oniy the given function values at

founth onder Runge Kutta method is

method of second order
differcnt pont
tedious,
s
we
, Numeri al
11 3 Second order Runge-Kutta
dy
method (for ffrst
AIM. To solve dx -f(x, y) given y(x)=Yo
Proof. By Taylor scries, we havec,
v(r+h) =y (*) +hy()
+)+O(h'
Differentiating the equation (1) w.r.t. x,
df of dy
+
-f,+yJ,=Í, +,
Using the values of y and y" got from (1) and
y(a+h) -y () =hf+, hU,+] + 0u'
(3), in (2), e

Ay = hft;,+)+ow)
Let Ajy= k, =fu, y). Ax =hf(u, y
A,y=k,=hf(r+ mh,y +mk)
and le Ay = ak, + bk,
constants to be determined to get the
where a., band m are better acous
of Ay. of h.
Expand k, and Ay in powers
two variables, we h
Expanding k, by Taylor series for
k,=hf(r+mh, y +mk,)
mh
mh + mk 2!
2

mh +mk,
mhf, +mhf,+ 2!
SInce k, =

powers of h
=hft mh (f,+f)+ higher
Substituting k, k, in (7),
hy= abf +bl af+m (, +f) +O(h)
-(a+b) hf+ bmh (f; +ff) +O(h)
Equating Ay from (4) and (10), we get
 w/utton ofr Ordinary Dferential Equations
38|
mh'+) higher powers of h
subslitut1ngA. A, in
(7), (9)
AYahf

bmh' 4, +f) +oh')

and (10), we get ..10)
iquatingAy from (4)
and bm =2
..01)
Now we have only two equations given by (1) to
unknons a, b, m. solve for three
From a+h=1,N= | - b nd also m=
2h using (7),
Ay =(1- k, + bky
where k=hf(x, y)
h hf
k, =h, 2b "y+
25
Now Ay = y (x+ h)- y (x)
h hf
y(x+ h) = y(*) + (1 - b) hf+ bhfx 26
y+
25
1., Yn+=y, +(1- b) hfon, y,)
h
+bhf|x, t 26 2b
From this general second order Runge-Kutta
formula, setting
a=0, b =1, m= 2 we get the second order Runge-Kutta algorithm as

kË = hf (x, y)

ky = Second order R.K. algorithm

and Ay = K where h=A.
Since the derivations of third and fourth order Runge- Kutta
algorithms are tedious, we state them below for use.
The third order Runge-Kutta method algorithm is given below:
k, =hf(x, y)
,h,y ) Third order

ky = h (x+ h, y+ 2k,- k) R.K. algorithm

and Ay=k, +4k, + k,)

382
The fourth order Runge-Kutta method algorithm
problems unless
otherwisc nentioncd.
Iis Numerieat
k, = hf(, y)

ky =hf(uth, y
+ k) Four
R.K.thalgororditerhm
+k4)
and Ay=(k,+2k, +2k,
+ Ay
y (r+h) =y (r)
dy
solve dx -f(., y). y )= o
Working Rule: To

Calculate k, =hf(o o) 1

+k,)
ky=hf(+h,yo
1
2k, +2k, tk4)
and Ay=¿(k,+
where Ar=h.
Now y, = Yo+Ay.
from (ij, y)) and repeating the process, We
Now starting
(, y) etc. method,
order Runge-Kutta
Note 1. Insecond
h 1

h
:.=Yo +Ayo =.
method.
This is exactly the Modified Euler
second order is nothing but be
So, the Runge-Kutta method of
Modified Euler method.
alone, then the fourth orda
Note 2. Iffu, y)=f), ie., only a function x
Runge-Kutta method reduces to
kË=hf()
 - the area of y=f(r) between X= X and

Le,
equal intervals of length by

Ay reduces to the area by Simpson's one.

thethree methods, (2nd order, 3rd
Xhird*thnçwe ith
Simpson': One )

In
all

values of Kc, ky
arder and h
aresame. Therefore, one need
Nale

doing by all the three methods.

while
t
repeat otrhdeer) the.
Apply the fourth order Runge- wortk
1.
fanple
y'=xty y(0) = 1. Kutta method to
(Ap. 1992)find
that
mentioned in
Since h is not
the
solotion.

