[sehli]

Hello Everyone,
I have a question about the calculation of normalised residuals in OpenFOAM. What are the advantages in comparison to other residual types (e.g. maximum residuals)? You can find in the code:

// --- Calculate normalisation factor
normFactor = gSum(mag(wA - pA) + mag(source_ - pA)) + matrix_.small_;
initialResidual() = gSumMag(b - Ax)/normFactor;

Am I right, information about very few high residual values in the domain get lost through the summation over all cells?

Thanks