Transcription

SSC01-VI-5A Systems Engineering Tool for Small Satellite DesignAllan I. McInnes, Daniel M. Harps, Jeffrey A. LangVehicle Concepts DepartmentThe Aerospace CorporationM4/922, P.O. Box 92957Los Angeles, CA 90009-92957Phone: (310) 336-1871Email: [email protected] Charles M. SwensonDepartment of Electrical and Computer EngineeringUtah State University4120 Old Main HillLogan, UT 84322Phone: (435) 797-2958Email: [email protected] The growing popularity of small satellites for applications of all kinds has lead to a marked increase inthe number of requests from customers of The Aerospace Corporation for studies involving small satellites. Theexisting design tools used by the Corporation for concept evaluation of large spacecraft have, in many cases, proveninadequate for these small spacecraft studies. As a result, Aerospace is developing a systems engineering tool tosupport the conceptual design of small satellites.The Aerospace Corporation’s small satellite systems engineering tool utilizes a spreadsheet-based approach toefficiently track information regarding the mass, power, and volume of the satellite subsystems. This subsysteminformation is derived through a variety of means, including analytical relationships, iterative solvers, and databasesof components appropriate for small satellites. Physics based models for such factors as solar illumination andexternal torques have been incorporated into the tool to aid in the analysis of the design.In addition to data tracking, the spreadsheet approach used makes it easier for a concurrent engineering methodologyto be applied to the design process. This means the effects of a change in one subsystem are immediatelypropagated to the other subsystems, and system-level effects are more easily identified. The end result is a tool thatfacilitates rapid systems-level concept evaluation and trade-space exploration in support of the small satellite designprocess.This paper describes The Aerospace Corporation’s small satellite systems engineering tool. The approachunderlying the tool, as well as an overview of the implementation, relationships between the subsystems, and theflow of information are presented.spectrum. Both civil and military space programs havelaunched research and development efforts focused onsmall satellites. Examples include NASA’s SpaceTechnology 5 Nanosat Constellation Trailblazer, andthe Air Force Research Laboratory’s TechnologySatellite of the 21st Century (TechSat 21) andMightySat programs, which seek to test and provetechnologies and architectures. Further advances insmall satellite capabilities are being driven by researchinto new technologies such as microelectromechanicalsystems (MEMS)1. For example, The AerospaceCorporation’s Center for Microtechnology isIntroductionA growing number of future space missions, for eitherprogrammatic or technical reasons, require small, lowmass, low-cost satellites. Interest in these new missionconcepts is encouraged by the perception that a smallsatellite can be both capable and low-cost. As a result,more small satellites are being formally studied at theconceptual stage of many civil, commercial, andmilitary programs than have been in the past. Figure 1clearly illustrates the impressive growth in the numberof small satellites launched over the last two decades,particularly at the smaller ( 25kg) end of the size1Copyright 2001 The Aerospace CorporationAllan I. McInnes15th Annual AIAA/USU Conference on Small Satellites

investigating a concept for future satellites, in which theentire spacecraft is fabricated on silicon, using acombination of MEMS components2.USA Small Satellite LaunchesCivil, Commercial, MilitarySatellites Launched25 Small satellites often have fixed solar arraysinstead of sun-tracking solar arrays Small satellites often do not have deployables Small mass leads to reduced thermal inertia Small size leads to reduced power generationand storage capabilities Volume can be tightly constrained Surface area can be at a premium Little historical data is available at the lowerend of size spectrum (there have beenrelatively few programs, and those are notalways well-documented), making parametricresource estimates difficult Small satellites use smaller components, newtechnologies (e.g. MEMS), and non-traditionalvendors0 to 25 kg2025 to 50 kg50 to 100 kg15100 to 200 989198819871986198519841983198219811980-YearFigure 1 Growth in small satellite launchesIt is well known that decisions made in the conceptphase of a program can determine approximately 70%of the cost of a program3. The increase in smallsatellite launches, and the planned inclusion of smallsatellites in so many future programs, indicate a needfor systems engineering tools to aid in the conceptualtrade studies for these programs. These tools must beappropriate to small spacecraft and the newtechnologies from which they will be composed. TheAerospace Corporation (hereafter referred to asAerospace) is presently working to