Author |
Message |
Sudhakar Jaganathan (sudhakar)
New member Username: sudhakar
Post Number: 1 Registered: 12-2007
| Posted on Monday, December 17, 2007 - 08:58 am: | |
Hello Everyone i am new to FLEX so this might be primitive, how can i implement symmetry or periodic Bc's in Flex.Follwing is equations equations dx(c*s^2*dx(s))= dt(s) boundaries region 1 start(0,0) value(s)=0.92 line to (w,0) value(s)=0 line to (w,5) mesh_spacing=1 value(s)=0 line to (0,5) value(s)=0 mesh_spacing=100 line to close I want to specify fluid concentration at bottom which intrudes into domain with symmetry on sides |
Robert G. Nelson (rgnelson)
Moderator Username: rgnelson
Post Number: 1025 Registered: 06-2003
| Posted on Monday, December 17, 2007 - 03:06 pm: | |
1. Your equation diffuses only in X. Symmetry on the X boundaries means nothing goes anywhere. If you want to diffuse in Y (having given a boundary difference in the Y direction), you need a Y-diffusion term. Div(c*s^2*grad(s)) will diffuse both ways. 2. Simple Natural(S)=0 on the sides provides mirror conditions, resulting in a system that is 1D in Y. Unless there is some asymmetry in things you add later, the solution will be independent of X. 3. You can use Periodic boundary conditions in X. See "Periodic Boundary conditions" in the Help Index.
|
Sudhakar Jaganathan (sudhakar)
New member Username: sudhakar
Post Number: 2 Registered: 12-2007
| Posted on Monday, December 17, 2007 - 05:33 pm: | |
Thank You Dr.Nelson I included y diffusive term and i am able to put periodic bc's,manual was helpful. I have one more question in the reactangle , if i specify s=1 at bottom, periodic bC's at right and left sides and If i put natural(s)=0 at top what does this mean in my case. I got this wonderful software only two days ago i am really enjoying working with it,programming is so simple even a non-programmer like me could use it without any difficulty. |
Robert G. Nelson (rgnelson)
Moderator Username: rgnelson
Post Number: 1027 Registered: 06-2003
| Posted on Monday, December 17, 2007 - 10:21 pm: | |
Uniform source at the bottom and periodic (or, equivalently, natural(s)=0) on the sides means the problem is really one dimensional in y, not two dimensional in X,Y. Natural(s)=0 on the top is a reflective (zero flux) boundary condition. So S will diffuse from bottom to top, reflect, and fill up the box to a uniform s=1.
|
Sudhakar Jaganathan (sudhakar)
Junior Member Username: sudhakar
Post Number: 3 Registered: 12-2007
| Posted on Monday, December 17, 2007 - 11:40 pm: | |
Thankyou again, probelm i am simulating is similar to infiltration into soil, where bottom end is saturated at s=1 and other end open, infiltration takes place due to a constant capillary force specified as 'c' and fills up the entire domain to s=1. does the zero flux bc's i specified at other end appropriate |
Robert G. Nelson (rgnelson)
Moderator Username: rgnelson
Post Number: 1028 Registered: 06-2003
| Posted on Tuesday, December 18, 2007 - 03:13 pm: | |
FlexPDE is a general purpose tool for creating solutions of systems of Partial Differential Equations. PDE Solutions Inc cannot take the responsibility of deciding whether a particular system of equations and boundary conditions is the "correct" model for any given user's application. This is a decision the user must make for himself. We can advise on what various FlexPDE constructs mean or do, and we can point out facilities that exist in FlexPDE for various uses. We cannot tell you if your system is the "right" one.
|