P20101: 3u Cubesat Flight Control

Preliminary Detailed Design

Table of Contents

Team Vision for Preliminary Detailed Design Phase

What did the team plan to do during this phase?

What did the team actually accomplish during this phase?

Feasibility: Prototyping, Analysis, Simulation

Sail Folding

The team started with three feasible methods to fold the sail for deployment.

1. The previous year’s method of folding into 4 rectangles

2. Similar to LightSail 2, fold into 4 triangles into compartments

3. Similar to NEA Scout, roll the sail

We did preliminary feasibility evaluations and tests on each of these methods:

In order to test the first method, we made a 1U sized box out of cardboard and placed the sails inside to check the fit. After folding the sail according to the last year’s team, the pieces fit, but had to be forced into the 1U size. The amount of shoving required to fit the sail would make packing the sail into the cubesat a challenge.

Image:Photo Gallery/foldingtable.tiff

Lightsail2 used the method described below. The sail parts were folded to 1U size horizontally, then half of the sail part was folded vertically towards the center following a triangular prism shape. These newly folded sails fit into a compartment of that shape.

Image:Photo Gallery/prismsail.tiff

The result of the first tries folding show that it was difficult to get the sails nicely folded since ours already had creases in it. Testing out using an iron to smooth out the fold showed immediate damage to the sail, even in the lowest setting.

In order to test the fit of the sail, a 1U box was 3-D printed with a cross into the middle to divide the space into 4 triangular compartments. The sail did not nicely fit into the compartment but could fit if squeezed in. This was not deemed a very feasible method for us to use.

Image:Photo Gallery/prismexample.tiff

The spindle method was tried with two different folding configurations. The sail quadrant was folded in half, then folded into strips as shown below.

Image:Photo Gallery/folding1.tiff

This first spindle method was difficult to fold, but was very easy to roll onto the spindle and compress the air out of it. Deployment was tested using two sail pieces and did not unroll very smoothly.

Image:Photo Gallery/folding2.png

public/Photo Gallery/sail_folding.gif

The second spindle method proved easier to remove air pockets and reduce the space take up by the folded sail. It was much more successful during deployment tests, and we were able to test deployment with all 4 sail quadrants. folding each quadrant took about 5 minutes with two people, so this method will be fairly quick to test multiple trials with.

Simple testing of the spindle deployment method is shown below:

public/Photo Gallery/deployment.gif

The sail had a rip, which caused one side to fail early on, but then recovered.

After these feasibility tests, we have chosen to use the second spindle method of folding the sail in half and then into vertical strips.

Tools for the sail folding will be needed. In particular, we have discovered that a long piece of metal that is the length of the sail side and the width of each fold section is useful to crease the sail while folding and keep folding consistent. For sail packing with a spindle, being able to roll outside the CubeSat would be ideal. If this cannot happen, an attachable handle tool on the top of the spindle would need to be designed.

For the second method, we can approximate the diameter of the sail roll to make sure it would fit in the CubeSat form. If we assume the thickness of tinfoil, 0.016 mm and neglect the fact that the sail is not the same thickness at all points of the folded strip due to the triangle shape, then add a factor of safety on the thickness of 1.5 to account for bend radius, we can plot a trend for the width vs. roll diameter. There is not an analytical solution to this problem, so each point was evaluated at intervals of 1 cm.

Image:Photo Gallery/widthvsroll.jpg

We can tailor our needs for the size of the CubeSat, to keep the whole sail deployment system down to 1.5U.

Sail Deployment

With the selection of a sail packing method, we were able to start making changes to the deployment system to try and reach our goal of 100% deployment.

During our first test of the deployment mechanism, after the booms were unrolled halfway they stopped unrolling and we could hear the sound of the tape measure folding inside of the spindle. From what we could see, it seemed that the booms were expanding off of the spindle inside the deployment mechanism and causing an increase in the friction. In order to fix this, The deployment cover was re-designed with tensioners added to keep the booms tight against the spindle. In the next design phase we will test this modification to see if it fixes the problem.

Another design decision we had to make was whether to have the sail or the deployment mechanism on top.After weighing the pros and cons, it was decided that it was much more feasible to put the sail on top for easy access when packing the sail and easier manufacturability.

The change of the sail folding to include a spindle also required the addition of one to our design. The spindle will be free-spinning as opposed to attached to the reel because the boom and sail will unroll at different speeds.

Deployment Sensor

After concept selection, we decided upon a new sensor for detecting deployment as opposed to the previous design. The main selection criteria focused on the issue of incompatibility of the last sensor, an opto-interrupter, with new types of sail booms. The new solution is designed to be robust, small, and adaptive to different boom technologies. The SpectraSymbol "Flex sensor," based a force sensitive resistor, will be used to sense the decrease in radius of the boom spindle reel as the booms are deployed. It is mounted on an already existing piece of hardware, the tensioners, with no additional manufacturing or design change overhead.

public/Photo Gallery/flex_sensor_22.jpg

The sensor has been tested in the lab to show a linear response. It is known to be affected greatly by temperature, meaning that a pre-calculated value will not be used, but instead a calibration will be made before deployment, and then the deployment status based on the linear slope of the data gathered. This avoids biasing errors due to temperature and environment.

public/Photo Gallery/flex_sensor_diagram.PNG

It has also been determined that the desired deployment accuracy of 5% can be met using this method, as opposed to the previous method, in which increasing accuracy would result in weakening the booms.

