ME 490B:

Engineering Design Senior Project

Professor Simpkins

Spring 2015

Department of Mechanical Engineering

SDSU Thruster Project

Caesar Amparo
Phong Huynh
Randy Marles
Olumayokun Olagbemi
Blake Ramuno
John Schmidt

Project Sponsor: Mike McCrory (Knuckledragger Designs)

TABLE OF CONTENTS

INTRODUCTION / PROJECT GOALS…………………………………………..2

FIXTURE AND FABRICATION
LOAD CELL
BATTERIES / CONNECTORS……………………………………………………3
SYSTEM LAYOUT AND MOTOR CONTROLS
SITE SURVEY
ANALYSIS
SOLIDWORKS FLOW SIMULATION ANALYSIS
ANSYS COMPUTATIONAL FLUID DYNAMICS (CFD) SOFTWARE
REFERENCES

1
INTRODUCTION / PROJECT GOALS

One of the most interesting and innovative extracurricular organizations on campus is

the SDSU Mechatronics club. A competition is held annually where the Mechatronics
team competes against teams from other Universities and High Schools to design and
build an autonomous submarine that runs an obstacle course. The Mechatronics Club
design includes multiple thrusters in order to generate the movement required.
Although the thrusters are performing their function, there is currently no way to
determine whether each thruster is operating at its maximum potential.

The thruster project was developed to explore the field of underwater propulsion. The
main goal of this project is to design and fabricate a portable test fixture for the Seabotix
BTD150 thruster the current SDSU Mechatronics Club could use to analyze the thruster
performance before competitions, during and after competitions.

As part of the Mechanical Analysis, simulation software will be used to help analyze the
flow trajectories of the thruster while running in normal operating conditions. An overall
thrust force for the thruster will try to be determined with the simulation software. This
thrust force result will be compared to the published data for the Seabotix BTD 150
Thruster as well as the test data obtained from the thruster test fixture.

Figure XX: ​Seabotix BTD 150 Thruster

2
TEST FIXTURE AND FABRICATION

The test fixture is designed for testing thrusters from the Mechatronics RoboSub on a
dock or pool edge to allow for transverse thrust along a pool wall. It is lightweight and
able break down into three main assemblies, and two cases to hold the control panel
and batteries.

The main assembly is the fixture with the thruster and load cell. It consist of a long
pivoting beam with 10:1 ratio from the thruster to the load cell. This portion is attached
to the main body, or legs, of the fixture using clevis pins. The pins allow for easy
breakdown and to make it easy for the team to switch to another thruster for testing
without maneuvering the entire test fixture out of position. The control panel/battery
tower is bolted down the main body to assure stability during tests runs.

The fixture is entirely made of 6061 aluminum, which is widely used for its corrosion
resistance and weldability. The legs feature weather resistant EDPM rubber strips to
prevent slippage and also prevent damage to pool surfaces during test.

During fabrication, we learned how to weld using a TIG welder, produce parts using a
waterjet cutting machine and use a fully automated band saw.

3
LED Bracket - This bracket will hold led lights, chip, and its electrical wires. This will be
made of aluminum. Drawing was done using solidworks and 3D printed. This old
bracket was made by the Mechatronic Team. We will be using this design because it
produced the stability we are looking for in terms of how well it connects to the Thruster
we are currently using.

4
Figure XX: ​LED Bracket

5
BATTERIES / CONNECTORS

In order to give the test fixture power needed had to find a way that was portable and safe. This is why we
chose to power the texture with batteries. With batteries we would not have to worry about running long
wires, also we would be able to power the fixture outside without the worry of outlets.

Sizing Battery

The first step in creating the battery box was sizing the batteries. I first had to find the highest amount of
voltage being drawn from each device. I found that the maximum amount of voltage would come from
the thruster at 19V, because of this I had to find batteries of over 19V. After this I calculated the total
current to 4.393 amps.

Components Voltage Current (amps)

Thruster 19 V 4.25 A
Arduino Yun 5V .0465 A
Arduino Uno 5V .0465 A
Load Cell 5V .05 A
Table

Once I found the parameters I was then able to choose batteries appropriate for the fixture. The fixture
needed 19V so I choose to run two 12V batteries in parallel to have a total voltage of 24V. After choosing
the appropriate maximum voltage I then had to choose the amount of amp hour the batteries would need.
To do this I used ​equation a​ to calculate the battery life and from this I was able to choose the right
amount of amp hours. With an amp hour of 7.5 and a 50% duty cycle I calculated that the batteries would
run for 2.44 hour.

Battery Life = Battery Capacity in Milli amps per hour / Load Current in Mill amps * 0.70

Equation a​ [6]

With ​equation b​ I was able to calculate a duty cycle of 50%. I approximated the run times off of the
lengths of the approximate time each test would take. We averaged out that each test would table about

6
five minutes. With this we choose to calculate a duty cycle of 50%. By choosing a high duty cycle we
would be able to prepare for any unforeseen testing situation.

