O The Runge-Kutta Methods This method was devised by Runge, about the year 1934 and extended py Kutta a few years later. Therefore we call this method as Runge-Kutta method. Unlike any of the methods, explained in the preceding two sections the increments of the function are calculated once for all by means of a definite set of formulae. Here a set of formulae are given without proof for solving a differential equation of the form & f ( y) under the initial condition y (9) = Yo: Let # denote the length of the interval between equidistant values of x. The various types of formulae according to their order are given below. | Second order Runge - Kutta Method y If the initial values are xq, Yo for the differential equation = f (x, y) then the first increment in y viz Ay is computed from the formulae ki hf (o> Yo)4.38 sessssssssstsscessststensessnstnesnsnassseesenenseunt NUMERICAL METHODs ‘ a ky = Af |xot720tQ Ay = k Nowx, = x9 +h, y) =o + Ay The increment in y for the second interval is computed in a similar manner by means of the formulae, ky = Af Gyr, : A MY. ky = Aflatzntz)> Ay = hk, ~ and so on for the succeeding intervals. i Q Third order Runge - Kutta Method The third order Runge-Kutta method is designed by the following formulae. ky = hf (%,y0), h ky { = nslortwett) ( / Ay = hf %o+h, yo + 2ky-k), ‘3 * Now the first increment in y viz Ay is computed from ! 1 Ay = [ky + 4ky +k] Now x1 = x9 +h, yy =yq+ Ay The increment in y for the second interval is computed in a similar manner by means of the formulae. AL = AS Guy), . : hk ky kh = as(ast.y,+4) > y= hf Qy+hy+2ky-h), ; : AY = & [ky + 4hy + hs], and so on for the succeeding intervals, NNgoth VALUE PROBLEMS FOR ORDINARY DIFFERENTIAL EQUATIONS g Fourth Order Runge-Kutta Method : : | This method is most commonly used in practice. Unless and otherwise gated, Runge-Kutta method means only Fourth Order Runge-Kutta method. Let & = f @&, y) be a given differential equation to be solved ynder the condition y (xo) = yo. If h be the length of the interval between equidistant values, then the first increment in y is computed froth the formulae. ~ ky = hf.@o¥)> = atlart grat ky = hf\xotz.YtQ) > a h kh “t ts ~ rilaehveB) : .. ky = hf Hoth yo* hs) a “ Ay = FE (ky + ky + 2s + a) a. 7 Nowx, = x9+h,91;=¥o + Ay The increment in y for the second interval is computed in a similar manner by means of the formulae. - ky = hf Gpy> - h ok = rila+BontZ), ky = h ky ky = nian), ky = Af (ey thy th)» Ay = F (ky + 2h + 2 + ha) ‘ad so on for the succeeding intervals. \ ue . h nge- Note : If o is a function of x alone then the Fourth Order Rung Kutta method reduces to Simpson’s Rute. - For, ky = ASG) uUnrr4 @_. NUMERICAL METHODS 4.40 woes by = Aseo4) kth 