Preliminary Design Review

Transcription

1 Preliminary Design Review Purdue University Thermodynamic Experimental Microgravity Platform May 1, 2002 Purdue University Luca Bertuccilli Christopher Burnside Javier Lovera Thomas Martin Timothy Sanders Stephanie VanY

2 Executive Summary 1

3 Table of Contents 1.0 Mission Statement 2.0 Mission Objectives 3.0 Satellite Description 4.0 Concept of Operations 5.0 Major Design Requirements 6.0 Orbit Selection 7.0 Launch Vehicle Integration 8.0 Spacecraft as a System 9.0 Payload Subsystem 10.0 Attitude Dynamics and Control Subsystem 11.0 Track, Telemetry and Control Subsystem 12.0 Command and Data Handling Subsystem 13.0 Power Subsystem 14.0 Thermal Subsystem 15.0 Structures and Mechanisms 16.0 Summary of Spacecraft 17.0 References 18.0 Appendices 2

4 1.0 Mission Statement Every aspect of our society is infused by space technology in some way. Our dependency on information is more evident everyday as our economy moves toward globalization and information technologies. This dependence drives us to provide better methods to harness space assets. As our abilities grow and we capitalize space more we must optimize our techniques, and invent new ones, to meet the growing demand our customers and beneficiaries. This movement towards optimization is a trend we see in many industries. The most prevalent is the commercial aircraft industry. The current state of the art airliners are optimized designs of former aircraft and are of proven design. Over the last 30 years aircraft have not changed in general characteristics, but have become safer, more economical and better in many other ways. Commercial spacecraft designs are entering a similar era. The first 40 years of spacecraft testing and operation have given rise to acceptable methods to produce spacecraft. The next 40 years will see an increase in efficiency and monetary return of space assets. It is this movement towards optimization, which requires refinements in our methods. An area where refinement of techniques is occurring is in the propellant management and gauging field. Although this is not the only field, it is an important and integral part of each spacecraft in orbit. The goal of propellant gauging is creating and refining methods to report how much propellant is left in the spacecraft as the satellite performs its duties in orbit. Several methods of gauging are currently in use. One in particular uses the thermal properties of the propellant to compute the volume of fluid left in the tank. The technique measures the length of time it takes a change in temperature to occur with a known amount of heat energy added to the tank. The idea can be simplified to heating a cup of water in the microwave oven. Say you want to heat a cup of water to boiling. If the cup is full it will take longer to boil, but if the tank is almost empty you will wait a much shorter time to have boiling water. The spacecraft uses a similar technique to gauge propellant remaining in the tank. Purdue University is proposing a satellite whose primary purpose is the study of thermal propellant gauging techniques. Presented in this document is a preliminary design for the satellite. We have affectionately named the satellite PuTEMP, or Purdue University Thermodynamic Experimental Platform. Throughout this document we refer to it as PuTEMP. This design document covers the each of the necessary subsystems and the designs of each. We have included much information in the appendix, which demonstrates the trade studies and other work needed in the design of this satellite. 3

5 2.0 Mission Objectives 2.1 Primary Mission Objective The Primary Mission Objective is to collect data to assist in the generation of more accurate computer simulations used in industry for propellant gauging. Through the development of a student-built micro-gravity experimental platform an accurate satellite fuel tank model will enable onboard readings accurate volume measurements through temperature gradient. 2.2 Secondary Mission Objectives As one of the most highly regarded Aerospace Engineering Schools in the country, it is important for Purdue to take the lead in satellite design. The successful launch of a student build satellite will place Purdue in the spotlight and validate its reputation for producing quality engineers. PuTEMP will also be used as an additional AmSAT relay satellite after it has completed its primary mission objectives. 4

6 3.0 Satellite Description 3.1 Satellite Systems Overview PuTEMP is designed to be a small satellite capable of being launched as a secondary payload. The satellite bus is a frame design using angle-iron made of 7075 Aluminum. Cross-members and a shelf-structure provide further structural support and attachment points for the various subsystem components. Total S/C weight is 50.0 kg with exterior dimensions of 30x30x60 cm. Attitude is controlled through gravity-gradient stabilization with magnetic torquer assistance. Intended orbit is dusk-dawn sun-synchronous with an inclination of 98.6 o at an altitude of 800 km. Transmitters (x2) Receivers (x2) Nadir Antennas (x2) 30cm Simulated Propellant Tank 60cm 3.0cm 35cm 1.25cm Figure 3.1 Satellite Internal Layout Ariane 5 Launch Ring Attachment 5

7 4.0 Concept of Operations 4.1 Overview The PuTEMP Concept of Operations consists of four distinct mission phases. The on-orbit portion of the mission is most important to achieving the mission goals. It is also the largest phase in relation to time. However, the proceeding to phases must obviously be completed successfully before on-orbit operations begin. 4.2 Mission Phases Launch Preparation PuTEMP is designed to require a minimum of handling in preparation for launch. The payload tank is a sealed structure requiring no filling or other maintenance (ref. Sec. 9). No subsystem components will need to be exchanged or added prior to launch. The batteries will need to be charged and a systems level checkout performed to insure proper operation of all subsystem components. This checkout procedure will be built into the CPU memory creating an automated checkout that can be performed quickly. The S/C is required to be totally inert (no radio-transmissions) during countdown and flight. This means all onboard communications equipment most be turned off prior to vehicle integration. Before vehicle integration, the S/C will be placed in an autonomous mode. Once placed in this autonomous mode, the S/C will no longer be able to receive commands. The function of this autonomous mode will be discussed in the next section. Integration into the launch vehicle will be relatively simple. The equipment required to attach the S/C to the launch vehicle is provided by the launch vehicle manufacturer and will already have been integrated into the S/C structure. Only one external connection will be supplied to the S/C from the launch vehicle and will consist of an umbilical link to provide a trickle charge to the batteries Spacecraft Deployment During countdown and flight the S/C will be inert. Deployment of the S/C is accomplished by the use of a small spring ejection system integrated into the attachment ring provided by the launch vehicle manufacturer. While the speed of the ejection can be specified from 1 to 4 m/s, the attitude pointing at orientation cannot. As a result, there is no guarantee that the S/C antennae will be in the proper orientation to receive commands from the ground station. The S/C must then operate autonomously to achieve the proper orientation. The S/C will be placed in the autonomous mode before vehicle integration. When the S/C detects that the umbilical link has been separated, indicating deployment from the launch vehicle, an automated process will begin: 6

8 1. All TT&C components turned off for launch will be powered on allowing the S/C to send and receive signals 2. Gravity Gradient Boom will be deployed and the proper orientation attained to allow communication with ground station 3. Begin transmitting orientation and S/C health 4. Await commands from ground station Orbital elements will be determined by data provided by NORAD. Based on these elements and worst-case estimates of the time needed to achieve gravity gradient stabilization, the ground station will know when and where to begin looking for the satellite On-Orbit Operations Once the S/C has been acquired by the ground station normal S/C operations can commence. The function of the satellite is to conduct payload experiments. Therefore, the on-orbit Concept of Operations is centered around the payload and its requirements. An operating cycle centered around the payload requirements needs to be defined as the payload will carry out repeated experiments with little change in demands on other S/C subsystems. Unlike most LEO satellites that are constrained by their orbital period and the associated power subsystem requirements for battery charge/discharge, the orbit of PuTEMP is such that a simplified power scheme can be used (ref. Sec. 13.1). Instead, PuTEMP is constrained by the period required between experimental runs. As will be discussed in Section 9, the payload propellant tank requires a fixed cool-down period. Being constrained by this cool-down period means that an operating cycle can be defined as the time it takes to conduct an experiment and allow for tank cooling. Having defined the operating cycle the specific functions within the cycle can be laid out as follows. With an experiment just completed and primary power depleted: 1. Begin recharging primary power source 2. Transmit experimental data to ground stations as ground track allows 3. Insure full charge of primary power 4. Insure required cooling of payload propellant tank 5. Reorient S/C in preparation for next experimental run 6. Conduct Experiment This cycle can be repeated as many times as needed within the lifetime requirements of the S/C. Operating Modes and power requirements are such that Steps 2-5 can take place in parallel (ref. Sec. 13.2). Recharge time will be affected but only to a small degree End-of-Life It is envisioned that once the primary mission of PuTEMP has been completed, it can be turned over to AMSAT. The design of the TT&C and C&DH subsystems is such that converting the 7

9 spacecraft to amateur satellite operation is desirable. The modulation scheme in place on PuTEMP to directly interface to AMSAT ground stations. 8

10 5.0 Major Design Requirements 5.1 Customer Attributes - Philosophy Design requirements for PuTEMP have been defined with the goal of providing an appropriate platform with which to conduct the planned payload experiments. Consideration was given to the specific needs of the payload, the imposed requirements or the launch vehicle, and the desire to produce a student built satellite at Purdue University. The needs of each are interconnected and dependent upon one another Payload PuTEMP s primary mission is to successfully complete the payload experiments. The testing environment must meet all requirements laid out by the experiment team based on the data they wish to gather. Failure to acquire data within the limits set for the payload would mean a failure of the primary mission. Details of the experiment are presented in Section 9. Because conducting payload experiments is the primary mission, reliability is key. Every design decision is judged on its impact to overall system reliability Launch Provider The S/C will have to be placed into a orbit with proper orbital characteristics necessary to achieve the mission goals. PuTEMP will be launched as a secondary payload and as such will have to comply with all requirements placed upon it by the launch vehicle. These requirements include limits on mass and dimensions as well as structural loads and inert package for launch Purdue University As PuTEMP will be built by Purdue University as a student built satellite, this imposes constraints on the design. Being a university satellite, the design must also be within the realm of what is available to Purdue University. Cost becomes a consideration due to the fact that limited funds are available for a non-revenue generating satellite. If, for example, a structural component cannot be fabricated at Purdue and no supplier can be found within the cost constraint, that component is simply not an option for the design. In some cases, a trade-off between reliability and cost must be made. 5.2 Engineering Requirements Specific engineering design requirements are laid out in the following table. The origins of every requirement are driven by the customer attributes described above. Further discussion of each specific requirement is provided in subsequent sections. Table 5.1 presents the design requirements: 9

11 Launch Vehicle Requirements S/C shall require minimum ground handling prior to launch S/C shall meet required dimensions and launch condition environments Payload SPT shall be 7.5 cm radius by 25 cm in height Shall have the ability to raise tank temp. 10deg Shall have the ability to store data from experiment Shall have the ability to retrieve data from tank experiement Oscilations of tank shall be under deg/sec and spin below rad/sec Shall have the ability to determine if s/c, and tank, are oscilating greater than 5 deg/sec ADCS Shall have the capbability to control s/c within required limits of SPT Shall orient the Spacecraft to be Nadir- Pointing TTC S/C shall have the ability to communicate with the Purdue University ground station at least once per day Metric for meeting Requirement Minimal ground handling S/C axial frequency >90Hz lateral frequency >45Hz Axial limit load factors of -7.5 to +5.5 Laeral limit load factors of -6.5 to +6.5 Accurate Tank Size in s/c model Power capability Number of Heaters/Thermisters CPU capabilities Communication with ground station oscilations within criteria or ability to control s/c within criteria measurement of Established by payload, discussed above orbit selection Requirement Met X X X X X X X X X X X How Requirement is met Only s/c system checkout and battery charge is required prior to vehicle integration Analysis of structure with Ansys Analysis of structure with Ansys Accurate tank size Batery sized for experiment run Number accepatble for heat distribution and sensing CPU memory storage of 4 Meg 6am-6pm sunsynchronous orbit, two passes over ground station a day motion of s/c controlable within criteria with magnetorquers, Sun sensors detect oscilations as small as 5 deg/sec. Magnetorquers create large enough dipole to control satellite within at most 3 deg/sec. 4-meter boom-4-kg Tip Mass gravity gradient stabilization. 5 and 6 ma^2 dipole moment magetic torquers 4-meter boom-4-kg Tip Mass gravity gradient stabilization. sun-synchronous oribt at 98.6% inclination, altitude of 800 km, RAAN of deg 10