Duty Cycle (time on)/ (time powered down) *100

Equation b

Battery Material

In order to stick to the budget I had to choose a material for that battery that was not be too expensive and
would fit into our small dimensions we had. Since we wanted to stick to our goal of portability I had to
find a battery material that is compact. The first choose was lithium ion because you can get a great deal
of power from lithium ion with out sacrificing space. Although, I found out that lithium ion was hard to
find and was very expensive. The lowest price I found for a battery with our desired parameters was about
$110. I also found that lithium ion does not react well with water if not contained properly. This is why
ultimately I ended up choosing lead acid. Lead acid is bigger and I found two 12V 7.5 amp hour batteries
for $29.9 and two battery chargers for $45.9 from ​www.batterymart.com​ . The price of the two batteries
and the two chargers ended up being a fraction of the price of the lithium ion battery. Also, lead acid is
safe and it can handle water and abuse a lot better than lithium ion. Since lead acid is bigger and heavier
we chose to use this to our advantage as a counter weight to secure our fixture to the dock.

Battery Box

In order to contain the batteries and the changers we choose to use a Pelican iM2200 case. This case
would protect the batteries from the water also it would prevent the batteries from shorting out. The
batteries and chargers are then connected to a four-pole double throw toggle switch, making one direction
on and the other direction charge. By doing this it allows everything to be connected and things can be
take in and out of the battery box as needed. Foam was used to hold the batteries in place and a hole was
cut at the bottom of the box to improving ventilation to prevent overheating.

Figure (Battery Box)

Connectors

7
We chose to use conxall connectors to make all of the connection on the fixture. Conxall connectors
protect the wires from shorting with one another, and it also makes the fixture more users friendly and
compact. The amount of pins used for each item was chosen by the number of connections each item had.
For example, the battery box had two connection so we chose to use a there pin connection. We generally
chose to use have an extra pin so that each item would have different connections. By doing this we added
a factor of safety because the user will be force to make the correct connections. Each item has its own
unique connection so there is no chance of the wrong connection being made. Although, the connectors
add a convenience factor for the user the main benefit it provides is that the connection can only be made
in one direction. This makes it very difficult for the system to short, adding a high amount of safety.
Another, consideration that was made was the positioning of the male and the female connections. In
most cases the female connection we chosen to be at that panel of the control box. Although, from the
battery box to the control box we choose to use a male connection at the control panel this makes it
difficult for the batteries to short, it also allows the connection to be only made in one direction.

Items Pin # Female Connectors Male Connectors

Battery Box 3 4182-3SG-300 3182-3PG-3DC
Thruster 2 4182-3SG-300 3182-2PG-3DC
Load Cell 6 7282-6SG-300 7282-6PG-300
RPM Sensor 4 7282-4SG-300 7282-4SG-300
Table

Figure​ (Connector Layout)

8
Figure​ Description (Batteries to the left and chargers are to the right)

9
LOAD CELL

In order to measure accurate readings from the thrusters output a S type load cell
(​3138_0 - S Type Load Cell (0-100kg) - CZL301C)​ will be used in order to gather data. A
load cell is an instrumentation that senses forces. This is done through a transducer by
converting mechanical input force into an electrical output signal. Specifically, the
S-beam load cell measures strain gauges(small resistor patterns) through both
compression and tension.

Figure XX: ​3138_0 - S Type Load Cell (0-100kg) - CZL301C

However, electrical signal output from a load cell and requires a special instrumentation
that can amplify its signal. The most commonly used amplifier to amplify a load cell
signal is a INA125. The amplifier provides complete bridge excitation and precision
differential-input amplification on a single circuit.

Figure XX: ​Instrumentation Amplifier with Precision Voltage Reference

10
 Figure XX: ​INA125 Pin Configuration

In order to amplify the voltage signal of the load cell, wires has to be connected by
following the schematic in ​Figure xx​. It was also determined that the signal that we
obtain is dependent on the value of our resistor.

FigureXX​: INA125 Schematic

11
 FigureXX​: load cell, INA125 amplifier, and arduino connections

The toughest part in making the load cell work is learning how to read a schematic and
making proper connections to the breadboard in order for the load cell to function. Also,
another challenge was making sure that the amplifier does not burn out, given the
amount of experience in making circuit connections it was bound to happen.

12
SYSTEM LAYOUT AND MOTOR CONTROLS

As much as we may not have liked it, there was no way of getting around the
electronics aspect of the project. The devices we need to use in order to accomplish our
project goals all require some sort of electrical development. We determined that the
best way to manage all the inputs and outputs of the system was to use Arduino
microcontrollers. Arduinos are a powerful tool in that they allow the user to write code
and upload the information directly to the board via a USB connection. Also, the code
language is basic and is similar to C.

