Example: marketing

Second Order Linear Differential Equations

CHAPTER12 SecondOrderLinearDifferentialEquations a relationinvolvingvariablesx y y y . Asolutionis a functionf x suchthatthesubstitutiony f x y f x y f x givesanidentity. Thedifferentialequationissaidtobelineari f it is linearinthevariablesy y y . We have alreadyseen( )how tosolve firstorderlinearequations; ( )y ay by g x whereaandbareconstants,andg x is a differentiablefunctionofx. ,wesaw thata firstorderequationhasa one-parameterfamilyofsolutions,andthatth especificationofaninitialconditiony x0 y0uniquelydeterminesa , ,andnumbers y0 y 0, there isa uniquefunctionf x which solvesthedifferentialequation( )andsatisfiestheinitialconditionsf x0 y0 f x0 y tocompletelysolve equation( )whenthefunctionontherighthandsideis zero:( )y ay by 0 Thisis calledthehomogeneousequation. Animportantfirststepis tonoticethatiff x andg x aretwo solutions,thensois thesum;infact,sois any linearcombinationA f x Bg x.

12.1 Homogeneous Equations 177 Now unfortunately a quadratic equation does not necessarily always have two real roots, so we have to examine the cases separately. Case of two real roots. If the discriminant a2 4b 0, then there are two real roots, and it is straight-forward to find the solution of the corresponding initial value problem.

Tags:

  Linear, Value, Second, Order, Differential, Equations, Second order linear differential equations

Information

Domain:

Source:

Link to this page:

Please notify us if you found a problem with this document:

Other abuse

Advertisement

Transcription of Second Order Linear Differential Equations

1 CHAPTER12 SecondOrderLinearDifferentialEquations a relationinvolvingvariablesx y y y . Asolutionis a functionf x suchthatthesubstitutiony f x y f x y f x givesanidentity. Thedifferentialequationissaidtobelineari f it is linearinthevariablesy y y . We have alreadyseen( )how tosolve firstorderlinearequations; ( )y ay by g x whereaandbareconstants,andg x is a differentiablefunctionofx. ,wesaw thata firstorderequationhasa one-parameterfamilyofsolutions,andthatth especificationofaninitialconditiony x0 y0uniquelydeterminesa , ,andnumbers y0 y 0, there isa uniquefunctionf x which solvesthedifferentialequation( )andsatisfiestheinitialconditionsf x0 y0 f x0 y tocompletelysolve equation( )whenthefunctionontherighthandsideis zero:( )y ay by 0 Thisis calledthehomogeneousequation. Animportantfirststepis tonoticethatiff x andg x aretwo solutions,thensois thesum;infact,sois any linearcombinationA f x Bg x.

2 Thus,onceweknow two solutions(they mustbeindependentin thesensethatoneisn t a constantmultipleoftheother)wecansolve y 0 y 0 4 y 0 1 Now, weknow thatcosxandsinxaresolutionsoftheequation ,sowetrya solutionoftheformy x Acosx Bsinx. Evaluatingatx 0, wefindthatA ,gettingy x Asinx Bcosx,andevaluatingatx 0, wefindB x 4 cosx thaty1 cosxis thesolutionwiththeparticularinitialvalue sy1 0 1 y 1 0 0 andy1 sinxis thesolutionwithy1 0 0 y 1 0 1 . Thenthesolutionwithinitialvaluesy 0 andy 0 is( )y x y 0 cosx y 0 y 0 withgiveninitialvaluesy 0 y 0 Nowexande xaresolutionsofthisdifferentialequation, sothegeneralsolutionis a t have aseasya timefindinga solutionlike ( ),sincethesefunctionsdonothave theinitialvalues1 0;0 1 respectively. Howeverif weintroducethefunctions( )coshx 12 ex e x sinhx 12 ex e x thesedohave therightinitialvalues:( )cosh0 1 sinh0 0( )ddx coshx sinhx ddx sinhx coshxso cosh 0 0 sinh 0 ,thesolutiontoourproblemis( )y x y 0 coshx y 0 sinhx Thisparticulardifferentialequationcomesu psooftenthatit is importanttorememberthesefunctions,coshx sinhx, calledthehyperbolicfunctionsandtheirbasi cproperties:equation( )and( )cosh2x sinh2x 1 Becauseof( )thesefunctionsparametrizethestandardhyp erbola(andit is forthisreasonthattheyarecalledhyperbolic functions).

3 We now bea rootoftheequation( )r2 ar b 0 Thenerxis a solutiontothehomogeneousequation:( )y ay by 0 Equation( )is calledtheauxiliaryequationof thedifferentialequation( ).To verifythepropo-sition,lety erxsothaty rerx y r2erx. Substitutingintoequation( ):( )r2erx arerx berx erx r2 ar b 0if andonlyif r is a rootoftheauxiliaryequation. unfortunatelya quadraticequationdoesnotnecessarilyalway shave two realroots,sowehave realroots. If thediscriminanta2 4b 0,thentherearetwo realroots,andit is 6y 5y 0 y 0 4 y 0 1 Theauxiliaryequation,r2 6r 5 0 hastherootsr 1 5, soe xande ( )y Ae x Be 5xwithderivativey Ae x 5Be 5x Evaluatingatx 0, wehave 4 A B 1 A 5B. Solvingthispairofequations,wegetA 19 4andB 3 4, sooursolutionis( )y 194e x 34e functionx x t satisfiesthedifferentialequation( )x 2x 15x 0 Underwhatconditionsonthevaluesofxatt 0 willthisfunctiondecayto0 ast ?

