DRAFT-Sat
Preliminary Design Review
2003-2004 Senior Projects
University of Colorado
Aerospace Department
Objective

Provide a low budget system that can remove
orbital perturbations from a satellite trajectory.




System must compensate for perturbations in 3 axes
causing the satellite to move 1cm or more from an
orbital trajectory.
System must be self contained.
System must be testable both in microgravity aboard
the KC-135 and in 1g of gravity on the surface of the
Earth.
Applications in preventing orbit decay and
improving satellite position knowledge.
2
Background

2002/2003 Drag Free Team
Tested a 2-D prototype.
 Verified the prototype could hold the position
of the system to within 1cm for 10 seconds on
an air table.
 CO2 propulsion system with a thrust of 0.7 N
to control the prototype.
 Data processing done on a PC in MATLAB.
 Bang-Bang control law with derivative gain.
 Powered by external source.

3
STRUCTURE
Chris Erickson
4
Structural Design
Structural Design Drivers:
•High strength / weight
-Minimize mass for air table testing.
•Robust enough for handling
-Safely below yield strength during handling and assembly.
•Meet G-loading and drop requirements for KC-135
-Meet KC-135 flight requirements of:
9 g forward loading.
3 g aft loading.
6 g downward loading.
2 g lateral loading.
2 g upward loading.
Sustain 4ft drop in 0.75 g environment.
•Ease of Machining
5
Architecture Comparison
-Compare architectures using structural design drivers.
Due to unknowns with component weight, g-load and drop tests will
not yield sufficient comparisons between structures at this point of the
design.
Instead, the structures will be compared by their robustness to
crushing, a good measure of overall strength and handling
robustness.
-Use FEM models of structures for comparison.
Load Models with 200 lb on their upper surfaces.
Constrain vertical movement of lower surface.
6
Comparison:
Study is done using
6061-T6 Al
Comments:
Minimum
FOS
0.063
0.58
Octagonal Ring is nearly 10x
stronger.
(based on yield strength)
Weight
0.88 lb
0.91 lb
Nearly identical weight.
Machining
Complexity
Complex
Simple
Octagonal Ring is much
7
simpler to machine.
Structural Materials
Material Design Drivers:
•High strength & stiffness to weight ratio.
•Easy to machine.
•Readily available.
•Low cost.
The options to choose from are:
•Aluminum Alloy
•Plastic
•Steel
•Titanium Alloy
8
Structural Risks
Options
*
*
Comments
Cubic Structure
Lots of machining, poor
performance. May not be able
to meet mass limitations with
this structure.
Octagonal Ring Structure
Easy to machine, high
performance, light weight.
Risk
Performance
Cost
Materials:
Aluminum
Low density, high stiffnessweight, easy to machine, readily
available.
Plastic
Easy to machine, high
performance, light weight, low
stiffness.
Steel
High Density, difficult to
machine.
Titanium
High cost, difficult to machine.
9
Octagonal Ring Assembly
Conceptual Assembly of Proposed Design
10
SENSING
Ryan Olds
11
Sensing Options
Options
Comments
Accelerometer
Temperature drift, small size,
readily available, needs extra
hardware and software.
Digital Surveillance Camera
Digital Interface, small size, no
documentation or information
from manufacturer.
ActivRobotics Pan-Tilt-Zoom
Camera
Large size, has built in color
tracking capability.
CMU cam / CMU cam 2
Heritage with project, reusable
software, small size, on board
color tracking.
Risk
Performance
Cost
12
CMUcam2


Heritage
 CMUcam1 used by 20022003 Drag Free team to
successfully control a 2-D
plate to within 1cm.
 Much of the software is
already written and tested.

