P20101: 3u Cubesat Flight Control
/public/

Detailed Design

Table of Contents

Team Vision for Detailed Design Phase

What did your team plan to do during this phase?

What did your team actually accomplish during this phase?

Progress Report

Plan to Accomplish by Detailed Design Review

Tasks Accomplished as of 11/26

Remaining Tasks and Owner

New Decisions

Prototyping, Engineering Analysis, Simulation

Deployment Testing

We continued our testing of the sail deployment with some modifications to the deployment mechanism. The first test was with the new deployment cover and tensioners designed at the end of the last phase. We tested with just the booms, and while deployment was smoother at first, they quickly got stuck and were unable to extend further. The tensioners seemed to be putting too much force on the spool, and the motor was not able to overcome this force to deploy.
11/15 Deployment test with 3D printed cover and tensioners

11/15 Deployment test with 3D printed cover and tensioners

Next, we tested with no deployment cover at all. With this change, we were able to get the booms fully deployed. Based on this test, we decided to drop the tensioner idea and instead modify the current cover to work for us.

11/15 Deployment test with cover removed, successful 100% boom deployment

11/15 Deployment test with cover removed, successful 100% boom deployment

We next tested the deployment of one sail with our protoype spindle. The motor did not provide enough force in the extending booms to unroll the spindle. When we gave the spindle some help, the sail managed to deploy to 100%. When we took a closer look at the spindle assembly, it was determined that the cause of the spindle sticking was too much friction on the outer tube as a result of screwing it directly onto the deployment cover. Washers were added to prevent the free-spinning outer tube from contacting the deployment cover

11/15 Deployment of one sail with some help

11/15 Deployment of one sail with some help

In the next deployment test 2 sails were used, but when the booms tried to exted they were not stiff enough to pull the sails. From this test, we learned that the sail must be spiralled in the opposite direciton of the booms.

Correct sail spiral shown in blue with boom spiral shown in red

Correct sail spiral shown in blue with boom spiral shown in red

We went back to deploying with one sail, and noticed that the booms were again unable to extend with the sail. When only one sail quadrant is used, the forces on the booms are uneven, so one boom is unable to extend. You can see in the time lapse below that the top boom is bending instead of extending as the sail pulls on it in the direction that it easily buckles.

Timelapse of deployment test on 11/21

Timelapse of deployment test on 11/21

This should not happen when using all 4 sails, so in our last test we used all 4 sail quadrants. the first time, the sail was rolled in the wrong direction, so deployment was unsucessful. We re-rolled the sails and tried to deploy again, but one of the sail quadrants was attached wrong and made an extra wrap around the sails. This led to another unsuccessful deployment where we had to interfere and fix the sails.

Timelapse of deployment with all 4 sails on 12/2

Timelapse of deployment with all 4 sails on 12/2

12/2 Deployment test with all 4 sails, unsuccessful due to weak booms and attachment

12/2 Deployment test with all 4 sails, unsuccessful due to weak booms and attachment

Ice Rink Construction for Testing Deployment

The main deployment test will take place in MSD-II to prove deployment in a friction-less environment. To achieve close to friction-less conditions for the boom and sail deployment, a 45 foot by 35 foot ice rink was constructed in Jarrett's backyard. It is fit with a six inch tubing hole in the center for the CubeSat to sit in. The expected time at which it will be frozen and ready is January 2020. The deployment test, which is detailed further in the Test Plans sections, will take place in February 2020.

Ice Rink under construction

Boom Stiffness

In our deployment tests, we noticed that the booms had some trouble with pulling out the sail. Part of this was due to friction in the sail spindle, but the booms are also fairly flimsy. As this was an ongoing problem in our tests, we looked into alternative booms. One alternative option we considered was two tape measures bonded along one edge. This boom shape was also used in Lightsail and Nanosail D, therefore it has proven flight history. The second shape we tested was two booms bonded face to face, which is used on NEA Scout.

In order to test the stiffness, we took 2' sections of the steel booms and taped them into the desired shape. We them clamped these sections to a table and used weights to test their stiffness. The tables below show the results of this testing.

To test the feasibility of manufacturing these booms, we tested spot welding the booms, taping, and fusion. The steel of the booms were so thin that spot welding just created holes and did not hold up. Fusion help up for some bending, but started to break after repeated bending and pulling. This would possibly hold for a few tests, but wuld not stand up to repeated testing and use, and is not reliable enough for flight. the taped section held up well for one shape, but might have more trouble if the whole boom was taped.

Deployment Tools

The team saw the need to develop tools to help with deployment testing. A long flat metal piece was found and marked to be used to fold the sail. A hand tool to attach onto the boom spool and spacer to keep the spool disengaged from the gears of the motor were designed. This allows a faster way to reel in the booms once a deployment test is done.

Two handles attach to the sail spindle in an effort to make rolling the sails easier as well.

