How to solve error in pdepe

I'm using pdepe to solve a parabolic pde in (x,t). I'm getting an error for times bigger than 0.3, though it works fine for smaller values. Is there a way to fix this?

function [c, b, a, D] = reac1(r, t, C, DCDr)

n = 1; 

D_0 = 1; 

C_0 = 100;

c = 1;

b = D_0*((C/C_0)^n)*DCDr;

a = -C*(1-C); %reaction term f(C)

%-----------------------

function value = ic2(r)

value = dirac_approx(r);

%-------------------------

function [d] = dirac_approx(x)

epsilon = 0.01;

d = epsilon^(-1)*max(1 - abs(x/epsilon), 0);

%--------------------------

function [pl, ql, pr, qr] = bc2(rl, Cl, rr, Cr, t)

pl = 0;

ql = 1;

pr = 0;

qr = 1;

%------------------------

clear all

m = 0; 

r = linspace(0, 0.2, 100);  

t = linspace(0, 0.4, 100);

C = pdepe(m, @reac1, @ic2, @bc2, r, t);

Warning: Failure at t=3.243157e-01.  Unable to meet integration tolerances without reducing the step size below the
smallest value allowed (8.881784e-16) at time t. 
> In ode15s (line 730)
  In pdepe (line 289)
  In reaction1 (line 9) 
Warning: Time integration has failed. Solution is available at requested time points up to t=3.232323e-01. 
> In pdepe (line 303)
  In reaction1 (line 9)

Thanks!

[Torsten]

Without seeing your code, the only thing we can say is: Check your equations and their implementation in pdepe.

Best wishes

Torsten.

[María Jesús]

I've edited the code into the question

[Torsten]

You missed to include "dirac_approx".

Best wishes

Torsten.

[María Jesús]

Sorry; it is there now.

Answer 1

[Torsten]

More and more of C is produced within the domain according to the sourceterm -C*(1-C).

This causes the solution C to become larger and larger and in the end approach infinity as t->00.

So back to my first advice: Check your model for physical sensefulness.

Best wishes

Torsten.