Group 2
Field Grace
Matt Gildner Rachel Price
Mission Objective Tree
y travel to and survey
y Harvard Bridge
g Pylons.
y
Goal: Autonomously
Objectives
Rugged Vehicle
Fly Planned Path
Flyy Parallel to Wall
Photo‐Survey Bridge
Support Tasks
Analyze Vehicle Capabilities
Construct Support Structure
Analyze Sensors and Motors
Navigate Bridge with GPS
Cl d L
Closed‐Loop
C
Controll
Navigate Bridge with Sonar
Record Images of Bridge Pylon
Solar Feasibility
Spec/ Select Solar Panels
Project Division
y Vessel Structure ‐ Physical Components
y Sensors – GPS, Sonar, etc.
y Propulsion ‐ Motors
y Control – Data Collection, Commands
y Solar ‐ Panel Selection/Integration
y Environment – Waves
Student D
•
•
•
•
Vessel Modifications
Buoyancy
Weight and Trim
Final Design
Desig
V
l Modifications
M difi i
Vessel
Motivation: needed survivable & rugged vessel
y Functional Requirements:
y Stable
y Maneuverable
y Sensor mounts
y Rugged design
PT 2017
Photo of the Pro Boat Miss Elam 1/12 Brushless RTR
removed due to copyright restrictions.
Pro Boat Miss Elam
Spirit of the Challenge
Final Design
y Trimaran
y Stability = pontoons
y Maneuverability = motors mounted outboard
y Sensor mounts = metal sheet + hull structure
y Rugged = strong, non‐corrosive materials
P
Si
l l i
Pontoon
Size C
Calculations
V = Awaterline
h ~ Akite
h
li •h
ki •h
Streamline : d1/d2 ~ 3
d2
d1
V = .5(d1)(d2)(h)
= .5(6”)(18”)(5.75”)
= 310.5 in3 > 303 in3
Mold to shape pontoons
H ll Stability
St bilit
Hull
Metacentric height v. Draft
120
100
80
Draft = 3.5 in
GM = 44.2
44 2 in
60
40
20
0
0
0.5
1
1.5
2
2.5
Draft (in)
3
3.5
4
4.5
Hull Trim
Trim Moment as a Function of Draft
450
1in Draft
2in Draft
3in Draft
3 5i (T
3.5in
(Tow T
Tank)
k)
250
• Loaded with battery and
pontoon assembly only:
- Freeboard = 0 in !
50
-4
-3
-2
-1
0
1
2
3
4
-150
-350
-550
Trim (in)
• Port-Starboard trim with symmetry
• Loaded with extra 3.5 lbs at
bow:
- Freeboard = 0.75 in
Student B
•
•
•
•
Pontoon Wave Forcing
Wave –Frequency Resonance
Transportation Stress Hazards
Crashworthiness‐‐
Crashworthiness
Crashworthines
Spring‐Mass Resonance Model
y Single cantilever beam w/point loading : Euler Bernoulli Equation
d 4u
w( x) = F ⋅ δ ( x − L)
EI 4 = w( x) where :
F
dx
y Mass –Spring System: F = −k ⋅ u
y Resonant
R
t ffrequency:
k
ω=
m
y Vessel Natural Frequency:
q
y ω = 83Hz
u=
k=
3EI
L3
y Far higher than frequency of water waves!
y Near frequency of transportation vibrations
y Switch
h bolt
b l attachments
h
to delrin
d l wingnuts
3EI
L3
Wave Forcing on Pontoon
• B = 0.075m,
0 075m T = 0.1m
0 1m
• Worst‐case waves :
ω= 1Hz A= 0.3m
• Deep water waves:
k=
F FK = − ∫ ∫ p nds
Sw
l 2 ρage − kT cos(ωt ) sin( kb)
FZ =
k
FzMax
L 2 ρage − kT sin( kb)
=
= 166 N
k
ω2
m
~ 0.1
• Surface pressure integration:
Fx = 2 Lρag (1 − e − kt ) sin(ωt ) sin( kb)
FxMax = 2 Lρag (1 − e − kt ) sin(kb)
Fx Max = 0.44 N
Implications of wave forcing
y Forcing moment: 50Nm
y Horizontal force on pontoon is negligible Danger zones:
y Vertical force causes moment
y Reaction force on boat structure: 277.8N
y Bend in pipes
MR
y Stress calculated using σ =
I
y Max stress: 8.8 x 106 Pa
y Yield stress of aluminum: 4 x 108 Pa
y Force on bolts
• Originally, load distributed over 6 bolts
• Total force on each bolt is 46.3N
y
y
Strut bend
Structural bolts
If stress is too high, the internal
structure could be ripped
completely out of the boat!