The arduinos consist of digital and analog pins, 5V outputs, and grounds. The pins can
either be used as inputs or outputs, reading or sending any increment of voltage
between 0 and 5. However, there is only a limited number of pins available, so we
needed two to accommodate all of our components. This also allowed us to keep
certain aspects of the system separate, such as the motor controls and the display/load
cell/ RPM feedback device. The entry level arduino model is the UNO, which was
decided was best fit for the motor controls. The second arduino was the YUN model,
which was best suited for the RPM feedback, load cell, and LCD display due to its ability
to establish a wireless connection with other devices. The goal was to use this feature
to relay collected test information to a computer in order to organize the data in a way
that was useful to the Mechatronics team.

As the brains of the electronics, the arduinos were positioned at either end of our
perfboard and attached to headers for easy removal and replacement. One of the
drawbacks of the arduino boards is that they are limited to an input and output of around
5V. Since our thruster runs at a maximum of 19V, an external power supply and driver
was needed to provide sufficient input. We were able to get a motor driver that is
tapered specifically for the arduino and plugs in as a shield (a separate board that lays
on top of the arduino and so it has direct connections and allows access to unused
pins). The batteries were connected directly to the motor driver through its input voltage
ports, allowing for sufficient power to be supplied. However, the 26V from the batteries
was too much to power the arduinos which require 5V. In order to solve this problem,
we added a 5V step down regulator to the list of hardware. This chip was able to take in
the voltage from the batteries and reduce it to the 5V needed to power both of our
arduinos independently from the USB connection.

13
The necessary controls of the system directly correlate to what it is that is being
accomplished by the system. When deciding upon what to include, we kept the end goal
in mind which was to measure the output force of the thruster, the rpm, voltage, and
current. Keeping track of the voltage and current across the motor is important so that
you can keep an eye on the power running through the thruster in case there is a
system error. Additionally, these readings can help us keep track of battery life and give
us an estimate as to what capacity the motor is running at.
One of the main concerns we wanted to address with this project was to determine if the
output force was the same for the thruster when running in forward and reverse. Also, a
way for the user to increase the power supplied to the motor is necessary. Last, we
wanted to be able to turn the controller on or off independently from the battery box.
With all of these inputs needing to be addressed, we listed the controls needed:
On/Off
Speed
Forward
Reverse
For the on/off, a simple toggle switch was chosen. A potentiometer was used to control
the speed since it allows for small adjustments and variable output. Initially, the forward
and reverse commands were going to be through momentary pushbutton switches, but
a lack of time pushed for using what was available which were two light up rocker
switches. However, a lack of being thorough resulted in the selection of rocker switches
that lit up on AC current, and out whole system is DC. This led to the use of
independent LED lights that corresponded to their respective rocker switch’s states.
In an effort to eventually become waterproof, we decided to use a Pelican IM2200 case
to house the control panel. Also, being able to close the case adds protection to the
hardware. An aluminum plate was chosen to be the face of the control panel based on
availability and ease of use. A 3D model of the hole layout and overall size was drafted
and sent to the water jet on campus to fabricate to the specifications.

14
Having a larger case and plate made prototyping the controller a lot easier since there
was more room to spread things out.
The next step was to begin implementing the controller features. The first piece of
hardware to get implemented was the LCD screen. Since most LCD screens are similar,
it was very easy to find a layout and diagram we could follow. The LCD consisted of
mostly logic connections but also a potentiometer to control the brightness of the
display.

After we verified the functionality of the diagram on a breadboard, the connections were
transferred to a more permanent perfboard. Perfboard is different from breadboard in
that there are no underlying connections, but there are metal terminals that we can
solder headers to in order to make connections. In order to keep troubleshooting simple
and things easier to change if needed, we utilized headers (small posts) and plugs
when possible. We made a harness for the LCD screen that plugged in on both ends.
The next feature to be implemented within the system was the speed controlling
potentiometer. Luckily, the motor driver we purchased came pre programmed with
commands for running the motor at certain speeds. The voltage passed through the
driver to the motor was on a scale of 0-400, with 0 being completely off and 400 being
all the voltage supplied to the driver. Considering that the potentiometer is basically a
variable resistor, an easy way to tell the motor driver to run the motor at speeds
corresponding to the potentiometer dial was to translate the amount of voltage running
through the potentiometer to the 0-400 scale. We were able to run the 5V from the
arduino, through the potentiometer, and then into a digital pin that was set up as an

