The detail of the Fourth order, least squares scheme
 

Dear friends:
In OpenFOAM, there is a "Fourth order, least squares scheme" for the gradient computation.

Could you please give me some references about this.
Regards

Yes. If you take field values in cell neighbours, you can fit a second order function - either for gradient evaluation or interpolation.

If you take neighbouring values AND gradients, you can fit a fourth order function.

Hope this helps,

Hrv

P. S. The volume and surface integrals are still second order - life is not perfect...

Thank you very much! Could you please give me some reference paper or book. I still cann't get it. I have goolged it but nothing found.