RMAX, RDEP, QCUT, QGAM tags

$$\begin{eqnarray*} RMAX &=& \mbox{maximum radius for non local projection operato... ...mization} \\ QGAM &=& \mbox{high q-value for optimization } \\ \end{eqnarray*}$$

These tags control the real space optimization of the pseudopotentials [47], and the extend of the non local projection operators. If no real space optimization is selected QCUT must be zero. The default values are:

RMAX	=	RCORE
RDEP	=	RCORE
QCUT	=	-1, automatic real space optimization
		default cutoff
QGAM	=	2*QCUT

If real space optimization should be done, QCUT must be set to the energy cutoff, which will be used in VAMP. Anyway here QCUT has to be supplied in 1/a.u. (i.e. as a inverse length) and can be calculated from the cutoff energy using the formula

$$\sqrt { E_{\rm cut} /13.6058)}.$$

If any wrap around errors are omitted in VAMP, QGAM can be 3*QCUT, but if the 3/4 rule is used for setting up the FFT meshes (see section 8.4) QGAM must be 2*QCUT. To get accurate real space projection operators RMAX has to be somewhat large than RCORE, usually 1.25*RCORE is sufficient. After the optimization the projection operators have been changed between QCUT and QGAM, the new projection operators are written to the file POTCAR in real and reciprocal space. This means that slightly different results might be obtained if the real space optimization has been done even if the projections are evaluated in VAMP in real space (LREAL=.F.). If the unmodified reciprocal projection operator should be written to POTCAR set QCUT to a negative value.

Finally there is a default optimization build into fourpot3, which can be selected by QCUT=-1. In this case the pseudopotential is optimized for the 3/4 FFT meshes, QCUT is set according to the default cutoff ENMAX and the RMAX is set to RCORE*1.3.

For further reading we refer to [47].