SERVO TUNA 3 Research Summary 2005

1
SERVO TUNA 3Research Summary 2005

• Research Director
• Dr. Afzal Suleman
• University of Victoria
• Mechanical Engineering Department
• Victoria, BC

2
Hydrodynamic Propulsion

• Starting in Feb 2005 the original servo tuna was
  rebuilt to continue the research with the model
  shown below.

3
The servo motors were enclosed in aluminum boxes
with o rings for the moving shafts. The boxes
were filled with silicone jelly so that water
cannot enter even if the seals fail. Lexan was
used for the viewing windows.
4
Shown is the complete navigational system on a
small circuit board.
5
All Fish Parts shown include body, battery,
controller, and tail assembly.
6
Water Proofing Tricks The waterproofing of the
electronics was accomplished by stuffing the
circuit board plus the tips of the wires into a
pill bottle filled with silicone jelly. The
wires emerged from a hole in the top. We made
sure that the jackets of the wires were also
buried in silicone so that water would not be
able to travel up the insulation jackets of the
wires. The fish was flooded with water and
there were no leaks. The electronics worked
well. This is a very easy way to protect
electronics from water.
7
Thrust Tank Tests The fully assembled fish was
tested in a thrust tank. The force generated was
60 grams. The power consumption was 6 watts.
Interestingly the best swimming pattern was
produced by an algorithm that simply runs a sine
wave through the motor array. Theoretically the
best way to do it is by having the sine wave
amplitude decrease with each motor as you go from
tail tip to body. In the water the wave
effectively was like this even though we were
running a straight sine wave pattern. It would
appear that motors closer to the body are not
capable of sustaining the full travel of motion.
Thus the theoretically correct pattern appeared
because of the physical forces of the water on
the fish. This was a surprising discovery.
8
Free swimming test This was the first free
swimming test and there were a few problems
relating to the buoyancy. We forgot to bring
the lead ballast weights to the lake for the
test. This was to be a surface test to avoid
losing the robot in the lake. The front was too
buoyant and we used a rock to weigh it down.
In spite of the problem the fish did swim well
and was able to turn left and right. We will
repeat the test after carefully adjusting the
ballast weights.
9
A fisherman on the dock thought it was cheating
...
10
Movie of Free Swimming Fish run file
short_swim.avi
11
Future Work 1/ Add proper ballast so that it is
slightly positively buoyant. 2/ Add skin to tail.
3/ Add dive plane fins to the sides to assist
in moving deeper. The microcontroller supports
two more servo motors. 4/ Do a towing tank test
so that the speed of travel is known. 5/ Add
remote control via sonic commands of the
telephone keypad. The microcontroller can do
this if the sensors are added. 6/ Add an
ultrasonic sensor so that it can move
automatically away from obstacles.