Laplace aaa transforms 1 Learning outcomes When you have completed this Programme you will be able to: ‘© Obtain the Laplace transforms of simple standard expressions ‘¢ Use the first shift theorem to find the Laplace transform of a simple expression multiplied by an exponential ‘© Find the Laplace transform of a simple expression multiplied or divided by a variable ‘© Use partial fractions to find the inverse Laplace transform © Use the ‘cover up’ rule ‘© Use the Laplace transforms of derivatives to solve differential ‘equations ‘© Use the Laplace transform to solve simultaneous differential ‘equations ‘Prerequisite: Engineering Mathematics (Fifth Edition) Programme 26 Introduction to Laplace transforms 47Programme 2 Introduction The solution of a linear, ordinary differential equation with constant coefficients such as the second-order equation af") + Of) + FH) = s(t) ‘can be solved by first obtaining the general form for the expression F(t). This general form will contain a number of integration constants whose values can be found by applying the appropriate boundary conditions (see Engineering Mathematics, Fifth Edition, Programme 25). ‘A more systematic way of solving stich equations is to use the Laplace ‘transform which converts the differential equation into an algebraic ‘equation and has the added advantage of incorporating the boundary conditions from the beginning, Furthermore, in situations where f(t) represents a function with discontinuities, the Laplace transform method can succeed where other methods fail. Laplace transform techniques also provide powerful tools in numerous fields of technology such as Control Theory where a knowledge of the system transfer function is essential and where the Laplace transform comes into its own, Let us see what It is all about. or a more detailed introduction see Engineering Mathematics, Fifth Edition, Programme 26.) Laplace transforms ‘The Laplace transform of an expression f(t) is denoted by L{f(t)} and {s defined as the semi-infinite integral Ufo} [rive “ae @ ‘The parameter s is assumed to be positive and large enough to ensure that the integral converges. In more advanced applications s may be complex and in such cases the real part of s must be positive and large enough to ensure convergence. In determining the transform of an expression, you will appreciate that the limits of the integral are substituted for f, so that the result will be an expression in s. Therefore up) = [> fide *ar— re) Make a note of this general definition: then we can apply itLaplace transforms 1 49 Sowehave L{f(t)}=[- fide ar = F(s) @ Example 1 To find the Laplace transform of f(t) = a (constant). La) = [ae ae = (S £(0-1}=% (s>0) 2) Example 2 To find the Laplace transform of f(t) = e (a constant). As with all cases, we multiply f(t) by e* and integrate between t = 0 and t = oo. Let) = [teat flewoar Jo Finish it off. -He}=5 >a) 8) So we already have two standard transforms Uy" and Yea)50 Programme 2 Note that, as we said earlier, the Laplace transform is always an expression in s. Now for some more examples BW a. To find the Laplace transform of f(t) =sinat. We could, of course, apply the definition and evaluate stn) = [sate using integration by parts. However, it is much shorter if we use the fact that el = cos +jsind so that sind is the imaginary part of e”, written (e!) ‘The function sinat can therefore be written ¥(e#!) so that Lina) 1090) af oetar = [ehh a Aah} ae) 4553} ‘We can rationalise the denominator by multiplying top and bottom by. wo sala . shia) __a . Esinat) = 9 {SAH = Lsinat} = 3" (@) We can use the same method to determine L {cos at} since cos. is the real part of eM, written (e*). ‘Then L{cosat}Laplace transforms 1 54 L{cosat} = >* © ws =aTa Because shia) s