$EWALD group       (relevant for all-EFP runs with PBC)                         
   This group controls evaluation of the electrostatic                          
energy of EFP calculations by means of the Ewald sum                            
formulae.  This gives a more accurate evaluation of these                       
long range interactions than the minimum image convention,                      
which sums only up to a distance of one box, centered on                        
each particle.  Ewald sum formulae are not used for the                         
other, shorter range interactions in the EFP model, such as                     
exchange repulsion and polarization, which are always                           
evaluated by the minimum image convention.  This group is                       
relevant if and only if a periodic box is defined in the                        
$EFRAG input group.                                                             
IFEWLD = a flag to activate Ewald sums for electrostatics                       
         The default is .FALSE.                                                 
LEVEL  = 1 means Ewald sum charge-charge interactions only,                     
           which is the default if IFEWLD is turned on.                         
       = 2 charge-charge, charge-dipole, dipole-dipole                          
       = 3 charge-charge, charge-dipole, dipole-dipole, and                     
           charge-quadrupole terms should be Ewald summed.                      
TNFOIL = a flag to select tin foil boundary conditions,                         
         which uses a metallic continuum past the cutoffs,                      
         instead of a vacuum.  The default is .TRUE.                            
BETA   = parameter for the direct summation, in 1/Bohr.                         
         It should be 1.7/cutoff.  Cutoffs are specified                        
         in $EFRAG, with the periodic box sizes, use a                          
         cutoff in units Angstrom in this formula, as the                       
         value 1.7 includes the conversion factor.                              
         The default=0.2                                                        
KMAX   = number of reciprocal vectors in each direction.                        
         This should be kmax >= 3.2L/cutoff, where the                          
         radial cutoff, and box side L are both given in                        
         your $EFRAG.  The default=10                                           

generated on 7/7/2017