12 (launch day dependent) S/C antennae shall operate on zero gain S/C telecommunications operate on existing modulation techniques S/C uplink and downlink frequencies are pre-existent BER of 10^-5 Command and Data Handling S/C shall have the capability to store all SPT experiement information S/C shall have capability to convert experiement information to downloadable format for ground staion Data storage reinforced for radiation Power Shall have capability to supply power to the SPT for experiments while providing power to all other S/C subsystems: Thermal S/C shall maintain interior temperature within required limits of 0-40 degc: Structure and Mechanisms Load bearing stucture shall support the SPT and all other subsystems within configuration constraints Shall meet launch vehicle frequency and laoding requirements Pointing requirement for the antenna not very stringent use of pre-existing AMSAT communication technology frequency used is already allotted for in the spectrum Use of pre-existing technology for the modem CPU memory at least 256KB Simple technology for use with the Purdue ground station Data storage hardware must be reinforced and existing in order to minimize complexity Battery sizing criteria (maximum based on Raising SPT temperature 10 deg in 15 minutes) Thermal equilibrium maximum and minimum S/C temperatures All components must fit within satellite dimensions of the launch vehicle 33cmx33cmx60cm S/C axial frequency >90Hz lateral frequency >45Hz X X X X X X X X X X X omni-directional antennae Gaussian Minimum Shift Key (GMSK) modulation Uplink of 146 MHz, downlink of 437 MHz Use of GMSK modem by Spacequest data storage capacity of 256KB plus an additional 8 MB allotted for with extra space (experiment really needs approximately 4 MB) AX.25 protocol Use of radiationhardened hardware by Spacequest Battery is sized to 20.94W-hr capacity to run S/C during an experiment Analysis has shown that equilibrium temps are within specified limits All componenst fit within the required satellite dimensions S/C natural frequency with internal masses of 92 Hz 11

13 S/C shall be manufacturable at purdue laboratory facilities Propulsion Axial limit load factors of -7.5 to +5.5 Laeral limit load factors of -6.5 to +6.5 Materials shall be managable by Purdue University student S/C shall require no propulsive device X X X Frame and panel material yield and ultimate strength greater than launch loading with a maximum deflection of 0.009cm Aluminum 7075 material is used No propulsive devices Table 5.1 Engineering Requirements Breakdown 12

14 6.0 Orbit Selection 6.1 Requirements Orbital selection impacts PuTEMP in two main areas, power and TT&C. Power requirements are such that maximum sun exposure is desired. As will be discussed further in Section 6.4, TT&C requires specific orbital criteria to allow communication with the Purdue Ground Station. Orbital altitude is a significant concern for both power and TT&C because of the increased transmission power required at higher altitudes. Fortunately, the requirements of the two systems are not mutually exclusive allowing one orbit to fully satisfy the requirements of both systems. 6.2 Selection Criteria The selection criteria were based on the orbit trades of the mission as a whole. PuTEMP s mission, which is centered the payload, rules out the consideration of a constellation. So all trades were performed with a single satellite. The following criteria will be used in the orbit selection: Minimum eclipse time (<50% of orbital period) Communications with Zucrow (Purdue University, i 40 ) Perturbation effects (lifetime effects) Accessibility of specialized orbits 6.3 Eclipse Time Minimizing TT&C transmitter power requires the S/C to be in a relatively low oribt. This rules out orbits at the Lagrangian libration points and lunar crossing orbits. Some orbits that could yield minimum eclipse time with a reasonable altitude are limited to: polar, Sun-synchronous, Geosynchronous, and Molniya orbits Polar Orbits Polar orbits are defined by their inclination being fixed at 90 with respect to an Earth-fixed reference frame. Because of the J2 geopotential effects, the right ascension of the ascending node will change at a rate dependent on altitude and eccentricity of the orbit. This would only allow for certain times of the year where the sub-solar point is within the orbit plane. The worst case would be when the plane intersects the solar terminus. The eclipse time here would also be dependent on the altitude and eccentricity of the orbit Sun-synchronous Orbits Sun-synchronous orbits are characterized by their nodal precession rates being set to /day (same as the angular velocity of the mean sun), their eccentricities being nearly zero, and their orbits being retrograde rather than direct. This parameterizes the choice of inclination as a function of the chosen altitude. This matching makes the nodal precession rate constant and zero with respect to the sun. The upper limit for this altitude is around 6000 km. With proper 13

15 selection of RAAN it is possible to attain a dusk-dawn orbit with very small eclipse time depending on an appreciably high altitude Geosynchronous Orbits Figure 6.1 Sun-synchronous Orbit Geosynchronous orbits are characterized by their orbital periods matching the Earth s average spin rate (15 /hr). This fixes the orbit radius at 42,160 km and the inclination equal to zero. There are only two zero eclipse periods during the year, the summer and winter solstices Molniya Orbits These highly elliptic orbits are characterized by their rates of change of the argument of perigee being zero. This matching fixes the inclination to 63.4, the eccentricity to 0.75, the semi-major axis at 26,600 km, and their apogee set in the northern hemisphere. Since their orbits remain in the northern hemisphere for about 92% or their orbital period, this greatly reduces the eclipse time. Figure 6.2 Molniya Orbit 6.4 Communications with Ground Station 14

16 The location of the ground station is 40 2 N, 86 5 W, which means if we are in a LEO, our inclination will need to be at least 40 in order to pass over the Zucrow Ground Station. If we are in a GEO, then our slant range is much larger and thus will place additional constraints on our power and C&DH subsystems to be discussed in Section13.0 and Section Perturbations If we limit our search to a LEO we can neglect the third body effects of the moon and sun. For altitudes below GEO, the geo-potential perturbations, namely J2, dominate the gravitational perturbations. For altitudes of about 800 km and below, the atmospheric density increases such that the atmospheric drag would dominate the solar radiation pressure effects. 6.6 Accessibility to Specialized Orbits This criterion is dependent upon launch vehicle capabilities and available launch sites. Wherever the launch site is located, the absolute value of the latitude must be less than or equal to the inclination of choice. If it is greater than the inclination, then a launch window will not be available. The ideal case is to be less than the inclination because this gives two launch windows per day. The launch window here is only constrained by the desired inclination of the LEO and the launch vehicle plane change capabilities. Small errors in the launch azimuth (measured in clockwise direction from north pole) can give large errors in final inclination. These azimuth angles are also constrained at the launch site for safety reasons. Therefore the choice of orbit inclination will ultimately determine where PuTEMP can launch. 6.7 Altitude Trades Before choosing a final orbit type mission altitude must be defined. In order to do this an altitude trade was performed. The results can be seen Table 6.1 below. Characteristic Allowed Range (km) Comments Launch Capability < 1800 Launch Vehicle Limit Radiation <= 800 Below Van Allen Belt Communications > 600 In genera higher is better. Table 6.1 Altitude Trades for PuTEMP Mission PuTEMP decided to choose an altitude of 800 km. This allows TT&C transmitter power to be kept to a minimum. Also, the majority of sunsynchronous orbits have historically been placed in this altitude. 6.8 Orbit Type Selection Given that the nominal altitude is fixed at 800 km the Molniya orbit can be ruled out. With this in mind, the decision remained whether to choose between the remaining specialized orbits or design a different one. Maximizing the sun exposure for the power system is a priority means the ideal orbit type would be the Sun-synchronous orbit described above. Therefore for a Sun- 15

17 synchronous orbit at an altitude of 800 km with a period of minutes (14.3 revs/day) was selected. This will fix the inclination as seen in figure 6.3. i = 98.6, Alt. = 800 km Figure 6.3 Sun-synchronous inclination determination. Now that inclination has been determined the orientation of the orbit normal with respect to the sun needs to be defined. To do this a particular launch date will need to be selected with respect to the position of the sun along the ecliptic plane with respect to the Vernal Equinox. Once this has been determined, RAAN can be set to this angle plus 90, which will initially point the orbit normal towards the sun. Since it is a Sun-synchronous orbit this orientation will remain fixed allowing perpetual sunlight for the duration of the mission. The code used here can be found in Appendix A. Selecting April 25, 2002, as a baseline data and choosing a launch site located at 5 N, 52 W (Kourou, French Guiana), the RAAN needed for this orbit was determined to be with a launch azimuth of at the ascending node and at the descending node. Disregarding the orbit plane change capabilities of the launch vehicle, this gives PuTEMP a launch window of LST at the ascending node and LST. This selection is the ideal case. In the event that the required azimuth, or RAAN could not be met, then the analysis would have to be reversed. Given a non-optimal launch event, the eclipse time would no longer remain at zero and thus would have to be evaluated from this new RAAN. 16

18 7.0 Launch Vehicle Integration 7.1 Selection Criteria PuTEMP desires the capability to launch on a wide range of launch vehicles with only minor modifications to launch vehicle interfaces. This flexibility will greatly decrease the lead-time necessary to prepare the S/C for launch on a specific vehicle. As a result, PuTEMP will be available to fly as soon as a slot opens on any launch vehicle within the design limits. It is hoped that this will allow PuTEMP to fly on the cheapest available launch vehicle. To determine the launch vehicle design requirements a number of launch vehicles were first investigated (Ref. App. B). Only vehicles with previous histories of launching small, secondary payloads were examined. Four items specifically were considered in vehicle selection: Available orbits Maximum sun exposure desired for power, LEO desired for TT&S S/C mass limits The payload has minimum mass that can be used to determine overall minimum S/C mass S/C size limits The payload has minimum dimensions that can be used to determine overall minimum S/C dimensions S/C structural design loads and limits Highest possible structural loads and limits chosen 7.2 Launch Vehicle Imposed Requirements Combinations of the various vehicle mass and size limitations were considered to give a hybrid requirement that would meet the requirements of several different launchers. After reviewing the available launch vehicles presented in Appendix B, three candidate vehicles were selected: Ariane5, Delta II, and Space Shuttle Hitchhiker. Each system provides LEO orbits with the Delta II and Ariane5 offer the additional advantage of Sun- Synchronous orbits. The restrictions imposed by each launch vehicle were considered. The resulting dimensions and mass restrictions allow PuTEMP to be launched by all three systems. Maximum Satellite Mass - 68 kg Maximum Satellite Width (square) cm Maximum Satellite Height cm Figure 7.1 Ariane5 While 68kg was chosen as the mass limit, 40kg is being used as the design-to limit. This should insure that any mass growth will not reach the overall limit of 68kg. Each vehicle imposes different structural loads and limits on the S/C. The Ariane5 limits were chosen because they were the highest. Designing to the Ariane5 limits means that the S/C will automatically meet the limits of the other vehicles. Because PuTEMP is designed to the Ariane5 requirements, it is considered the preferred launch vehicle. 17