Diffractive Element Sensor

To measure the angular velocity applied to the CubeSat on its Z-axis by the diffractive elements - as is the goal of the experiment - a sensitive gyroscope is required. The following sensor (MPU-6050) was readily available at RIT, and was tested to determine if this small angular velocity could be measured at the desired precision.

public/Photo Gallery/angular_velocity_testbed.gif

A testbed was constructed using a motor with a known rate of rotation per applied voltage, 0.0875 radians/sec/V. The controller mounted on the motor is a custom designed embedded system intended for remote data-logging purposes. It runs from battery power and logs data to an SD card. This controller applied a rotation to the entire unit at a rate from 20% to 100% over the course of 250sec.

public/Photo Gallery/gyroscope_test.PNG

The experimental data confirmed that after filtering, the signal can be interpreted to get instantaneous angular velocity without the need for integration, which would not work with such a low acceleration. Another important note is that the gyroscope is capable of collecting the correct measurements regardless of the initial conditions of the system. The body can be rotating or disturbed prior to power-up.

Serial Interface

The serial interface decided upon for both communication to the main Flight Computer and to the sensors and external memory was SPI. The MSB(Most Significant Board) is going to be a slave to the Flight computer and will be able to recieve 8-bit commands from the Flight Computer. The sensors and memory will be slaves to the MSB.

public/Photo Gallery/SPI_setup.png

public/Photo Gallery/SerialInterfaceDiagram.png

Reaction Wheel Calculations


The governing equations for the ADACS system govern the relative motion of rigid bodies. These equations are shown in the figures below.

public/Photo Gallery/eqn142.PNG public/Photo Gallery/eqn143.PNG

For these preliminary calculations we impose an angular velocity to stay sun-normal through out the heliocentric orbit, and the external torques are only from solar pressure and are of the order 10^(-11) Nm. Following an example in the reference text "Orbital Mechanics for Engineering Student" (Howard, 2014), we arrive at the following equations that govern the motion of the reaction wheels throughout the orbit.

public/Photo Gallery/ADACS_equations.PNG

The reactions wheels 1, 2, and 3 are aligned with the x-, y-, and z-axes respectively. The preliminary calculations show the maximum spin of the reaction wheels will have to be approximately 6000 rpm. The plot below shows the spin of the reaction wheels as a function of time over the course of the year.

public/Photo Gallery/PDR_ADACS_Plot.png

Drawings, Schematics, Flow Charts, Simulations

Mechanical CAD Models

In this design phase, simple mechanical models were created to get an idea of space requirements and start to plan out the detailed design. Below is a picture of the initial cubesat model:

public/Photo Gallery/preliminarycadmodel.png

We are re-designing the structure of the cubesat so that it will be able to fit standard cubesat boards. This will allow us to purchase pre-made boards and an ADACS without modifying it or having to custom make one. The previous hinge design took up critical space, so we will be replacing it with auto-opening hinges available for purchase. They will serve the same function and take up much less space. The bright red object in the model is the placement of the new spindle int he top section of the cubesat.

All CAD models for this project, as well as their drawings in PDF format, are located here.

Software Flowcharts

The Software Diagrams can be found in the public EDGE repository.

public/Photo Gallery/MSBSoftwareArchitecture.png

public/Photo Gallery/FlightComputerSimulatorArchitecture.png

Electrical Architecture

The detailed electrical schematics for the Mission Specific Board (MSB), as well as the Raspberry Pi to PC104 adaptor board can be found on EDGE in PDF format here.

Proposed Flight Ready CubeSat

public/Photo Gallery/3U_Structure_Flight.PNG

MSD-II Deliverable Prototype CubeSat

public/Photo Gallery/3U_Structure_Deliverable.PNG

PC/104 CubeSat Bus

To standardize the boards, and ensure they can be connected to any COTS CubeSat purchased components whose manufacturers follow the PC104 standard, the following pinout diagrams have been devised. Maximum allowable voltages and currents, as well as signal types are specified.

public/Photo Gallery/Bus_Pinout_H1.PNG public/Photo Gallery/Bus_Pinout_H2.PNG

PC/104 PCB Final Form Factor

public/Photo Gallery/PCB_Form_Rev2.PNG

public/Photo Gallery/PCB_Stack_RevB.PNG

public/Photo Gallery/PCB_Stack_RevB2.PNG

Mission Specific Board Architecture

The Mission Specific Board, abbreviated as MSB, is the newly scoped in board, which provides all functions specific to this solar sailing CubeSat mission. These were derived directly from the transform diagram and brought down to specific electronic subsystems. The MSB is intended to be capable of being mounted onto any standard CubeSat COTS boards to perform its mission.


public/Photo Gallery/mission_specific_board_bd.PNG


Bill of Material (BOM)

The initial bill of materials is based on needs for the prototyping phase only. We have yet to estimate the final cost of the project, until we are certain the methods being used for sail deployment. This BOM is grouped by categories relating to the full system diagram of the CubeSat.

public/Photo Gallery/proto_BOM.jpg

Test Plans

ADACS Test plan

Deployment Test plan

Raspberry Pi Flight Computer Test

Testing strategy: Simulate mission lifetime

Risk Assessment

We have updated our risk management to detail new risks introduced by the designs created in this phase. Also, as a result of de-scoping, many old risks associated with components that are not specific to the mission, have been removed.

public/Photo Gallery/updated_risk.PNG

Design Review Materials

Plans for next phase

Our Gantt chart has been updated to reflect specific plans of individual team members as well as the full team in the detailed design phase.

Updated Gantt Chart

public/Photo Gallery/DDR_gantt.png

Team Detailed Design Phase Goals







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