hi I tried to solve the 2D burger's equaion using the approximate riemann solver of roe. the coding was done in matlab. the actual solution involves a shock. I got a solution matrix will elements as NaN. which i found meand Not a Number.(indeterminate). to find the bug i solved for one time step and found the values of flux went to infinity at many points. Is this normal? Is it because of shock? And can i get a proper solution if i use flux limiters? i tried flux averaging, which failed too.

Hi,

Check your denominator in the diffusion part of the interface fluxes i.e : alpha(j)*(u_j -u_j-1)., alpha = dg/du. You might not have implemnted this step properly. Try first order initially. Dont jump to flux limiters in the first instance.

Regs, Dominic

Hi,

If it a non uniform grid check your grid metrics as well.