[mandeepcfd]

Dear all,
i am trying to write udf for grid motion.i need to know the component of the concentration gradient in the direction normal to the moving boundary. to take into account the mesh non orthogonality fluent computes the gradient component in the normal direction as

((phi1-phi0)/ds)*(A.A)/A.es + cross diffusion term

the second term can be evaluated using a macro " boundary secondary gradient source".
my doubt is with first term
(A.A)/A.es , ds can be found using macro "boundary face geometry".
but phi 1 for a boundary face is not defined. for internal face phi 1 ,phi 0 represent the value of scalar (conc in this case)at cell on either side of face.
SHOULD I USE THE VALUE AT FACE CENTRE FOR PHI1?
ALSO IN THAT CASE WHETHER I SHOULD USE PHIFACE-PHI0 OR PHI0-PHIFACE
ANY KIND OF HELP IS HIGHLY APPRECIATED
thanks