4 Theauxiliaryequationr2 2r 15hastherootsr 3 t Ae 3t Be5t. ThiswilldecayatinfinityonlyifB , evaluatingxandx at0 givesustheequations( )x 0 A B x 0 3A 5B SettingB 0, theconditionbecomesx 0 3x 0 Ifthediscriminanta2 4b 0,thentherootsaretwo complex conjugatenumbers i i .Let s firstlookat thecasey y 1 0 are i, andwe d like to saythatthesolutionsarethefunctionseix e ix. Thisdoeswork,andallthealgebrainthecaseof realrootsworksjustaswellinthiscase,oncew ehave ,rememberequation( ):thegeneralsolutionofy y 0 is( )y x y 0 cosx y 0 sinxIfy x eixis torepresenta solutionofthisdifferentialequation,wehav ey 0 e0 1,andy 0 ie0 i, sowemusthave( )eix cosx isinxNoticethatif wedifferentiatethisexpression,weget( ) sinx icosx i cosx isinx sothisexpressionis consistentwiththedifferentiationrulefort heexponential:( )ddxeix ieix Chapter12 SecondOrderLinearDifferentialEquations17 8 Infact,definingthecomplex exponentialby( )is , if wesubstitutetheMaclaurinseriesforallthef unctionsin( )wegetanidentity:( ) n 0 ix nn!

5 N 0 1 nx2n 2n ! i n 0 1 nx2n 1 2n 1 ! complex number i if wedefinetheexponentialfunctionas( )e i x e x cos x isin x , ofcourse,weareinterestedonlyin shownis thatif i aretherootsoftheequationr2 ar b 0,thenthefunctionse i xsolve thedifferentialequationy ay b ,whichgivesusthedesiredtwo theauxiliaryequationforthedifferentialeq uation( )y ay b 0hasthecomplex roots i , theneverysolutionofthedifferentialequati onis oftheform( )Ae xcos x Be xsin x e x Acos x Bsin x Insolvinginitialvalueproblems,wecanworkw iththecomplex solutionsorsolutionsoftheform( );usuallythelatteris x t ofx 2x 0 Sincetherootsoftheauxiliaryequationr2 2 0 are i , thegeneralsolutionis( )x t Acos t Bsin tIt is easytoseewhatthisfunctionlookslike bydefining( )C A2 B2 arctan B A Then( )becomes( )x t C cos cos t sin sin t Ccos x Ccos x Thusthegraphofx x t is a simplecosinecurve ofamplitudeC, andperiod2 , shiftedtotherightbythephase.

6 ( ). 2 1 5 1 1 0 5 0 5000 50 5111 51 522 y x ofy 2y 5y 0, withtheinitialvaluesy 0 2 y 0 2r 5 0 hasthesolutionsr 1 2i. Thusthegeneralsolutionisy e x Acos 2x Bsin 2x . To solve forAandBusingtheinitialvalueswemustfirst differentiatey:( )y e x Acos 2x Bsin 2x e x 2 Asin 2x 2 Bcos 2x Substitutingtheinitialvaluesgivestheequa tionsA 2 A 2B 1,whichhasthesolutionsA 2 B 1 ( )y e x 2 cos 2x 12sin 2x Caseofa doubleroot. If thediscriminanta2 4b 0, thentheauxiliaryequationhasonerootr, tryy uerx, whereuis a , thedifferentialequationis( )y 2ry r2y 0 Substitutingthisyintheequationwegetto( )y 2ry r2y erx u 2u r ur2 2r u ur r2u erxu 0 Thusu Ax theauxiliaryequationforthedifferentialeq uation( )y ay b 0hasonlytherootr, theneverysolutionis oftheform( ) Ax B 4y 4y 0, withinitialvaluesy 0 2 y 0 Ax B e2x, withderivativey 2 Ax B e2x Ae2x.

7 Substitutingtheinitialconditionsgivesthe Equations ( )2 B 1 2B A ThusA 5 B 2 andtheansweris( )y 5x 2 e2x 17 PSfragreplacements00112345670 20 40 60 20 40 60 500400 300200 1000010020030040050012345610 50 1 0 50 5 ,oneusuallywantstohave ,theyalsodeterminetheirbehavior. Exceptfortheidenticallyzerosolution,alls olutionsgrow Exceptfortheidenticallyzerosolution, negative anda positive root. Allsolutionsgrow exponentially, exceptforthemultiplesoftheexpo-nentialwi ththenegative Thisis thecaseoftheequationwitha 0 andb 0, whichwecanwriteas( )y 2y 0 Aswesaw ,thegeneralsolutioncanbewrittenas( )y x Acos x Bsin x ory x Ccos x anoscillationofperiod2 , amplitudeCandphase ( ).Complexroots. Inthiscasetherootsareoftheform i andthegeneralsolutioncanbewritteninthefo rm(followingthepreviousdiscussion)( )y x Ce xcos x 0, thisgivesanoscillationwithexponentiallyi ncreasingamplitude,andif 0, thisgivesanexponentiallydampedoscillatio n.

