08.04 - Runge-Kutta 2nd Order Method For ODE

RUNGE-KUTTA 2ND ORDER METHOD 08.04.
Chapter 08.04
Runge-Kutta 2nd Order Method
for
Ordinary Differential Equations
Runge-Kutta 2nd order method is a numerical technique to solve ordinary differential
equation of the form
= f ( x, y ), y (0 ) = y 0
dy
dx
Only first order ordinary differential equations can be solved by using Runge-Kutta
2nd order method. In other sections, we have discussed how Euler and Runge-Kutta
methods are used to solve higher order ordinary differential equations or coupled
(simultaneous) differential equations.
How does one write a first order differential equation in the above form?
Example 1
+ 2 y = 1.3e − x , y (0 ) = 5
dy
dx
is rewritten as
= 1.3e − x − 2 y, y (0 ) = 5
dy
dx
In this case
f ( x, y ) = 1.3e − x − 2 y
Another example is given as follows
Example 2
+ x 2 y 2 = 2 Sin (3x ), y (0) = 5
dy
ey
dx
is rewritten as
dy 2 Sin(3 x) − x 2 y 2
= , y (0) = 5
dx ey
In this case
2Sin(3x) − x 2 y 2
f ( x, y ) =
ey
Euler’s method is given by
y i +1 = y i + f ( xi , y i )h (1)
where
W:\mws\gen\08ode\mws_gen_ode_txt_runge2nd.doc
RUNGE-KUTTA 2ND ORDER METHOD 08.04.2
x0 = 0
y0 = y0
h = xi +1 − xi
To understand Runge-Kutta 2nd order method, we need to derive Euler’s method from
Taylor series.
2 3
y i +1 = y i +
dy
(xi +1 − xi ) + 1 d 2y (xi +1 − xi )2 + 1 d y3 (xi +1 − xi )3 + ...
dx xi , yi 2! dx x , y 3! dx x , y
i i i i
yi +1 = yi + f ( xi , yi ) + f ' ( xi , yi )( xi +1 − xi ) + f ' ' ( xi , yi )( xi +1 − xi ) + ... (2)

1 2 1 3
2! 3!
As you can see the first two terms of the Taylor series
y i +1 = yi + f ( xi , y i )h
are the Euler’s method and hence can be considered to be Runge-Kutta 1st order
method.
The true error in the approximation is given by
f ′( xi , yi ) 2 f ′′( xi , yi ) 3
Et = h + h + ... (3)
2! 3!
So how would a 2nd order method formula look like. It would include one more term
of the Taylor series as follows.
yi +1 = yi + f ( xi , yi )h + f ′( xi , yi )h 2
1
(4)
2!
Let us take a generic example of a first ordinary differential equation
= e − 2 x − 3 y , y (0) = 5
dy
dx
f (x, y ) = e−2 x − 3 y
Now since y is a function of x,
∂f ( x, y ) ∂f ( x, y ) dy
f ′( x, y ) = + (5)
∂x ∂y dx
∂
= (e − 2 x − 3 y ) +
∂x
[
∂ −2 x
∂y
]
(e − 3 y ) (e −2 x − 3 y )
= −2e −2 x + ( −3)(e −2 x − 3 y )
= −5e −2 x + 9 y
The 2nd order formula for the above example would be
y i +1 = y i + f (xi , y i )h + f ′(xi , y i )h 2
1
2!
yi +1 = yi + (e − 2 xi − 3 yi )h + (− 5e − 2 xi + 9 yi )h 2
1
2!
However, we already see the difficulty of having to find f ′( x, y ) in the above