L{cosat} = afte} — Recapping then: L{1} = 5 ety = Usin2e} 5 Lost} ua} =1 1 eS 2 $ Linay= x2 Gi— Leosat) = zag Example 4 To find the transform of f(t) = t* where 1s a positive integer. By the definition L{¢") = [fee Sat, Integrating by parts ue") KS), + [Petar fren] +2 [ertetat ‘We said earlier that in a product such as fe the numerical value of s Is large enough to make the product converge to zero as t + 20 fe» suet ee ‘You will notice that IK ete de is identical to f te* dt except that lo lo fe Felt (7 ‘Thisis a reduction formula, and if we now replace m by (1 — 1) we get | . =0 at 6 nis replaced by (n—1) “y= [ee then I and the result (6) becomes I,52 Programme 2 weed mH ee -- (next line) So finally, we have matte mee ned nod y But fy =L{}=1(1} =4 2 yp ate Den ayn 320) at ®) 1()= and with m=0, since O!=1, the general result includes 1{1} =1 which we have already established. Example 5 Laplace transforms of f(t) = sinh at and f(t) = coshat. Starting from the exponential definitions of sinh at and coshat, ie. sinhat =}(e—e-*) and coshat =}(e +e) ‘we proceed as follows. (@ f(t)=sinhat. L{sinhat) f "sinh at e~* at =ahe- eters dp olf fee et lo Complete itLaplace transforms 1 53 Lsinh at) = 54 @ Because etek tian ler -comedS5-S5) if 11). a aca seal oe :, L{sinhat} = (9) z-8 © f(t) =coshat. Proceeding in the same way L{cosh at} = L{cosh at} = 5+ w@ =e coat) =} et semetaradf fet stata 2. “let emmy apa a = s-a*s+a 2. L{coshat} = 20) So we have accumulated several standard results: La} Le") =e Usinat) = Lcosat) = Le") = Sa L{sinhat} = 5% pi Uooshat) = 35 ‘Make a note of this list if you have not already done so: it forms the basis of much that is to follow.54 Programme 2 Ga) ‘The Laplace transform is a linear transform, by which is meant that (1) The transform of a sum (or difference) of expressions is the sum (or difference) ofthe individual transforms. That is LC®) £500} = LAF(O} EL8(O)} (2) The transform of an expression that is multiplied by a constant is the constant multiplied by the transform ofthe expression. That is 1K} = KLEF(O} Note: Two transforms must mot be multiplied together to form the transform of a product of expressions ~ we shall see later that the product of two transforms is the transform of the convolution of two expressions Example 6 (@ Let +1) =1 {26} +141) = 2Lfe"} +10} _2 24541 “s1' 2" 84D (©) L{2sin 3 + cos 3¢} = 21 {sin 3¢} + L{cos 3t} Bs 546 FEO REO FES (0) L{4e + 3.cosh 4t} = 4L{e**} + 3L {cosh 4t} Ge 16) So 1. L(2sin3t+ 4sinh3t 2. {Se +cosh 2t} 3. L{8 422 441} = ww 18243), 6-45-20 1 Fis 49 +4546} ustoar ! G-ae-9)' The working is straightforward. L{2sin 3t + 4sinh 3t} = 2. 18(s% +3) ‘#81 65% 4s~ 20 C-ae-4Laplace transforms 1 35 3. oP 42P art} = 3423 alte a 44546) uid ets ‘We have been building up a list of standard transforms of simple expressions. Before we leave this part of the work, there are three important and useful theorems which enable us to deal with rather more complicated expressions. Theorem 1 The first shift theorem C5) The first shift theorem states that if L{f(#)} = F(s) then L{e*f(} = Fis +a) Because L{e~*(t)} = £ That is Le“ f(0} = Fs+a) ‘The transform L{e-*/(t)} is thus the same as L{f(t)} with s everywhere in the result replaced by (s-+2) For example L{sin2t} = 575 eapije star =f" fie dt = Fs +a) 2 2 then Le sin2t} = Sag apes Similarly, L{@} Because L{t? {Pe} L{QPet} is the same with s replaced by (s ~ 4). wa Here is a short exercise by way of practice. Exercise Determine the following. 