Hi Himalaya Singh,

The problem is that the initial T is zero, so having the conductivity equal to 2*T leads to a zero (singular) matrix and nothing can be computed. You seem to be setting the initial conditions correctly, but those are not used if pb.set_solver(nls) (nls = Newton solver) is used - you would need a time-stepping solver instead.

Maybe you can just use 2*T+ "a small number" to make the problem regular?

r.

On 10/14/23 08:14, Himalaya singh wrote:

Hello, sfepy community

I am new to this tool, and I have been exploring it for some time. I got some issues related to Nonlinear diffusion's equation problem.

It was working with 1+2*T**4 as in example however changing the non linear term to 2*T is not working and tried other functions also.

As I took 100mm*100mm domain and left boundary I fixed it to 25 degree celsius(gamma1) and Right boundary(gamma2) as 500 degree celcius.

Initial condition is taken as 25 degree celcius I took a nonlinear function as 2T, I tried a lot, however the solution was reaching to infinity, I mean it was showing (sfepy: linesearch: iter 2, (nan < nan) (new ls: 1.000000e-01)).

I attached the file below For Your Reference.

Any lead or help will be highly appreciated Thank you for your help!

Himalaya singh

Machine Design Department

IIT Madras