CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM Programming & Development (https://www.cfd-online.com/Forums/openfoam-programming-development/)
-   -   why div(grad) is not equal to laplacian (https://www.cfd-online.com/Forums/openfoam-programming-development/187977-why-div-grad-not-equal-laplacian.html)

JasonWang3 May 19, 2017 05:57
why div(grad) is not equal to laplacian
 
Hi all

I did a test about div(grad(Ux)) and laplacian(Ux), but got different results. Why did it happen?

However, div(mesh.Sf()*snGrad(Ux)) is equal to laplacian(Ux).

I am looking for a way to calculate the second derivatives of Ux dux2dy2. I am not sure grad(grad(Ux).component(vector::Y))..component(vec tor::Y) is right or not.

JasonWang3 May 19, 2017 09:15
I test using icoFoam and 'cavity' in the tutorial.

Did anyone meet the same problem?

randolph July 2, 2018 12:46
Quote:
Originally Posted by JasonWang3 (Post 649563)
I test using icoFoam and 'cavity' in the tutorial.

Did anyone meet the same problem?
div(grad) is not equal to Laplacian in the discrete sense, although they are equal in the continuous domain. You can show this by comparing the D2 matrix with the D1*D1 matrix. I know this is true in FDM. I think also apply to FVM.


All times are GMT -4. The time now is 17:13.