1. Lfe**cosh3t} 4. L{e*cost} 2. L{2e*sin3t} 5. L{@*sinn2t} 3. Late) 6. L{te} Complete all six and then check with the results in the next frame56 Programme 2 aw Here they are, 1. L{cosh 3t} = 5* *. L{e* cosh3t} mrs _ se Sees 5 3 2. Lsin3t} = 555 ~ L(2esin3t} = me “scat a ua 48 2 Mae aot ; s-2 4. Hoos} = 35> Ueteost) =P 5. Lsinh2t} = 3 6 Le} = LEO} =FO9 then L{EF(O} =F) Because Hero} =|" rinertar= |" rio(—S) at SP peter That Lif 0} = -F(s) For example, L{sin21} = 37-5 4/2) & tana 8 (g24) and similarly, L{tcosh3t) =Laplace transforms 1 57 Because L{t cosh 3t} ey ‘We could, if necessary, take this a stage further and find L{1® cosh 3t} af e+9 Le =L{t(tcosh peo 1{f cosh 3¢} = L{t(teosh 31) 3{e3} _2s(s? +27) (F-9)> 4 Likewise, stating with L{sin4¢} = 35g Ltsin at) = cece and EP sin at} = 8s 8(3s? - 16) @) (+16? appiing 10) ~ 4 (F9) i ach ewe ‘Theorem 2 obviously extends the range of functions that we can deal wth, So, in general if L{f()} = Fl), then Ler(o) = YGF) Make a note ofthis in your recnd book58 Programme 2 (GD Theorem 3 Dividing by ¢ IE LF(O} = 09 then {0} o}- f F(o)do provided Lim (M42) exists, To demonstrate tis we start from the righthand side of the result note the dummy vat [Ltiow-[{[froeserhar tht's the ent at 5 an expression in s which - f [L poetarae comesttom he lower imi 0 Jos of the integral so the vari- "rol, ) dag able of integration, which Isabsorbed during the pro- 3 cess of integration, 1s © changed to ¢. Notice also ( 8 £ ros thane Increhonge the ff} order of integration. a This ule Somenhat ested in se, sine its aplabe only if Lim (C2) exis indeterminate cass, we we Lp a find out. Lets ty couple of examples ‘Example 1 Dasemine {884} Fist we test Lim {S824} . (F ein} -{3} top and bottom separately and substute #0 in the result to ascertain the limit ofthe new expression. im 292) — rm {22°84 a, tat the limit exists and the theorem can therefore be applied. 4 terre {88 = = fn 5 aretan($) = arctan (2) By L'Hopital’s rule, we differentiate So L{sinat} Notice that arctan (2) + arctn(2) seen from the igureLaplace transforms 1 59 Example 2 =s2t} the limit exists (EES) Apply Hopital’ rule. ri weet whats Lin {= 22!) as a Ty ficcosaty _,(2sinzt . Lim {2=882P py 2802) 09 ni ext ur cosan) <2 gS S44 ‘Then, by Theorem 3 HEM} -L tae Complete it Let us pause here for a while and take stock, for we have met a number of results important in the future work.60 Programme 2 1. Standard transforms fo Lf) =F(s) eo sina cosat sinhat coshat PS t t (na positive integer) 2 Theorem 1 The first shift theorem. EL FCO) F(s), then L{e*f(¢)} = Fis +a) ‘Theorem 2. Multiplying by t HEL{F(O) = FE), then L{Hf(O)} a Reon Theorem 3. Dividing by t If L{F(®)} = F(s), then iff (oy -[ F(o) de provides tim {HO} exists, eo let Now let us work through a short revision exercise, so move on a Determine the Laplace transforms of the following expressions. 1 2 3 4 5 sin3t cos2t o ot sinh 3t 6 tcoshat 7 #344 en T 9 ecosat 10 fsint Complete the whole set and then check results with the next frameLaplace transforms 1 6 Here are the results. [26] 1 2 7 3 8 4 9 Bre 62-2 5 yo 8-2 ey It is just a case of applying the standard tranforms and the three theorems. Now on to the next piece of work Inverse transforms Here we have the reverse process, Le. given a Laplace transform, we have to find the function of t to which it belongs. For example, we know that re ‘the Laplace transform of sin at, some can oy ter) = inate mba ning the inverse transform and not a reciprocal. WIL barns {T= fe eg} eonns @ fg} ae CD @ nf} an) ri{}aa Jacoese @ refs!) saat PB. wert, ‘Therefore, given a transform, we can write down the corresponding expression in t, provided we can recognise it from our table of transforms. >»62 Programme 2 But what about £~ a}? he ety ot spin curio anda earns va consteing 1-H{=21}, 4 nappens tt we can wie +1 1,2 reg 8s the sum of two simpler functions —' +75 which, of course, makes all the difference, since we can now proceed weet retary which we Immediately recognise as SB eee are called the partial 1 The two simpler expressions <5 and == and the ability to represent a complicated algebraic fraction in terms of its partial fractions is the key to much of this work. Let us take a closer look at the rules. Rules of partial fractions 41 The numerator must be of lower degree than the denominator. This is usualy the case in Laplace transforms. If itis no, then we first divide out. 2. Factorise the denominator into its prime factors. These determine the shapes ofthe partial fractions. 