develop such tools.These differences mean that although the process usedto design small and large satellites is similar, the toolsrequired to support the process are different.Small arge SatelliteSun-trackingsolar arraysNewtechnologiesFixedsolar arraysSmall Satellite Systems EngineeringSystems engineering is concerned with the overallperformance of a system for multiple objectives (e.g.mass, cost, and power). The systems engineeringprocess is a methodical approach to balancing the needsand capabilities of the various subsystems in order toimprove the performance of the system. The size,volume, and mass constraints often encountered insmall satellite development programs, combined withincreasing pressure from customers to pack morecapability into a given size, make systems engineeringmethods particularly important for small satellites.Low densityLarge volumeDeployablesSurplussurface areaFigure 2 Small Satellites vs. Large SatellitesConcurrent Engineering MethodologyTraditional design methodology is a sequential,multidisciplinary process, and as such, has severaldisadvantages.Often, one subsystem cannot bedesigned until the results from another subsystem areavailable. Communication of design data from onesubsystem specialist to another can be complex andtime-consuming. Thus, due to the time required tocomplete a design iteration, the number of iterationsthat can be performed is very limited4.Spacecraft systems engineering is an established andwell-understood discipline. However, many of thestandard tools and techniques used to performconceptual design of spacecraft contain implicitassumptions that are based on the characteristics oflarge satellites.This is a problem, since thecharacteristics of a small satellite can differ from thoseof traditional large satellites (Figure 2) in a number ofways:In an effort to improve upon the traditional sequentialapproach to design Aerospace has developedcentralized design processes (Figure 3) based on aconcurrent engineering methodology. Using this designprocess a systems engineer works with subsystemspecialists to generate simplified subsystem design2Allan I. McInnes15th Annual AIAA/USU Conference on Small Satellites

algorithms.The systems engineer compiles theinformation into a single spreadsheet-based tool knownas a Concurrent Engineering Model (CEM). Thesystems engineer then uses the spreadsheet model toquickly design a spacecraft and examine possibledesign trades. Because the models and information foreach subsystem are linked within the spreadsheet, theeffect of any change is instantly seen by all of thesubsystems. All of the subsystem requirements areconsidered simultaneously. The trade space that can beexplored in a given amount of time is greatly expanded.The design cycle can be shortened from months to daysor even hours.Figure 4 Comparison of design processesThe concurrent engineering approach to spacecraftdesign can also be used in a distributed mode, in whichspecialists operate individual subsystem spreadsheetsthat are linked via a network (Figure 3). This kind ofdistributed real-time design process has beensuccessfully used in Aerospace’s Concept DesignCenter4 (CDC), as well as in other similar facilities suchas JPL’s Project Design Center (PDC). The distributedprocess has the advantage that the subsystem specialistsremain in the loop during design iteration, allowingmore complex subsystem design algorithms to be used.As a result, a distributed process can achieve a higherlevel of design fidelity than is typically available withinthe framework of a centralized design process.Small Satellite Concurrent Engineering ModelSpacecraft conceptual designs prepared by Aerospaceare used to support feasibility studies, program costestimates, trade space explorations, and technologyinsertion studies. Aerospace has developed severalsystems engineering tools to support spacecraftconceptual design tasks, including a number of CDCteams, and a variety of CEMs. However, these toolsare usually intended for designing large satellites, andthus incorporate assumptions that make the tools lesseffective in the small satellite regime.In an effort to improve Aerospace’s small satellitedesign capabilities, development of the Small SatelliteConcurrent Engineering Model (SmallSatCEM) wasinitiated. The SmallSatCEM project aims to produce atool that will aid systems engineers in performingconceptual-level design studies of small satellites. TheSmallSatCEM is intended for use as a tool to generatepoint designs in support of conceptual design studies.Additionally, it is hoped that component and subsystemdesigners will find the tool useful in estimatingperformance requirements for new small satellitecomponents that are in the development phase.Figure 3 Types of design processSmallSatCEM GoalsThe primary goal of the SmallSatCEM project is todevelop a useful tool that will allow satellite systemsengineers to rapidly design and analyze a small satellitebus. A secondary goal is to develop