Requirements
 5V DC Power Supply.
 Draws 200mA.
 RS-232 or TTL serial
communication with a
microcontroller.
 Needs to be 2 inches from the
object it is tracking.
 Tracked object must have a
sharp contrast with surroundings
(ex. Red on White).
Specifications
 4 - 17 frames per second.
 2.25" wide x 1.75" high x 2" deep.
 Up to 160 x 288 Resolution.
 B/W Analog video output
 SX52 Processor (33% faster than
previous system).
 Can operate between 1200 and
115200 baud.
 Wide angle lens can increase 13
FOV
to 55.
Proof Mass / Cavity Requirements

Proof Mass

Needs to be visible to
the camera.




Contrasts with
Background.
Must be at least 5% of
camera FOV.
Must be capable of
movement in 3 axes.
Cannot be perturbed.

Cavity



Needs to be
transparent so that the
proof mass is visible.
Must provide adequate
light so the proof mass
is visible.
Must allow the proof
mass to displace more
than 1cm.
14
Proof Mass / Cavity

Proof Mass



Bright Color (Red) to contrast
with cavity.
1cm diameter (allows proof
mass to be 10%-20% of
camera FOV)
Cavity

5cm sided cube.



2cm 5cm
Allows proof mass to
displace as much as 2cm.
Clear plastic walls allow the
camera to see the proof
mass.
LEDs
Transparent
Walls
2cm
5cm
Error Box
15
Camera Orientation for 3-D Color
Tracking
CMUcam2 resolution (160 x 288)
55
55
5cm
5cm
5cm
Zres = 92 pix/cm
55
Zres = 4 pix/cm
Yres =51 pix/cm
z
Xres = 92 pix/cm
y
Yres =51 pix/cm
Xres = 92 pix/cm
16
x
Design Issues and Risks
ISSUES/RISKS
BACKUP PLAN
1.
CMUcam2 is a new product
and might have bugs.
1.
Use original CMUcam that was
used last year.
2.
CMUcam had problems
running in obscure modes.
2.1
3.
CMUcam might have trouble
seeing the proof mass
because of lack of light, glare,
shadows, etc.
3.
Use LEDs inside the proof
mass cavity.
4.
CMUcam won’t be available
until late October. Could be
delayed.
4.
Use original CMUcam that was
used last year.
New documentation has been
found to deal with this problem
if it occurs.
2.2 Contact professors at
Carnegie Mellon University
who work with the camera.
17
DATA ACQUISITION
Mike Cragg
18
Microcontroller Requirements

Data Rate


Processor Speed


1280-5440 Hz
Memory


128-544 Hz
13600 bytes
Interface


RS-232
TTL
19
Control Flow Chart
CMOS
Image Array
CMOS
Image Array
CMUcam2 (2)
Buffer
Buffer
CMUcam2
Vision Board
CMUcam2
Vision Board
RS-232
or TTL
RS-232
or TTL
Microcontroller
Microcontroller
RAM
PC
RS-232
TTL
Solenoids (6)
TTL
20
Microcontroller Options

Motorola



Atmel AVR
Microcontroller


Expensive
Complicated

Motorola
AVR
Inexpensive
Microchip PIC
Microcontroller

Cost Performance Risk
PIC
Inexpensive
Firsthand knowledge
21
Risks/Off Ramps

Not enough RAM



Insufficient Processor
Speed



Will need umbilical to
download data directly to
PC  Not self contained
Will still be able to function,
but no verification

RAM


Processor


Umbilical
Lower data rate
Interface

Utilize RS232 and TTL
connections
Slow reaction time
Limited control
Interface

Unable to download data to
PC
22
PROPULSION
Katie Dunn
23
Propulsion Flow Chart
Propellant
Piping
Fluid Regulation Device
Piping
Power
Fluid Release Device
X6
Data
Piping
Nozzle
X6
24
Propulsion System Requirements