3 A linear factor (5+a) gives a partial fraction A= where A is a constant to be determined. gives A,B 4. Arepeated factor (s+ a7 gives 4 ; ms (4a) Bes Gray Gear 5 Similarly (5 +a)® gives 4 —,_B_,_€ a ear ae 6 A quadratic factor (52 + ps +g) gives ae 7 Repeated quadratic factors (s? + ps+q)° give PseQ_. RS4T SEPS+q (Re pstgy s-19 GFD6=5) So has partial fractions of the formLaplace transforms 1 63 ana 8-481 (+ 3)5—2) Be careful of the repeated factor. a, B c a 543° (6-2) 6-27 has partial fractions of the form .. Let us work through the various steps with an example. Example 1 rodetemine Hf 81} (@) First we check that the numerator is of lower degree than the denominator. In fact, this isso. ‘ Ss+1 Ss+1 (b) Factorise the denominator 3°45 = GN ay (© Then the partial fractions are of the form We therefore have the identity S41 A, B 43 If we multiply through both sides by the denominator s? —s— 12 = (s—4)(s +3) we have Ss+1=A(s+3)+B(s—4) ‘This is also an identity and true for any value of s we care to substitute = our job is now to find the values of A and B. ‘We now substitute convenient values for s (@) Let (6—4)=0,i0.s=4 2, 21=A(7) +80) © Let s+3)Example 2 vetemine-{2¢=8}. Working as before, f(t) = ase Because 95-8 LO) = Fy (@) Numerator of first degree; denonominator of second degree. ‘Therefore rule satisfied. 9-8 A, B © Se=A st 5-2 (© Multiply by s(s 2). -. 95-8 = A(s~2) +Bs. (@ Puts=0, -8=A(-2)+B(0) -. A=4. (© Puts-2=0,4e.8=2, 10=A(0) +B) -. B=s. fem if 2 } 4 ese" Example 3 S188 +41 Express. Fs) ==—155+41 in partial fractions and henc press Fly) = EASES in part tions and hence determine its inverse transform. Ss =1Ss+41 <5 + has partial fractions of the form (6+ 2\6-3)Laplace transforms 1 65 A,B, Co (6) se 6-3) ‘Now we multiply throughout by (s + 2)(s 3)? and get P-155-441 = ASP +Bls-+2)(6—3) +1542) Putting (6-3) 0 and then (s+ 2) = 0 we obtain ras asdandont G7 Now that we have run out of ‘crafty’ substitutions, we equate coefficients of the highest power of s on each side, i.e. the coefficients of s*. This gives ... 1=A4B 2. 1=3+8 Gy) Sa15s441 3 (6+2)(s-3)? $42 eG} 3e# and 2c" Gy) But what about wf z 3}? a we semember int {3} = 7 9) and that by Theorem 1, if L{f(€)} = F(s) then L{e“/()} = F(s +a). 1 1 ; Grays ke sx mith s replaced by (6-3) Le. a= 3. rf ahe (s-3) af? =158 +41) _ yon {meth} vacates66 Programme 2 Example 4 af 4k = 5546 Determine I {anes} ‘Notice that this time we have a quadratic factor in the denominator 48-5846 A Bs+C Gee eH 41 44 49° 554 6=A(S? +4) +(Bs+O(5+1). (@) Putting (+1) =0,le.s=-1, 15=54 2. A=3 (©) Equate coefficients of highest power, Le. s? 4=A4B 5. 4=34B 2. B= (© We now equate the lowest power on each side, i.e. the constant 6=44+C 3 6=124+C 7. C Now you can finish it off. f(t) = w® f(t) = 3e* + cos sin Because a. s 6 WO Sat Rea ya F(t) =3e + cos 2t—3sin2¢ 6 The ‘cover up’ rule While we can always find A, B, C, etc., there are many cases where we can use the ‘cover up’ methods and write down the values of the constant coefficients almost on sight. However, this method only works when the denominator of the original fraction has non- repeated, linear factors. The following examples illustrate the method.Laplace transforms 1 67 Example 1 B —2 By the ‘cover up’ rule, the constant A, that is the