Failure of internal boat structure
y Internal structure to which pontoons are
mounted is held onto hull with epoxy
y Approximate that most stress on epoxy is
in the shear direction
y Epoxy much weaker in ‘peel’ forcing
y Peel forcing unlikely: hull not pulled outwards independent of plate
y Shear force per unit area of epoxy contact: 1.5
1 5 x 104 Pa
y Shear strength of epoxy: 1.4 x 107 Pa
Transportation and Handling
y Most likely transportation method is to carry by frame
y Potential problems due to weight of boat:
•Total weight of boat hull and components: 26lbs
•Stress on epoxy << than rated shear stress (8,000 PSI)
•Original 6‐bolt design: 19.3N per attachment point
• Internal boat structure failure possible
• Design modified to reduce these issues
• Number of bolts increased to 12
Pontoon Arm Collision
y Impact creates moment around central attachments
y U‐bolts on plate – act as moment constraint
y Peel forces on epoxy
y Crash 1 pontoon while boat travelling at 1 ms‐1 forward
y Relevant for mission and for vessel transportation
y Results:
y Moment around center – 15.87 Nm
y Force per U‐bolt
U bolt – 31.24
31 24 N (U
(U‐bolt
bolt max.
max rating: 1935 N)
y Average force on epoxy – 89.25 N
y Total peel strength of epoxy – 91.44 N/mm epoxy peeled
Student E
•
•
•
•
Reading GPS Data
GPS Test Results
Result
Reading Compass Data
Readingg Sonar Data
Global Positioning System
y Standard Format:
GPRMC,135713.000,A,4221.4955,N,07105.5817,W,4.29,258.17,310809,,*16
y GPS shield is used on Arduino MEGA
y 1 signal
i l hit/second
hit/
d
Courtesy of Arduino.cc. Used with permission.
Trial 1
Trial‐2
Trial‐3
GPS Output
y Imaginary
X‐Y
I
i
X Y coordinate
di t system
t
y Sailing pavilion as the origin
y GPS latitude and longitude processing:
y Decoded into x‐y coordinates
y Sent as an input to control system
Photo by ladyada on Flickr.
A i Compass
C
OS5000 3
3‐Axis
y Primary Navigation
y Low Noise
y Works under the bridge
y Faster refresh rate than GPS
y Outputs to Control System
y Heading: 0 to 360 degrees
y Pitch: ‐90 to +90 degrees
y Roll: ‐180 to +180 degrees
WR 1
LV MaxSonar WR‐1
• Range 0‐255 inches
• Analog and serial output
Photo of the MaxSonar WR1
http://www.active-robots.com/products/sensors/maxbotix/WR1-large.jpg
removed due to copyright restrictions.
– Analog accurate 1” of serial output
– Maintain moving average off anallog output
•
•
•
•
•
•
Fully waterproofed
M
Mounted
t d to servo
Run at pre‐set heading until wall is detected
Wall detection @ 3.05
3 05 m,
m
Accurate reliable readings @ 2.13 m
Safety buffer of 1.52
1 52 m from wall
Student A
• Selected Sonar
• Wall Finding
Findin
• Wall Following
• Control Architecture
Mi i Pl
i
Mission
Planning
z
Mission stored as an array of way points
z
z
Sample way point: { x, y, heading, speed, range, mode }
Modes describe behavior of boat
Test Mission
z
z
z
z
Unreliable GPS and sonar data
Reliable compass data
Internal clock, counting time
Mission composed of target headings, and times at
which to change way points
Wall Finding and Following
z
Two modes:
modes
z
z
Mode 1: Travel at set heading until wall is detected
Mode 2: Hold constant distance from wall,
wall assumed to
face a known heading
M d 2
Mode
Mode1
Data Collection
z
Serial
z
z
z
Analog
z
z
z
GPS: Position
Compass:
p
Heading,
g, Pitch and Roll
Sonar: Range
Solar: Voltage Output
Internal
z
z
Time: Since program start up
Way point and error
Data Logging
z
Logging constrained to <1 Hz
z
z
Logged to an onboard netbook
z
z
z
Higher speed causes serial buffers to fill
Data stored via Realterm
Formatted for MATLAB importing
External GPS data stored on SD card
Student C
• Propulsion and Speed Control
• Thrust Tests
Test
• MATLAB Simulations
• Real‐
Real‐World Tests
Propulsion System Design
y 2 Thrusters
y 1 per outer hull
y Differential thrust for yaw
y PWM control
t l with
ith Arduino
Ad i
and Speed Controllers
y 12V DC Trolling Motors
Speed Controller Selection
y Pro Boat 40A Waterproof ESC
y Limited PWM frequency
q
y range
g
y Incompatible with Arduino PWM
y Victor 884 ESC
y Compatible with Arduino
y Not Waterprooff
Photos removed due to copyright restrictions.