the tool in such away that it can be used widely within Aerospace. TheSmallSatCEM is aimed at designing single spacecraft,and thus will not include capabilities such asconstellation design. However, the SmallSatCEMcould be used to derive a spacecraft design suitable forsome pre-determined system architecture andconstellation geometry.Aerospace has a dual approach to conceptual spacecraftdesign, with the choice of design process depending onthe needs of the customer (Figure 4). The CDC is usedwhen higher fidelity and direct interaction with thecustomer is desired. Specific trades are developed indetail. A CEM is used when a rapid answer is needed,a broad trade space is desired, and lower fidelity isacceptable. CEMs are usually developed for a specifictype of mission (e.g. GEO communications spacecraft),and then modified and reused for later design studies.3Allan I. McInnes15th Annual AIAA/USU Conference on Small Satellites

There are a number of characteristics that contribute tothe usefulness of a systems engineering tool5, such as Relevance to the study being performed Credibility in the eye of the decision maker Responsiveness of the model Transparency User friendlinesslanguage so, in principle, any functionality of anexternal program can be duplicated. This is being done,although only to the level of fidelity that is needed forconceptual design. Another advantage of using Excel isthat it is widely known and flexible enough that theSmallSatCEM can be expanded or customized by a useras needed. Unlike compiled software, the use of Excelwith Visual Basic makes the internals of the tool easilyaccessible, and thus checking equations or updatingmodels is very easy.The design of the SmallSatCEM addresses thesedesirable characteristics in various ways.The SmallSatCEM is conceived as a single user tool,and it is intended that the user be able to complete adesign without the need for intervention by a subsystemexpert (although consultation with experts wouldobviously be beneficial, and is encouraged). To helpthe user rapidly specify a new spacecraft design theuser interface layout of the spreadsheets is consistentthroughout the Excel workbook.Diagrammaticrepresentations of subsystem models are included to aidthe user in understanding how the tool functions andhow the required user inputs are used to specify thesubsystem design.Since the tool is intended for use in small satellitedesign studies, a conscious effort has been made toavoid the inclusion of modeling assumptions that arerelevant only to large satellites. The SmallSatCEMinstead includes databases of components appropriatefor small satellites, and physical models (such as solarillumination of fixed solar cells) that support the typesof analysis needed for small satellite design.Additionally, attributes that are important in designingsmall satellites, such as volume and surface area, aretracked and reported. The development team isimplementing standard models and equations6,7 wherethey are appropriate, and validating the models againstother tools wherever possible.Workbook ArchitectureThere are two major processes in an iterative designcycle: the progression from requirements throughdesign to a design state, and the analysis of a specifieddesign to determine its performance relative torequirements (Figure 5). This design, build, and testcycle can be applied both at the level of the wholespacecraft, and at the level of a single subsystem.Spacecraft studies at the conceptual level may involveone or both of these processes, depending on the goalsof the study.The subsystem models that are being implemented inthe SmallSatCEM are intended to be simple enough todescribe a design at the conceptual level, and yetprovide sufficient detail to isolate the major systemdrivers within the spacecraft design.TheSmallSatCEM team has made an effort to keep themodels as general as possible in order to ensure that thetool is reasonably flexible. As an example, theastrodynamics model is not tied to a particular planet(the tool presently supports Earth and Mars as centralbodies), and allows for elliptical orbits of arbitraryinclination.At the conceptual stage of a project the systemsengineer is often faced with the task of designing apossible top-level architecture for a spacecraft todetermine project feasibility. A typical approach to thistask is to use the requirements of the payload to selectsome known hardware components (i.e. make systemdesign decisions), thus leading to a spacecraft design.The arrow at the top of Figure 5 illustrates this process,in which the requirements lead, via design decisions, toa design “state”. The state specifies the components ofeach subsystem, as well as mission details for thespacecraft, as determined from the design process.The SmallSatCEM is being implemented in MicrosoftExcel as a single workbook, and makes extensive useof Visual Basic to extend the capabilities of Excel.The decision to use Excel and Visual Basic was madefor several