coefficient of 2, is found by temporarily covering up the factor sin the denominator of (5) and finding the limiting value of what remains when s (the factor covered up) tends to zero. We know that F(s) = SS ae partial fractions of the form 4 + Therefore = coeticient of $= Lim {957} — 4, thats A = sn s-2 Simiaty ,the coeftcent of ys obtaned by coveing up the factor (s ~ 2) in the denominator of F(s) and finding the limiting value of what remains when (s — 2) ~0, that is s+ 2. ‘Therefore B = coefficient o Lim {°°=8) = 5, Tat is B=. So that gs-8 4 5=2)~s Another example Example 2 sy $417 A,B Cc FO = Gage ae *s-1 A: cover up (s~ 1) in Fs) and find urlerajeca} 6 ‘Similarly B: B c: c= pin ttt?) aD 5-642! BS Kt 2 O= Tat s-3 s— So f(t) =€% +26 —3¢ Every entry in our table of standard transforms gives rise to a corresponding entry in a similar table of inverse transforms. Let us tabulate such a list.68 Programme 2 & Table of inverse transforms 7 f a 1 -at “ale ira # | @ | aposuveintesen 3 fT | (a postive integer) | A | crapostive ints a sinat cosat sinhat coshat ‘Theorem 1 ‘The first shift theorem can be stated as follows. If F(s) is the Laplace transform of f(t) then F(s +a) is the Laplace transform of e-*f(t). Here is a short revision exercise. Exercise 1. Find the inverse transforms of 1 5 Bs4+4 @ % © Sp © 2 Express in partial fractions 225-416 - + |g Salst6 ® Brie DHery wrn6-2 3 Determine on fcemeas © eames}Laplace transforms 1 4e%, set, © 3cos3e+4sinat 24 4 ©) SFT 3-3" GaP () 26" + 3cos2t + sin 2t Solution of differential equations by Laplace transforms To solve a differential equation by Laplace transforms, we go through four distinct stages (@) Rewrite the equation in terms of Laplace transforms. (b) Insert the given initial conditions. (©) Rearrange the equation algebraically to give the transform of the solution. (@) Determine the inverse transform to obtain the particular solution. ‘We have spent some time finding the transforms of a variety of functions of f and the inverse transforms of functions of s, i. we have largely covered steps (a) and (4) of the above list. However, to write a differential equation in Laplace transforms, we must obtain the ax ex transforms of the derivatives SF and $5. Transforms of derivatives Let f"(t) denote the first derivative of f(t) with respect to f, F"(t) denote the second derivative of f(t) with respect to f, etc. Then LEP()= [ef ae by definition. Integrating by parts ur} = ere), = FP rot-sesy ar When t= co, e~#F(t)70 Programme 2 a yt~t~:~S Because s is positive and large enough to ensure that e~ decays faster than any possible growth of f(t). ~ LF ()} = FO) + sL{f(O} Replacing f(t) by f’(t) gives Le" ()} =. ws) Lef"()} =F (s) ~ sf(0) — 7°) Because Writing L{f (8) = -F(0) + sL{f(O)} Lf"(O} =F) +SL{F(O} (0) + s(—F (0) + SL{F()}) (s) as usual, we have Lif (0) Lf (6)} = Fs) Lf (O; (s) — F(0) Lf" (O} = ?F(s) — sf(0) — f'(0) We can see a pattern emerging Lp} = ww Lf"()} = FFs) —F(0) — 70) - F"0) Alternative notation We make the working neater by adopting the following notation. Let x= (0) and at £=0, we write ae i.e. F(0) =%0 ie. f) =m te. f"(O)=x ete My he ("(0) =H Also we denote the Laplace transform of x by %, ie. X= L{x} = L{f(t)} = F(s). So, using the ‘dot’ notation for derivatives, the previous results can be written eseLaplace transforms 1 nm L(x =x (es) LG} = 8-2 1G} = 8x—s10— Lf} = 82 — 2x9 — 5x1 — x In each case, the subscript indicates the order of the derivative, 0. dx Lee, xy =the value of $= at ¢ Notice the pattern of the results. Lfi}=.. 