method. What Runge and Kutta did was write the 2nd order method as
yi +1 = yi + (a1k1 + a 2 k 2 )h (6)
where
k1 = f ( xi , yi )
k2 = f ( xi + p1h, yi + q11k1h ) (7)
This form allows one to take advantage of the 2nd order method without having to
calculate f ′( x, y ) .
So how do we find the unknowns a1 , a2 , p1 and q11 . Without proof,
equating Equation (4) and (6) , gives three equations.
a1 + a2 = 1
1
a2 p1 =
2
1
a2 q11 =
2
Since we have 3 equations and 4 unknowns, we can assume the value of one of the
unknowns. The other three will then be determined from the three equations.
Generally the value of a2 is chosen to evaluate the other three constants. The three
1 2
values generally used for a2 are , 1 and , and are known as Heun’s Method,
2 3
Midpoint method and Ralston’s method, respectively.
Heun’s method
1
Here a2 = is chosen, giving
2
1
a1 =
2
p1 = 1
q11 = 1
resulting in
⎛1 1 ⎞
yi +1 = yi + ⎜ k1 + k2 ⎟h (8)
⎝2 2 ⎠
where
k1 = f ( x i , y i ) (9a)
k 2 = f ( xi + h, y i + k1 h ) (9b)
This method is graphically explained in Figure 1.
y Slope = f ( xi + h, yi + k1h )
yi+1, predicted
Slope = f ( xi , yi )
Average Slope =
1
[ f (xi + h, yi + k1h ) + f (xi , yi )]
2
yi
x
xi xi+1
Figure 1. Runge-Kutta 2nd order method (Heun’s method)
Midpoint method
Here a2 = 1 is chosen, giving
a1 = 0
1
p1 =
2
1
q11 =
2
resulting in
yi +1 = yi + k2h (10)
where
k1 = f ( xi , yi ) (11a)
⎛ 1 1 ⎞
k 2 = f ⎜ xi + h, yi + k1h ⎟ (11b)
⎝ 2 2 ⎠
Ralston’s method
2
Here a 2 = is chosen, giving
3
1
a1 =
3
3
p1 =
4
3
q11 =
4
resulting in
1 2
yi +1 = yi + ( k1 + k 2 )h (12)
3 3
where
k1 = f ( xi , yi ) (13a)
⎛ 3 3 ⎞
k 2 = f ⎜ x i + h , y i + k1 h ⎟ (13b)
⎝ 4 4 ⎠
Example 3
A ball at 1200K is allowed to cool down in air at an ambient temperature of 300K.
Assuming heat is lost only due to radiation, the differential equation for the
temperature of the ball is given by
dθ
= −2.2067 × 10 -12 (θ 4 − 81 × 108 )
dt
Find the temperature at t = 8 minutes using Runge-Kutta 2nd order method. Assume
a step size of h = 4 minutes.
Solution
dθ
dt
(
= −2.2067 × 10 −12 θ 4 − 81 × 10 8)
( )
f (t , θ ) = −2.2067 × 10 −12 θ 4 − 81 × 10 8
Per Heun’s method given by Equations (8) and (9)
⎛1 1 ⎞
θ i +1 = θ i + ⎜ k1 + k 2 ⎟h
⎝2 2 ⎠
k1 = f (ti , θ i )
k 2 = f (ti + h, θ i + k1h )
i = 0, t0 = 0, θ 0 = θ (0) = 1200
k1 = f (t 0 ,θ o )
= f (0,1200)
= −2.2067 × 10 −12 (1200 4 − 81 × 10 8 )
= −4.5579
k 2 = f (t 0 + h,θ 0 + k1 h )
= f (0 + 240,1200 + (− 4.5579)240)
= f (240,106.09)
= −2.2067 × 10 −12 (106.09 4 − 81 × 10 8 )
= 0.017595
⎛1 1 ⎞
θ 1 = θ 0 + ⎜ k1 + k 2 ⎟ h
⎝2 2 ⎠
⎛1 ⎞
= 1200 + ⎜ (− 4.5579 ) + (0.017595 )⎟240
1
⎝2 2 ⎠
= 1200 + (− 2.2702)240
= 655.16K
i = 1, t1 = t 0 + h = 0 + 240 = 240, θ1 = 655.16 K
k1 = f (t1 ,θ1 )
= f (240,655.16)
= −2.2067 × 10 −12 (655.16 4 − 81 × 10 8 )
= −0.38869
k 2 = f (t1 + h, θ1 + k1 h )
= f (240 + 240,655.16 + (− 0.38869)240)
= f (480,561.87 )
= −2.2067 × 10 −12 (561.87 4 − 81 × 10 8 )
= −0.20206
⎛1 1 ⎞
θ 2 = θ 1 + ⎜ k1 + k 2 ⎟ h
⎝2 2 ⎠
⎛1 ⎞
= 655.16 + ⎜ (− 0.38869 ) + (− 0.20206 )⎟ 240
1
⎝2 2 ⎠
= 655.16 + (− 0.29538)240
= 584.27 K
θ 2 = θ (480) = 584.27 K
The results from Heun’s method are compared with exact results in Figure 2.
The exact solution of the ordinary differential equation is given by the solution of a
non-linear equation as
θ − 300
0.92593 ln − 1.8519 tan −1 (0.0033333θ ) = −0.22067 × 10 −3 t − 2.9282
θ + 300
The solution to this nonlinear equation at t=480 is
θ ( 480) = 647.57 K
1200
h=120
Temperature, θ(K)
Exact
800
h=240
400
h=480
0
0 100 200 300 400 500
-400
Time, t(sec)
Figure 2. Heun’s method results for different step sizes
Using smaller step size would increases the accuracy of the result as given in Table 1
and Figure 3 below.
Table 1. Effect of step size for Heun’s method