19 8.0 Spacecraft as a System 8.1 Layout and Arrangement 8.2 Component Packaging Figure 3.1 Major S/C Elements 8.3 Mass Budget Mass budget margins were chosen based on the present state of each subsystems design. Most subsystems are using a margin of 1.1 because in these cases actual components have already been selected. The power subsystem remains at a higher level of design with some components remaining to be selected (ref. Sec. 14). While payload components are selected, a margin of 1.2 will continue to be used to insure integrity of the experiment. Mass budget breakdowns are presented by subsystem in Figure 3.2 and Table

20 Payload* 9% TT&C 6% C&DH 1% Power 5% Structure 58% AD&C 21% Figure 3.2 Mass Budget Percentages Subsystem Mass [kg] Margin % Total Payload* TT&C C&DH Power AD&C Structure Total 50.0 Table 3.1 Mass Budget Values 8.4 Mass Moments of Inertia The mass moments of inertia are presented below. Mass moments are important for the AD&C subsystem and for the structure. The effects of mass moments on AD&C are discussed in detail in Section 10. The launch vehicle. Ixx=102.7 kg/m2 Iyy=102.8 kg/m2 Izz= 0.9 kg/m2 19

21 20

22 9.0 Payload 9.1 Summary This portion of the preliminary design document details the scientific payload of PuTEMP. The experiment description for both the primary and secondary experiments can be found in section 1.2. Section 1.3 details the equipment design. The sensor interface to the Command & Data Handling subsystem is present. Finally, in section 1.4, a brief overview of the current state of the payload subsystem is presented. 9.2 Experiment Description The mission of PuTEMP is testing methods for propellant gauging techniques. To accomplish this goal a series of experiments have been designed Primary Experiment Thermal Propellant Gauging (TPG) The primary experiment tests a method currently in use on commercial communication satellites. Spacecraft currently on orbit have heaters and heat sensors in place on the propellant tanks. When the current generation of spacecraft were launched, the only purpose of the heaters and sensors was to maintain a comfortable temperature for the propellant. Over time a technique of measuring temperature changes on the surface of the tank was developed to find an estimation of how much propellant was currently in the tank. Several good estimation schemes are currently in use for propellant gauging. A successful method is using a propellant budget. After each thruster firing, the amount of fuel consumed in the maneuver is subtracted from the amount previously known. Records are kept which show the depletion of the fuel over the lifetime of the satellite. The problem with this method is twofold. First, some lifetimes are over a decade long. Tank boil-off, thruster aging, and other effects create uncertainty in how much propellant moves through the nozzle of the thruster. This makes the method satisfactory for beginning of life measurements, but end of life is more difficult. This leads into the second problem. Propellant gauging is especially important for end of life operations. We must know when to move the spacecraft from its orbit so a new one can replace it. As time progresses with this method, it becomes more difficult to judge the amount of fuel remaining in the tank. For this reason, interest has been building to find more efficient methods of propellant gauging techniques. This is why PuTEMP has been designed. Currently no known spacecraft have been flown for the purpose of large-scale fluid experiments. The goal of the primary experiment is to place a simulated propellant tank into an environment similar to a full-scale craft. This propellant tank contains a known fill fraction similar to an end of life scenario. The tank is surrounded with temperature sensors to record a temperature profile of the tank. Because of the necessity to keep the spacecraft as simple as possible the tank is allowed to reach an equilibrium temperature with the internal volume of the spacecraft. (The thermal environment description can be found in section 8.0 of this report.) Once the tank reaches equilibrium and the spacecraft is in a stable operating mode, a temperature increase is 21

23 induced into the tank via a series of heaters located on the end caps of the tank. The heat energy is added for a specified amount of time. After this time has elapsed, the heaters are turned off. Throughout the heating process, the sensors have been registering the temperature at various points on the outside wall of the tank. This data is stored onboard the spacecraft until an appropriate time for download to a ground station. After the experiment has been run for the full length of time, the tank is allowed to reach equilibrium once again with the spacecraft. After this has occurred, the experiment can be run again. There is an alternate method to running the experiment. We can run the experiment with a specified amount of temperature change at a specific location. For example we could turn the heaters on and time how long it takes the temperature of the central perimeter area of the tank to heat up by 10 degrees C. Both methods are acceptable and can be employed with the same hardware. This method would require a more autonomous vehicle, but is not much more complicated because of the ability to upload new programs to the flight computer. The principle behind this experiment is a simple everyday phenomenon. A full spacecraft tank will take longer to heat up than a spacecraft tank that is almost empty. This is the same principle when making coffee in the microwave. A full cup takes a long time to heat in the microwave, but a cup half full takes a shorter amount of time. Operational spacecraft use this method to compute the fill fraction. The experiment will already have a known fill fraction. So we can verify and refine the current calculations in use by industry Secondary Experiments Other experiments are possible as well. For PuTEMP, we were faced with a decision early in the development cycle. We had a choice of choosing from a variety of experiment options. We could only have a single experiment to run and design the spacecraft to meet this need. Looking out further though, other experiments similar to the TPG experiment could be performed with minimal addition to the hardware. We had a choice of designing our spacecraft to meet the objective of having several primary experiments. This seemed like a logical choice until we realized by making them all primary we needed to make sure each was a complete success. Because of the time constraint for this design this was not feasible. The methodology we are working with is having a single primary experiment, the TPG. This placed the emphasis of the mission on making the primary experiment successful. Included in our design is the possibility to perform numerous secondary experiments. The hardware is not designed with the secondary experiments as critical to the mission. If the secondary experiments are not successful or not performed the mission is still a complete success if the primary experiment returns useful data. The possible secondary experiments are overviewed in the following sections. No detailed analysis or design is performed on these options. Thermocapillary Baseline Experiment Thermocapillary dynamics is a poorly understood phenomena. The effect of thermocapillary dynamics on satellite propellant gauging techniques is an even more complicated problem. This experiment looks to understand the dynamical system of fluid motion due to thermocapillary dynamics. To accomplish this we allow the fluid to freeze completely solid inside of the 22

24 spacecraft tank. A complete heating cycle is performed to bring the frozen body back to the liquid state. By measuring the difference between the temperature profiles of the primary experiment and the frozen profile itself, the thermocapillary effect can be found. P-V-T Gauging Experiment A method used at present to determine propellant volume is based on using the pressure of the tank pressurant. This pressure when coupled with the temperature of the tank can give the volume of the gas in the tank. From this gas volume the propellant volume can be computed. In commercial spacecraft this method involves external methods due to the difficulty in creating propellant safe measurement devices compatible with Hydrazine or other propellants. Because of the use of water in our experiment, we can place a pressure transducer inside the tank. We believe this will result in a new benchmark for gauging external pressure measurements. Cyclical Heating Experiment A constraint on many space missions is the minimization of power required to perform the mission. One possible way to do this is to reduce the power requirements to keep propellant tanks at a constant temperature. The cyclical heating experiment is an investigation into using a modulated heating cycle on the tank. By operating at a specific frequency we believe we can reduce the power required to heat propellant tanks. Other experiment options are available. Dr. Collicott from Purdue University in the Aeronautical & Astronautical Engineering department has given these experiment descriptions to the PuTEMP team as guidelines to the design of the spacecraft. 9.3 System Design Because of the decisions we made at the beginning of the design process, we are requiring only a single primary payload, but the design is open for other secondary experiments, which require minimal modifications to the spacecraft. This simplified our spacecraft design to fit the time frame to complete a preliminary design. The primary component of the payload subsystem is the Simulated Propellant Tank (SPT) Simulated Propellant Tank (SPT) In designing the simulated propellant tank constraints were discovered that limited the design options. These constraints bounded the tank design and quickly allowed us to converged on a tank design. The follow list is the constraints assumed in the design process. 1. The tank had to fit comfortably in 30 cm. X 30 cm. X 60 cm. This is the maximum volume the spacecraft can have. 2. The tank must accurately simulate end of life conditions for a communication satellite. 3. The experiment must be performed in a short enough time span to minimize exterior influences on the data (e.g. heat loss through insulation). 4. The spacecraft bus limits the available power for the payload 23

25 5. The tank must be able to handle at least 2 bar (atmospheres) of both implosive and explosive pressure. 6. The experiment fluid must simulate Hydrazine, and must be relatively safe to handle. 7. The tank must be easy to manufacture and durable enough for space flight. 8. The largest possible tank size is the best to simulate a full-scale tank. The design of the simulated propellant tank is driven primarily by the power available to run the heaters. Very early in the design, we understood available electrical power would dominate the design of the payload. This is because we have such a small exterior surface area to gather power. This strong design driver bounded the fluid fill fraction we could use in the experiment. We also require a fill fraction similar to an end of life scenario. Our choice was to set the fraction at 25% full. This value was set as the high amount. Smaller fill fractions will lower the power required. For the remainder of the trade studies we assumed a 25% fill fraction. Another factor in designing the tank is the pressure requirement. We arbitrarily set the pressure inside the tank. Because we are not releasing fluid at any point in the mission we could have a low internal pressure. We decided upon using a 1 bar internal pressure once on orbit. This simplifies manufacturing and integration on Earth because the tank can be open to the atmospheric pressure while filling and testing. Once in orbit the 1 bar pressure would be maintained because the tank is sealed during manufacturing and integration. The tank pressure is maintained with Nitrogen gas. Nitrogen was selected because it is inert with most substances we use and is very cheap to purchase. We set the tank thickness to mm (1/8 in). This is for exterior loads on the tank as well as ease of manufacturing. Experiment fluid is distilled water. Water has similar characteristics to Hydrazine. Water is the simulate for many experiments run by Purdue University in the Vomit Comet program run by NASA. Water is safer than Hydrazine and no special handling requirements will be needed for launch vehicle integration. Because of the surface area constraint, only a small amount of power can be generated. We started by looking at the power required to heat the tank in a specific amount of time by a certain temperature. We used this as the primary guide to sizing the tank. For example, say we wish to heat the tank in 15 minutes. What power is required if we want to get a change in temperature of 10 degrees? We ask this same question for various heating times. With this information in hand we can define what the tank size must be. Figure 9.1 shows an example plot from the trade study we did to find an acceptable temperature range. This plot shows the heat energy vs. temperature changes for various experiment run times. With this information in hand the tank dimensions and other physical characteristics quickly converged to an acceptable solution. The dimensions of the tank are cm in diameter and cm in length (6 x 12 in.). 24

26 Heat Energy Q (W) min 30 min 45 min 60 min delta T (deg C) Figure 9.1 Carpet plot of energy calculation Aluminum is the construction material. It was chosen for its compatibility with the fluid; Aluminum will not rust in the presence of water. Aluminum is also easy to machine and is cheap to purchase. The exact type of Aluminum to be used is not specified here. The tank is surrounded by blanks of multi layer insulation. The buildup of the experiment tank is as follows. A tank with perimeter vanes is the inner core. On the immediate exterior of the tank lie the heat sensors and heater strips. Also the mounting hardware is located here. On top of this equipment is the multi-layer insulation. The only exposed portion of the tank is the support structure and the wiring coming from the sensors. Figure 9.2 shows this buildup graphically. The diagram is a cut away section of the tank wall. Included is the layout of the tank vanes. 25