7 3 & Now, at long last, we can start solving differential equations. Solution of first-order differential equations example 1 save the equation 224 gen hata 0 ‘We go through the four stages. (@ Rewrite the equation in Laplace transforms, using the last notation LQ} =% La} = 1{4} = ne aee)”t”té‘" 5 = $s? — 235 +26 “G-6-26-3) (@ Now for partial fractions 52-235 +26 G-DG-DE-H s-1 5-2-3 S82 — 235 +26 = A(s~2)(s~3) + B(s~ 1)(s—3) + C(s—1)(s~2) So that A= ..eccsseeeeei ; C= =e to @ As you see, the Laplace transform method can be considerably shorter than the classical method which requires (@) determination of the complementary function (©) determination of a particular integral (© obtaining the general solution, before (@ aniving at the particular solution by substitution of the initia Conditions in the general solution. >76 — Programme 2 Here is another example. Example 2 @x ae Solve £* — 4x = 24.cos2t given that at ¢ = 0, x=3 and (a) In Laplace transforms (63) (#2~s%-x) 42-4, +4 (©) Insert initial condition, ie. x9 = 3; x1 = 4 8x-3s-4-49= sya . — 2as . (aaa as 4 PS S84 4? 4365-4 16 (o xa BS HAS +365 + 16 © (PFA =2HE+2) Expressed in partial fractions, this becomes 35? 445? 4365 416 = (As + BY(s— 2)(5-+2) + C(s? +4)(5-+ 2) + D2 +4)(5—2) Putting (s 2) 2, gives C Putting (5-+2) ~2, gives D=2 Equating coefficients of s* and also the constant terms gives A = ~3 and B=0.Laplace transforms 1 7 4c + De cos 2t @ Now let us solve another equation, this time using the ‘cover up’ rule, Example 3 Solve £+ Si + 6x = 4t, given that at 0, O and i=0. AAs usual we begin (32% — sto — x1) + 5(s% x0) + 6% = 4 4 =O; m=O 4, (PHSS+OR=Z 4 2- FGF DET) ‘The # in the denominator can be awkward, so we introduce a useful trick and detach one factor s outside the main expression, thus rlarraives} “i ftsratia} Applying the ‘cover up’ rule to the expressions within the brackets 1f4d 1 +3 Now we bing tenes ak int he ld a2 td and the second and third terms can be expressed in simple partial fractions so that which can now be simplified into 21 51 578 Programme 2 Sgn kg ot 9 There are times when a quadratic coefficient of ¥ cannot be expressed in simple linear factors. In that case, we merely complete the square converting the expression into (s+k)?ta?. Let us see such an example. Example 4 Solve ¥—2i + 10x =e, given that at ¢=0, x= 0 and x We find the expression for ¥ as before. @) =a > GF = BF I) Because (62x —s%0 — a1) — 2(5% — a0) + 108 2 — 29k + 10% (@-25+10)=1+ 45 s-1 G=ME= BHI Expressing this in partial fractions Evaluate the coefficients. { } 1052" F470, Because s-1 Bs+C ($—2GF= 25 +10) ~ (S—2) * F—Bs +10 A(S? — 25+ 10) + (5 2)(Bs +O) Put (5-2)=0,ies=2 1=A(4—4+10) f] O=A4B WA-2C -. 20=2Laplace transforms 1 79 Now we have to find the inverse transforms to obtain x. The first 1 s-10 term + is easy enough, but what of 520? The denominator will not factorise into simple linear factors; therefore we complete the square in the denominator and write it as, s-10__s~10 F=Bs+10 GaP +9 and then we improve this still further and write it in the form (s—1)-9 s-1 S-Y=9 We are quite happy with this, for —S—} (= 49 oem 149 Pest with s replaced by (s~ 1), which indicates an extra factor ef in the final function of t (Theorem 1). is merely eed f a s-1 8 so zat f4 so, Bes (1 +9 ae} x X= 7p (Et ~ e cos at + 3e sin 3t} Just try one more like this one ae Solve + i+x=e given that at f=0,x=0 and i= 1. We find the expression for % as before. ‘ S42 ws + +s+ 1) Because (32% — sto — x1) + (8% — 40) + 1 FT Where xo = Oand x: so that : 1 Se-tysttee ty therefore 1 see x(@4s41)=14 2, 52 WPts4 altaya giving 542 Ge +s Expressing this in partial fractions Bvaluate the coefficients80 Programme 2 542 A, Bsa CESS S41 eset so that S42=A(2 +541) + (Bs +C\(s +1) Put s+1=0, that is s=~1 then 1=A(=141) sothata=1 [°] o=a+e so that B= —1 [CT] 2=A+C —sothatC=1 Therefore 1st sel @ys41 Completing the squares inthe second term gives s-1 Feed sol s+} are FT Ey one s-1 st aH G++) a= Sth rye) +9"+(4) so that x a) [ay ST ep cysLaplace transforms 1 81 Before we leave this topic, the same general approach can be employed for solving simultaneous differential equations. Let us see an example in the next frame Simultaneous differential equations [ez] Example 4 Solve the pair of simultaneous equations pose ityset given that at f= 0, x=0 and y=0. (a) We first express both equations in Laplace transforms. 