reasons. A primary driver is that the use ofExcel allows the SmallSatCEM team to rapidly developa reasonably uncomplicated, user friendly, selfcontained tool that can be easily distributed throughoutAerospace. The temptation to link with externalprograms has been expressly avoided because of thepotentially limited availability of these programs toother users. Visual Basic is a full programmingRather than proceeding from requirements to a design,a different question is often posed at the conceptualstage of small satellite projects: given a volume andmass constraint, what can be done with a smallspacecraft of a given configuration, or what type ofpayload can be supported? A common variation on thisquestion is: given a small satellite configuration, what Microsoft Excel and Microsoft Visual Basic are trademarks of theMicrosoft Corporation.4Allan I. McInnes15th Annual AIAA/USU Conference on Small Satellites

DesignVisual Basic Analysis ToolsState RequirementsConfiguration:size, shape, etcMass propertiesSolar cell &batteriesSurface thermalpropertiesOrbit and attitudeTelemetry ratesetc Payload powerPayload pointingThermalconstraintsCoverageTelemetry needsPointing accuracyetcModelStateSubsystem Design SheetsAnalysisRequirementsFigure 6 Information flow within theSmallSatCEMFigure 5 Iterative Design Cyclesubsystem miniaturization or new technologies need tobe developed to make this system possible? Both ofthese questions are examples of a situation in which thestate of a spacecraft is specified or hypothesized, andthe requirements that can be supported by the spacecraftmust be determined. The arrow at the bottom of Figure5 represents this process of analysis, in which someaspect of the spacecraft or subsystems is simulated todetermine performance.The SmallSatCEM is being implemented such that itrepresents the spacecraft subsystems in terms of alimited set of pre-defined units (components orfunctional blocks), the parameters of which can bespecified by the designer. Once state of the spacecraftdesign is defined in terms of these parameters a varietyof analysis tools can be applied to simulate theperformance of the design during mission operations.These analysis tools are largely implemented in VisualBasic, and answer specific questions about issues suchas power production, disturbance torques, or telemetryaccess times.A systems engineering tool for conceptual design mustfacilitate both the design and analysis processes in orderto be flexible enough to deal with a large variety ofdesign problems.With this in mind the Excelworkbook that comprises the SmallSatCEM has beendeveloped to explicitly reflect the cycle portrayed inFigure 5, and, as a result, incorporates a worksheetwithin the workbook dedicated to tracking the designstate of the spacecraft. Orbit and payload requirementsare captured on a separate worksheet, and flow fromthere to the worksheets used to describe eachsubsystem. The information flow within the workbookis depicted in Figure 6, which helps to illustrate the wayin which the workbook structure maps to the iterativedesign cycle.Figure 7 Component selection from adrop-down menuThe workbook uses separate worksheets to describe,design, and analyze each subsystem. The worksheetseach contain a specific section intended for use as adesign tool, the outputs of which contribute tospecifying the state of the spacecraft. The design toolsconsist of menu selections from databases ofcomponents (Figure 7), historical models built up fromexperience, and computation chains of physical models.The worksheets being implemented as part of theSmallSatCEM are: Payload & Mission, unications (including telemetry, tracking &control), Attitude Determination & Control, Power,Mass Distribution and Properties, Thermal, ModelState, Database, and Cost. A brief description of eachworksheet will follow.5Allan I. McInnes15th Annual AIAA/USU Conference on Small Satellites

Payload & MissionThe payload is the purpose of the mission; this makesthe payload requirements the driver for the entiredesign. Up to five generic payloads may be specifiedby providing system requirements such as mass, power,volume, slew requirements, stability, data rates, and soon. The payload requirements are passed on to eachsubsystem sheet, where they are used as a guide forsubsystem design decisions. Also specified on thePayload & Mission sheet are the orbital elements forthe initial, operational, and disposal orbits, and thedesired operational attitude.ConfigurationThe size and geometry of the spacecraft bus is specifiedon the Configuration sheet (Figure 8). Spacecraftgeometry is limited to a right cylindrical polyhedron,for which the designer specifies a height, diameter, andnumber of sides. Up to 8 deployable “panels” can bespecified in terms of size, location, and tracking mode.These panels can be used to simulate deployable solararrays, antennas, thermal radiators, or gradient booms,depending on the specified panel shape.PropulsionThe Propulsion