27 Fig 9.2 Tank Structure and MLI Installation Sensor placement is critical to returning good data. Because of the choice of computer hardware, we have available to us up to 24 analog thermistors. A thermistor is a temperature sensitive resistor. Figure 9.3 shows the relative placement of each sensor on the tank. The hemispherical end caps each receive 5 sensors. 1 sensor is place at the tip (or as close to the tip as possible because of filling hardware and other attachments). The remaining 4 sensors are placed near the seam connecting the hemisphere to the cylindrical section of the tank. The sensors are placed 90 degrees from each other at each quadrant of the hemisphere as you are looking down on it. Each hemisphere is identical in the placement of the sensors. Fig 9.3 SPT Shape and Sensor Placement On the cylindrical portion of the tank 12 sensors are in found. The central set, as seen in the Fig 9.3, is located at the same positions as the hemispherical end caps. The rows immediately to each side of the center are also at the quadrant points, but have been rotated by 45 degrees. This allows a systematic coverage of the tank surface. For redundancy purposes, the sensors are wired to different circuits. Each circuit has sensors located at all points on the tank. This will insure that if one sensor circuit fails, the temperature profile for the entire tank will still be available, albeit at a lower spatial resolution. The sensor electronics is explained in section Electronic Interface Returning useful data is the ultimate goal of this mission. Temperature information is the primary data being transmitted to the ground station. To get this information from the tanks sensors requires a sophisticated analog to digital conversion process. A large amount of data needs to be processed in order to get a temperature value to the experimenter on the ground. We choose to perform a good portion of the processing onboard the spacecraft. The data from the sensors is analog in nature. We take direct sensor readings and convert it to digital format for storage in the system memory. A particular A/D converter on board the spacecraft can handle up to 8 different sensors at 10-bit conversion accuracy. Because of the small power requirement of 26

28 the conversion processors a second layer of redundancy is in place as well. We also choose to do redundancy in components because the PIC controllers are not space rated hardware. To avoid making the mission too complex and expensive, we chose to make the system from non-radiation hardened equipment, but add the extra layer of redundancy. A voting system is created which verifies the sensor information and makes sure accurate data is being culled from the A/D converter. Fig. 9.4 shows the proposed schematic for a single sensor circuit. In the diagram the sensor bank is connected to the A/D converters. The A/D converters have 8 pins available; each sensor is connected to a pin. The sensor select line is responsible for telling the hardware which sensor we take data from. The data can only be read sequentially so we must prioritize and step through each sensor. Fortunately, it takes very little time to complete this step. From the A/D converter the digital temperature information is fed into a voting computer. This computer checks the data for inconsistencies with previous data and with the other side of the redundancy tree. If this data passes the test it is put on the bus and sent to the CD&H for storage. Fig 9.4 Single Temperature Sensor Circuit The specific hardware design for the data acquisition is based on PIC microcontroller technology. We feel this gives a low cost approach to solving the problem of large amounts of hardware and software required by the payload. 9.4 Preliminary Subsystem Characteristics The parameters of the Simulated Propellant Tank are summarized in Table 9.1. In this table some of the information is provided in both SI and English units are present for convenience. The fluid properties of water also included on the right. This set of tables was used directly in the trade studies. 27

29 English (in) SI (m) Value Units Diameter Tank Volume m^3 Length Fill Fraction 0.25 Radius Fluid Volume m^3 Height Fluid Density 1000 kg/m^3 Thickness Fluid Cp 4182 J/(kg*K) Value Units English (in) SI (m) Wall Volume m^3 Thickness Wall Density 2800 kg/m^3 R_outer Wall Mass kg/m^3 R_inner Wall Cp 920 J/(kg*K) Table 9.1 Simulated Propellant Tank Parameters For the current design of the data acquisition system we need 22 sensors located in 3 separate circuits. A total of 6 A/D converters is necessary for the 2 layers of redundancy. Also 6 separate voting machines are required as well. Each of these could be a simple PIC microcontroller. The interfacing technique is a simple RS-232 or similar serial interface. The final mass and power requirements are tabulated in Table 9.2. The maximum power draw of the system is about 45 Watts. This corresponds to the payload running the experiment and the heaters running at maximum power. The minimum power requirement corresponds to using no heat addition and only using 4 sensors to keep a constant temperature profile of the tank. Value Units Payload Mass 2.5 Kg Maximum Power Required 45 Watts Minimum Power Required.5 Watts Table 9.2 Final Payload Subsystem Parameters 28

30 10.0 Attitude Determination and Control (AD&C) 10.1 AD&C Subsystem Requirements The AD&C subsystem was designed around requirements of the payload and the TT&C subsystem. The payload does not require a specific orientation of the satellite in its orbit, however it does specify spacecraft oscillations to be below a frequency of deg/sec and the S/C must operate at a spin rate lower than deg/sec (ref. Sec. 9). These correspond to the worst-case attitude required for S/C operation and are set in order to obtain relevant data from the payload. TT&C subsystem requirements, nonetheless, require a specific orientation. One face of the spacecraft must be nadir pointing in order to provide a line of sight for communications. Given a small satellite design requirement, power capability and acquisition play a large role in the satellite design due to the limited surface area for solar panels. These low power capabilities drove the need for an AD&C system as power-economical as possible; the PuTEMP team also constrained the AD&C to be as simple as possible Attitude Determination Attitude determination is needed in order to operate successfully the attitude control devices. Also to ensure the spacecraft to be within the attitude requirements dictated by the payload, and therefore ensure the relevance of the date obtained from the experiments. For these reasons, a successful attitude determination is a must in the mission. Horizon sensors might be a simpler device for attitude determination in an earth pointing spacecraft, however, this type of sensors were not incorporated into the spacecraft due to volume, mass, and budget constraints. The volume constrain comes from the fact that the horizon sensor could not be placed in the nadir pointing face of the spacecraft because Antennas, receivers and transmitters, for telemetry, and supports for that payload were also located in this face of the spacecraft, therefore, the limited space on the face required for a successful operation of the Horizon sensor leaded into investigation on other ways of solving the problem of attitude determination. Two kinds of devices were considered. These devices are sun sensors and magnetometers. The sun sensors give information on the angle changes up to accuracy of 0.5 deg, and the magnetometer gives information on the strength of the magnetic field, so that the magnetic torquers are effectively operated. One of the concerns with using sun sensors was the range of angular velocities at which the spacecraft would be operating. In order to successfully sense angular velocities of the magnitude of 0.1 Hz with the sun sensors, the extreme angles must be greater than 0.5 deg. The accuracy at which the spacecraft records the oscillations is driven by the speed of the onboard CPU, which is responsible of the interpretation of the sun sensors data. The CPU is fast enough and can therefore process any angle changes sensed by the sun sensor. After the spacecraft is ejected from the launch vehicle, an automatic process of boom deployment will take place. After this event, the spacecraft obtains a nadir pointing position. In 29

31 order to overcome the problem of a constant changing angle with respect to the sun due to an earth pointing spacecraft, a detailed knowledge of the ephemeris tables of the spacecraft is needed. With this knowledge the attitude determination algorithm is able to have into account the changes in angles of the spacecraft with respect to the sun due to the rotation of the body axis with respect to a sun centered axis, and therefore give accurate information about the attitude of the spacecraft itself with respect to an earth centered inertial reference frame. In order to be able to provide information about the attitude at any given time during the orbit, an arrangement of four sun sensors, each corresponding to a side of the satellite that may be in contact with the sun, is used Control Among the control devices studied were momentum bias wheels, magnetic torquers, and gravity gradient stabilization. In order to meet the team constraints the momentum bias wheels were disregarded for their complexity and power consumption. Passive control was decided to be the best option. Gravity Gradient Stabilization met most of the requirements and constraints, and therefore became the most viable option for attitude control. This type of control is known to work in several small satellite missions (e.g. Surrey Small Satellites). It ensures a stable spacecraft with a nadir pointing face, however it presents intrinsic oscillations that need to be damped out by other means. In order to damp out the oscillations inherent in the gravity gradient stabilization other control devices are required. To this end magnetic torquers are included in the attitude AD&C subsystem. They can be handled in such a way that the moment generated is always opposing the motion of the spacecraft and in this way damp out oscillations. This is possible by changing the current direction so that the dipole moment is reversed as needed. The CPU of the spacecraft is fast enough so that the speed at which the current changes are done is effective. The magnetic torquers were sized so that the magnetic dipole moment would generate a torque capable of controlling any type of perturbations, and to handle any kind of emergency maneuvers commanded from the ground station, also to maintain the spacecraft within the oscillation and spin ranges specified by the payload requirements. This is clearly shown in Appendix D, Figure D Subsystem components selection and sizing Attitude Determination 4 2-Axis Sun Sensors deg of accuracy -2 orthogonal axes -100mW of maximum power consumption -+/-12V of Power Supply -0.3 kg -Product of Surrey Satellite Technology (SSTL) 1 IM-102 Magnetometer 30

32 -2 orthogonal axis -<100mW of Power Consumption -5 V of Power Supply kg -Product of Ithaco Space Systems Attitude Control 1 SSTL-Weitzmann 4-m Deployable Boom -5 A of current for >10msec for Pyro-cutter actuation -2.2 kg excluding tip mass -A 4 kg tip mass, which includes the magnetometer -Product of SSTL 1 Mt-5-2 Magnetic Torquer 2-5 am Linear Dipole Moment Watts of Power Consumption -5 V of Power Supply -0.3 kg -Product of Microcosm Space Mission Engineering 1 Mt-6-1 Magnetic Torquer 2-6 am Linear Dipole Moment Watts of Power Consumption -5 V of Power Supply kg -Product of Microcosm Space Mission Engineering The previous hardware was selected because it meets the requirements, it is space qualified, and is known to work in other small satellite missions Predicted performance The spacecraft is predicted to be stable within the allowable ranges of oscillations dictated by the payload. Several configurations of boom length-tip mass guarantee the latter. A trade study involving the c.g. location of the spacecraft with different types of boom lengths and tip masses arrangements can be found in Appendix D, Figure D.1. From this chart it is clear that a longer boom with a heavier tip mass will translate the c.g. location farther out in the boom, and increase the spacecraft moments of inertia with respect to the x and y axis. However, there is a trade-off between the magnitude of the moments of inertia and the capability to damp out the oscillations inbuilt in the gravity gradient stabilization. It is observed from different configurations that heavy tip masses with long booms generate a non-optimal stiffness towards the magnetic torque, resulting in an extremely long time necessary to damp out oscillations. An example of these trade studies is shown in Appendix D, Figures D.4, where a 13 kg tip mass and a 6-meter boom and a 4-meter boom-4kg tip mass were simulated. There are also trade-offs between the magnitude of the disturbance torques (e.g. solar pressure, aerodynamic disturbances) and the magnitude of the gravity gradient torques. A difference of at 31

33 least an order of magnitude is desired so that any disturbance torqueses are overcome by the gravity gradient stabilization. The critical configuration that met this requirement was a 4-m boom-4 kg tip mass arrangement, and therefore became in the most viable option (Appendix D- Figure D.3). This configuration is incorporated for the previous reason and for its fast response towards the magnetic torquers control. This is clearly shown in the Figures 10.1 and Appendix D, Figure D.4, displays more information about the oscillations about the different axis of the satellite for the 4-meter boom- 4kg tip mass configuration and a 6-meter- 13kg tip mass configuration. This plots clearly display the increasing stiffness towards the magnetic torques generated by a heavier configuration. Figure 10.1 Angular velocity [deg/s] in the Roll Axis Without Magnetic Torquers for a 4-meterboom-4 kg tip mass Ixx= [kg*m^2] Iyy=125.4 [kg*m^2] Izz 0.7 [kg*m^2] From Figures 10.1 and 10.2 and the figures displayed in Appendix D, the range of angles in which the spacecraft will be operating can be specified. It can be observed that for the most part, this range lies above the sensitivity ranges of the sun sensors (i.e. above 0.5 deg). After actuating the magnetic torquers for a considerable amount of time (e.g. 2 or more orbits) the oscillations are no longer within the range of sensitivity of the sun sensors. 32