Provide enough thrust to control the system
Provide control in three axis
Build to KC-135 requirements
Must be able to withstand the pressure needed
to obtain the required thrust
25
Design Alternatives
Options
Fluid Regulation
Device
Fluid Release
Device
Comments
Regulator
Used to regulate the pressure as it
comes out of the propellant tank.
Orifice
Extra manifold for holding
depressurized gas would be
needed.
Our
switch/valve
design
Design a valve to allow fluid to flow
to nozzle using a switch to trigger
the actuator.
Solenoid Valve
Uses electromagnetism to open and
close a valve allowing fluid to flow in
and out.
Risk
Performance
Cost
26
Design Alternatives, cont’d
Options
Propellant
Piping
Comments
CO2
CO2 is stored at high pressures in
liquid form, usually in a small canister
of some kind. It is an inert gas that
will not harm the users.
Air
Air is stored in it gaseous form in
large containers. It is an inert gas
that will not harm its users.
Propane
Vapor pressure at room temperature
is 96 PSI, making the need for
pressure regulation in the system not
needed.
Copper
Copper is heavy, but holds pressures
as high as 2300 psi with a 0.035 in
wall thickness
Plastic
Plastic tubing is not very heavy, but is
not very strong.
Risk
Performance
Cost
27
The Solution
Thrust must be
able to counter
a 1 cm/sec
initial velocity
Structure
mass is 5kg
or less
Thrust must be able to
control the position of
the system to within 1
cm
Model thrust using kinematics physics
equations
Graph of force
vs time
Steve Graph
Range of thrust the system must produce 0.1 N – 0.7 N
Choose
solenoid based
on thrust range
Characterize
power of
solenoid
Choose
batteries to
supply power
Pick material to
hold specific
pressure28
Risks and Outstanding Issues

Miscalculation of thrust
range



Drives many components
of system
Use regulator on apparatus
so it can be adjusted
(within a certain range)

Basic model did not
include


Friction of air table
Pressure losses in piping
Inability to meet specified
power consumption
requirement for solenoids

OFF RAMP-use external
power
29
POWER
Stephen Levin-Stankevich
30
Cameras(2)
5V @ 200mA
5V Regulator
Batteries
COTS
Data
Acquisition
System
5V @ 200mA
On/Off
Switch
V >~ 8V
I < ~ 1A
High V
High I
Regulation
Circuit
Propulsion
Solenoid
Actuators
31
Power Budget
Component
Current (mA)
Voltage (V)
Power
(W)
Duty Cycle
(%)
Camera (2)
400
5
2
100
2
Data Acquisition
200
5
1
100
1
Solenoids(6)
3000
12
36
20
7.2
Total
Operating
Power (W)
10.2
32
Batteries

AA rechargeable batteries
Provide simple interface
 Low cost, lightweight, and easily obtainable
 Easily adaptable if system requirements
change


Propulsion design may require alternative

If high current and voltage cannot be supplied
through a capacitor circuit to the solenoids a
Li-Ion battery may be required for high current
draws.
33
Batteries – Trade Study Table
Options
Manufacturer
Comments
COTS
AA
Batteries
Duracell,
Energizer,
etc
1.5V each may be put in series
for required V. Low cost and
easy interfacing makes for
good design choice
COTS
AA
Rechargeable
Radio
Shack
1.25V each similar to standard
AA except may be recharged
to reduce overall cost.
Li-Ion Laptop
Computer
Battery
Computer
Manufacturer
V up to 14.4V and long lifetime
are advantageous. High cost,
weight, and possible difficult
interface are disadvantages
External Power
Supply
Risk
Performance
Cost
If analysis shows charging
solenoids with internal supply
is not feasible an external
supply may be used. (OffRamp)
34
SOFTWARE
Stephen Levin-Stankevich
35
Control System Block Diagram
Camera Sensors
X,Y,Z pos
File on
microcontroller
Error
Estimation
In Software
Control Law
(trade study)
Thrust
Feedback Loop
Courtesy of 2003 DFS Team
36
Control System
Last year’s results show the bang-bang
controller provides accurate results.
 Primary work will be developing the control
software for use with the microcontroller
 Goal is to reduce dead-band size by
implementing smaller thrust
 Analog control of pressure for a P-D
controller will be a project on-ramp.

