eigenvalue problem Log Out | Topics | Search
Moderators | Register | Edit Profile

FlexPDE User's Forum » User Postings » eigenvalue problem « Previous Next »

Author Message
Top of pagePrevious messageNext messageBottom of page Link to this message

Fernando Buezas (fernandob)
Member
Username: fernandob

Post Number: 4
Registered: 02-2007
Posted on Monday, March 12, 2007 - 04:45 pm:   

Dear Nelson:
please
doest exist anyway to solve eigenvalues's problem like L(u) + f(lambda)u = 0 ?

For example f(lambda)=lambda^2
thankyou
Top of pagePrevious messageNext messageBottom of page Link to this message

Robert G. Nelson (rgnelson)
Moderator
Username: rgnelson

Post Number: 788
Registered: 06-2003
Posted on Monday, March 12, 2007 - 05:24 pm:   

FlexPDE solves the system A*u + Lambda*B*u = 0
But you can subsequently compute
Lambda1 = sqrt(Lambda)
to solve the effective system
A*U + Lambda1^2*B*u.
Top of pagePrevious messageNext messageBottom of page Link to this message

abbasali
Member
Username: abbasali

Post Number: 7
Registered: 04-2010
Posted on Sunday, October 17, 2010 - 08:54 am:   

Hi dear Dr. Nelson

I am a mathematics student and I have derived some abstract mathematics result about an eigenvalue problem. A simple form of this problem is as follows

TITLE 'nonlinear eigenvalue problem'

select
modes=1
ERRLIM =.001
VARIABLES
u
DEFINITIONS
v
r1=3
r=10
e1=3
e2=0
k1=1* USTEP(r1-(x-e1)^2-(y-e1)^2 )
k2=1*USTEP(10-(x-e2)^2-(y-e2)^2 )




EQUATIONS

Div(grad(u))-v*u=(-k1*lambda*u)-((k2-k1)*(lambda^2)*u)



BOUNDARIES

REGION 1 'blob1' { the outter blob }
v=.7
START(r+e1,0)
VALUE(u)=0


ARC(CENTER=e1,e1) ANGLE=360

REGION 2 'blob2' { the embedded blob }
v=.1
START(r1+e2,0)

ARC(CENTER=e2,e2) ANGLE=360

PLOTS
surface(u) painted
SUMMARY
REPORT(INTEGRAL(u^2,'blob2'))
REPORT(INTEGRAL((k2-k1)*u^2,'blob1'))
report(lambda)
END


I would like to know is this type of equations which depends linearly on the eigenparameter in a subset of the domain and quadratically elsewhere , after some simplifications, models any physical phenomenon?

regards.
Top of pagePrevious messageNext messageBottom of page Link to this message

rgnelson
Moderator
Username: rgnelson

Post Number: 1416
Registered: 06-2003
Posted on Wednesday, October 20, 2010 - 03:27 pm:   

FlexPDE does not know how to do nonlinear eigenvalue problems. It solves the system A*u - Lambda*B*u = 0

In your case, the Lambda^2 will be treated as if it were Lambda.

In Buezas' system, there was only a single reference to lambda^2, so it could be replaced by a redefined lambda1.

Since your system includes both first and second powers, this kind of redefinition will not reduce the system to a single power of lambda.

There may be ways to redefine the system to do what you want, but I don't know what they are right now ....

Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action:

Topics | Last Day | Last Week | Tree View | Search | Help/Instructions | Program Credits Administration