Deployment Sensor Feasibility

It is a key customer requirement that the solar sail’s deployment status be reported. The deployment motor and mechanism must be stopped when the full boom deployment of two meters is reached. This is also so that in the case that deployment fails, useful data can still be gathered by computing the area affected by solar pressure. The engineering requirement most important to this sensing is a resolution greater than or equal to 5% deployment, which is a resolution of 10 centimeters of boom length.

Original Photogate

The previous MSD team attempted to make deployment measurements using a photogate, also known as an opto-interrupter. This devices functions by outputting a signal when its emitter light is blocked from the view of its detector. The team had intended for these photogates, GP1A57HRJ00F, to detect painted on strips on the boom as it passed in between the sensor. This sensing method has three major drawbacks that conflict with our requirements, leading to it being revised.

Flex Sensor

The flex sensor was intended to solve the above issues by taking a direct measurement of the boom spool radius, which is proportionally related to the total length of deployed boom. It operated by varying resistance based on the angle of tensioners pressing onto the spool. This was a non-invasive method that could make a measurement with a predicted resolution of 1.25%.
Intended mounting for flex sensors on boom spool tensioners

Intended mounting for flex sensors on boom spool tensioners

However, the tensioner design was removed after it was discovered that the way the booms had been severely restricted in the previous design was the main issue with their stalling during deployment. When all contacts between the cover and the boom were removed, subsequent deployment tests have not experienced any stalling of the boom deployment.

Force Spring Sensor

The force spring sensor would operate in the same way as the flex sensor, and directly measure the radius of the boom spool as the boom deploys. The concept would have used a force sensitive resistor, commonly found in load cells, in combination with a spring that would decompress as the radius decreases. As the radius decreases, the force applied to the sensor decreases proportionally at the rate of the spring constant. The accuracy of this sensing method is acceptable for the application.
Force Sensitive Resistor measures spool radius to related to % deployment

Force Sensitive Resistor measures spool radius to related to % deployment

Ultimately, this sensing method was ruled out due to not meeting standards for risk tolerance. The main concern is method of adhering the spring to the rest of the structure could not be determined. Even if time was spent to design a more rugged sensor and spring enclosure, it was still deemed too risky because the mechanism can likely sustain damage during violent vibrations such as when the satellite is carried in the rocket launch.

Capacitive Sensor

The radius of the boom spool could also be measured by a non-contact method of capacitive sensing. Since the general equation for capacitance is inversely proportional to distance, a sensor could be situated next to the spool and measure its radius as the distance between the sensor and spool increases. This works because the tape measure spool is electrically connected to spacecraft chassis ground. To determine the capacitance, a frequency is passed through the capacitive sensor and then the resulting shift in phase is measured. This phase shift is measured by a phase detector circuit, also known as a double frequency mixer, against a reference signal. One big advantage of this deployment detection method is that it is independent from temperature offsets in space, since there is no atmosphere, so the only the permittivity of free space, e0, is used.
Capacitive Sensor measures spool radius to related to % deployment

Capacitive Sensor measures spool radius to related to % deployment

However, when the actual distances the sensor must measure were taken into account, especially after the boom spindle was resized to be larger, the sensor cannot measure the full range. The dynamic range of the sensor can be changed by modifying the reference circuit and reference resistance in the phase shifter, also known as the tuning filter. Using MATLAB, an interative solution was run for one million data points over the full range the sensor would need to detect, which is between 1mm and 23mm.

To detect the phase shift, the value of the phase must be kept between -80 degrees and 0 degrees. The equation governing this operation is based on the arctangent of the filter values over the distance. An iterative solution showed that no selection of values could achieve the full range, only small portions of it, which is unacceptable for deployment measurement.

Capacitive Sensor distance measurements plotted by range of Tuning Filter Coefficients

Capacitive Sensor distance measurements plotted by range of Tuning Filter Coefficients

Reflective Sensor

Assuming that Kapton tape can be applied to the tape measure, which is acceptable for space applications, a reflective method can be used. As Kapton pieces pass by the sensor, QRE1113, it emits an infrared beam, and then its infrared sensor outputs an analog value based on the reflectivity of the surface passing by. By monitoring this analog value, the sensor can be calibrated during flight, and minimize the risk of other light sources blocking the measurement.
Diagram of Reflectance Sensor Operation

Diagram of Reflectance Sensor Operation

A new PCB was designed to mount two sensors on next to the boom spool. It also includes the attachment for the deployment motor, so that the harness from the MSB up to the deployment structure can be kept all in one ribbon cable and minimize space.