Please see:
Pro Boat Waterproof ESC with Reverse 5-12V 40A
VEX Robotics Victor 884 + 12V Fan
Thrust Test
y Test conducted in water tank
y Thrust was measured at
differentt motor
diff
t voltages
lt
y Data fit to linear curve
y Maximum vehicle
thrust=2X28.17N=56.3N
y Minimum voltage=1.5V
g
Force=Sensor Force X R2/R1
Thrust Test Data
C t l System
S t
d MATLAB Simulation
Si l ti
Control
and
y Closed‐loop proportional control
y heading and speed using sonar, GPS, and compass
y Mission set
set‐points
points determine control behavior
y Way‐point Traveling
y Static Heading Following
y Wall Following
y Boat and control system modeled in MATLAB using ODE45
y M, J, B(lateral)
B(lateral), B(rotational)
y Various set‐points were used as inputs to adjust the gains
Si l ti
Simulations
Simulations
Experiments
Experiments
Student F
Characterize wave spectrum
of Charles River
Ri
River
Setup
Flow rate
meter
computer
LabPro
LoggerPro
Real Term
Pressure
transducer
Arduino
microprocessor
River Wave Data
Case 1
y Fair weather
y No Wind
y No river activity
Case 2
y Raining
y Steady wind
y No river activity
y Case 3
Case 4
y Raining
y High winds
y No river activity
y Test
T tD
Day
y Fair weather
y Slight wind
y High boat traffic
River Wave Data
Case 1
y Fair weather
y No Wind
y No river activity
Case 2
y Raining
y Steady wind
y No river activity
y Case 3
Case 4
y Raining
y High winds
y No river activity
y Test
T t Day
D
y Fair weather
y Slight wind
y High boat traffic
D t
Data
Heights and Wave Periods
Average 1/3 wave height = 0.7174 cm
Average period length = 1.1487 s
River Wave Data
Case 1
y Fair weather
y No Wind
y No river activity
Case 2
y Raining
y Steady wind
y No river activity
y Case 3
Case 4
y Raining
y High winds
y No river activity
y Test
T t Day
D
y Fair weather
y Slight wind
y High boat traffic
D t
Data
Heights and Periods
Average 1/3 wave height = 1.3295 cm
Average period length = 0.5450 s
River Wave Data
Case 1
y Fair weather
y No Wind
y No river activity
Case 2
y Rain
y Moderate wind
y No river activity
y Case 3
Case 4
y Rain
y High winds
y No river activity
y Test
T t Day
D
y Fair weather
y Moderate wind
y High boat traffic
D t
Data
H i ht and
dP
i d
Heights
Periods
Average 1/3 wave height = 1.4242 cm
Average period length = 0.8082 s
River Wave Data
Case 1
y Fair weather
y No Wind
y No river activity
Case 2
y Rain
y Steady wind
y No river activity
y Case 3
Case 4
y Rain
y High winds
y No river activity
y Test
T t Day
D
y Fair weather
y Slight wind
y High boat traffic
D t
Data
Heights and Periods
Average 1/3 wave height = 3.5856 cm
Average period length = 0.7936 s
W
S
Wave
Spectra
Conclusions
y No real current
y Waves driven by two sources
y Wind
y River activity
y Frequency
F
ffaster in
i presence off wind
i d source
y Need more data!