37
Control System – Trade Study
Table
Options
Manufacturer
Comments
Bang-Bang
Control
N/A
Applies on/off control to thrusters at
each sampling time. Negligible power
difference. Accomplished in 2-D
previous year. Consumes more gas.
Uses a dead-zone to prevent noisy
output.
Proportional
and
Derivative
Control
N/A
Applies proportional control to thrusters
based on position and velocity
feedback. More accurate control
possible. Less gas consumption.
Requires analog/digital pressure
regulator adding circuit complexity.
Requires redevelopment of control law
from previous year.
Risk
Performance
Cost
38
Testing and Verification
Chris Erickson
Testing Options

Pendulum Setup


Spring Mass Setup


Attach the structure to a spring system in all 3 axes
and verify position control.
Air table Setup


Suspend the structure from a tether and verify
position control.
Test position control 2 axes at a time on the air table
(test 2 axes and flip structure)
KC-135

Test position control in a micro-gravity environment.
40
Testing Options Trade
Options
Comments
KC-135
Microgravity environment allows
control in all 3 axes to be verified at
once. Design must conform to all
NASA safety requirements.
Air Table
Provides testing and verification in 2
axes at a time. Near Frictionless
surface allows the mass to slide on a
2-D plane. This test environment was
demonstrated by the 2002-2003 Drag
Free team.
Pendulum
The system oscillates excessively and
is not an ideal environment for a test.
Large perturbations from gravity
present.
Spring
Wild oscillations occur making this test
environment unattractive. Large
perturbations due to spring forces.
Risk
Performance
Cost
41
Project Management Plan
William Lumbergh
Prof Penina Axelrad
Office: ECAE 159
Phone: (303) 492-6872
Prof Steve Nerem
Drag Free Spacecraft
Project Manager
Ryan Olds
Office: ECAE 100
Phone: (303) 492-6721
Safety Engineer
Stephen Stankevich
Software
Stephen Stankevich
Katie Dunn
Mike Cragg
Ryan Olds
Mechanical Design
Engineer
Chris Erickson
Structure
Chris Erickson
Instrumentation
Engineer
Mike Cragg
Data Acquisition
Mike Cragg
Professor
Advisory
Board
Chief Financial Officer
Mike Cragg
Propulsion
Katie Dunn
Mike Cragg
Power
Stephen Stankevich
Sensing
Ryan Olds
43
DRAFT-Sat
Work Breakdown Structure
1.0 Management
1.1
Schedule
2.0 Systems
2.1
Integration
of
subsystems
3.0 Testing
4.0 Software
3.1 Test
planning
4.1 Extend
existing
code to 3
dimensions
5.0 Structure
6.0 Data
Acquisition
7.0 Propulsion
6.1 Select
microcontroll
er
7.1 Select
propellant
and
solenoids
8.1 Select
power
source
9.1 Select
sensors
8.2 Supply
power to all
subsystems
9.2 Proof
mass and
cavity
2.2 Design
requirement
s
3.2 System
and
subsystem
testing and
verification
4.2 Improve
performance
of PD
controller
5.2 House
all
subsystems
6.2 Program
microcontroll
er
7.2 Thruster
model
1.3
Financial
2.3 Trade
Studies
3.3
Technical
reports
4.3
Translate
MATLAB to
C language
5.3 CG
placement
6.3 Interface
with sensors
and
propulsion
7.3
Propellant
piping
system
5.4 Machine
structure
6.4 Store
data in
memory
1.4 Team
organization
9.0 Sensing
5.1 Select
materials
1.2 Task
managemen
t
3.4 Conform
to all KC135
requirement
s and
regulations.
8.0 Power
9.3 Interface
with
microcontroll
er
Schedule
45
Cost Estimates
Subsystem
Part
Quantity
Unit Cost
Cost
$0.00
$0.00
Software
N/A
Structure
Aluminum
TBD
~$1000.00
PIC Microcontroller
2
$10.00
$20.00
PIC Board
2
$50.00
$100.00
Misc Hardware
TBD
Solenoid Valves
6
$50.00
~$300.00
Pressure Regulator
1
$200.00
~$200.00
Nozzles
6
$50.00
~$300.00
CO2 Canisters
TBD
$5.00
~$50.00
Piping
TBD
$50.00
TBD
Batteries (Li-Ion)
TBD
$150.00
~$150.00
IC Parts
TBD
$5
$50
CMUcam2
2
$109.00
$218.00 46
Total =
~$3000.00
Data Acquisition
~$200.00
Propulsion
Power
Sensing
Open Issues
Propulsion System Thrust Sizing and Part
Selection
 Microcontroller Setup
 Battery Selection