15
input. In the lowest dial position, the resistance is the highest so no voltage passed,
whereas when the dial was open the resistance was very little so all of the voltage
passed. The input scale on the digital pin is from 0 to 1024, with 1024 being 5V. The
map command was used in the arduino code to convert the voltage seen in the digital
pin (0-1024) to the motor driver scale (0-400). Lastly, the converted value was then
declared to be the speed the motor was set to. Since this conversion takes place within
the loop of the code, the value is constantly being updated allowing for full spontaneous
control of the motor speed.
After the speed control was developed and transferred to the perfboard, the next
controls were the forward and reverse direction. Since the motor driver was already
programmed to see positive values as forward and negative values as reverse, all we
needed to do was develop a way to translate a switch to mean positive and another
switch to mean reverse. To accomplish this, we decided to use the rocker switches as
on/off switches for 5V from the arduino to run through to individual input pins. When set
as an analog read, the arduino interprets a 5V input as 1 (or HIGH) and no input as 0
(or LOW). 5V was then hooked up to both switches and the pins they were connected to
declared as either forward or reverse pin. From there, it was a matter of writing the logic
for the arduino to know which pin meant what. Basically, when the red switch was
turned on, the arduino read the corresponding pin as HIGH, and when that pin was
HIGH it caused the value set by the potentiometer to either be passed directly through
or to be multiplied by -1. In addition, a check is needed to ensure that if both switches
are on, then the system would shut off, and that when either of the switches are off, the
motor is off. This was accomplished through a series of if/else statements. Then, plugs
were made for each switch, the logic for the LED lights were implemented (included
within the logic statements), and the connections were made on the control panel.

#include "DualVNH5019MotorShield.h"

DualVNH5019MotorShield md;

const byte POTENTIOMETER = A0;

const byte CONTROL = 9;

int reading;

int i;

int setspeed;

const int rvsbuttonPin = 3;

const int fwrdbuttonPin = 2;

16
const int fwrdsignal = 13;

const int rvssignal = 12;

int fwrdbuttonState = 0;

int rvsbuttonState = 0;

void setup()

Serial.begin(9600);

md.init();

pinMode(fwrdsignal, OUTPUT);

pinMode(rvssignal, OUTPUT);

pinMode(fwrdbuttonPin, INPUT);

pinMode(rvsbuttonPin, INPUT);

pinMode(CONTROL, OUTPUT);

void loop()

fwrdbuttonState = digitalRead(fwrdbuttonPin);

rvsbuttonState = digitalRead(rvsbuttonPin);

if (fwrdbuttonState == HIGH && rvsbuttonState == HIGH)

{setspeed = 0;

md.setM1Speed(setspeed);}

17
else if (fwrdbuttonState == HIGH)

reading = analogRead(POTENTIOMETER);

i = map(reading, 0, 1024, 0, 400);

setspeed = i;

md.setM1Speed(setspeed);

else if (rvsbuttonState == HIGH)

reading = analogRead(POTENTIOMETER);

i = map(reading, 0, 1024, 0, 400);

setspeed = -i;

md.setM1Speed(setspeed);

else if (rvsbuttonState ==LOW)

{ setspeed = 0;

md.setM1Speed(setspeed); }

else if (fwrdbuttonState ==LOW)

setspeed = 0;

md.setM1Speed(setspeed);

Serial.println(fwrdbuttonState);

18
 The last input on our list was the system on/off switch which was added last as
we hooked the perfboard and its components up to the battery source. The toggle
switch was connected in line with the positive cable leading into the system.
Since the control panel serves as the headquarters for the entire system, we
believed it would be the best place to attach all of the system components (thruster,
load cell, battery, and RPM feedback). Holes were cut out for each component, and
connectors were mounted so that each component could be removed when needed.
After many hours spent troubleshooting, re soldering connections, and developing code,
the control panel was laid to rest on 4 L brackets mounted inside the pelican case.

Being involved with the electronics aspect of the project allowed us to shed light
on a dark area of our Mechanical Engineering education. Electrical Engineering is a vital
part of our digital age although it is one of the least liked subjects in our curriculum. In
this project, we were able to gain great hands on experience with soldering, cutting and

19
stripping wires, and making connections with plugs. Also, being able to interpret
electrical schematics and diagrams was essential in transferring our components to the
more permanent perfboard. This part of the project was particularly rewarding since we
were able to take things that weren’t preprogrammed to perform a function and find a
way to create a desired output. We can watch lectures and take notes all we want, but
nothing compares to the learning gained by using some solder and watching a motor
speed up or slow down based on the layout you developed.

20
SITE SURVEY

Site (1) Lake Murray

Lake Murray is one of the two locations that we are planning on testing our Seabotix
BTD 150. This is a perfect testing site because it is open to the public. The depth of the
water is greater than 15 feets, which exceeds our requirements. Overall we felt this
should be one of the location we can do testing on when ever we need testing.

21
 Figure XX: ​Lake Murray Dock

Site (2) Aquaplex

Aquaplex is the second location that we are planning to do the testing on. It meets our
basic requirements due to water depth being greater than 15 feet. Aquaplex is not open
to the public but only to members of the school. There is many rules and regulations
where we have to follow in order to use this location. Lifeguard needs to be on site and
appointment only.

