Second Fick's law with nonconstant diffusion coefficient

Click For Summary
SUMMARY

The discussion focuses on solving a modified version of Second Fick's law in nondimensional spherical coordinates, where the diffusion coefficient varies with time and radius. The equation presented is dC/dt = (1/rho^2)*d/d(rho)(rho^2*D(rho,t)*dC/d(rho)). Participants concluded that an analytic solution is not feasible, and suggested using the method of lines (MOL) to convert the equation into a set of ordinary differential equations (ODEs) for numerical solution. FORTRAN was recommended as a suitable programming language for implementing the numerical solution using an automatic stiff integration package.

PREREQUISITES
  • Understanding of Second Fick's law and its applications
  • Knowledge of partial differential equations (PDEs)
  • Familiarity with the method of lines (MOL)
  • Proficiency in FORTRAN programming language
NEXT STEPS
  • Research numerical methods for solving partial differential equations (PDEs)
  • Learn about the method of lines (MOL) for converting PDEs to ODEs
  • Explore automatic stiff integration packages available in FORTRAN
  • Investigate alternative numerical tools for PDE solutions, such as MATLAB or Python libraries
USEFUL FOR

This discussion is beneficial for mathematicians, physicists, and engineers working with diffusion processes, as well as researchers seeking numerical methods for solving complex partial differential equations.

mabiondi
Messages
2
Reaction score
0
Hello everybody.
I should solve a modified version of second Fick's law in nondimensional spherical coordinates; t is the time and rho [0,1] the nondimensional radius. In this equation the diffusion coefficient is vraiable with t and rho.

The equation is the following:

dC/dt = (1/rho^2)*d/d(rho)(rho^2*D(rho,t)*dC/d(rho))

Initial conditions:
C(0,rho) = a*rho^4 + b*rho^2 -(a + b);

Boundary conditions:
dC(t,0)/d(rho) = 0
C(t,1) = 0

Is it possible to solve such an equation by Laplace transforms?
Alternatively, is there a user friendly tool for numerical solution of a PDE like this?

Thanks in advance, cheers.
Marco
 
Engineering news on Phys.org
I don't think an analytic solution exists. It would have to be solved numerically. I would just convert it to a set of ODE's using the method of lines, and then solve the equations numerically in FORTRAN using an automatic (stiff) integration package.

Chet
 
  • Like
Likes   Reactions: mabiondi
Thanks a lot for the suggestion, I really appreciate it. I'll try to develop a routine using MOL.
Cheers,
Marco
 

Similar threads

  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 1 ·
Replies
1
Views
609
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 29 ·
Replies
29
Views
2K
  • · Replies 0 ·
Replies
0
Views
1K
Replies
3
Views
2K
  • · Replies 6 ·
Replies
6
Views
2K
  • · Replies 11 ·
Replies
11
Views
4K
Replies
0
Views
1K