47
Questions?
48
Appendix
49
Sensors and Actuators

Actuators
Propulsion


Fluid regulation device
Fluid release device

Sensors
Control System


Propulsion


Camera
Fluid regulation device
Power

Voltage monitor
50
Background
Control System
 Propulsion

51
Control System
Previous Work

Last year’s team designed a bang-bang
controller with a derivative gain.

The controller was run through MATLAB on a PC
interfacing with output ports.
52
Propulsion System Background

2002/2003 Team








Propellant: CO2
Fluid Regulation Device: GO Regulator with gauge
Fluid Release Device: Solenoid
Nozzle:
Amount of Thrust: 0.7 Newtons
Four thrusters for 2-axis control
Interaction with Bang-Bang control system
Thrusters remained on until system was physically
back inside dead band
53
KC-135 Regulations
Structure
 Power
 Propulsion

54
Structure Requirements
9 g forward loading.
 3 g aft loading.
 6 g downward loading.
 2 g lateral loading.
 2 g upward loading.
 Sustain 4ft drop in 0.75 g environment.
55
Power Safety Concerns

All batteries must be of the dry-cell or gel-cell
type


Wiring shall be copper and meet sizing
requirements


Most standard batteries fit this category
Small power consumption of our system makes this a
minimal concern
An accessible power “kill” switch must be placed
on experiment structure.

A switch will be added to effectively cut-off battery
supply to system.
56
Propulsion Safety Requirements




Max allowable working pressure (MAWP) must be
greater than a factor of 4 compared to material
ultimate strength and a factor of 2 when compared to
the yield strength
Must have a pressure relief system set to no larger
than 10 percent above MAWP to prevent over
pressurization
Pressure gages must be sized to indicate a minimum
of 150 percent and maximum of 200 percent of
systems MAWP
Pressure regulators must have pressure relief
mechanism internal to gauge and properly calibrated
57
Fluid Regulation
Device


Withstand loads of CO2 canister
pressure: ~6250 kPa (~900 psi)
for CO2
Have a gauge so operators can
monitor/change regulation
pressure
Fluid Release
Device



Nozzle


Withstand pressures from the
solenoid
Disperse air in a direct fashion, 90
deg from surface of structure
Withstand pressures given
from fluid regulation device
Obtain a signal from
microcontroller of when to
actuate release of fluid
Must provide sufficient
pressure to nozzle to provide
the correct amount of thrust to
apparatus
58
Propellant




Supply min N - max N
of thrust
Contained within the
structure
Enough propellant for
30 seconds of
thrusting
Safe for student use
Piping


Withstand pressures
in regulator portion
and solenoid portion
Connect to fluid
regulation and fluid
release device
without leaking
59
Structure
60
Cubic Architecture
Minimum Factor of Safety:
0.063
Structural Weight:
0.88 lb
(based on 6061-T6 Al)
Factor of Safety Distribution
(based on yield strength)
61
Octagonal Ring Architecture
Minimum Factor of Safety:
0.58
Structural Weight:
0.91 lb
(based on 6061-T6 Al)
Factor of Safety Distribution
(based on yield strength)
62
Structural Materials
Aluminum Alloys:
3 main alloys:
•2xxx – least expensive, used in aircraft structures.
(Yield = 255 MPa for 2025-T6)
•6xxx – Easier machined than 2xxx, stronger than 2xxx, can be
anodized. (Yield = 275 MPa for 6061-T6)
•7xxx – Most expensive, highest strength, can be anodized.
(Yield = 505 MPa for 7075-T6)
Our choice: 6xxx aluminum unless higher
strength is needed, then 7xxx aluminum.
63
Sensing
64
Accelerometers


