CFD Online Discussion Forums - terms with the second derivative

Dear Foamers,
I need to discretize/calculate terms which contain second derivative (not Laplacian), as follows:
1.
$\frac{\partial }{\partial y} \left( \frac{\partial f}{\partial x} \right)$
2.
$\frac{\partial }{\partial x} \left( \frac{\partial f}{\partial x} \right) - \frac{\partial }{\partial y} \left( \frac{\partial f}{\partial y} \right)$

3.
and non-linear term with first derivative:
$\left( \frac{\partial f}{\partial x} \right)^2 - \left( \frac{\partial f}{\partial y} \right)^2$

For now the only way which works for me is to calculate twice the gradient of the function:

HTML Code:

gradf = fvc::grad(f); \\ it gives (f_x, f_y)

gradgradf0 = fvc::grad(gradf.component(0)); \\it gives (f_xx, f_yx)

gradgradf1 = fvc::grad(gradf.component(1)); \\it gives (f_xy, f_yy)

then term 1. is:

HTML Code:

gradgradf1.component(1); \\ f_xy

term 2. is:

HTML Code:

gradgradf0.component(0) - gradgradf1.component(1); \\ (f_xx - f_yy)

term 3. is:

HTML Code:

gradf.component(0)*gradf.component(0) - gradf.component(1)*gradf.component(1); \\ (f_x)^2- (f_y)^2

but unfortunately second derivative calculated as above gives me non-smooth results...

Is there any other/better way to compute terms like the one above ?

Thanks
ZZM