34 Figure Angular velocity [deg/s] in the Roll Axis With Magnetic Torquers for a 4-meterboom-4 kg tip mass Ixx= [kg*m^2] Iyy=125.4 [kg*m^2] Izz 0.7 [kg*m^2] This is not a concern, if the sun sensors are not giving any useful data of attitude determination, then the spacecraft is known to be stable, or within a very small range of oscillations. The plots previously mentioned anticipate the ability of the spacecraft to meet the requirements imposed by the payload. They clearly show oscillations well within the required ranges, and within the constraints of the attitude determination components. Actuation of the Magnetic Torquers will be done until the sun sensors are no longer able to sense perturbations. This will ensure a high percentage of stability above the limits constrained by the payload, and will therefore ensure a long enough time to execute experiments with optimal conditions. 33

35 11.0 Tracking Telemetry and Control (TT&C) Requirements The Tracking Telemetry and Control (TT&C) subsystem is comprised of various components, each of which will be discussed in the following section. TT&C is an extremely important subsystem for our mission, for it will allow the commands to be processed and forwarded to the ground station. The requirements set forth for the TT&C subsystem are: Low power consumption (the only component allowed to exceed 5 Watts is the antenna) Simple usage and failsafe redundancy Ability to perform at a data rate of at least 9600 bits per second (bps) Use of existing communication protocol for simplicity (preferred protocol is existing AX.25 protocol) In the completion of the link budget, have a satisfactory time margin with which to communicate to Earth Due to requirements in the Concept of Operations (ref. Sec. 4) and the arrangement of our antennae, the potential may arise for the spacecraft to be released from the launch vehicle in an unusual attitude. This could result in the two antennae (transmit and receive) not being oriented towards Earth. The scenario to fix this has been to require that the software included in the TT&C will begin a timer when the S/C is released. If the spacecraft has not received a communication signal from the ground station for a predetermined period of time, and the ADCS system senses that the spacecraft is in an unusual attitude, an automatic control system will begin to function, thereby reorienting the spacecraft such that the antennae will be facing the Earth. Thus for a brief, albeit critical period of time, the spacecraft could be operating autonomously. As an additional note, this automatic control mechanism will enable the spacecraft to reorient itself, but it will not begin the experiment until a ground signal has been sent telling the spacecraft to do so TT&C Software The AX.25 Protocol is a communication protocol based on packet radio techniques, where information is transmitted in small packets (hence the name) instead of continuously as in many other telecommunication transmissions. It is a point-to-point transmission passed over Local Area Networks (LAN), and requires a direct line of sight in order for communication to occur. It was decided to use this type of protocol as opposed to other such protocols as TCP/IP since this was a pre-existing protocol used in ground-to-satellite communication setups, such as AMateur SATellite (AMSAT). The format of this software is as follows. For unnumbered information, the format is: Flag Address Control FCS Flag /560 Bits 8 Bits 16 Bits

36 Table 11.1 Unnumbered Information Format Here the flag is a chacteristic 8-bit sequence that is not repeated again throughout the message, and is unique to the beginning and ending of the message frame. The address indicates the source of the information and the destination of the information. The control portion indicates the type of information being passed. The Frame Check Sequence (FCS) is equivalent to a parity check for the data, and is used to verify that the data has not been corrupted. For the case when numbered information is passed through, the frames appear as follows: Flag Address Control PID Information FCS Flag /560 8 Bits 8 Bits N*8 Bits 16 Bits Bits Table 11.2 Numbered Information Format Here the PID is the protocol identifier, and is used to define the type of protocol being used (Appletalk, Flexnet, ARPA Internet Protocol, etc.). The information is of course the actual information being passed through. All the other terms are as for the earlier case. For more information, please see AX.25 Link Access Protocol for Amateur Packet Ratio TT&C Components The components used to meet these requirements are the GMSK Modem, UHF Transmitter, VHF Receiver, and two patch antennae. The link budget will be shown after a description of the components GMSK Modem Figure 11.1 GMSK Modem The Gaussian Minimum Shift Keying (GMSK) 2 Modem used for this experiment has a Bit Error Rate (BER) of 10-5, which is a standard requirement for most space-based applications. The modem operates on 130 mw of power, and can attain a data rate transmission of 9600 bps, 1 AX.25 Link Access Protocol for Amateur Packet Ratio. Version 2.2, Revision 11 Nov 1997 Tucson Amateur Packet Radio Corporation. 2 GMSK is a type of modulation where the phase of the carrier is varied by the modulating signal, which is pseudorandom. This type of modulation is different from others, such as BPSK (Bi-Phase Shift Keying) and QPSK (Quadri-Phase Shift Keying) in that these latter modulations vary the phase of the carrier by 180 or 90 degrees respectively, in a known fashion. 35

37 though it could be higher. Further, this GMSK modem is specifically designed for AX.25 packet transmissions as made in AMSAT, and thus is useful since it is a pre-existing concept UHF Transmitter Figure 11.2 UHF Transmitter The UHF Transmitter operates on a range of 0-7 W of RF power. It can support data rates of bps, which is much higher than those required by the group. In the operating temperatures, the transmitter has a frequency stability of 5 ppm, which is a very high degree of precision. It also contains a 12-bit Digital-to-Analog (D/A) convertor, which is sufficient for the experiments that we are running VHF Receiver Figure 11.3 VHF Receiver The VHF receiver operates on lower power requirements than the transmitter, since it does not need to transmit data, namely on the order of 80 mw for the temperature operating range of the spacecraft. The Intermediate Frequencies (IF) used by the receiver allow for efficient data demodulation, these frequencies being on the order of 455 KHz. The front-end noise figure of this receiver is less than 1 db, which is optimal since it means that we will be maintaining much of the original signal without significant noise addition. This was an important parameter in the choosing of the receiver, since the incoming signal from Earth will be heavily corrupted by external sources of noise, and the group does not need additional noise to be added by the receiver. This receiver can operate comfortably on a 12 V DC, as will be supplied by the batteries Patch Antennae 36

38 Figure 11.4 Patch Antenna The antenna that the group decided upon was the patch antennae available from Seaveyantenna Inc. The group decided on the patch antennae for simplicity and for avoidance of any oscillatory perturbations upon deployment of a boom-style antenna. Furthermore, the patch antennae that were picked are omni directional and thus the pointing requirements for these antennae are not as stringent as for the case of a directional antenna, with a non-zero gain. Thus, the use of these antennae will provide excellent communication Link Budget It was paramount to define the link budget for this group s mission, and this was done immediately. It was determined that historically, the uplink and downlink frequencies used for these types of missions were 146 and 437 MHz respectively. Thus, these values were picked immediately, so as to not reinvent the wheel on frequency allocation. If other frequencies had been chosen for this experiment, allocation and permissions would have been required to perform the experiments. The next series of calculations made was the data rate required by the experimental setup. It was found that the passage over the ground station was a worst-case 10 minutes. Upon verifying the total amount of data to be taken in the experiment with the payload, it was determined that even with a worst-case scenario of a ten-minute pass, a data rate of 9600 bps would have been sufficient. If for some reason, the operational reality is that it takes longer than ten minutes to transfer the data at this rate (which is also supported by the Purdue ground station), a control algorithm can be implemented allowing the remainder of the data to be transmitted on the next pass. By using the BER requirement of 10-5 and using a GMSK modulation, it was found that a Eb/N0 3 ratio of 10 db-hz was sufficient for our mission. 3 Eb/N0 is defined as the Energy per Bit (Eb) divided by the noise spectral density (N0). 37

39 Figure 11.5 BER as a function of Eb/N0 Since the experiment is at an altitude of 800 km, the propagation path was determined to be 800 km. The transmitter line losses were taken empirically from data from earlier missions, and were found to be 1 db for most cases. When all was finished with the link budget was found, and the group had a margin of 12dB for the downlink portion, and a 22dB margin for the uplink portion. See the following figure for the link budget. Item Symbol Downlink Uplink Frequency F GHz XTR Power P W 4 50 XTR Power P dbw 6 17 XTR Line Loss Li db -1-1 XTR Antenna Gain Gt dbi 0 12 EIRP EIRP dbw 2 28 Propagation Path Symbol km Space Loss Ls db Pointing/Polarization Loss La db RCV Antenna Gain Gr dbi System Noise Figure NF db 5 5 System Noise Temperature Ts K Data Rate R bits per second Eb/N0 Eb/N0 db-hz BER BER 1.00E E-05 Required Eb/N0 Req. Eb/N0 db-hz Efficiency db -1-1 Margin db Table 11.3 Link Budget for PuTEMP 38

40 11.5 System Overview The system block diagram can be seen below and it incorporates the C&DH and TT&C subsystems. It can be seen that the antennae (that will be on the external side of the spacecraft) will be connected to the receiver and transmitter. The received signal will then go to the modem (which will demodulate the signal) and send it to the flight computer. The flight computer will then be responsible for distributing the information to the relevant subsystems. Note that the payload will be distributing the data directly, which will then be low-level processed and distributed to the flight computer that will then send it through the modem out to the transmitter. Ant. 1 UHF XTR VHF RCVR Note: dashed lines indicate incoming signal, solid lines indicate outgoing signal. GMSK Modem 12V Ant. 2 Data Collection Device A/D Converter Flight Computer and Data storage To payload To ADCS Figure 11.6 Block Diagram for the system 39

41 12.0 Command and Data Handling 12.1 The Requirements The Command and Data Handling (C&DH) subsystem is comprised of various components, each of which will be discussed in the following section. C&DH is an extremely important subsystem for our mission, for it will allow the commands to be executed and data to be processed for the experiment. The requirements set forth for the C&DH subsystem are: Low power consumption (less than a maximum of 5 Watts per component); Simple setup and usage (alternatively, small packaging, as allotted for by the satellite requirement); Stability to provide the necessary data rate for information transmission of the experimental data; Storage of the data until downlink (data amount on the order of Megabytes (MB)) C&DH Components The components used to meet these requirements are the FCV-53 Flight Computer (and associated software) and an 8-MB memory board FCV-53 Flight Computer Figure 12.1 FCV-53 Flight Computer The FCV-53 Flight Computer meets the low power and simplicity requirements. It operates on 3.3 V DC with a 16-bit microprocessor for central processing. It contains forty (40) individually addressed Input/Output (I/O) components. The memory, as described in the specifications, is continually checked and corrected, thereby fixing any bit errors associated with any radiation exposure. This reinforces the point of using off-the-shelf hardware, which already has the issue of radiation incorporated in the design of the component. Furthermore, this flight computer has an external port for the addition of RAM disk, which will be important as the group will be using additional external memory for the storage of data MB Memory Board 40