22
 Figure XX: ​Aquaplex Pool

ANALYSIS

Axial Flow Pumps (AFP)

For investigation and research of thruster propeller and shroud design, we were tasked
to research centrifugal pumps and ducted propellers. There are 3 main types of
centrifugal pumps; Radial Flow, Axial Flow and Mixed Flow. For this project, the focus
shifted towards Axial Flow Pumps (AFP) as this is the closest in design to the design of
the Seabotix BTD 150 Thruster. Below is a schematic of a typical AFP.

Figure XX: ​Schematic of an Axial Flow Pump [1]

A typical AFP consists of the following components:

● Propeller​ (axial impeller) – Propels the fluid in a certain direction (to generate
thrust for forward momentum in a thruster design)
● Cylindrical Casing​ – Contains the flow and directs the fluid
● Motor Drive Shaft​ – The propeller is attached to the motor drive shaft, the motor
drive shaft is what turns the propeller.
● Bearings​ – Holds the shaft and allows the shaft to rotate freely with little
resistance.

An AFP has a propeller that runs inside the cylindrical casing. Pressure is developed by
the flow of the fluid across the propeller blades and the fluid is pushed in a direction
23
parallel to the shaft of the pump. This allows the fluid to enter the impeller axially as
well as discharge from the pump nearly axially. In the thruster design, this discharge is
what drives the thruster and generates forward momentum. The shroud of the propeller
is what theoretically acts as the cylindrical casing even though it is not fully enclosed
like the AFP is.
Advantages of an AFP:

● High flow rate to head ratio – this means that the discharge of the pump is
relatively high compared to the lift it can overcome. Typically an AFP can pump
3 times more of a fluid at lifts of less than 4 meters.
● Easily adjustable for peak efficiency – The AFP can be easily adjusted to run at
peak efficiency at low flow/high pressure and high flow/low pressure by changing
the pitch on the propeller.

Disadvantages of an AFP:

● Cavitation – this is the process of formation of vapor bubbles when subjected to

reduced pressures at constant ambient temperature. This can cause noise,
damage to pump components as well as decreases the efficiency of the pump.
[2]

24
Ducted Propellers

A ducted propeller, also referred to as a Kort nozzle, is a propeller encased in a

non-rotating nozzle or shroud. The propeller spins inside the shroud and creates flow of
a fluid. The shroud acts like the cylindrical casing of the AFP and directs the flow of the
fluid axially which produces the thrust to propel the object forward. Below is a
schematic of a ducted propeller.

Figure XX: ​Schematic of an Ducted Propeller [3]

A ducted propeller consists of the following components:

● Propeller​ – Propels the fluid in a certain direction through the shroud

● Ducted Shroud​ –Directs the fluid flow
● Motor ​– Drives the propeller
● Motor Shell​ – Contains the motor and protects it from fluid. Serves as an
obstruction the fluid flows past which reduces the thrust output.

At lower speeds, typically less than 10 knots (5.14 m/s), using a ducted propeller
increases efficiency, offers better course stability and the propeller is less vulnerable to
damage from debris However, at higher speeds the efficiency is reduced, course
stability is worse and there is an increase in cavitation. As mentioned in the AFP
section of this report, increased cavitation can cause noise, damage to the components
and decreased efficiency. Overall, if operating at low speeds, less than 10 knots, the
ducted propeller seems to be the best choice for design in a thruster. [4]

25
26
Thrust Force Analysis

One of the main objectives for this project is to test the existing Seabotix BTD150
Thruster that is already being used on the RoboSub. In order to do this, an
understanding of thrust and the math behind is very important. Below is some research
done on how thrust is determined theoretically and some of the aspects involved. There
are many factors that contribute to the force applied by the thruster and a lot of it has to
do with the propeller design. There is also a pressure drop across the propeller that
must be accounted for. Determining the velocity of the fluid generated by the propeller
is also a consideration that must be analyzed. The speed of the fluid is what creates
the thrust that pushes the vehicle forward.

Thrust – ​A mechanical force generated by the engine to move the vehicle through the
air/water. [5]

● Used to overcome drag force and to overcome the weight of an aircraft.

● Generated by engines through some kind of propulsion system.
● Propulsion system must be in physical contact with a working fluid (water, air).
● Generated through the reaction of accelerating a mass of gas.
● Thrust is generated in the opposite direction of the accelerated gas.

Propeller Produced Thrust​ – Using a motor to turn a shaft attached to a propeller

used to produce thrust. [5]

● Propeller acts like a rotating wing creating a lift force by moving through the air.
● The accelerated gas (working fluid), is the surrounding air that passes through
the propeller.
● Can have anywhere from 2 to 6 blades.
● Blades are usually long and thin.
● A cut through the blade perpendicular to the long dimension will give an airfoil
shape.
● Because of the tangential velocity, the blades are usually twisted from hub to tip.
● The angle of attack of the airfoils at the tip is lower than at the hub.