P20101-021 PCB with mounted reflectance sensors

P20101-021 PCB with mounted reflectance sensors

Orbital Calculations

The orbital perturbations that we are most concerned with are J2, solar radiation pressure, and atmospheric drag. Atmospheric drag should take over once the satellite altitude is less than 500km. Our initial orbit is around 620km. This section was done with the assistance from chapter 12 of "Orbital Mechanics for Engineering Students 3rd Ed." and supplemental MATLAB code.
public/Photo Gallery/reference.jpg

public/Photo Gallery/reference.jpg

Using Guass’ variation of parameters method, the differential equations for solving change in orbital elements over time is:

public/Photo Gallery/guassgeneral.jpg

public/Photo Gallery/guassgeneral.jpg

Solar Radiation Pressure Facing Sun: The acceleration in the rsw frame:

public/Photo Gallery/radiation1.jpg

public/Photo Gallery/radiation1.jpg

he differential equations for solving change in orbital elements over time is:
public/Photo Gallery/rad2.jpg

public/Photo Gallery/rad2.jpg

The predicted results of 1 year:
public/Photo Gallery/radiationresult.jpg

public/Photo Gallery/radiationresult.jpg

J2 Perturbation: The acceleration in the rsw frame:

public/Photo Gallery/J1.jpg

public/Photo Gallery/J1.jpg

he differential equations for solving change in orbital elements over time is:
public/Photo Gallery/J2.jpg

public/Photo Gallery/J2.jpg

The predicted results of 1 year:
public/Photo Gallery/Jresult.jpg

public/Photo Gallery/Jresult.jpg

Results from adding the two perturbations together:

public/Photo Gallery/addresult.jpg

public/Photo Gallery/addresult.jpg

Atmospheric Drag Perturbation: The acceleration in the tangential velocity direction:

public/Photo Gallery/drag1.jpg

public/Photo Gallery/drag1.jpg

The differential equations for solving change in orbital elements over time is:
public/Photo Gallery/drag2.jpg

public/Photo Gallery/drag2.jpg

In Progress:

ADACS Model

The ADACS model is still in progress. We will report the current state and the path forward. The math for the ADACS analysis is properly laid out in the PDF linked below.

ADACS Analysis Mathematics

Currently, this model is undergoing verification. The first step in this process is to make sure the results make physical sense. The easiest way to see this is by calling all body velocities and accelerations except the body velocity in the x-axis. As stated in the PDF, the body is experiencing a constant angular acceleration in the z-axis. Since we are demanding, in this case, that the body angular velocity be zero (with an Earth inertial reference frame) the results should show the third wheel, the wheel in the z-axis, increasing over the course of the mission life. The graphs below show the momentum wheel velocities and the angular momentum of the spacecraft.

Momentum Wheel Angular Velocities

Momentum Wheel Angular Velocities

Spacecraft Angular Velocities

Spacecraft Angular Velocities

The wheel angular velocities are massive, and for good reason. Through one orbit the angular velocity of the spacecraft along it's z-axis increases massively. Obviously, the momentum wheels will not be able to cope with this increase beyond reason. Most momentum wheels cap at 6000 rpm for a given spin inertia of the disk it holds to store momentum. However, for this experiment we will not be fighting the spin in z-axis. While this will not saturate the third wheel the angular momentum of the entire spacecraft will increase due to the Solar Radiation Pressure (SRP), thus making it more "difficult" to control the attitude of the spacecraft. This will in turn eventually cause the wheels to saturate to stay Sun pointing throughout the mission, and coping with any orbital perturbations that may arise. This increase in angular momentum can be seen in the second graph. These results are calculated using MATLAB and Simulink. The simulink used for this analysis is displayed below.

Momentum Wheel Angular Velocities

Momentum Wheel Angular Velocities

ADACS Analysis MATLAB Script (NOT FINAL)

ADACS Analysis Simulink File (NOT FINAL)

A preliminary PID control law for one wheel in relation to rotation of one axis of the satellite has been identified. The input to the system is a spacecraft reference velocity, the output is the modeled actual velocity. The simulink file is shown:

public/Photo Gallery/sim1.jpg

public/Photo Gallery/sim1.jpg

public/Photo Gallery/sim2.jpg

public/Photo Gallery/sim2.jpg

Correct parameters have not been found yet, but an example curve with estimate parameters:

public/Photo Gallery/examplecurve.jpg

public/Photo Gallery/examplecurve.jpg

Design: Drawings, Schematics, Flow Charts, Simulations

Mechanical Design

A new frame has been designed to allow easier testing and allow PC104 boards to fit. The new structure consists of rails made of 1/16" wall aluminum angle with feet on each end. The top and bottom plates are made of 0.25' aluminum, and the top plate is in the shape of an x to allow for the triangles on the side panels to slot in. This frame was designed for ease of testing and manufacturability, while still being structurally sound.

New Cubesat Design

New Cubesat Design

The sailand deployment mechanism take up almost 2U together, while the electronics stackup takes up less than half a U. This leaves 2.805" of space for a future ADACS system to use.