42 Figure 12.2 FM-16 Memory Board Shown here is the FM-16 Memory Board, which is similar to the FM-8 Memory Board that the group is going to use for the extra data storage capabilities. This board meets the requirements of simplicity and the necessary capabilities of data storage. It has a power consumption of 10 mw at idle and operates on 3.3 V DC. It is specifically made for attachment to the FCV-53 Flight Computer, and therefore special arrangements need not be made to connect the two components Space Environment on the Electronics The space environment is known to be extremely harsh for electronic components due to the radiation present. This can be visually seen below in Figure. 4 It can be seen that at an altitude of 800 km, the proton flux (where the energy of the protons is greater than 1 MeV) is found to be greatest along the region of the South Atlantic Anomaly (SAA). This radiation dose does not prove fatal, however, even though the concentration seems extremely elevated for the location analyzed. The shielding of the reinforced Spacequest components is sufficient to ensure that the radiation will not harm the electronics. Please see the Section?? for discussion of safety concerns for radiation Required Software Figure 12.3 Radiation at 800 km altitude 4 Courtesy of the analysis of the Canadian Space Telescope (MOST); this graph pictorially shows the 1 MeV proton flux at an altitude of 800 km. 41

43 The software required to operate the C&DH will be off-the-shelf equipment provided by Spacequest. The primary concern for the software will be to: Collect data Via an Analog-to-Digital (A/D) converter, convert the data into discrete sets Process the data to the central processor Send the data to the data storage facility Store the data to memory, until the command comes to download the data, at which point the job will be transferred to the TT&C subsystem (please see the next section) 12.6 Trade Studies The main trade study that was made was with the payload group to investigate the amount of data that would be required to be stored on board the spacecraft, prior to downloading to the ground station. It was discovered during the orbit selection phase that ideal orbit of 98.7 at an altitude of 800 km would provide an ideal time of passage of approximately 14 minutes. The design procedure for the data rate were made assuming a passage of 10 minutes and a data size of approximately 4 MB, and thus it was discovered that the pre-existing hardware that could provide a similar data rate was for a downlink transmitter operating on 9600 bps. The requirement was made that the uplink receiver work on a similar data rate, in order to accommodate commands from the ground station. It was discovered furthermore, that in order to accommodate a realistic BER of 10-5 a realistic data rate of 9600 bps for a GMSK modulation was required. 42

44 13.0 Power 13.1 Power Subsystem Requirements The requirements placed on the PuTEMP power subsystem are somewhat different from that of most satellites and result in a unique Concept of Operation (ref. Sec. 4). For most LEO satellites, one power cycle is defined by single orbit and the amount of sunlight and eclipse time. The primary power source, normally solar arrays (SA), must power all S/C subsystems and charge the secondary power source (batteries) during the sunlight portion of the orbit. During eclipse, when the SA is not generating power, the secondary source provides the necessary power. After each eclipse, the secondary power source is drained and must be recharged before the next eclipse period. PuTEMP is different in that it does not define a power cycle to be one orbit with the coinciding sunlight/eclipse periods. Instead, one power cycle is defined as the time it takes to prepare and conduct one experimental run of the payload. The payload requires a fixed period of time between experimental runs (to allow sufficient cooling of the tank fluid) that is longer than a single orbit. Because of this the secondary power system does not need to be fully recharged every orbit. Rather, charging can be accomplished over several orbits. Also unique to PuTEMP is the definition of Primary and Secondary Power. Power requirements during payload operation are 5.4 times above any other operating mode (ref. Sec. 13.3). This creates a large need for power but only for a limited period of time (15 minutes per experimental run). Studies discussed in Section indicated that solar arrays would not be able to provide sufficient power based on available solar panel array area. To insure enough power can be provided to the power-hungry payload, secondary (rechargeable) batteries were chosen as the primary power source. Solar arrays are then used as a secondary power source to charge the batteries after experimental runs as well as to provide power to all S/C subsystems Power Subsystem Budgets Power Budget Before exact details of the power system could be specified power required by individual S/C subsystems had to be determined. Initial power estimates were based on historical percentages. It was determined from the outset that the payload would consume the largest portion of S/C power. Based on simple calculations of energy transfer necessary to achieve a given temperature rise in a given fluid, the payload requirements were established (ref. Sec. 9). As expected, a significant amount of energy is required to raise the temperature of the experimental fluid. Every other subsystem was scaled to the payload power requirements based on historical data. As the various subsystem components were refined and specific hardware was selected, power requirements became more refined and margins were reduced. Table 13.1 presents the final power budget for PuTEMP. All items shown are actual flight components. For the majority of components, a margin of 1.2 was used. The 20% margin takes into account possible variations from the manufactures specifications and any losses that might not have been accounted for. 43

45 Subsystem Component Quantity Ideal Power (Watts) Margin Required Power (Watts) Voltage (V) CD&H CPU Bus Antenna Transmitter Receiver Modem Uplink Receiver Secondary Receiver Payload A/D Converters Heater Sensors AD&C Magnetometer Magnetic torquers Sun Sensor Table 13.1 Power Budget Margins for the tank heaters and AD&C components were given greater margins. The payload is designed to minimize heat transfer out of the fluid tank but cannot be totally isolated from the rest of the S/C. Some heat will be lost from the tank and heaters. It is thought that a 50% margin would account for any heat transfer losses. This is a significant margin because the heaters account for 78% of the required power. For the AD&C components, a margin of 1.4 will be used to account for the remaining uncertainties in the AD&C design. As AD&C requirements become more refined these margins can be reduced Mass Budget The mass budget remains a theoretical total at this point. Certain components have yet to be selected and so most be estimated. The battery weight is fairly certain and contains a 20% allowance for packaging and wiring. Similarly, the solar cell weight is accurately known but does not take into account a coverglass or any interconnections. Charge Controller mass is estimated to be 20% of the battery mass. Based on preliminary component studies this appears to be a reasonable number. Power Regulation and Control is over-predicted based on the mass of existing shunt regulators and blocking diodes. Harnesses and cabling are estimated to be between 5-25% of the total power system mass. Because PuTEMP is a relatively small satellite a value of 10% was chosen. The mass breakdown is presented in Table Overall mass is given a margin of 1.4 to insure that the mass predictions are not underestimated. Battery Weight 1296 g Solar Cell Weight 82 g Harnesses and Cabling 138 g Power Regulation and Control 100 g Charge Controller 259 g 44

46 Total Mass: 2624 g Table 13.2 Power Mass Budget 13.3 Operating Modes PuTEMP has several different operating modes that are used during different phases of the mission (Table 13.3). Experimental Mode is the primary mode of operation and the one demanding the most power. For all modes, the C&DH components are maintained at full power. This is necessary to insure the S/C can receive commands from the ground station. Based on full antenna and transmitter powers, the S/C will be able to transmit data during any mode. This is advantageous because it allows data transmission during every ground station pass, even if the S/C is in Recharge or Reorientation Modes. Subsystem Experimental Mode Transmission Mode Recharge Reorientation Component Mode Mode C&DH CPU x x x x x Bus x x x x x Antenna x x x x x Transmitter x x x x x Receiver x x x x x Modem x x x x x Uplink Receiver x x x x x Secondary Receiver x x x x x Payload A/D Converters x x x x Heater x Sensors x x (4) x (4) x (4) Attitude Magnetometer x Magnetic Safe Mode torquers x Sun Sensor x Total Power: 83.76W 13.38W 13.38W 15.62W 11.08W Time to Charge: 11.80Hr 11.80Hr 12.65Hr 10.93Hr Table 13.3 Operating Modes Reorientation will generally be performed before the beginning of an experimental run (ref. Sec. 10). The Safe Mode will only be entered when a problem is detected with the S/C. Time to Charge the batteries is shown and is based on an analysis performed in Appendix G. These recharge times are below the time required between experimental runs. Therefore, the payload limits the number power cycles Power Subsystems 45

47 A general overview of the Power Subsystem is presented in Figures 13.1 and 13.2 below. There are four major power subsystems: Power Generation and Collection, Charge and Discharge Control, Energy Storage, and Power Conversion and Regulation. The Power Generation and Collection uses the secondary power source, SA, to convert solar radiation into useful electrical energy. This electrical energy is fed to the S/C subsystems and the primary power source. The batteries, which constitute primary power, must be charged in a controlled way to insure proper battery operation throughout the duration of the mission. Improper charging of the batteries could result in degraded performance and possible loss of primary power. Power coming from either the primary or secondary power sources must be properly regulated so that each subsystem component receives power in the form it requires. Further details are presented in the following sections. Power Generation and Collection Secondary Power Charge and Discharge Control Voltage/Temp Sensing Power Conversion and Regulation Quasi-Regulated Energy Storage Primary Power Power Distribution Subsystem Loads Figure 13.1 Power Subsystem Design Current Sensor Charge Controller Temp Sensor Voltage Sensor Current Sensor Secondary Power - SA Primary Power Regulation and Distribution Figure 13.2 Power Subsystem Layout Primary Power - Batteries 46

48 Rechargeable (secondary) batteries are used as the primary power source. As discussed in the Section 13.1, the batteries are responsible for providing all necessary S/C power while in Experimental Mode. Battery selection is based on two primary requirements, the required battery capacity and the output voltage. Looking at Table 13.3, the power required during the Experimental Mode is watts. Taken over the total period of the experiment, this works out to be W-Hr of capacity. Consequently, the batteries must provide a capacity of at least W-Hr. Also, the majority of S/C subsystem components require an input voltage of 12 Volts. Given this, a battery output voltage of 12 volts would allow an unregulated bus during battery discharge. Batteries are made up of individual cells. Each cell has a given capacity and output voltage. Depending on the cell type chosen, individual cell capacities and voltages are well below the those required by the S/C. To achieve the proper output voltage from the batteries, individual cells must be strung in series. Each cell in series adds to the total available voltage. Once enough cells are strung in series to achieve the proper voltage, this arrangement is defined as a battery. To achieve the proper total capacity (20.94W-Hr) individual batteries are then strung in parallel. While any battery selected must meet the required capacity and voltage, there are other considerations that have to be taken into account when selecting a battery type. For our satellite, the desire for reliability and low-cost are major design drivers. Taking these factors into account, Nickel Cadium type batteries were chosen. NiCd batteries offer the advantage of a long and successful history of use in satellites. This extensive heritage has resulted in a very reliable battery with a wide range of commercially available configurations. The disadvantages of NiCd batteries are a low specific capacity compared to other battery technologies and a low depth-of-discharge (DOD). The low specific capacity means that NiCd batteries will have a larger mass compared to other battery types. However, the mass of the batteries turns out to be only a small percentage of the total mass and therefore not a significant design concern. The low DOD is not a significant issue because the mission will consist of a low number of charge/discharge cycles (< 1000) and a short (< 1 year) on-orbit life. Based on cycle life this means a DOD of 60% can be used with margin to spare. Specific NiCd battery type selection was based again on cost and reliability. After reviewing several small satellites battery choice, Sanyo Cadnica series battery cells came to the forefront. In fact, Cadnica cells are used by Surrey Satellite Technology Ltd., for their commercially marketed small satellite battery packs. Several different Cadnica batteries were examined to determine which cell would provide the most efficient design (Appendix G). An attempt was made to try and minimize the mass and number of individual batteries required to meet the capacity requirements. The individual cell voltage for all Cadnica series batteries is 1.2V requiring 10 cells to achieve the 12V requirement. In the end, the Sanyo Cadnica N-1700SCR was chosen. The important data for the cell is presented in Table The N-1700SCR offers the advantage of being a fast-charge cell. If necessary, the cell can be charged at a high-rate. 47