10 + 4) ky = hf(xoth) . ay = tf Feee)+ 2420 +) 2f(20#3) +/+ »| G) [reo + a(x +f) +fG0+ ] ich i ined by applying Simpson’s hich is the same result as would be obtaine: , Rule in the interval x9 to x9 + h if we take two equal sub-intervals of idth 2 width > . Now we shall illustrate this method in the following examples. ( A AX exaup.e1 6 — : Find the values of y(1.1) using Runge-Kutta method of the third |S order and (ii) Runge-Kutta method of the fourth order given that dq art et. J Given y= Yty ie, SOY) = Pty And also given that xp = o=l = i) yo = Land h=0.1 To find y (1.1) using third order Run, Now k= AF (X,Y) ge-Kutta Method : = hly? +094] = ODL +A] = 0.1) 7 k= 0.2 hy IL a0+d yok] Z @ Una/BLEMS FOR ORDINARY DIFFERENTIAL EQUATIONS ......+0e-ss00e00+ 441 ponerse F ky)2 h ky = al (e+) +(xo+ +4) (v0 4) | 2 0.2\2 0.1 02 - on[ (1+) +(e )(O+¥) | 4 = (0.1) [1 +01) + (1 +0.05) (1 +0.1) ]- ky =: 0.2365 ky = hf (xoth, yo+2ho~h) = h[(yot2h- hy + (x0 +A) (yo + 2k - 1) 1 = (0.1) {C1 + 2(0.2365)- o2Pt(1+01) = (0.1)[0.273y + .1)(1.273)] = (0.1) [3.0208 ] key = 0.30208 . 1 chy = ¢ lh t4h+ bl ca z [ 0.2-+ 4(0.2365) + 0.30208 ] ‘ 1 A = gf 1.44808} c Ay = 0.24135 yy = YorAy = 140.24135 ; y(L1) = 1.24135 To find y (1.1) using Runge-Kutta Method of Fourth order: Here x) = 1, youl and h=0.1 [1 +2(0.2365)-0.21} Now k, = hf (0.0) oe M h (90° + x0 yo) = OD [P +] =O.D2) ky = 0.2 A= ofl nate nt |«il Goodale) Oo) i - anf DOP) = (0.1) [(1.1 + (1.05) (1-1) ] ky = 0.2365 = ky = af nti nee |] = af (we) «(~e8) (or) « oof (AR) e(149) (14285) = (0.1) [ (1.11825) + (1 + 0.05).(1.11825) ] = (0.1) [2.4246 ] oo . ) ky = 0.24246 Ne, . ky = hfQoth, works) - = (0.1) [ (mths +(x +h) (yo +hs)] = (0.1) [ (1.24246)? + (1.1) (1.24246) ] = (0.1) [ 1.5437 + 1.366706 ] ; (Or = (0.1) [2.91041] : { / ky = 0.29104 - PS 1 “Ay = é [ky +2 ky +2 hy + ky] : = [0.2 + 2(0.2365) + 2(0.24246) + 0.29104 ] £1 1.44896 7 je ycby f oe 5 0.24149 OY = Yot Ay = 1+0.24149 . yy = 1.24149“ . SC sprvik. VALUE PROBLEMS FOR ORDINARY DIFFERENTIAL EQUATIONS p EXAMPLE2 @ p>- By applying the fourth order Runge-Kutta Method find fromy’ =Y— XY (() =2 taking f= 0.1. y-(02) SU) = . a rt 7O Given y! = y-x. 7 vor ie, f(x,y) = y-x ve and y (0) = 2 Le. x9 = 0, yo = 2 and k= 0.1. We know that the fourth order Runge-Kutta formula for finding the first increment in y viz Ay is given by 1 : by = Gly +2hg+2h +h) where ky = hf (Xo, yo) . h kh) ky = dot nia) . y= ng{ 0+ B yt 2). 7 )— ky = hf (xo +h, yo+ ks) wi. 2. ky = (1) 09-40) = 012-9) t n=O v[(o0+B)-(-0+8)] A ox [(2+52)- (o+%)| += (0.1) [2.1-0.05] = 0.205. t= en|(w+)-C"9)] - oof (2+%#8)-(-¥)] — = (0.1) [2.1025 - 0.05] =. 