8 Massmontheendofa verticallyhangingspring,andthensetit inmotion;how canwedescribethesubsequentmotion?Aswehav e ,thespringis s secondlawofmotion,thisforceisma, whereais theaccelerationofthemassm. Lettingxrepresentthedownwarddisplacement fromequilibrium,wehavea x , andif thespringconstantisk, thisgivesustheequation( )mx kx orx kmx 0 Letting k m, thishasthesolution( )( )x t Ccos t whereCand have tobea ,whenmis measuredinkilogramsandxinmeters,thenforc eis measuredinnewtons, smallerscale,mis ingrams,xincentimeters,forceindynes,andk indynes/meter. However, intheBritishsystemit is customarytorefertotheweightwoftheobject( inpounds), w g, whereg 32ft/sec2is theaccelerationdueto gravity. Finally, intheBritishsystem,thespringconstantis massof10g hangsfroma springwithspringconstantk 0 4 extendedanadditional8 ,give 0 x 0 8 x 0 alsohave 0 4 10 0 04 0 ( )x t Ccos 0 2t We get,fromtheinitialconditions( )8 Ccos 0 0 2 Csin so 0 andC 8, andtheequationofmotionis( )x t 8 cos 0 2t We couldhave concludedthismorequickly, byobservingthattheinitialconditionstellu sthatx 8whenthevelocityis 0, sothemaximumextension(theamplitude) configurationalreadyinmotion,andwhenwema ke ourobservation(attimet 0),themassis displaced12cmdownwardandis travelingdownwardat a velocityof1 ,theequationhasthegeneralform( )x t Ccos 0 2t andtheinitialconditionsgive( )12 Ccos 1 0 2 Csin We solve forCand ( )

9 Ccos 12 Csin 5 AddingthesquaresofbothequationsgivesusC2 122 52 169,soC 13,anddividingoneequationbytheothergives ( )tan 5 12 sothat 0 126 andtheequationofmotionis( )x t 13cos 0 2t 0 126 a 10lb. objectis hungfroma springwithspringconstantk=9lbs/footandth enisgivenaninitialvelocityof24ft/sec,wha tis themaximumextentofthespring?Herem 16 32andk 9, sowehave thespringequation( )12x 9x 0sox Acos 3 2t Bsin 3 2t . Theinitialconditionsx 0 0 x 0 24leadtoA 0 B 8 ( )x t 8 2 sin 2 2t whosemaximumvalueis 8 2 , letusreturntoourspringwithspringconstant kandmassm, andsupposethatit is insertedina viscousfluidwhichimpartsa 0 betheconstantinthisproportion,weseethate quation( )is replacedby( )mx kx qx orx x 2xwhere k mand q m. Therootsofthisequationare( )r 2 4 22 If 2 , thentherootsarebothrealandnegative, andthereis 2 , thenwehave complex roots, 2 0, ,if wewanttohave a gooddampingeffect(asfor shockabsorber)weshouldbesurethat is sufficientlylarge;thatis,thatthefluidis systemconsistingofa springina viscousfluidis installedsoastoabsorbtheshockona shockis sustainedwhenthesystemis inequilibriumimpartinganinstantaneousvel ocityof100 take fortheamplitudeoftheoscillationtobereduc edto1 cm?

10 Thebasicdifferentialequationis 100x 600x 2500x 3 4i, sothegeneralsolutionis( )x t e 3t Acos4t Bsin4t Nowatt 0 x 0 x ,wefindx t 25e 3tsin4t. Nowthemaximumamplitudesofthisdampedvibra tionoccuratthevaluest k 8,forkanoddinteger. Hereis a tableofthefirstfew , lessthan1 cm. returnnow tothegeneralsecondorderequationwitha nonzerorighthandside;( )y ay by g x yp x is a particularsolutionoftheequation( ).Theneverysolutionis oftheformy yp yhwhere yhis a :( )y y x 2 y 0 4 y 0 2 It is easytoseethatyp x x 2 is a oftheform( )y x x 2 Acosx BsinxTo findAandBweusetheinitialconditions(atx 0):( )4 0 2 Acos 0 2 1 Asin 0 Bcos 0 givingus4 2 A 2 1 B, soA 2 B 1, andthesolutionisy x 2 2 cosx 1 3 cos 2x solvesthedifferentialequationy y cos 2x findthesolutionwithinitialvaluesy 2 1 y 2 3 Chapter12 SecondOrderLinearDifferentialEquations18 4We know thesolutionhastheformy 1 3 cos 2x Acosx Bsinx.


Related search queries