49 Cadnica N-1700SCR Nominal Capacity 1.7 A-hr Nominal Voltage 1.2 V Maximum Discharge Voltage 1.35 V Minimum Discharge Voltage 1.15 V Discharge Current 2.6 A (4C) Charge Temp (fast) 0-45 degc Discharge Temp (-)20-60 degc Weight 54 g Diameter 2.3 cm Height 4.3 cm Table 13.4 Battery Cell Specifications To meet the capacity requirements, only 2 individual battery will be needed. The overall battery power characteristics are presented in Table It should be noted that this configuration creates a power margin of 11.1%. As noted in the figure, the calculations include a 60% DOD and a total system loss of 5%. The 5% system loss takes into account losses through wiring and the various control systems. All told, the battery mass is just over 1 kg. Battery Requirements System Max Power W System Capacity W-hr Number of Cells (Series) 10 Nominal Bus Voltage 12 V Maximum Bus Voltage 13.5 V Minimum Bus Voltage 11.5 V Capacity (Individual Battery)* W-hr Number of Batteries (Parallel) 2 Total Capacity** W-hr Total Weight 1080 g Total Volume cm^3 Margin 2.32 W-hr % *Includes a DOD of 60% **Includes a 5% System Loss Table 13.5 Battery Power Characteristics The overall mass of the batteries was assumed to be 20% larger than the accumulated mass of the bare cells. This takes into account the additional mass of the battery packaging and wiring interconnects between the various cells. These numbers are also based on an operating temperature of 28 0 C. Any variation from this temperature will result in decreased performance. In general, high or low temperatures adversely affect the performance of NiCd batteries. Testing and matching of cells will have to be performed. The data presented above is based on average values provided by the manufacturer. Individual cell characteristics will vary slightly from the manufacturer s values. The exact charge/discharge characteristics for the cell will have to be determined to insure proper charging takes place. It is important to insure proper matching 48

50 of cells. If cells with different capacities are placed in the same battery, charging could be adversely affected as one cell might reach full charge before the others Secondary Power The only practical means available for generating power for a small satellite is photovoltaic cells. Photovoltaic cells operate by converting solar radiation into electrical power that can be used by the S/C subsystems. In most satellites, the SA functions as the primary power source (ref. Sec 13.1). However, it was thought that for PuTEMP a SA would not be able to provide the necessary power in Experimental Mode. The design requirement of high-reliability and low-cost meant that the use of deployable solar panels was not a viable option due to their added complexity. This limited the solar panels to a body-mounted design. With spacecraft dimensions already set, initial calculations based on solar cell current density suggested that body mounted solar panels would not be able to provide the required power for the payload. Therefore, batteries were chosen as the primary power source and the SA will be used as the secondary power source. The power generated by a solar array is determined by: Solar cell type Surface area available to mount solar panels Inherent degradation of the solar cells over the lifetime of the S/C. Solar cell degradation can have a significant effect on the performance of the SA. Radiation from the space environment is the major contributor to solar cell degradation and is a function of the orbit selected and the on-orbit time. As a result of this degradation the End-of-Life (EOL) performance of the solar panels most be taken into account. Solar cells must be arranged in a similar fashion to the batteries. Due to the charging characteristics of the NiCd batteries, the SA must provide a total voltage of 1.45V for every cell in the battery pack. This means the SA must provide a total voltage of at least 14.5V. To achieve this individual solar cells are wired in series until the additive voltage of the individual cells equals the total required voltage. Much like batteries, this series string of cells is called a solar panel. The parallel connection of the solar panels into an array results in the overall power of the solar array. Two major types of solar cells are available for use on small satellites. Until recently, silicon based cells were the industry standard. Silicon cells offer efficiencies in the ~15% range. Recently, Galilium Arsenide (GaAs) based solar cells have been increasingly used. While they are more expensive to manufacture, triple-junction InGaP/GaAs/G cells can offer efficiencies of ~23% with increased radiation tolerance over Si cells. The disadvantages of GaAs based cells are increased weight and higher operating temperatures. Several different solar cells were examined to determine the performance variations (Appendix G). Based on the higher efficiencies, it was no surprise that the GaAs based cells produced significantly higher power outputs than the Si cells. It is important to note that the highest power 49

51 output achieved was 51.6 W. This is well below the 83.8 W required during Experimental Mode thus providing justification for the selection of batteries as the primary power source. The Si cell achieved a power output of 24.4 W, roughly a fourth of the power required to run in Experimental Mode. As discussed above, degradation had to be considered when determining the solar array performance. Power outputs for each solar cell type are based on EOL performance with an assumed radiation fluence equivalent of MeV e-/cm 2 corresponding to approximately one year of radiation exposure in LEO. For the Si cells, the radiation degradation results in 91% of BOL power available at EOL. Solar cells are also affected by their operating temperature. As the temperature of the cell increases, the efficiency is reduced. An operating temperature of 60 0 C was assumed. The corresponding loss in power was 16% and was taken into consideration. The Concept of Operations (ref. Sec. 4) is such that a certain amount of flexibility exists in the time required to recharge the batteries. The GaAs cells would be able to charge the batteries quicker due to their higher power outputs. The Si cell provides 11.1 W of power above what is required by the S/C subsystems in Recharge Mode (ref. Fig. 13.3). This translates to an excess current of 720mA and an equivalent charge rate of ~0.4C. Based on charging considerations discuess in Appendix G, this corresponds to a recharge time of 11.8 hours. This means that the Si cell will have no problem charging the battery within the time required by the Concept of Operations. As a result, the Si cells will still be able to meet the system requirements at their reduced power output. Given that both the Si and GaAs cells will meet the power requirements, other factors have to be considered. The advantages of GaAs with respect to radiation degradation have been taken into account by the EOL power. Si cells offer the advantage of a lower operating temperature, which is advantageous from a thermal perspective. But perhaps the largest factor to consider is cost. Si cells are lower-cost than GaAs cells because the technology is more mature and the manufacturing is easier. With cost taken into account, the Si cells will be used. The cell and array characteristics are shown in Table Spectrolabs Silicon K4702 Solar Cells Current Density (mp) 36.8 ma/cm^2 Voltage (mp) V Efficiency (mp) Weight g/cm^2 Thickness 0.02 cm Cell Area 27.5 cm^2 Table 13.6 Solar Cell Specifications Solar Cell Requirements Number of Cells (Series) 27 Individual Panel Voltage V Total Panel Area 1485 cm^2 Individual Panel Current A Individual Panel Power W Number of Panels per Side 2 Total Power per Side W 50

52 Total Power per Side* W Final Voltage* V *Includes losses due to radiation and temperature Table 13.7 Solar Panel Characteristics The solar panels will be placed on the four largest faces of the S/C. The cell area of 27.5cm 2 was chosen to match the other cells being compared. Spectrolabs is able to make cells up to 8x8 cm. A minimum of 27 cells must be placed on each face to insure the proper output voltage. Arranging the solar cells onto each face then becomes a matter of tile placement. Given the dimensions of each face and using a solar cell dimension of 7.6x3.7cm, two solar panels were able to be placed on each face. The packing factor for this arrangement is 0.79, or 79% of the total face area. As the satellite rotates along its Nadir pointing axis, solar panels on each face will move into and out of sunlight. At most, only two of the four faces will see sunlight at any given time. Solar cell output is proportional to the cosine of the angle each panel makes to the sun. Thus, the power output of each face will vary as a cosine wave. Total power output will then be an additive function of each face s output. This is shown in Appendix G. The most important result shown in Appendix G is that the minimum power output by the solar array is when one face is normal to the sun and all other faces are shadowed. Therefore, this minimum power case was used when determining charge times and power available to S/C loads Charge and Discharge Control Charge control is necessary to insure safe and reliable operation of the batteries throughout their lifetime. In general for NiCd batteries, charge control becomes increasingly important as the charge rate increases. As the battery nears full charge a proportionately larger amount of energy is wasted creating heat and reaction gases inside the cell. Higher charge rates increase this effect. When high-rate charging is used, the cell cannot handle over-charging as the battery will over pressurize. A charge controller is then required to reduce the voltage when full charge is neared and to stop charging when battery capacity has reached 100%. A wide variety of charge control methods exist with varying levels of sophistication and control. The type of feedback received from the battery and how it is used determines the type of control method. Perhaps the simplest methods for feedback control involve measuring the batteries temperature or voltage. The important considerations for each are presented below: Temperature Sensing: Fast-charge cells can be controlled simply with temperature monitoring as their temperature does not rise much until full charge is approached Temperature rise for a specific cell has a direct relationship to charge current Voltage Sensing: 51

53 Voltage can be used to sense cell charge, but varies with temperature, age, history, and construction For fast-charge cells, when the temperature begins raising rapidly (fully charged) the voltage will begin to decrease from its maximum Individual battery characteristics must be well understood to allow successful use of voltage sensing control Each method presents its own advantages and disadvantages. The variations that can affect the cell voltage and temperature make each control method somewhat unreliable when used in isolation. However, the combination of voltage and temperature sensing results in one of the most reliable forms of battery charge control. Voltage-Temperature Cutoff (VTCO) is a control scheme that turns off charging when either a voltage or a temperature limit is reached. VTCO results in high charge level without compromising battery life and reliability. Three major components are required for a VTCO feedback charge controller (ref. Fig. 13.5). The heart of the VTCO is a Charge Control Unit (CCU) which receives information from the voltage and temperature sensors. The CCU is programmed to terminate charging when the heating rate or peak voltage limits are met. The CCU can be programmed to take into account several different variables affecting charging. The other two components are the voltage and temperature sensors which feed into the CCU. Component selection has not been completed but is under investigation. The temperature sensing will be accomplished by thermistors identical to those used by the payload (Ref. Sec. 9). This will simplify the design of the system by requiring fewer A/D converters. Four thermistors will be required, two on each battery pack. The second thermistor serves as a backup on each battery pack. A specific voltage sensor has not been selected but these devices are commercially available and simple in design. More consideration will have to be given to CCU selection. The Phillips Semiconductors TEA1102 series of Fast charge ICs has been investigated. The TEA1102 is a board-mounted chip with T/ t and peak voltage detection. The chips are ~1 cm 2 and represent a minimum mass addition to the overall S/C Power Conversion and Regulation Power Conversion and Regulation is accomplished through a Direct-Energy-Transfer (DET) scheme. A DET system operates by dissipating excess current provide by the solar arrays but not needed by the S/C loads. This dissipation is accomplished through the use of shunt regulators which convert the excess current into thermal energy which can be radiated into space. The DET system offers a simple (reliable), low-mass design. The shunt regulators are selfcontained units with no outside control required. Because only excess power is dissipated, the DET system is highly efficient. The use of shunt regulators results in a quasi-regulated bus design. A quasi-regulated bus works by maintaining a constant voltage during battery charging and an unregulated voltage during discharge. Solar cell output is based on an I-V curve with voltage output proportional to the current demand. As loads are removed from the system, the voltage output from the SA shifts downward. The shunt regulators are connected in parallel with the SA (Ref. Fig. 13.5) and serve 52