Figure XX: ​Schematic of a Typical Airfoil

27
Angle of Attack​ – The angle between the chord line and the flight direction. [5]

● Has a large effect on lift.

● Chord Line – The straight line drawn from the leading to the trailing edges of the
airfoil.
● For small angles, lift is related to angle - greater angle = greater lift
● For larger angles the relation is much more complex.

PROPELLER THRUST EQUATIONS

Simple Momentum Theory [5]

Thrust ​T​ generated by the propeller disk is equal to the pressure jump ​Δp​ multiplied by
the propeller disk area ​A​:

F = ΔpA
Propeller Disk – Imaginary disk made by the propeller rotating.

Use Bernoulli's equation to relate the pressure and velocity ahead of and behind the
propeller disk, but not through the disk. Ahead of the disk the total pressure ​p​t0​ equals
the ​static pressure ​p​0​ plus the dynamic pressure 0.5ρV​0​2

p​ = p​ + 0.5ρV​
t0​ 0​ 0​
2

where ​V​0​ is the velocity of the aircraft.

Downstream of the disk

p​ = p​ + 0.5ρV​
te​ 0​ e​
2

where ​p​te​ is the downstream total pressure and ​V​e​ is the exit velocity.

At the disk itself the pressure jumps

Δp = p​ - p​
te​ t0

Therefore, at the disk

Δp = 0.5ρ[V​ - V​ ] e​
2​
0​
2​

28
Substituting the values given by Bernoulli's equation into the thrust equation, we obtain

F = 0.5ρ A[V​ - V​ ] e​
2​
0​
2​

Propeller Analysis [5]

The amount of thrust depends on the ​mass flow rate​ through the propeller and the
velocity change through the propulsion system.

F = [ṁV] - [ṁV] ​
e​ ∞

Through the propeller disk area, the mass flow rate is

ṁ = ρV A p​

Substitute this value for the mass flow rate into the thrust equation to get the thrust in
terms of the exit velocity, entrance velocity, and velocity through the propeller

F = ρV​ A[V​ - V​ ]
p​ e​ ∞​

Using ​Bernoulli's equation​ upstream and downstream of the propeller

F = 0.5ρA[V​ - V​ ] e​
2​
∞​
2​

Combining the two expressions for the thrust ​F​ and solving for ​V​p

V​ = 0.5[V​ - V​ ]
p​ e​ ∞​

The airspeed through the propeller disk is simply the average of the free stream and exit
velocities.

Static Thrust ​– The amount of thrust produced by a propeller which is located

stationary to the earth. [5]

Determining the thrust produced by the propeller

T = (π/4)D​2​ρvΔv (1)

Where T = Thrust, D = Propeller diameter, v = velocity of fluid at the propeller

Δv = velocity of fluid accelerated by propeller, ρ = density of fluid

29
Common practice states that velocity of the air at the propeller is v = 1/2 Δv of the total
change in air velocity giving

T = (π/8)D​2​ρ(Δv)​2 (2)

Power absorbed by the propeller from the motor

P = (TΔv)/2 → Δv = (2P)/T (3)

Substituting into equation 2

T = [(π/2)D​2​ρP​2​]​1/3
Solving for Seabotix BTD150 Thruster

T = [(π/2)D​2​ρP​2​]​1/3

T = [(π/2)(0.076m)​2​(1000kg/m​3​)(80W)​2​]​1/3 ​=

T = 38.72 N
Convert to kilogram force
T = 4.0 kgF

Results of Theoretical Analysis

After applying this equation to the specifications we received from Seabotix, the
theoretical force of the thruster is approximately. The measured force from Seabotix’s
test data is approximately 2.9 kgF (28.44 N). Assuming that our equation does not
account for interferences that are unknown to us at the moment, our result is
acceptable. This is approximately a 30% difference but there are several unknowns
that are not accounted for and more research needs to be done in order to determine
what these unknowns are and how to account for them in these calculations.

30
SolidWorks Flow Simulation Analysis

Since this project is focused around helping the RoboSub team test the existing thruster
and develop an improved thruster, we have been working with the Mechatronics Club
on certain tasks. One of the tools we were able to obtain from the RoboSub team is a
license for the SolidWorks Student Engineering Kit. Included in this kit is the Flow
Simulation add-on which can measure various different parameters of the flow of a fluid
(or gas) across a body. This software can measure and simulate fluid flow, fluid forces
(in our case, thrust) and heat transfer through a solid body.

Flow Simulation is a very complex software that involves a lot of training and practice.
We were able to set up a training session with GoEngineering this semester and
learned a lot about how to set up a simulation. Due to the late start of the projects in
semester 1, we were only able to receive this training the last week of lectures. As of
the end of semester 1, we were unable to run a successful simulation. There are many
parameters that must be set up correctly and if the set up is incorrect, the simulation will
not run. The GoEngineering instructor tried to help us figure out the bugs in our system
but he was unable to do so.