Step size, h θ (480) Et ∈t %
480 -393.87 1041.4 160.82
240 584.27 63.304 9.7756
120 651.35 -3.7762 0.58313
60 649.91 -2.3406 0.36145
30 648.21 -0.63219 0.097625
800
Temperature, θ(480)
600
400
200
0
0 100 200 300 400 500
-200 Step size, h
-400
Figure 3. Effect of step size in Heun’s method
In Table 2, the Euler’s method and Runge-Kutta 2nd order method results are shown
as a function of step size,
Table 2. Comparison of Euler and the Runge-Kutta methods

Step size, θ ( 480)
h Euler Heun Midpoint Ralston
480 -987.84 -393.87 1208.4 449.78
240 110.32 584.27 976.87 690.01
120 546.77 651.35 690.20 667.71
60 614.97 649.91 654.85 652.25
30 632.77 648.21 649.02 648.61
while in Figure 4, the comparison is shown over the range of time.
1200
1100
θ(K)
1000 Midpoint
Temperature,
Ralston
900
Heun
800
700 Analytical
600
Euler
500
0 100 200 300 400 500 600
Time, t (sec)
Figure 4. Comparison of Euler and Runge Kutta methods with exact results
over time.
How do these three methods compare with results obtained if we found f ′( x, y )

directly?
Of course, we know that since we are including first three terms in the series, if the
solution is a polynomial of order two or less (that is, quadratic, linear or constant),
any of the three methods are exact. But for any other case the results will be
different.
Let us take the example of
= e − 2 x − 3 y , y (0 ) = 5 .
dy
dx
If we directly find the f ′( x, y ) , the first three terms of Taylor series gives
yi +1 = yi + f ( xi , yi )h + f ′( xi , yi )h 2
1
2!
where
f (x, y ) = e−2 x − 3 y
f ′(x, y ) = −5e−2 x + 9 y
For a step size of h = 0.2 , using Heun’s method, we find
y (0.6) = 1.0930
The exact solution

y ( x ) = e−2 x + 4e−3 x
gives
y (0.6) = e−2(0.6 ) + 4e−3(0.6 )
= 0.96239
Then the absolute relative true error is
0.96239 − 1.0930
∈t = × 100
0.96239
= 13.571%
For the same problem, the results from the Euler and the three Runge-Kutta method
are given below
Table 3. Comparison of Euler’s and Runge-Kutta 2nd order methods

y(0.6)
Exact Euler Direct 2nd Heun Midpoint Ralston
Value 0.96239 0.4955 1.0930 1.1012 1.0974 1.0994
∈t % 48.514 13.571 14.423 14.029 14.236

08.04 - Runge-Kutta 2nd Order Method For ODE

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

08.04 - Runge-Kutta 2nd Order Method For ODE

Uploaded by

Copyright:

Available Formats

RUNGE-KUTTA 2ND ORDER METHOD 08.04.

yi +1 = yi + f ( xi , yi ) + f ' ( xi , yi )( xi +1 − xi ) + f ' ' ( xi , yi )( xi +1 − xi ) + ... (2)

However, we already see the difficulty of having to find f ′( x, y ) in the above

Figure 1. Runge-Kutta 2nd order method (Heun’s method)

Figure 2. Heun’s method results for different step sizes

Table 1. Effect of step size for Heun’s method

Figure 3. Effect of step size in Heun’s method

Table 2. Comparison of Euler and the Runge-Kutta methods

while in Figure 4, the comparison is shown over the range of time.

How do these three methods compare with results obtained if we found f ′( x, y )

The exact solution

Table 3. Comparison of Euler’s and Runge-Kutta 2nd order methods

You might also like