sheet is divided into transfer propulsionand on-orbit propulsion sections. These systems aredesigned independently, using drop-down menus toselect the thruster type and quantity for each system.The change in velocity ( v) requirements are calculatedfrom the orbit parameters defined on the Payload &Mission sheet. The propellant mass and tank size isthen determined using an iterative solver.Figure 8 Specifying a configurationCommunicationsThe Communications subsystem is divided intotelemetry, tracking & control up and downlinks, a datadownlink, and crosslinks. The hardware that compriseseach link is designed independently via databaseselection of components. A link analysis tool calculateslosses, gains, power, antenna sizing, efficiencies, and soon.Command & Data HandlingThe Command & Data Handling subsystem sheetallows required data rates, compression ratios, andground station contact duration information to beentered. From this information, storage, processing,and memory requirements are derived. Databaseselections can then be made for the Processor, Memory,Data Storage, and Input/Output interface needed tomeet the derived requirements.Attitude Determination & ControlThe attitude determination and control system(ADACS) is designed by selecting sensors andactuators from drop-down menus. Visual Basic code isused to simulate the disturbance torques acting on thespacecraft over one orbit (Figure 9). Using spreadsheetcalculations the maximum values of the disturbancetorques and the accumulated angular momentum arecomputed, and compared to the capability of theselected components. Control torque and/or thrust levelrequirements are also computed, based on the slewrequirements of the selected payloads.6Allan I. McInnes15th Annual AIAA/USU Conference on Small Satellites

modeling and sizing of the attitude control system. Thebus structural material can also be specified on thissheet, via a database-driven drop-down menu. Giventhe mass distribution information, the centers of massand mass-moments of inertia are calculated for eachpanel and for the entire spacecraft.Atmospheric Torques2.0E-031.0E-030.0E que (N m)-2.0E-03ThermalThe Thermal sheet treats each panel, body surface, andinterior zone as a node. Conduction coupling factorsare assigned between nodes. A thermal analysis canthen be performed to determine the temperature of thedifferent nodes. This analysis is compared to theminimum and maximum temperature requirements forthe electronics and other components. Based on theseanalyses, additional surface area for radiating heat, ordeployed radiators, can be assigned.TheSmallSatCEM team is considering temconfigurations to accommodate studies that are not yetat the level of detail required by the present .0E-03-8.0E-03True Anom aly (deg)Figure 9 Atmospheric disturbance torquesPowerThe Power sheet allows the designer to select the typeof solar cells to be used for body-mounted ordeployable arrays, as well as the battery type and busvoltage. Based on the types of solar cells and batteriesselected, the solar array area and the battery massneeded to meet the power requirements of thespacecraft are computed. Since it cannot always beassumed that all of the solar arrays will be directlyfacing the sun, the Power sheet includes an analysis toolthat simulates the solar illumination for each of thespacecraft body surfaces and deployable panels (Figure10).Model StateAll entered and calculated information from eachworksheet is linked to the Model State sheet. Asmentioned previously, this sheet contains the state ofthe spacecraft design. The Model State is the source ofdata for all of the analysis tools, as well as anysubsystem sheets that require information on thepresent spacecraft design. This arrangement ensuresthat design data is consistent throughout the model.TopSurface Illum ination (One Orbit)DatabaseThe Database sheet is a collection of tables ofcomponent data for a variety of different components.These tables act as the source of data for the drop-downmenus used on the subsystem design worksheets. Thedatabases contain three categories of components:traditional, research, and future or non-existing. Atpresent, the data contained in these databases,particularly data on “traditional” components, requiresmodernization and population by components that areappropriate for small satellites.The ability toincorporate fictional components into the databasesallows technology insertion scenarios to be studied,while still retaining a clear delineation between real andprojected data.Direction CosineBottom1Side 10.9Side 20.8Side 30.7Side 40.6Panel 10.50.40.30.20.1004590135180225270315360True Anom aly (deg)Figure 10 Solar illumination for each spacecraftsurfaceCostThe Cost model makes use of various parametric costrelationships that are derived from Aerospace’s SmallSatellite Cost Model8,9. However, many future smallsatellites may use non-traditional space componentsthat cost relationships based on historical data are illequipped to model. The lack of testing, handling, andMass Properties & DistributionThe Mass Properties & Distribution sheet provides thedesigner with a way to distribute the component masseson the different surfaces and panels of the spacecraft, aswell as within several internal “zones”. This massdistribution information can then be used for thermal7Allan I. McInnes15th Annual AIAA/USU Conference on Small Satellites

