Author |
Message |
Harvey (jones)
New member Username: jones
Post Number: 2 Registered: 06-2006
| Posted on Tuesday, November 14, 2006 - 01:06 am: | |
We are using Flexpde version 5.0.10 3D. We are using this program to investigate the time dependent behavior of two coupled PDEs in one-dimension, We want to use periodic boundary conditions. If w is the period then can we define the region as Region 1 Start(-10) periodic(-x-w) line to (10) periodic(10+w) This doesn't work. Can you please tell us how to incorporate a periodic boundary condition into our problem. Thank you. |
Richard Williams (rew1000)
New member Username: rew1000
Post Number: 1 Registered: 01-2007
| Posted on Friday, January 19, 2007 - 01:52 pm: | |
I'm trying to do something very similar - can FlexPDE handle periodic boundary conditions in 1D? Thanks... |
Robert G. Nelson (rgnelson)
Moderator Username: rgnelson
Post Number: 739 Registered: 06-2003
| Posted on Monday, January 22, 2007 - 12:52 am: | |
We overlooked periodicity when implementing the adaptation to 1D geometry. Sorry. I will put it on the development list. Since 1D is less demanding of computer resources than higher dimensions, perhaps you could fake it by a chain of repeated units with imbedded boundary condition.
|
Nadeem Ansari (nadeem)
Member Username: nadeem
Post Number: 5 Registered: 11-2006
| Posted on Tuesday, March 06, 2007 - 11:19 pm: | |
Dear Dr. Nelson In the forum there was some discussion on periodic boundary conditions in 1-D. You mentioned: "Since 1D is less demanding of computer resources than higher dimensions, perhaps you could fake it by a chain of repeated units with imbedded boundary condition. " Could you please clarify what you mean by "a chain of repeated units with imbedded boundary condition. " |
Robert G. Nelson (rgnelson)
Moderator Username: rgnelson
Post Number: 782 Registered: 06-2003
| Posted on Wednesday, March 07, 2007 - 06:32 pm: | |
My phrase "with imbedded boundary condition" was an unfortunate lapse. What I mean is that periodic boundary conditions simulate an infinite assembly of like structures, cheek-by-jowl. So one approximation to an infinite assembly would be a "very large" assembly, side-by-side. The interior members of this set should give a solution close to the infinitely periodic one. Since 1D is not very demanding of resources, this "large assembly" should not stress the computer's capability. We will try to get 1D periodicity implemented in a maintenance release.
|