Ertyy(0)=| .: f y ) =X+y, Xh =0,

quest on, we take
j=01,x,=0:2
fourth order Runge-Kutta method, for the first
-1)=01 interval,
By
=(01)4p+ y) =(01)(0 +
k=hflo )=(
+thtJ=(0)f(005, 195,
- (0-1)(005 + 105) =01|

=(01)f(005, 1055) = (0- 1)(005 + 1-055) =0110%

=(01)f(0-1, 11105) =(0:l)(01+1105) =0-12105

=(01+022 +0-2210+ 012105) =0110341667.

01)=y, =yo + Ay= 1-110341667 1110342.
Now starting from (x, y) we get (x, y»). Again apply Runge-Kutta
Miim replacing (Xp yo) by y).
h=hfk, y)=(01)(41 +y) =(01l)(0-1 ++1-110342)
1 =01210342

4+;=(01)f015,1-170859)
H7859)=0-1320859
= (01)(0-15 +
84

(0 1)f(0s.
=(0
1)f(0 2, 1
=0 144298048 242980M8
(01)+6, +.2k, +2k, +ka)
(02)
y =y
v(02) =12428055 =l1103424!
Correct to four decimal places, y(0-2) =12428.
Example 2. Ohtainthe values ofy at x =01, 0,2
order (i) third order and
of (i) second (iii) fourth oTder
y'=-), given y (0) = )
diferential equation fo
Solution. Here. f(x. y) =-y, =0. yo =l, (MKI V
()Second order:
k,=hf(Wp yo) =(0-1)(- y) =-01
x=01,x,=0)
1

=-0|x 095 =-
0.905
)) =o t Ay=l-0095 =
y;=y (0-1) =0.905
Again starting from (0-1, 0-905) replacing (p ) by
k,=(01)f(Kj, )}) =(01)(-0-905) =-0-0905

=(01) ((0-15, 0:85975)] = (01) (-085975) =- 0085975

Ay =k,
.. y2= y(02) = y, + Ay=0-819025
(iü) Third order:
k, =hf(oy yo) =-01

k, =hf(,+ h, yo +
2k, - k)
-
(0)f(0,09) =(01)(-0-9) =-009
Ay= (kt 4k, +k,)
y =
 tns

k,= hf (u, th. y +2, h)

=(0 Df(02. 0828) = 9082%
EY
y, = + Ay =091+(4, 4, +b,y
- 091
+-091
6
0823366 -034)-0nn,
y (02) =
iDFourth order:
k, =hf ( Yo) =(01)f(0, 1)=-0|

kh=hf|othYotke =(01}f(005, 09525,

=-0(9525
ky= hf (xo +h, yo +k,) =(01)f(01, 090475)
=-0-090475

Ay =(k, +2k, +2k, +ka)

y=Y% +Ay= l +(k+ 2k, +2k, +ky)
y,=y (0-1) = 09048375
Again start from this ( };) and replace ( Yo) and repeat
k,=hfu, y}) =(0:1)-)=-009048375
1

-(0Df015, 08595956) = -)cdo! 08595956

=(0:1)f(0 15,0 86I8577 ) = -0 cdot 08618577

ky=hflzy th, ytky)

=(01)f(0:2,0-8186517) =-0-08186517
Ay=(-0
6 09048375 2x0-08595956
-2 x0-08618577-0-08186517)
 386

00861066067

};=y(0 2) -
081873089
Numeri al Mon
Tabular values are
Thirdorder Fourth order
Second order

0.91 0-9048375

9483714j
0.905
0-823366 0-81873089
02 0819025
more closer to exact values
values are
Fourth order
dy
Biven
Example 3. Compute y
(03) dx +y =0, +xy'
taking h=0/ using R.K
method offourth order (correct to 4 y(0)1
Solution. y'=-(y'+y) =fu,yl; 4, =0, Yo=1, h=01, deCimal:
,=02, x =03, y, =?
For Ist interval:

Yo)] = -01
k,=hf(go yo) =(0-1)(-(aot
k, =hf|*o-(0)f(005,
h 095)