The end goal with Flow Simulation is to run a successful simulation with the existing
Seabotix BTD150 Thruster as well as run a successful simulation with our new design.
After running the two simulations, we will compare the results and see if there are any
differences. We will use these simulations to adjust our design in order to create more
thrust and improve upon the seabotix design. Ultimately, Flow Simulation is a tool to
help us design and determine where we need to make adjustments to our new design.

31
SIMULATIONS

Flow Simulation of Random Downloaded Propeller:

Figure XX:​ SolidWorks Flow Simulation Flow Trajectory Figure XX:​ Random Propeller Used in
Analysis of a Random Propeller Simulation

The above propeller was downloaded from GrabCAD. In order to test out the software,
we started with this propeller. An attempt to run a flow across the rotating propeller was
attempted and ultimately was unsuccessful. This was more of a trial run to see if we
could get any results and actually run a simulation. As you can see above in the figure
on the right, the simulation did give us a flow trajectory, however the trajectory lines
were random and did not display an expected result. The trajectory lines are all over
the place and the velocity profile received was much lower than what should have been
received. Since this was just a test run, the results received were generally positive and
helped in setting up a more simple analysis.

32
Flow Simulation of Simple Sphere:

After experimenting with the random propeller, our sponsor advised us to turn it back a
little and experiment with a simple sphere. It was asked that we create a sphere and
run a simulation of this sphere without any rotating regions. Below is the results of the
multiple attempts of running simulations of this sphere.

Figure XX: ​Pressure Profile of Simple Sphere Figure XX: ​Velocity Profile of Simple Sphere

Once again we were able to run a successful simulation however the results obtained
still were not what we wanted. As you can see in the figures above, the trajectory
arrows all point inward towards the sphere. The flow direction was supposed to enter
one end of the cylinder and exit the opposite end. The flow in this simulation enters
from both ends of the cylinder and exits out the sides of the cylinder (or at least that is
how it appears). The sides of the cylinder are supposed to be the flow boundary and
nothing should be able to exit there. There was obviously something wrong in our setup
but we were unable to resolve the issue and this is what we were able to come up with
at this point.
There were some positives from this simulation however. The flow trajectory looks
good until it passes by the sphere. The flow appears to be traveling the expected route,
hits the sphere and flows around it. Also, the velocity slows as the flow hits the sphere
which is to be expected when a flow is obstructed. The pressure also decreases as the
flow hits the sphere which should also be expected. Even though this was not an
entirely successful simulation, there were some positives points we were able to learn
from.

33
Another attempt was made to run a simulation on a simple sphere. After several emails
to the GoEngineering instructor we met with the first semester and adjusting several
parameters of the simulation, the following velocity profile was received.

Figure XX: ​Velocity Profile of Simple Sphere - Attempt #2

As you can see, this simulation attempt actually ended up being worse than the first
attempt. There appears to be not control volume. The control volume is supposed to
be enclosed in the cylinder where all the trajectory lines should be contained. The
trajectory lines start way outside the cylinder and continue through and out the end.
The trajectory lines also enter from the cylinder wall from the top and bottom. Also, the
velocity has a magnitude of 0 m/s all the way through when the velocity was set at a
rate of 50 m/s. After emailing back and forth with the GoEngineering instructor, we
were unable to find the source of the problem after running through the setup.
There were two good things that came out of this result however. The first was we were
able to get the flow to move in one direction. The flow started from the left and moved
in the right direction. The second positive result was the fact that the flow all ended in
the same direction and moved out of the cylinder in a straight trajectory. This ended up
being a classic case of, “one step forward, two steps back.”

34
ANSYS COMPUTATIONAL FLUID DYNAMICS (CFD) SOFTWARE

After receiving unsuccessful results from SolidWorks Flow Simulation, and speaking
with several experts in the field, we were advised to move away from SolidWorks
simulation because of its unreliability. It was recommended that we use ANSYS
Computational Fluid Dynamics (CFD) Software. ANSYS is said to be the industry
standard for simulating fluid flow and is what all the industry professionals use.
ANSYS computational fluid dynamics (CFD) simulation software allows you to predict
the impact of fluid flows on your product throughout design and manufacturing as well
as during end use. The software's unparalleled fluid flow analysis capabilities can be
used to design and optimize new equipment and to troubleshoot already existing
installations. ANSYS fluid dynamics solutions give you valuable insight into your
product's performance.
Our sponsor was able to get in contact with Dr. Christopher Paolini, Instructor of
Computer Sciences at San Diego State University. Dr. Paolini was kind enough to sit
down with us and help us run a simulation using ANSYS. The following are the results
from the simulation that was run.