project was also supported by the United States AirForce Space Test Program.government oversight make non-space-qualifiedhardware less expensive to purchase and assemble. Asurvey of non-space industry electronic componentsmay be beneficial to assist in understanding how tobetter provide accurate cost relationships for nontraditional components.1.ConclusionsSmall satellites are becoming a popular choice for lowcost, rapidly developed space systems. The applicationof systems engineering methodologies to small satellitedevelopment will help to ensure that small satellites arenot only low-cost, but fulfill their mission objectives.Existing tools for satellite systems engineering tend tobe biased toward large spacecraft, and lack capabilitiesthat are necessary for small satellite design. As a result,The Aerospace Corporation is developing a systemsengineering tool, known as the SmallSatCEM, intendedto support small satellite design studies.2.3.4.5.The SmallSatCEM is implemented as a self-containedMicrosoft Excel workbook, with a Visual Basicbackend to handle complex tasks. The tool is intendedto support the classical iterative design cycle, withoutthe need to consult subsystem experts or gather datafrom external software. To this end, the workbookincludes small satellite component databases, as well asphysical models and analysis tools selected for theirrelevance to small satellite design tasks.6.7.8.Development of the SmallSatCEM is ongoing. TheSmallSatCEM team is working to complete andvalidate the existing SmallSatCEM design. Once theSmallSatCEM is fully implemented and in operationthe development team will begin planning for theextensions or improvements that will invariably arisefrom actual real-world experience with the tool.Additionally, portions of the SmallSatCEM are beingtransitioned to the CDC, further enhancing the CDC’ssmall satellite design capabilities.9.ReferencesHelvajian, H., Microengineering AerospaceSystems, The Aerospace Press, El Segundo CA,1999.Janson, S.W., “Mass-producible Silicon Spacecraftfor 21st Century Missions”, AIAA SpaceTechnology Conference & Expo, Albuquerque,NM, Sept 1999.Hammond, W.E., Space Transportation: A SystemsApproach to Design and Analysis, AIAA, RestonVA, 1999.Aguilar, J.A., Dawdy, A. and G. W. Law, “TheAerospace Corporation’s Concept Design Center.”Proceedings of the 8th Annual InternationalSymposium of the International Council on SystemsEngineering, July 26-30 1998.Shishko, R. and R.G. Chamberlain, NASA SystemsEngineering Handbook, SP-6105, June 1995.Wertz, J.R. and W. J. Larson, Space MissionAnalysis and Design, Microcosm Press, TorranceCA, 1999.Griffin, M.D., and J.R. French, Space VehicleDesign, AIAA, Reston VA, 1991.Burgess, E.L., N.Y. Lao, and D.A. Bearden,“Small-Satellite Cost Estimating Relationships.”Proceedings of the 9th Annual AIAA/USUConference on Small Satellites, Logan UT,September 1995.Small Satellite Cost Modelhttp://www.aero.org/software/sscm/As the development of the SmallSatCEM proceeds, thestructure of the tool is becoming much more complex.This has caused concerns about the maintainability androbustness of the workbook.The Visual Basicbackend, in particular, takes some effort to understand.It is hoped that an aggressive code cleanup anddocumentation effort will help to mitigate theseproblems.AcknowledgementsThe authors gratefully acknowledge the support of TheAerospace Corporation’s Corporate Research Initiativeand Engineering Methods programs, and the Center forMicrotechnology, in making this work possible. This8Allan I. McInnes15th Annual AIAA/USU Conference on Small Satellites

Biography for Allan I. McInnesAllan I. McInnes is a member of the Vehicle ConceptsDepartment at The Aerospace Corporation. He isprimarily concerned with the conceptual design ofspace vehicles, and the development of systemsengineering tools such as the SmallSatCEM. He hasbeen involved in several studies examining future smallsatellite concepts, as well as studies supporting theMILSATCOM, SBR and GMSP programs. In additionto his work on the SmallSatCEM, Mr. McInnes ispresently supporting a review of the design tools usedin the PDC, as well as being involved in the JPL MERfault protection effort. Prior to joining The AerospaceCorporation, Mr. McInnes developed av

Small Satellite Systems Engineering Systems engineering is concerned with the overall performance of a system for multiple objectives (e.g. mass, cost, and power). The systems engineering process is a methodical approach to balancing the needs and capabilities of the various subsystems in