Advantages
 No need for a proof mass.
 Small requirements on the
structure.
 Small size
Disadvantages
 Temperature drift will cause
inaccuracies for small g
loading without temperature
control.
 Onboard A/D conversion
needed.
 New software needed.
MJ Electronics
Digital Surveillance
Camera


Advantages
 High Resolution (420 lines)
 Wide Angle
 Small size (1.87” x 1.0” x
1.87”)
Disadvantages
 Very limited documentation.
 New software needed for
65
object tracking.
ActivRobotics
Pan-Tilt-Zoom
Camera
System


Advantages
 Color tracking
software is built in.
 Plenty of
documentation and
heritage with several
robotic devices.
Disadvantages
 Large size.
 Pan-Tilt-Zoom
functions are not
needed.
CMUcam/CMUcam2


Advantages
 Heritage and compatibility
with last year’s project.
 Color tracking function built
in.
 CMUcam2 has improved
resolution (160x288)
 Small size
Disadvantages
 Had trouble running in
obscure modes.
66
CMUcam2 System Architecture
CMOS
Image
Array
Digital
Image
Data
Cmd
CMUcam
Vision Board
RS-232 or TTL
Serial
Proof Mass X-Y
Position Data
(On board Color Tracking) Camera register
Initialization
Data
Microcontroller
(Uses data and control
Algorithm to activate
Thrusters)
200mA
5V DC Supply
67
Testing
68
KC-135 Environment


Microgravity
environment allows
control in all 3 axes to
be verified at once.
Design must conform
with NASA safety
regulations.

Regulations apply to
system structure,
power, and propulsion
subsystems.
69
Air Table Setup



Near Frictionless surface
allows the mass to slide
on a 2-D plane.
This test environment
was demonstrated by the
2002-2003 Drag Free
team.
The absence of spring
forces and pendulum
oscillations makes this
environment favorable.

Can only verify control in
2 axes at a time.
70
Pendulum Setup
The dynamic response of
a pendulum system was
analyzed in MATLAB.
Result:
 The system simply
oscillates and is not an
ideal environment for a
test.
 Large perturbations from
gravity present.
Dynamic Response of Pendulum System
0.02
Position
Thruster Setting (-0.005 = on, -0.01 = 0ff, -0.015 = reverse firing
Desired Position
0.015
0.01
dx (m)

0.005
0
-0.005
-0.01
-0.015
2
4
6
8
10
12
time (s)
14
16
18
20
71
Spring Setup
The dynamic response of
a 3 axis spring system
was also analyzed.
Result:
 Wild oscillations occur
making this test
environment unattractive.
 Large perturbations due
to spring forces.
x or y response
0.04
0.03
0.02
position (m)

0.01
0
-0.01
-0.02
0
5
10
15
time (s)
20
25
30
72
Thruster
POWER
INTERFACE
Solenoid
CMUcam 1
(2-D view)
Solenoid
Thruster
Solenoid
Pressure
Regulator
Thruster
CMUcam 2
5V Regulated Line
12V Regulated Line
Solenoid
73
Thruster
Thruster
SERIAL
INTERFACE
TTL Serial
Solenoid
CMUcam 1
(2-D view)
RS-232 or
TTL Serial
TTL Serial
Solenoid
Thruster
Solenoid
RS-232 or
TTL Serial
Pressure
Regulator
Thruster
CMUcam 2
TTL
Serial
TTL Serial
Solenoid
74
Thruster
Thruster
PROPULSION
INTERFACE
Solenoid
CMUcam 1
(2-D view)
Solenoid
Thruster
Solenoid
Pressure
Regulator
Thruster
CMUcam 2
Regulated Pressure
Solenoid
75
Thruster