Hi Ionut, I'm not sure why your script is not converging. However, I do have some queries below.
On Sun, Feb 8, 2009 at 4:17 PM, Ionut Vancea <[email protected]> wrote: >> fSecondDerivative = -exp(-faceVar) + (3 * b / faceVar**4) >> DiffusionTerm(coeff=fSecondDerivative) Have you tried the above? This seems like the most sensible approach for the second order part of the equation. > fDerivative = exp(-var)-(b / var**3) > #fSecondDerivative = -exp(-faceVar) + (3 * b / faceVar**4) > > diffTerm4 = - DiffusionTerm((h**3, 1)) > eq = TransientTerm() == -DiffusionTerm((h**3, 1)) + > (fDerivative.getFaceGrad() * h**3).getDivergence() > #eq = TransientTerm() == -DiffusionTerm((h**3, 1)) + > DiffusionTerm(coeff=fSecondDerivative) What happens when you uncomment this? Actually, what is h? Shouldn't "h" be "var"? -- Daniel Wheeler