=-01 [(005) (095) +095] =-0-0995

h
o*jotj4 -(09f(005,,0.95025)
=(0-1) [- (0-05 x0-95025+ 1)(0-95025) )
=-0-09953987 - 0-0995

ky =hf(o t+h, yo +k,)

= (0-1)f(O1, 0.9005) =-00982

y=1+¿-01+2(-00995) +2(-00995)-0-09821
6

y(0-1) = 0.9006
 of Ordimary
Diolfferentil Lquaton
n place (o 'o
) (0Df0 I,09000) epcat the pu
-0-0982

)f015,
=-0-0960
08S I5)
=(01)f(015, 0852%,
=-0-0962
+A, vy tk3) =(01)f(02,,0 8044)
=-0-0934

= A9006+-00982 +2x(-00960) +2hv

6

t2(-00962)
0)=08046
Aeain,starting from (X, y) in place of (o Yo) +(-00934)
=-0-0934. k, =-00902, k, =-0-0904.
k,
1
Ay =y, t(k, +2k, +-2k, +k,)
k,z-00867
6
y(03)=0-7144.
Bxample 4. Using
R.K. method of fourth order, find y(08) correct
-iecimal places if y' =y -xy (06) =1:7379 (April 1991)
Solution. Here, x =0-6, y, =:11-7379, h =0l,x=07,x=08
fuy)=y
ByR.K. method of 4th order
1
y) = Yot(k,+ 2k, + 2k, + k,) ..()

ere
k, =hf (ky. yo) =(01)f(06, 1-7379)
=(0 1) |1-7379 -(061=01378
h

(065)'|=01384
=(0-1)f(065, 1-8068)8) =(0 1)|18068
 388

(01)|1 8071 -(065)-0h94

(0D08764)
+A,) =(01) f(07,
(07')=0 386
Numerinl M.
Hence, using (
(07)=,= 7370+101378 +2(0 1384).
y(07)= 18763
To find );=y 08), we again start from

where
Now, ,=)tk, +2k, +2k, +k,l
k, =hfu,y) =(0)|
h
(X)=0|1=03M%.\816
07'
=(0 1) [1-9456- (0-75)]=0-1383
\9436,
h

= (0-1)f(075, 1-9455)

=(0-1) [1-9455 (0-75)]=0-1383

k, =hf(x, th,yj +k)
= (01)f(08, 2-0146)
=(01) (20146- (08)]=01375
Using (2),.
Y, =y (08) =18763 + (01386+ 2(1-1383) +2(1-1383) +0!:
=2-0145
Y;=y (08) = 2-0145
Example 5. Using Runge-Kutta method of fourth order, solu
dy '
dx given y (0)=lat x=02, 0-4. (MS. April '92)
(Anna Ap. 2005)
Solution. y=fk, )=y-, (Anna Nov. 2004)
Here X=0, h02, x, =02, x, =0:4,Yo =1
f(Mo Y) =f(0, 1)=1+0
k, = hf (iy, Yo) =(02) xl =02
k,=f othlot =(02)f01, 11)
 Differential Equations 389
Solutionof
Ordinary
121 -091
1-21+00
-02)/('+01)'
=0-1967213

-02r011joI97213
=(02)f(0-1, 1-0983606)
(10983606)-(001) =01967
=(0) (1-0983606)'+(001)
k,=hf(nth, n+k)
=(02)f(0-2, 1-1967)
(1967)-(02) =0-189
=(02)|96+(02)

Ay=1k, +2k, +2k,+k¡l)

-102 +2(019672) +2(1-1967)+01891)
6
=0-19598
Ay = 1 19598.
y(02) =y; =Yo+ (, yj) =(0:2, 1-19598).
from
Againto findy(04), start
Now, (1-19598)-(02)2|
=0-1891
k=if4, )=(02) (1-19598) +(02
1
, +¿k,=(02)f(03,1-2905)

(129055)?-(03) |
-02)| (129055) +(o03)
k,= (0:2)f(G3, 1-28572) =0-1793
=(02)f(04, y, +k4) =(02)f(04, 1:37528)
=0-1687

=10-1891
6
+2 (0 1795) +2(0-1793) +0-1687|
=0.!792
390

=y (0-4) =y, + Ay = 1
.. y,
Runge-Kutta method fc
1-14.
tial equations

numerically t