Space Allocation in the Cubesat

Space Allocation in the Cubesat

The spindle is made of a machined shaft, a tube, and ball bearings. There are threads for 4-40 screws on each end. It is symmetrical either orientation. There are slotted notched straights on the shaft so pliers can be clamped on it.

Cross-section view of the spindle assembly

Cross-section view of the spindle assembly

The deployment cover has been slightly modified by removing the side flanges and adding a countersunk hole. A 3DPDF of the model can be found here Detailed Design Documents\CAD\PDR Model\DDRModel.pdf

Software Design

Electronics Design

The full prototype electronics necessary to support all desired demonstration CubeSat functions have been designed. Refer to the main architecture specification document to view a detailed description of the stack-up and its electrical specifications.

Each board that will be manufactured, assembled, and tested here at RIT during MSD-II is listed in the following table. The datasheet provides interfacing, power, and specification information for each PCB. A schematic is provided, drawn in Altium Designer, that represents the actual circuit as laid out on the PCB.

Design Number Board Name Abbreviation Datasheet link Schematic link
P20101-010 Electrical Power System EPS Datasheet Schematic
P20101-011 Solar Array Simulator PVSim Datasheet Schematic
P20101-020 Mission Specific Board MSB Datasheet Schematic
P20101-021 Deployment Sensor and Motor DSM Datasheet Schematic
P20101-030 Raspberry Pi Emulated Flight Computer EFC Datasheet Schematic
P20101-031 Emulated Reaction Wheel ERW Datasheet Schematic

Bill of Material (BOM)

The full excel spreadsheet of the Prototype Bill of Materials is located on EDGE. This prototype BOM currently reflects the order planned to go out over the Winter Break between MSD-I and MSD-II, and includes all necessary parts for the build and test.

Categories that items are described by include:

BOM

Budget

The current budget is made up of a base $500 provided by MSD and the left over $251.55 carrying over from P19101 for a total of $751.55.

An application to the NASA/New York Space Grant Consortium was put in for a total of $2000, and is expected to receive information on whether or not the reward will be given by December 10th, 2019.

Considering the prototype BOM contains a total of $531.27, there is a remaining $220.28 left if the grant is not awarded. This has been determined to be enough for unaccounted testing purposes during MSD-II.

In the case that the grant is awarded, the following budgetary options were given for using the $2000 award:

Test Plans

The team has drafted new test plans based on the detailed designs done in this phase. These tests are subsystem tests intended to be taken on during the subsystems build and test phase in MSD-II.

Mechanical Test Plans

Reaction Wheel/Control Law Test

This test is planned to happen at the start of the next phase and will use one reaction wheel and brushed DC motor to test a control law and the electronics. Once this is working, we will scale up to using 3 wheels and motors with a more complex control law. After this test, we will have a proven control law for 3-axis control of a system with reaction wheels.

Sail Folding Procedure Test

This test will also happen in the beginning of the next semester. It will consist of folding the sail 3 times with assisting tools to finalize the procedure and get an average time needed to fold the sail. After this test, we will be able to make a document outlining the proper folding procedure that our customer and the next team can use to fold the sail.

Deployment Test

This test will happen in mid-late February 2020 in a home-made ice rink in Jarett's backyard. The deployment mechanism in a special frame will be sunk into the middle of the rink and then the rink will be smoothed out. Testing will be done on the ice to get as frictionless of a surface as possible. This test will be done with all 4 sails. After this test, we will hopefully have proven that our system reached 100% deployment un-aided with all 4 quadrants.

Electrical Test Plans

Each previously described system for the electrical architecture will be tested in the EE senior design lab to satisfy engineering requirements and electrical safety and quality metrics. All testing will be done under proper ESD safe conditions in the "Flat-Sat" format, in which all PCBs and equipment are laid out for testing. The image below describes the proposed test bench setup using all equipment currently available to the MSD team.
test bench

Deployment Sensor

Success: Under all conditions, 20 counts made.

Power Systems

Success: No anomalies in expected power draw found.

Mission Specific Board

Success: All interface impedances match predicted impedances

Emulated Flight Computer

Success: All interface impedances match predicted impedances

Emulated Reaction Wheel

Success: Zero error and maintains temperature under 70C.

Software Test Plans

Risk Assessment

Two new risks have been added that relate to the ADACS hardware and possible failures.

Design Review Materials

Plans for next phase

Gantt chart with our plans for MSDII

Gantt chart with our plans for MSDII

Team Goals by End of MSD-I

Team MSD-II Preparation

Individual Goals

Amber

James

Sarah

Charlie

Nick

Jarrett


Home | Planning & Execution | Imagine RIT

Problem Definition | Systems Design | Preliminary Detailed Design | Detailed Design

Build & Test Prep | Subsystem Build & Test | Integrated System Build & Test | Customer Handoff & Final Project Documentation