0.20525 ky = (0-1) 0% As) Got I o wa UN 4 -—— —4 SIA csecssenenseneenees! aesaseesasonecneeneeseneesenneer yl), oe. Ay 7K a w W (0.1) [2 + 0.20525 -0-0.1] 0.210525 sth 2h +2 +a i cf = do. 2 + 20.205) + 2 (0.20525) + 0.210525] i [0.2 +0.41+0.41 05+ 0.210525] 0.20517 v=o t AY. 2+0.20517 2.20517 Next we have to find y (0.2) =y2 =y, + Ay ~fe Yi Lo: "where Ay [= \¢ [ky +2 ky +2 ky + ka] @ Ung. 1 Now k, W a x th=0+0.1=0.1 Af) AD -%] (0.1) [2.20517-0.1] = 0.210517 - h ky os (x,+4, yt +2) [b-4)-G3)] on| (2205174 22105 )-(01 ae (0.1) (2.31042 ~ 0.15] = 0.21604 = h h ky) - t(aj+2, n+] ALbn8)- God] oi 2 )re ysTIAL VALUE PROBLEMS FOR ORDINARY DIFFERENTIAL EQUATIONS 4 : sessesennanssenene 45 conf (220s 42248) (, +ot)] (0.1) [2.31319 — 0.15] = 0.21632 ky = hf thy +h) = hl th)- Gi tA. : = (0.1) (2.20517 + 0.21632)- (0.1 +0.1)] | i = (0.1) [2.142149 - 0.2] = 022214 \ a cs Ay = & Uy t+ ky + 2ky + a Al 1 =6 [0.2105 +2 (0.21604) +2 (0.21632) + 0.22214] ; . =6 [0.2105 + 0.43208 + 0.43264 + 0.22214] } = 0.21622 on 8 KW dn tay ne = (2.20517 + 0.21622 yx = 2.42139 “4h Hence we haye the following table. { x 0 ot | 02 bh oy 2 |2.20517| 2.42139 : | + EXAMPLE 3 * ee ‘ Find the values of y(0.2) and y(0.4) using Runge-Kutta Method of : : od . turth order. with f= 0.2, given that “qr =V" + 73.9(0)=08- Given - y= ety ie, fosy) = VP *Y And also given that x9 = Yo= = 0.8 and h=0. a |To find y (0. We know that the fourth order Runge-Kutta formula to find the first | increment in where k, = Af (xq, ¥o) k, k i = .2\2 Cys 2 02[ (0+22) +(os+ 222882) $20 pe aw os eek CON OATES Yio kk fu °%* ks vasssssssus NUMERICAL METHOps | 2) 3 y viz. Ay is given by 1 Ay = G [ky +2 hy +2 hy + ha] =h [Nw ty ] = (0.2)[Y0+08 ] = 0.17889 = asl ott y+H | NCFCD W =” 0.18968 ( } i asl nti nt ] 1 CoO B 7 2 (0 +P + (os + 218968) = .2)[ V1 +0.8 + 0.09484 ] | = 0.19025 = AS Goth, yo+ky) = INGOTS GS FR) = AVOF02¥ 4087 0.19025) = (0.2)V0.044 0.99025 = 0.20300ee oo VALUE PROBLEMS FOR ORDINARY DIFFERENTIAL EQUATIONS 4.47 1 chy = 6 Ut 2h +2hg thy] 1 = & [0.17889 + 2(0.18968) + 2(0. 19025) + 0.20300 } 1 . = g( 1.14175] : : ~ Ay = 0.19029 2.0.2) = yot dy = 0.8 +0.19029 0.99029 y (0.2) To find y(0.4) Here x; = 0.2, yi =0.99029 and h=0.2 Now k = hf(,y) = h [Natty | = (02) [Y@2y¥ + 0.99029 ] ky = 0.20301 in = WL att ont +i] iL oF O5) (0.2) | (02 s o2y + (0.9029 R 020301 }I = (0.2) V0. 3) + 0.99029 + 0.10150 ky = 0.21742 = afl nt xt bye]o. Ay “ y2 2 y (0.4) W W sc cecescuaaestesccotesesetctsomcsnenecsa-orcsetersesseecesei NUMERICAL METHOng: f (0. ty (0.3y" + 0.99029 + 0.10871 | 0! 0.21808 hf@ath,yitks) AN (a thy + (i +h) hf O2+02) + (0.99029 + 0.21808) (0.2) 1.36837 0.23396 7 [i+ 2k +2kyt ky] 7 [ 0.20301 + 2(0.21742) + 2(0.21808)-+ 0.23396 J bf 1.30797 ] 0.217996 y+ Ay 0.99029 + 0.217996 1.20828 [| 04 | 0.8 | 0.99029