1D diffusion-reaction model across two-layered slab using pdepe

Question

Giuseppe Pontrelli 2014-9-17

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/155142-1d-diffusion-reaction-model-across-two-layered-slab-using-pdepe

评论： Giuseppe Pontrelli 2014-9-29

matlab_cent.pdf

I have a diffusion-reaction model across a two-layered slab (see attachment). Two kind of problems arise when I solve with MATLAB:

1) In each layer, we have a coupling between an ODE and a PDE: How do you cope this with MATLAB? I use PDEPE to solve the system but the BC's for one variable are lacking. Any idea?

2) Because of the interface conditions, the solution may be not continuous in the whole domain, so I believe I cannot solve in one domain with PDEPE. An idea is to solve in two separate domains with PDEPE in an iterative way, imposing conditions coming from the other domain. What do you think?

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Bill Greene 2014-9-28

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/155142-1d-diffusion-reaction-model-across-two-layered-slab-using-pdepe#answer_153172

I am doubtful that pdepe can be used to solve this system.

I suggest using finite differences to discretize in the spatial direction and then use ode15s to solve the resulting system of ODE.

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Giuseppe Pontrelli 2014-9-29

You are perfectly right, and I will do as you suggest. However, I want (to avoid the space discretization and solution of a linear system) to try first an iterative procedure as follows:

AT EACH TIME STEP:

0) make a guess of the solution in domain 1 and 2

1) solve the continuous PDE with pdepe in domain 1 imposing as BC the values coming from 2) (or 0 at the first iteration) 2) solve the continuous PDE with pdepe in domain 2 imposing as BC the values coming from 1)

iterate 1-2 till (hopefully) convergence. do you think it is feasible?

请先登录，再进行评论。