Scribd
Upload
24 views3 pages

Sol 13

nn

Uploaded by

shishantvimal999
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
24 views3 pages

Sol 13

nn

Uploaded by

shishantvimal999
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

SOLUTIONS OF SELECTED PROBLEMS OF EXERCISE 13

ANS 13.7
Euler’s explicit scheme is given by
U s 1  U s  t U s (0)

Let us assume that the vector of primary nodal variables is given by

U   { A}e ct (1)

where {A} is a vector of constants and c is a positive constant. Thus

U   c{ A}e  ct


 c{U } (2)

Therefore

{U }s  c{U }s (3)

Substituting the relation given in Eq. (3) in Eq. (0), we obtain

U s 1  U s  ct U s  (1  ct ) U s (4)

From Eq. (4), it can be understood that if ct is more than 1, then the magnitudes of primary

nodal variables at the (s+1)th time step will be more than those at sth time step. Thus, the

magnitudes of primary nodal variables will increase with time, which is contrary to the actual

behaviour of exponential decaying function [Eq. (1)]. In this situation, the scheme is said to

become unstable because the error keeps on increasing with each time step. On the other hand, if

ct is less than 1, then the magnitudes of primary nodal variables at the (s+1)th time step will be

less than those at sth time step. Thus, the magnitude of primary nodal variables will decrease with

time, this is consistent with the actual behaviour. Thus, the error remains bounded with time and

the scheme is called stable. Hence, this scheme will be stable below a certain value of time step

t , and will be unstable above that value. Such a scheme is called conditionally stable.

ANS 13.9

Given <(1/2), we have to show that


2
t 
(1  2  )

As per the scheme,


U s 1  U s  t 1    U s   U s 1  (1)

Let us assume that the vector of primary nodal variables is given by

U   { A}e ct (2)

where {A} is a vector of constants and c is a positive constant. Thus

U   c{ A}e  ct


 c{U } (3)

Therefore

{U }s  c{U }s ; {U }s 1  c{U }s 1 ;


Substituting the above in Eq. (1):
U s 1  U s  t   1    c U s   c U s 1 
or
1   ct U s 1  1  1    ct  U s
Thus,
1  1    c t 
U s 1   U s
1   ct 
For stability,
1  1    c t 
1
1   ct 
or
1  1    c t   1   c t 
This implies that
1  1    c t   1   ct  and - 1  1    ct   1   ct 
The first statement is obviously satisfied. The second implies
2
t 
(1  2  )c

Now, putting (2) and (3) in


 M U   [ K ]{U }  0
we get
c  M  A  [ K ]{ A}  0
Thus c is one eigenvalue. As we are finding out the critical time, in place of c, we can
take maximum eigenvalue. Thus,
2
t 
(1  2  )

You might also like