explanation on boundary condition matrix

where can i find an explanation about boundary condition matrix? especially what does each row in this matrix stands for. the documantation provided in the matlab help is not very clear.

thank you.

Answer 1

[Bill Greene]

Have you looked at this page?

http://www.mathworks.com/help/pde/ug/assemb.html

There is a general description of the matrix format along with two examples.

If you are defining boundary conditions in a MATLAB script, my own preference (and recommendation) is to write a "boundary file" (user-written MATLAB function) to define your boundary conditions instead of trying to create a boundary matrix. I find writing a boundary condition function to be much more straightforward than writing a boundary matrix. The approach is documented here:

http://www.mathworks.com/help/pde/ug/boundary-conditions-for-scalar-pde.html

Bill

[Dany]

still not clear enough, although the "pdebound" might work. but now a new question came up. when you use the "pdebound" function you get 4 matrices at the end:

[qmatrix,gmatrix,hmatrix,rmatrix] = pdebound(p,e,u,time)

but when you use the "assempde" function you have to use one boundary matrix. how does those 4 matrices replace/create the one boundary matrix needed for the "assempde" function?

thank you.

[Bill Greene]

The first argument to assempde (or pdenonlin, parabolic, etc) can be a function handle. For example:

b = @pdebound;
u=assempde(b, ...);

(Note, the function doesn't have to be named pdebound for this to work.)

Bill

[Florian Baumgartner]

When I have only Dirichlet boundary conditions, do I have to gibe "assempde" 4 matrixes (q,g,h,r) or just 2 matrixes (h,r) with the command

if true
  b = @pdebund
  u = assempde(b, ...)
end