Student G
•
•
•
•
Vehicle Requirements
Po
Pow
wer
er Output Te
TTest
st
System Integration
Feasibility Comparison
Compariso
Project Overview
y Tasked with feasibility assessment for solar power use
y Desired quantitative data
Assess Vehicle
R
i
Requirements
Select
Panels
Check Panel
Specifications
Integrate
S
stem
System
Full
System
System
Test
Vehicle Requirements
Component
Voltage
Current
Power
Accelerometer
6V
50 mA
0.300 W
Sonar
5V
50 mA
0.250 W
GPS
3.3 V
50 mA
0.165 W
Arduino Mega
5V
50 mA
0.25 W
Electronics Total:
0.965 W
Motors (Min)
(
)
12 V
0.25 A
3W
Motors (Max)
12 V
2.50 A
30 W
Vehicle Minimum Total:
3 965 W
3.965
Vehicle Maximum Total Total:
30.965 W
Correcting Specifications
y Panel output: Pout = η panel Edensity Apanel
y Standard Peak Edensity : 1000 W/m2 (noon @ equator)
y Boston Dec.
Dec Peak Edensity : 800 W/m2
y Ideal: 10am on sunny day
Sun Power Density
y 600 W/m2
1200
P(t) = P ⋅sin(
t − t dawn
t dusk − t dawn
⋅ Π)
1000
Power in W/m2
y Non‐ideal: 10am Cloudy
y 200 W/m2
y Power through day
day:
800
600
400
200
0
‐200
200 0
Standard
5
10
15
20
Time of Day in Hours
Boston Ideal
Boston Non‐Ideal
25
Suntech STP0055‐12
Measured Capabilities
y Sunny:
y Short Circuit Current : 0.21
0 21 A
y Measured Output: 2.8 W
y Cloudy
y Short
Sh t Ci
Circuit
it C
Current:
t 0.065
0 065 A
y Measured Output: 0.933 W
Sunny
Cloudy
30
Voltage (V)
Estimated Capabilities
y Peak Rated Power = 5W
y Peak Power (sunny)= 3W
y Peak Power (rainy) = 1W
Suntech Panel Power Curve
20
10
0
‐10 0
0.1
0.2
Current (A)
2
V
P = VI + =
R
0.3
System Integration
Physical System
Electronic System
R1:
170 Ω
> 21 V
>4V
Arduino
Arduino
Voltage
Logger
Power
Resistor
140 Ω
21 V
y Panels sit on motor mounting
y Servo available for optimization
Solar
l
Panel
Vout =
R2
Vin
R1 + R2
R2:
40 Ω
Feasibility‐ Electronics
Measured Cloudy Output
Electronic Power Consumption
Extrapolated Sunny Output
8
7
6
Power ((W)
5
4
3
2
1
0
0
50
100
150
200
250
Time (s)
300
350
400
450
500
Full Vehicle Feasibility
70
Extrapolated Sunny Panel Output
Vehicle Total Minimum Power Use
Actual Mission Power Use
60
Power (W
W)
50
40
30
20
10
0
0
50
100
150
200
250
300
Time (s)
350
400
450
500
Mission Video
y View from vessel
y View from shore
Final Evaluation
Solar
Feasibility
Rugged
Vehicle
Fly
Planned
Path
Survey
Bridge
Wall‐
Following
y Vehicle survived well
y Deployment & leaks
y Path followed
y Solar power sufficient
y No time for
f bridge test
y Capability Present
y Wall
W ll ffollowing
ll i untested
t t d
Future Mission Objectives
y Add full GPS navigation capabilities
y Use Sonar for wall following
y Remote data logging abilities
y Actively control solar panel angle
y Test in warmer drier weather!!
Special Thanks to:
Franz Hover
Josh Leighton
Harrison Chin
Charlie Ambler
MIT Sailing Pavilion
MIT Towing Tank
Mechanical Engineering Department
Center for Ocean Engineering
Ch
Chevron
Schlumberger
S hl b
Q
Questions?
i ?
N lA
hi
SSpreadsheet
dh
Naval
Architecture
y Displacement v. draft
N lA
hit t
SSpreadsheet
dh t
Naval
Architecture
y GM calculation
N lA
hi
SSpreadsheet
dh
Naval
Architecture
y Trim Calculation
Fully Constrained Pontoon Attachments
Leads to stress concentrations at joints
Previous Designs
Pressure Transducer:
Calibration Curve
slope = -0.366 cm/Volt
MIT OpenCourseWare
http://ocw.mit.edu
2.017J Design of Electromechanical Robotic Systems
Fall 2009
For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms.