Embed Size (px)
Citation preview
75 Fifth Street NW, Suite 312 Atlanta, GA 30308, USA voice: +1-404-592-6897
web: www.InterCAX.com email: [email protected]
Dr.DirkZwemer,InterCAXLLC
Modeling The Martian Abstract........................................................................................................................................................1
Introduction.................................................................................................................................................1
MartianSurvival-Abandonment.................................................................................................................2
MartianSurvival-Transit.............................................................................................................................7
FinalThoughts............................................................................................................................................11
AbouttheAuthor.......................................................................................................................................11
Abstract
Two scenarios from The Martian, an adventure story describing the challenges of a NASAastronaut inadvertentlymaroonedonMars,aremodeledinSysML,withspecialattentiontocapturingthedetailedparametricsfromthenovel.Thefirstsectionexplorestheagronomicsofgrowingfoodonaplanet without normal air, water, soil or plant life. The second sections describes transportationplanning for extended travel without a highly developed road infrastructure. The SysML parametricmodels are intended to offer some guidance on adding these features to complexMBSE systems, inadditiontowhateverentertainmentvaluemayresultfortheterminallyengineering-minded.
IntroductionTheMartian,anovelbyAndrewWeir(BroadwayBooks,2014),isagreatread.Partofitsappeal
toengineersisthattheastronauthero“shows”hiscalculationsonhowtosurvivehismanychallengesaloneontheplanetMars.Therealismandresourcefulnessofhisworkmakehisfatethatmuchmorerealtothereaderandheightentheemotionalimpactofthestory.
Unfortunately,theengineersinthenovelaren’tshownasusingmodel-basedsystemsengineering(MBSE).Intheinterestoffutureastronauts,IwilltrytocapturesomeofthelessonslearnedontheAres3missionintheformofSysMLparametricmodels(whiletryingtominimizespoilersforthoseyettoenjoythebookorthemovie).
Twoscenarios,capturedasusecasesinFigure1,willbemodeled.Inthefirst,theastronautplanstoextendthefoodstocksavailablebygrowingfoodinsidehisMartianbasecamp.Inthesecond,heneedstotravelalongdistanceacrosstheMartianlandscape,carryinghisenergysupply,hislifesupportequipment,andthefoodandwaterrequiredtoreachhisdestination.ThefiguresuseMagicDraw(NoMagicInc.)tocreateandParaMagic(InterCAX)andMathematica(WolframResearch)tosolvethemodel,butthesamemodelwillalsobemadeavailablefordownloadfromtheInterCAXwebsite(www.intercax.com)inseveralforms,astheyarecompletedandverified:
• EnterpriseArchitect(SparxSystems),solvedwithSolvea(InterCAX)• IBMRationalRhapsody(IBM),solvedwithMelody(InterCAX)• IntegrityModeler(PTC),solvedwithParaSolver(InterCAX)• MagicDraw18.0(NoMagic),solvedwithParaMagic(InterCAX)
Figure1Astronautobjectivesundertwosceanrios
MartianSurvival-AbandonmentThefirstscenarioissetattheAcidaliaPlanitialandingbaseoftheAres3mission.TheBDDin
Figure2showsthatthebaseconsistsofthemainhabitat,twomobileMarsrovers(eachwithapopupshelter),andthelandingstageoftheMAV(MarsAscentVehicle).Italsoincludesthefoodstocksexpectedtobeusedundertheoriginalmissionplan.
Figure2APBasestructureTheastronaut,abotanistamongothertalents,wantstogrowenoughadditionalfoodtosurviveuntilrescue.Hisanalysisfallsintofourparts:
1. Howmuchplantingareacanhecreateinsidethehabitableelementsofthebase?2. Howmuchwaterdoesthesoilrequireforplantgrowth?3. Howmuchfood(heplantspotatoes)canhegrowpersol(Martianday)?4. Howlongwillhisextendedfoodsupplylast?
I have structured these questions as four separate analysis blocks containing parametric models, asshowninFigure3.TheseanalysisblocksreferencetheAP_Baseblock,aswellaseachother,toaccessthenecessaryvaluesfortheircalculations,withoutrequiringchangestothepropertiesorparametricsofthestructureblocks.
Figure3Analysisstructure,MartianSurvival–Abandonmentscenario
Theonlyparametricsembeddedwithin theAP_Baseblocksare for the food supply, shown in
Figure4.
Figure4FoodSupplyPAR,calculatingbasefoodstocksbasedontheoriginalmissionplan(e.g.6peoplefor30sols).The<equal>stereotypelabelsonthebindingconnectorshavebeenelidedfromalldiagramsforclarity.
SoilAnalysisPAR(Figure5)Thismodelcomparesthearableareainsquaremeterscreatedinsidethehabitatandpopup
sheltersagainstavailablearea,ultimatelycalculatinganareaMOS(marginofsafety)value.TheMOSispositiveaslongastheareaplannedtobeuseddoesnotexceedtheareaavailable.Italsocalculatesthevolumeofsoiltobecreated,whichisusedbyWaterAnalysis.
Figure5SoilAnalysisPAR
WaterContentAnalysisPAR(Figure6)This part of the model calculates the water required to make the soil fertile with the stock
availableatthebase,andcalculatesanotherMOS.
Figure6WaterAnalysisPAR
PotatoGrowthAnalysisPAR(Figure7)
Thiscalculatestheaveragenumberofcaloriesgeneratedfrompotatoespersol.Thisignorestheinitiallagwhilethepotatoesaregrowing,sincethefinalvalueofinterest,thenumberofadditionaldaysoffood,isnotconcernedwithwhenthepotatoesareeaten,aslongasthefirstcropcomesinbeforetheinitialfoodstockisconsumed.
Figure7PotatoGrowthAnalysisPAR
Figure8SurvivalAnalysisPAR
SurvivalAnalysisPAR(Figure8)Thisfinalsectionofthemodeldeterminesthetotalnumberofsolthattheminimumdietary
needoftheastronautcanbemetbetweentheoriginalfoodstockandtheadditionalfoodgrown.Notethattheperiodincreasesifmorefoodcanbegrown,buttheanalysisbecomesinvalid(periodbecomesnegative)ifthefoodgrowneachsolexceedsthefoodconsumed.
ModelExecution
Figure9Instance01,beforesolving Figure10Instance01,aftersolvingUsingvaluesfrom(orestimatedfrom)theAres3mission,aninstanceoftheparametricmodeliscreatedanddisplayedintheParaMagicbrowserinFigure9(beforesolution)andFigure10(aftersolution).Potatoeswillextendtheastronaut’ssurvivalfrom400solsto1471sols(ifthingsgoaccordingtoplan).
MartianSurvival-TransitThesecondscenariosendstheastronautonanextendedtripacrosstheMartianlandscape.The
top-levelblockinFigure11,Caravan,consistsofthetworoverscoupledtogether,ladenwitheverthingnecessarytosurvivethetrip.Thereisatimelimit,bywhichthetripmustbecompleted,andfoodandwaterstocksmustbesufficient.
Figure11TransportStructureBDD,showingthecaravan,rovers,andkeycargofortheanalysis
Iuseembeddedparametricsinthispartofthemodelforthemassrollup.AlltheblocksshowninFigure11arespecializationsofasupertypeTravelElement,whichusesrecursion(Figure12).
Figure12ParametricdiagramforTravelElement.ThemassofeachelementisthesumofthemassofitsexplicitpartsplusabasemassTheastronautdriveseachMartianday(sol)until theroverbatteriesareexhausted,manuallyunpacksanddeploysthesolarpanelscarriedoneachrover,waitswhilethepanelsrechargethebatteries,andrestows the panels, all during Martian daylight. Additional charging time is required to run theoxygenatorperiodically,whichremovesCO2fromtheroverairsupplybeforeitreachestoxiclevels.
