed the need to develop a manifold for the individual canisters to run into.  It also allowed for easy replacement of the carbon dioxide when the system was emptied.

The buoyancy system operated by using some of the same components as the resurfacing system, but the buoyancy system used a more complex operating system.  The buoyancy system used three solenoid valves to control the flow of carbon dioxide, a one way check valve to exhaust the carbon dioxide to the water, a microcontroller to control the timing of the opening and closing of the three valves, the canister to provide the supply of carbon dioxide, and a bladder to create positive buoyancy.  The layout of the system has the first two solenoid valves in line with each other.  The third solenoid valve is connected with a tee where the flow is diverted between two paths, one that leads to the bladder, and the other path runs to the third solenoid valve and then the one way valve after it.  The system operates in several stages by turning on a switch located on the microcontroller.  With the switch turned on and the valve on the carbon dioxide canister open the system begins operating.  In the first stage of operation the first two solenoid valves are open with the third valve closed.  In this first stage the carbon dioxide flows through the first two solenoid valves and is then forced to inflate the bladder since the third valve in closed.  It is during this stage that buoyancy is generated as the bladder displaces water from the housing of the submersible robot.  In the second stage, the second valve closes and prevents the bladder from filling any more.  At this point the buoyancy reaches its’ maximum height.  In the next stage, the third valve opens with the second valve still closed so the carbon dioxide leaves the bladder and flows through the third valve and out the one way check valve.  The submersible robot begins to descend at this point.  The third valve then closes and the second valve opens again to allow the process to repeat thus creating an alternating buoyancy system.

Along with the resurfacing system the alternating buoyancy system created design requirements for the housing.  One requirement was that part of the robot housing had to be filled with water so that the water could be displaced by the bladder.  The other thing was that the buoyancy system added weight so the housing design had to account for the extra mass.

As has been stated above there were many considerations in designing the housing for the submersible robot.  The housing had to be durable and had to be able to withstand the pressure at 300ft so the team decided on PVC pipe for its strength.  The housing had to be transparent to allow the solar cell to absorb sunlight.  The housing had to have a wet/dry region in it to allow the buoyancy system to operate.  Due to the requirement that the sensor packages had to be removed when the robot was retrieved the housing had to be accessible so the team used a threaded PVC pipe housing which used threaded end caps for access.  Finally and most importantly the housing had to provide neutral buoyancy for the robot.  This was done by calculating the mass for all of the components of the robot.  Then the volume of water displaced by the housing was multiplied by the density of the water so that the actual available mass could be calculated.  Due to this constraint and the availability of parts the housing changed from a from a 6” diameter 6” long pipe to a 4” diameter 15” long pipe to 3 4” diameter 15” long pipes.  After finally settling on a housing design the team found that the mass of the submersible robot to be 22 lbs.  The team then found the allowable mass using buoyancy calculations to be 27.5lbs.  This gave us 5.5lbs for a sensor package to be installed inside the electronics housing.  The team then performed water testing to verify the calculations and to determine any sources of leaks.  The housings carried the calculated weight and were water tight.  The final design consisted of the electronics to be housed in one housing, three of the solenoid valves and the main tubing to be located in a housing, and the final solenoid valve, the one way check valve and the bladders for the resurfacing and buoyancy system to be in a wet/dry housing.  One possible area for improvement in the housing design is in the support brackets for the housings.  The team used banding wire to hold the housings together because it is strong and can be easily removed if required but it looks unprofessional.  For a prototype design it is acceptable but it could be improved upon. 

        The team was not able to complete all of its testing due to some unforeseen problems.  During the testing of the carbon dioxide system for leaks and proper operation of the solenoid valves, the team found that the flared fittings were leaking where they met the copper tubing.  To solve this problem the team JB welded the fittings to the tubing.  This stopped all the leaks but created many problems during assembly.  Once the robot was assembled certain components could not be removed again such as the tubing used in the carbon dioxide supply system.  Due to this the team was not able to retest the system again until it was fully assembled and found at that time that the solenoid valves were not working properly because the carbon dioxide pressure was too high and was forcing the valves open.  The teams’ solution would be to install a pressure regulator, but time did not permit.  This being the case the final test of the resurfacing system and the buoyancy system could not be run to verify that the systems were operating properly.  Overall the team reached satisfactory completion of its goals of developing a submersible robot.  The team addressed what it felt were the key objectives required to design a submersible robot.  The team found that many of the problems arose during assembly as there was no manual to follow so things had to be reconfigured to allow for the robot to be assembled.  Given more time and resources the team is confident that the robot would work properly.  The team found this project challenging but very worthwhile and while the team did not meet all the objectives the team still considers this project successful.