54 to maintain a constant load on the SA. The shunt regulators dissipate more current as S/C loads are switched off and the power is not needed. By maintaining a constant load on the SA, the shunt regulators maintain a constant voltage to the S/C bus. As a result, when secondary power is being used the bus voltage is regulated to the fixed operating point set on the I-V curve. The 14.5V output discussed in Section corresponds to the Peak-Power point on the I-V curve. No attempt to maintain a constant bus voltage while on primary power is made. NiCd batteries maintain a relatively constant output voltage throughout their discharge. Bus voltage will vary between 13.5V and 11.5V with a majority of time being spent at 12V. Only two S/C subsystem components (Magnetometers and Magnetic Torquers, both 5V) require an input voltage different than 12V. It is then simpler and more efficient to regulate the voltage specifically to these two components. The rest of the subsystem components should be able to operate within the range of voltages provided by the bus. Avoiding the use of bus-level voltage regulation during discharge results in a simpler design with no losses being incurred through the use of a voltage regulator. Reducing the voltage to the Magnetometers and Magnetic Torquers can be accomplished with DC-DC converts. These devices operate by simply stepping down an input voltage to a specified output voltage. Several different DC-DC converters were studied. The results are presented in Appendix G. All converters reviewed were simple, pin-type chips. The Gaia Converter MGDSI-04-F-C was chosen because its output current of 800mA was in the required range. The overall design is suitable for mobile use suggesting a rugged design suitable for the space environment. Its specifications are presented in Table Converter MGDSI-04-F-C Input Voltage (V) 9-18 Output Voltage (V) 5 Output Current (ma) 800 Efficiency.78 Dimensions (cm) 3.2x1.9x1.2 Operating Temp. Range ( 0 C) -40 to 70 Mass (g) 10 Table 13.8 DC-DC Converter Specifications When solar panels are not illuminated and generating current, they will accept a reverse current. To prevent this, each face of solar panels must be placed in series with a blocking diode. The blocking diode prevents current from flowing backwards into the solar panels. A similar blocking diode must be placed in parallel with the charge controller to prevent battery discharge during charging. 53

55 15.0 Structures and Mechanisms Z X Y X Z Component Representation Physical Component Dimensions [cm] (length, width, height) BlackBoxes Sun Sensors Blue Cells on Exterior Solar Cells 3.7x.001x7.6 Blue Interior Rods Magnetic Torquers Vert. R=0.75, H=32.5 Hor. R=0.4, H=24 Grey Frame Load Bearing Frame 30x30x60 Green SPT R=6, H= 25 Green boxes (top view) Transmitters 2.5x7.0x1.3 Orange Plate (on bottom) Attachment Plate for Ariane 5 Launch vehicle (AL 7075) 3.5x9.5x10.7 Purple Box CPU 1.0x14.0x16.0 Salmon Box (top view) Antenna 7.1x15x3.0 Dark Teal Boxes Receivers 7.3x8.3x2.8 White Boxes Batteries Yellow Box Modem 7.0x1.0x7.7 Brown box (not in view) Bus 8.3x2.8x7.2 Orange Side Panels Side Panels (AL 7075.T6) 30x0.127x60 Red Box (Tall) Gravity Gradient Housing 11.5x10.2x26.4 Figure 15.1 Satellite Internal Layout 54

56 15.1 Structural Layout and Design The PuTEMP structural layout and design can be broken into three primary categories: load bearing structure, support for the Simulated Propellant Tank (SPT), and support for all the other subsystems. Since the satellite s main purpose is the success of the SPT experiment many of the requirements for the structural layout are driven by this component. This section details the decisions made in the areas concerned with the satellite structure and internal component layout. A mass breakdown according to subsystem components is listed in Table A detailed breakdown by subsystem is available in Appendix?. Goal Mass (kg) 40.0 Actual Mass (kg) 45 Subsystem Mass (kg) Margin % of Total Payload* TT&C C&DH Power AD&C Structure Total Table 15.1 Subsystem Mass Breakdown (including margins) *Thermal is included in payload Designing a small satellite to be launched as a secondary payload automatically sets requirements as to the overall dimensions of the satellite. The launch vehicles have specified dimensions and mass requirements, which must not be exceeded. It is advantageous to design a satellite that can be compatible with as many launch vehicles as possible. This opens up many launch options and opportunities. The dimensions of 30x30x60 cm were chosen as design parameters. These constraints come from the Delta launch vehicle. These dimensions also allow the satellite to be compatible with the Ariane IV and Ariane 5 launch vehicles. The Ariane 5 launch vehicle was chosen as our primary design to launch vehicle because it meets the sunsynchronous orbit requirement and other pertinent factors (discussed more in Sections 6 and 7). Some of the Ariane 5 secondary payload requirements are not yet established. In this case the stricter of the other two launch vehicle requirements will be designed to. The load bearing structure design criteria for the three launch vehicles being designed to are listed in Table15.2. Ariane IV Ariane 5 Delta 6925/

57 Secondary Payload Dimensions Steady State and Dynamic Frequencies (Hz) 45x45x45cm 78x60x60cm 78x33x33cm Axial Lateral Axial Lateral Axial Lateral >31 >10 >90 >45 >35 >15 Limit Load Accelerations (in g's) -7.5 to to to to +6.0 Maximum Deflections N/A N/A N/A N/A N/A N/A Table 15.2: Bold values indicate design criteria for PUTEMP satellite. (Ref. 1,2) Ariane 5 launch vehicle is strictest for the secondary payloads frequency and axial loading stiffness requirements. This indicates yet another reason the Ariane 5 is chosen as the design to launch vehicle. The satellite in its currently described configuration, Figure (in the beginning) is shown to be compatible with the Ariane 5 launch vehicle attachment ring. An extra plate of removable material is fitted to the bottom of the satellite upon which the launch vehicle attachment ring is bolted. This specific launch configuration is also reflected in the inertia and mass values. Launch attachment configurations for the Delta and Ariane IV were not readily available and are therefore not investigated in this design. It is anticipated that satellite modifications for attachment to the other launch vehicles does not require any more of a redesign effort than the Ariane 5 and the extra plate of larger material. 56

58 Figure 15.2 Load Bearing Structure, side panels shown in orange for contrast The load bearing structure, as seen in figure 15.2 of PUTemp consists of the aluminum 7075.T6 angle iron frame, shelves to support the internal systems (primarily during the launch phase) and load bearing panels. Aluminum was chosen due to its ability to be easily manufactured per the requirement of Purdue laboratory capabilities. Angle Iron was chosen as opposed to bar stock material to decrease the weight with little loss in directional stiffness. As well, angle iron frames have been used on a number of small satellites previously designed, Ref. 3,4. The angle iron dimension options are chosen due to the manufacturing availability of aluminum angle iron and modal analysis requirements. The aluminum material itself has requirements on yield and ultimate stress and tensile forces. Table 15.3 outlines the material properties of Aluminum 7075.T6. As shown in Figure 15.2 the shelves are held in place by aluminum angle iron on the top and bottom surfaces of the shelves and attached to the frame structure on the opposite brace. The shelves are each 1cm thick, the reasoning for this is discussed in the modal analysis section. The T-bar cross beams can also be seen in Figure Density (kg/m^3) Young's Modulus (N/m^2) Yield Strength f (MPa) Thermal Expansion ((1e-6) m/m*k) Fatigue Strength (MPa) Ultimate Sheering Strength MPa Ultimate Table 15.3 Material Properties for Aluminum 7075.T6, (Ref S.M.A.D.) Yield MPa Comment Prone to *S.C.C 57

59 * Stress Crack and Corossion Per the launch vehicle requirements two types of analysis were necessary for the design validation prior to launch. Ansys finite element modeling software was used for both of the analyses. The first of the analysis conducted is dynamic load bearing conditions of the satellite during the launch phase of the mission. Historical information has shown that the launch vehicle frequency requirement tends to have a much larger impact on the satellite overall design than the static requirements. (Ref. 2) A solid volume model of the satellite was imported into Ansys from Ironcad and a natural frequency modal and static loading analysis are conducted on the frame. was conducted on the satellite with the entire bottom surface constrained (as thought it were attached to the launch vehicle. Figure 15.1 shows the imported model. The mesh size represents typical mesh grid used for the analysis. The side panels are not viewed here and were not attached to the structure for the Ansys analysis. When the side panels are included the number of elements created in the mesh are too large for the available package of Ansys to evaluate. The side panels are taken into account by noting their added mass is 0.64kg each. This mass is reflected in the mass budget for the satellite structural mass. Since this mass is not included in the modal analysis the provided frequencies are in fact slightly lower. The entire load bearing structure viewed here is aluminum 7075.T6. See Appendix I for material properties used. 58

60 Z Y X Figure15.3: Ansys structural model imported from Ironcad for Static Loading Analysis Dynamic Analysis The modal analysis was conducted using the frame structure and internal volume components. The internal volume components were imported into Ansys from Ironcad and given appropriate densities which corresponded to the volumes imported for the mass properity. The resulting Ansys frequencies for the satellite box structure are listed in Table As previously stated the side panels were not included in this analysis. Therefore the frequencies stated in Table 15.4 would in fact be lower than shown here. A reasonable method to estimate exactly how much lower the frequencies would be is not known, therefore it is left to the future detail design of the satellite structure to determine the frequencies more accurately. 59

61 Mode Frequency (Hz) Motion Axis of Mtion in/out along Z right/left around X up/down around Z twist/rotation around Z rotation/extend around Z Table 15.4 Ansys Modal Frequency analysis. To achieve the frequencies as viewed in Table 15.4 the modal analysis was conducted on individual satellite structural components before the entire frame was assembled and analysis conducted. For example the shelves were made 1cm thick because according to the Ansys results for the shelf Ironcad model mad eof Aluminum material and constrained on all four sides the natural frequency of this shelf is 97 Hz. The structural components were sized accordingly as well. The resulting sizes for the individual structural components are listed in Table Frame Structural Component Size [cm] Constraints Frequency [Hz] Shelf 30x30x1 four edges imobile 97 Shelf Support Constrained on one side 30x1x1 Angle Irons area sn on outer edges 116 T-Bar Cross Constrained on long side 0.75x1.0(web)x30 Supports edges and in middle 110 Frame *Total Thickness Constrained on Bottom Web Length 5.08 Surface 83 Table 15.5 Frame structural components sizes * Total Thickness does not include thickness of bottom surface where the mass of the antenna, receivers, and transmitters was lumped into this mass. From Table 15.5 it is appearant why each structural member was sized to a frequency well above the required 90Hz. When the mode shapes for the overall frame frequencies were observed it was evident which structural component was driving the frequency since the individual components had previously been characterized for their modal motion. This also aided in the raising of the frequency in by determining which modal direction was in need of strengthening Static Loading Analysis The Ansys static analysis conducted used distributed forces over the surfaces of the satellite where the internal components are located. The force magnitudes used were calculated based on the mass of the components and the maximum launch limit load valued, provided in Table The components placed on the bottom of the satellite (viewed as top here), i.e. antennas, receivers and transmitters, were combined as a single mass the size of the bottom surface and mass is bouleoned into the top surface of the satellite. This lumped mass approximation simplified the model. Some cases were conducted for point massed on the bottom surface and it is believed that the material stresses will not be entirely insignificant compared to the SPT stresses on the shelf. As is seen in Figure 15.4,the result for the shear stress in the XZ direction, 60

62 the middle shelf which shall carry most of the SPT load during the launch position carries the most stress as well. Figure 15.4 Example Ansys static analysis result for Shear stress in the XZ direction Maximum Deflection [cm] Stress [MPa] Minimum Maximum *Material Strength Margin (Maximum Magnitude) X Y Z Shear Stress [MPa] 61