The number of solar panels to be carried on the trip represents a trade-off between the powerproduced,themasstobetransported,andthetimespentunpackingandrepackingthepanelsateachstop.ThesethreevaluesarecapturedintheparametricdiagraminFigure13.
Figure13SolarPanelparametricanalysis
As inthefirstscenario,weplacemostofthescenariospecificanalyses inseparateblocksthatreferencetheCaravansystem(Figure14).
Figure14TransitAnalysisBDD,showingthestructureofthesecondscenarioanalyses
ThePowerAnalysisPAR inFigure15calculates twokeyparameters,how long it takes to fullycalculate the caravan batteries and what percentage of the time must be spent powering theoxygenator,relativetohowlongtheatmosphereremainshealthy.Thesecondfactorextendsthetotaltriptimebyafixedpercentage,asshowninFigure16.
Figure15ParametricdiagramforPowerAnalysis.Notethatbatterychargingtimeisaffectedbyatmospherictransmissionfactors,whichmayreducesolarpaneloutput.
ThetravelTimeAnalysisPAR(Figure16)usesthepowernumbersandthecaravanpropertiestocalculatetotaltriptime.Theconstraintcalculatingdistancetraveledbythecaravanperbatterycharge,
Figure16ParametricdiagramforTravelTimeAnalysis.marginTimeistheMOSforthecalculatedtriptimevs.timelimit
DistancePerCharge,isestimatedfromempiricaldatacollectedduringtheAres3mission.Thisdistance,dividedbytheeffectivetimeperchargingcycle(thesumofthedrivetime,chargetime,andsolarpanelunpack/repack time, extendedby thedaylight hoursper sol andoxygenatoroperation), estimates aneffectivevelocityincludingdowntime.Thetotaltripdistance,dividedbytheeffectivevelocity,resultsintimeTotal:sol,thetotaltimeforthetrip.
Foodandwatercalculationsaresimilartothoseinthefirstscenarioandarenotrepeatedhere.At the Survival Analysis – Transit block (Figure 17) level, the parametric diagramuses conditionals toreturnverdictsforthethreesurvivalissues,1iftheastronautsurvives,0ifhedoesn’t.
Figure17ParametricdiagramforTravelElement.Themassofeachelementisthesumofthemassofitsexplicitpartsplusabasemass
The instancevalues inFigure18andofferhope;all three factorsarepositive. Thegivensaretaken from theAres3 video logs. The caravancan travel almost100kilometerspre charge, taking4hours at themaximum speed of 25 kph. Recharging the batteries takes about 10 hours ofMartiandaylight,withanadditional1.4hoursforsettingoutandpickingupthesolarpanels.Thecaravanstopstopowerandruntheoxygenatorforaboutonesoloutofeveryfive.Overall,thetripof3200kilometersisexpectedtake47.55sols,withinthe50soltimelimitandwithsufficientfoodandwater.Thisanalysisassumesnoproblemswith terrainobstacles, equipment failure,weather, or anyother challenge thatMarscanthrowattheunsuspectingvisitor.
Figure18Instance02,beforesolving Figure19Instance02,aftersolving
FinalThoughtsAfull treatmentof theAres3missionwouldstartwithmodelingtheoriginalprofileandtreat
the scenariosexplored in thisTechNoteasextendedusecases. Furtherextendedusecases couldbedeveloped because (Spoiler Alert!) none of TheMartian’s plans work out as expected. This furtherdevelopment would highlight the common strength of SysML and the astronaut hero, the ability toadaptandre-purpose.Astheastronautconfigurestheelementsoftheoriginalmissiontoservehisnewobjectives, theSysMLmodeler reusesobject fromoldprograms innewways. I like to think that theNASAengineersworkingtorescuetheherowere,behindthescenes,usingMBSEtocomeupwiththerapidredesignstheycomeupwith.
AbouttheAuthorDr. Dirk Zwemer ([email protected]) is President of InterCAX LLC, Atlanta, GA and holdsOCSMPcertificationasModelBuilder-Advanced.
Forfurtherinformation,[email protected].
Copyright2015InterCAXLLC
