How to specify a NeumannValue in a system of non-linear partial differential equations?

I am trying to solve a nonlinear system of partial differential equations. If I assume DirichletConditions it all works well.

tmax = 100;
zmax = 1;
pDGL0=D[c[z, t], t] ==Sqrt[T[z, t]] D[c[z, t], z, z] - 100 D[c[z, t],z];
pDGL1 = D[T[z, t], t] == 0.001 D[T[z, t], z, z] - 0.01 D[T[z, t], z];
aW0 = c[z, 0] == If[0 < z <= zmax, 0, 1];
aW1 = T[z, 0] == If[0 < z < zmax, 300, 400];
rW0 = DirichletCondition[c[z, t] == 0, z == zmax];
rW1 = DirichletCondition[c[z, t] == 1, z == 0] ;
rW2 = DirichletCondition[T[z, t] == 400, z == zmax];
rW3 = DirichletCondition[T[z, t] == 400, z == 0];
dgln = {pDGL0, pDGL1, aW0, aW1, rW0, rW1, rW2, rW3};
sol = NDSolve[dgln, {c, T}, {t, 0, tmax}, {z, 0, zmax}, MaxSteps-> {50,Infinity}]


This leads to a solution like:

cLsg[z_, t_] := c[z, t] /. sol[[1]];
TLsg[z_, t_] := T[z, t] /. sol[[1]];
Plot3D[cLsg[z, t], {z, 0, zmax}, {t, 0, tmax}]
Plot3D[TLsg[z, t] - 273.15, {z, 0, zmax}, {t, 0, tmax}]


However when I try to set a NeumannValue

Derivative[1, 0][c][0, t] == 1;


at one end of the boundary namely z==0 I get an error message like

CoefficientArrays::poly: c 23536+100 c23537-c23538Sqrt[T]-NeumannValue[0.2,z==0] is not a polynomial.

and also

NDSolve::femper: PDE parsing error of {c23536+100\ c23537-c23538\ Sqrt[T]-NeumannValue[0.2,z==0],T23539+0.1\ T23540-0.1\ T23541}. Inconsistent equation dimensions.

The code creating the error message is:

tmax = 2;
zmax = 1;
pDGL0 = D[c[z, t], t]==Sqrt[T[z, t]] D[c[z, t], z, z]-100 D[c[z,t],z]+NeumannValue[1, z == 0];
pDGL1 =  D[T[z, t], t] == 0.1 D[T[z, t], z, z] - 0.1 D[T[z, t], z];
aW0 = c[z, 0] == If[0 < z <= zmax, 0, 1];
aW1 = T[z, 0] == If[0 < z < zmax, 300, 400];
rW0 = DirichletCondition[c[z, t] == 0, z == zmax] ;
rW1 = DirichletCondition[T[z, t] == 400, z == zmax];
rW2 = DirichletCondition[T[z, t] == 400, z == 0];
dgln = {pDGL0, pDGL1, aW0, aW1, rW0, rW1, rW2};
sol = NDSolve[dgln, {c, T}, {t, 0, tmax}, {z, 0, zmax}]


The problem might be that there is a coupled system of non-linear pdes. When I tried to give the NeumannValue explicitly by

Derivative[1, 0][c][0, t] == 1;


this often gave the error message that it is consistant with its inital value or it said that the number of boundary conditions is insufficient. How can I fix this problem? Is there a better way to formulate the NeumannValues?

• Welcome to Mathematica.SE! I hope you will become a regular contributor. To get started, 1) take the introductory tour now, 2) when you see good questions and answers, vote them up by clicking the gray triangles, because the credibility of the system is based on the reputation gained by users sharing their knowledge, 3) remember to accept the answer, if any, that solves your problem, by clicking the checkmark sign, and 4) give help too, by answering questions in your areas of expertise. – bbgodfrey Sep 17 '15 at 14:49
• What do you try to state with the NeumannValue? This: Derivative[1, 0][c][0, t] == 1; ? – Eisbär Sep 17 '15 at 15:14

You may do it by MethodOfLines. Try this:

 tmax = 100;
zmax = 1;
pDGL0 = D[c[z, t], t] ==
Sqrt[T[z, t]] D[c[z, t], z, z] - 100 D[c[z, t], z];
pDGL1 = D[T[z, t], t] == 0.001 D[T[z, t], z, z] - 0.01 D[T[z, t], z];
aW0 = c[z, 0] == If[0 < z <= zmax, 0, 1];
aW1 = T[z, 0] == If[0 < z < zmax, 300, 400];
rW0 = c[zmax, t] == 0;
rW1 = c[0, t] == 1;
rW2 = T[zmax, t] == 400;
rW3 = T[0, t] == 400;
nV = (D[c[z, t], z] /. z -> 0) == 1;
dgln = {pDGL0, pDGL1, aW0, aW1, rW0, rW1, rW2, rW3, nV};
sol = NDSolve[dgln, {c, T}, {t, 0, tmax}, {z, 0, zmax},
Method -> "MethodOfLines"]


There will be several warnings, but the solution, nevertheless, looks reasonably:

    cLsg[z_, t_] := c[z, t] /. sol[[1]];
TLsg[z_, t_] := T[z, t] /. sol[[1]];

Row[{Plot3D[cLsg[z, t], {z, 0, zmax}, {t, 0, tmax},
AxesLabel -> {"z", "t", "c      "}, ImageSize -> 250], Spacer[20],
Plot3D[TLsg[z, t], {z, 0, zmax}, {t, 0, tmax},
AxesLabel -> {"z", "t", "T   "}, PlotRange -> All,
ImageSize -> 250]}]


giving this:

I mean that the boundary conditions are fulfilled in the solution. You may then further try to improve the MethodOfLines by methods you may find in this tutorial.