1 (7 yo) -¥ i (8-40) +7 = (©) Then we insert the initial conditions, x9 = 0 and yp = 0. Q si (© We now solve these for % and j by the normal algebraic method. Eliminating 7 we have st -2s-1 GrHe-H 25-1 yee TH Representing this in partial fractions gives x82 Programme 2 Because S21 A Ses HET aI 2, P= 25-12 Als+ (241) +Bs—1)(2 +1) + (S-1)(s+ (s+) zg Putting s=1 and s=~1 gives A =~} and B (Comparing coefficients of s? and the constant terms gives C= 1 and D=1. stl ¥a1 ww x= — |e Jett cost-+sint We now revert to equations (1) and eliminate x to obtain 7 and hence y, in the same way. Do this on your own. yen ww yade +}et— cost +sint Here is the working. S428 P= GING TD s-1 S47 o. 425-1 = AGS +1)(62 +1) + Bs — (2 +1) + (s—1)(s+1)(Cs+D) Putting s= 1 and s=~1 gives A=4 and B=}. Equating coefficients of s? and the constant terms gives C. D=1. Land a sy 347Laplace transforms 1 83 So the results are 1 x23 (¢ +e%) +sint + cost =sint + cost — cosht yaZ(E +e") +sint cost =sint cost +cosht 1 X=sint +cost—cosht; y-=sint—cost-+cosht Simultaneous equations are all solved in much the same way. Here is another. Example 2 Solve the equations 2) 6 +3x=0 34-3x-2=0 siven that at t= 0, x=1and y=3. Expressing these in Laplace transforms, we have 2(sP — yo) - 67+ 3¥=0 w Bis ~ x0) ~ 38-27 =0 ‘Then we insert the initial conditions and simplify, obtaining, (@) To find x a 38+ (25 6)7=6 2) x (s-3) (5 3)(3s~3)x ~ (28 ~ 6)7 = 3(s~3) Adding, [(s~3)(3s~3) +32 = 35-946 1. (Bs? 12s + 12)¥ = 35-3 (@-4s44)z=s-1 s-1 A |B _AG-2)+B (=2P s-2° 6-2-2)" AG-2)+B giving A=1 and B=1 1 a ssatgege nee (b) Going back to equations (1) and (2), we can find y. y84 Programme 2 wS 6c 302) Because, eliminating x we get 9 A.B s-27 Pf “28-22 . 65-9 =Als—2) +B A=6; B=3 63 yeh 7 He aa} = y= HGH + 3c} Simultaneous second-order equations are solved in like manner. ‘Again, with all these solutions it is a worthwhile exercise to substitute the solution back into the differential equation to verify that the solution is correct. (86) Example 3 If x and y are functions oft, solve the equations i42e-y=0 j+2y-x=0 given that at t= 0, x0 = 4 y» = 2; x1 =0; 1 =0. We start off as usual with | (s2%~s%0 ~ x1) +28-y=0 and (89 — syo —y1) +23 Inserting the initial conditions, we have 89454227 =0 S-2s+29-¥=0 ‘Simplifying these we can eliminate 7 to obtain X and hence x.Laplace transtorms 1 8s a= Bcost + cov @ Because (8 +2); a) “x4 (2 +2)7=25 @ Eliminating J and simplifying gives 4s! 410s @FNE+S 454108 _ A+B Cs+D +S #43 45? +105 = ($2 +3)(As +B) + (s? + 1)(Cs+D) Equating coefficients of like powers of s [8] 4-440 1 A¥e=4 [cl] 0=3B+D -. 3B+D=0 Putting s= 1, ©. 2A+2B+C4D=7 Putting s 2A~2B4+C-D=7 Putting C = 4 — A and D = ~3B in the last two leads to A Bo vccceedBD haa)” neon + ca) To find y we could return to equations (1) and (2) and repeat the process, eliminating Z so as to obtain 7 and hence y. But always keep an eye on the original equations, the fist of which is E42x-y=0 Therefore, in this particular case, y = % + 2x. So all we have to do is to differentiate x twice and substitute = doost + cos(V#) = —3sint—V3sin( V3) i= —Bcost ~ 3cos(V3t) “. y= ~Beost~ 3c01(V31) + cost + 2c08(V3t) 2 y= Boost cos( v4) which is a good deal quicker. So, as we have seen, the method of solving differential equations by Laplace transforms follows a general routine. (a) Express the equation in Laplace transforms (b) Insert the initial conditions (©) Simplify to obtain the transform of the solution (@ Rewrite the final transform in partial fractions (©) Determine the inverse transforms and, by now, you are fully aware of the importance of partial fractions ‘That brings us to the end of this particular Programme. We shall continue our study of Laplace transforms in the next Programme. Meanwhile, be sure you are familiar with the items listed in the Revision summary that follows, and respond to the questions in the Can You? checklist. You will then have no difficulty with the Test Exercise and the Further problems provide additional practice.Laplace transforms 1 87 Revision summary 2 [ea] 1. Laplace transform L{f(t) 2 Table of transforms fe) Lf} = Fs) a a a sinat cosat sinhat cosh at aa sat e (na positive integer) 3 Linearity of the Laplace transform (a) The transform of a sum (or difference) of expressions is the sum (or difference) of the individual transforms. That is LAF) £8(0} = L{F(O) £Ls(0)) (©) The transform of an expression that is multiplied by a constant Is the constant multiplied by the transform of the expression. That is URE ()} = RLFC} 4 Theorem 1 Fist shift theorem If L{f (8) = F(s), then L{e~f()} = F(s +a). 5 Theorem 2 Multiplying by t IEL{F(O) = FG), then L¢(0} = - 2 (F(9)}- 6 Theorem 3 Dividing by t re x(pe9y = A(9, hen {FO} = [Hepa provided that Lim {LO} exists, 2 Inverse transform HEL{F(E)} = FE), then L-HF(S)} = F(t >88 Programme 2 8 Rules of partial fractions (@) The numerator must be of lower degree than the denomi- nator. If not, divide out. (b) Factorise the denominator into its prime factors. (© Alinear factor (5 +a) gives a partial fraction 4 where A isa constant to be determined. (@) A repeated factor (5 +a)? gives A.B c Similarly (5 + a)? gives 44. —® (© Simitaly (644) goes + at as (A quadratic factor (5? + ps +4) gives (g) A repeated quadratic factor (s* + ps +4)” gives Ps+Q Rs+T SPS” (24 ps+qy 9 The ‘cover up’ rule The ‘cover up’ rule often enables the values of the constant coefficients to be written down almost on sight. However, this, method only works when the denominator of the original fraction has non-repeated, linear factors. 10 Table of inverse transforms FO) re) a £ a 5 — et sea (ra positive integer) By the first shift theorem If F(s) isthe Laplace transform of f(t) then F(s-+a) is the Laplace transform of e-*f(t).Laplace transforms 1 89 11 Laplace transforms of derivatives Lx} {88} 10-20 de where x9 = value of x at t= 0 Li} = si ~ 19 —n ete. x x1 = value of Fat t=0, etc. 412 Solution of differential equations (@) Rewrite the equation in terms of Laplace transforms. (©) Insert the given initial conditions. (©) Rearrange the equation algebraically to give the transform of the solution. (@ Express the transform in standard forms by partial fractions. (©) Determine the inverse transforms to obtain the particular solution, 13 Simultaneous differential equations Convert the simultaneous differential equations Into simulta- neous algebraic equations by taking the Laplace transform of each equation in turn, Insert the initial values. Solve the simultaneous algebraic equations in the usual manner and take the inverse Laplace transform of the algebraic solutions to find the solutions to the simultaneous differential equations. (Check this list before and after you try the end of Programme test. On a scale of 1 to 5 how confident are you that vance Frames ‘© Obtain the Laplace transforms of simple standard expressions? to ys O OF O O DO No ‘* Use the first shift theorem to find the Laplace transform of a simple expression multiplied by an exponential? @}to Ys OF OF OF OF O Ne Find the Laplace transform of a simple expression multiplied or divided by a variable? Ys O O DO 0 D No