Above is the control volume that was set up for ANSYS simulation. In order to set the
thruster up for this simulation, the thruster had to be modeled and then the control
volume built as a cavity around the thruster.

35
The boundary conditions the simulation was run at are as follows:
● Analysis Speed (v) = 0.5 m/s
● Temperature of the fluid = 15°C
● Maximum Depth = 9.8 m
These boundary conditions were determined based off of the operating conditions of the
competition the UAV will compete in. These conditions are the averages that were
obtained from the Mechatronics Club.
With these boundary conditions, the hydrostatic pressure is calculated using the
following equation:

P = ρgh
Where density (ρ) = 999.97 kg/m​3​, gravity (g) = 9.81 m/s​2​ and maximum depth of 9.8 m

From this equation, the hydrostatic pressure at the max depth of 9.8 m is 96.1 kPa.
With the boundary conditions listed along with the hydrostatic pressure calculated for
the specified depth, the simulation is setup with the following parameters.

Figure XX: ​ANSYS Simulation Mesh of Seabotix BTD 150 Thruster

Above is the mesh that was created for the Seabotix BTD 150 Thruster. Dr. Paolini
instructed to make the model as simple as possible, limiting the any small features if
possible. This was difficult to do considering we wanted the simulation to be as close to
the real thing as possible. Because the model has lots of small features the mesh that
was created had 6,856,162 nodes. This is a lot of nodes and when the simulation was
run, this caused it to take over a week to run. This is why we were only able to run one
successful simulation with ANSYS.

36
 Figure XX: ​Velocity Input Boundary

Above is the velocity input boundary. The box surrounding the thruster is the control
volume. The flow enters the control volume from one direction. For the velocity input,
the speed the flow enters the control volume is 0.5 m/s as mentioned in the boundary
conditions. This velocity input boundary shows what direction the velocity is flowing in.

37
38
 Figure XX: ​Results of Simulation
The above figures show the results of the simulation. Since this is for a static thruster,
there is no rotating region. Dr. Paolini mentioned that if we wanted to run a simulation
with a rotating propeller blade, we would need a full year in preparation. So this
simulation focuses on the flow of the fluid over the thruster as the thruster moves
through the water.
The results of the simulation show areas of reversed flow. These areas or reversed
flow are where the flow of the fluid actually starts travelling in the opposite direction of
the incoming flow. Due to the geometry of the thruster, the trajectory of the fluid
changes. This can cause a resistance to the thruster moving forward and can cause
the thruster to slow. This means the thruster must work harder to maintain a constant
speed.
The main objective of the simulations was to use these simulations to help modify the
existing thruster to help improve the design and overall performance of the thruster. By
being able to simulate these areas of reverse flow, we could adjust the design to try and
eliminate these areas of reverse flow. Unfortunately we did not have time to redesign
the thruster. If we had gotten to this point, we could use this simulation to alter the
design of these areas of the thruster. If the areas of reverse flow could be reduced, the
overall thruster performance could be improved and a more efficient thruster could be
produced.
After all that was completed with the simulation, the end result fell slightly shorter than
where we would have liked to end up. The overall failure of SolidWorks Flow Simulation
was disappointing in that we spent a lot time (1.5 semesters) trying to learn something
that ultimately did not give any successful results. If we had learned that this software
was not a good resource earlier, we could have used this time to experiment with
ANSYS more. The one simulation we were able to run with ANSYS gave better results
than several simulations run in SolidWorks. I feel that if we had more time with ANSYS
and more simulations were run, we could have possibly improved upon the design of
the existing BTD 150 thruster. All in all, this was a valuable learning experience where
we learned a lot about simulation software and what is involved in this process.

39
Larger lessons learned

Keep the entire system in mind before you start putting things in place

Determine the use case of the product when designing (the test fixture wasn’t first
compatible with a pool)

Cutting

Soldering

Welding

A good schematic will go a long way

Writing reports

Contacting and coordinating with companies/people (Leverett)

Pin out/ Connector diagrams

Check and double check (heat shrink, connectors)

Check the polarity of wires before making live connections

System integration

40
REFERENCES

[1] ​http://www.propellerpumps.in/pages/products/axial-flow-pump/

[2] ​http://en.wikipedia.org/wiki/Axial-flow_pump

[3]​http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=7&ved=0CC8QF
jAG&url=http%3A%2F%2F4hv.org%2Fe107_files%2Fpublic%2F1386100085_2431_FT
0_1353043533_2431_ft145838_duct_pdf.pdf&ei=_6AyVeaPM4T8oASVxICgDw&usg=A
FQjCNG4q8tsOCu134X5bqG98DKhf4_40Q&sig2=R_WuSFvze8p8rFMl1TFc8Q

[4] ​http://en.wikipedia.org/wiki/Ducted_propeller

[5] ​http://www.grc.nasa.gov/

[6] ​https://www.easycalculation.com/physics/classical-physics/battery-life.php

41