Job Steps
In DMAREL - Default Parameter Changes
All of these directives are optional. They may be subdivided into three classes.
| Printed output | DUMP, PRIN, LATT, BASI, PLUT, MINI, BOND, TORS, MOLE, GEOM |
| Program control | IREC, ACCU, ACCM, DBAS, CHGC, SCAL, CLUS, SYMM, FDAT, ZVAL |
| Additional data file input/output | DBAS |
Printed Output
| Operand type | I |
| Operand name | IDUMP |
This is used in conjunction with the PRIN directive discussed below. IDUMP may have values 1 or 2. IDUMP = 1 produces output which may be useful if an error is suspected in the input dataset. IDUMP = 2 produces output useful for diagnostic purposes only.
PRIN
| Operand types | A | I | [ A | I | ... | ...] |
| Operand names | APRINT | MPRINT |
The APRINT name defines certain sections of the program for which additional printed output is required. The integers MPRINT define exactly what is printed. There are currently 17 print options for DMAREL. Each of these will now be described in detail.
LATT
-
- 0
- No output
- > 0
- The reciprocal lattice vectors are printed in the output.
-
- 0
- Print:
- Basis species as input
- Origin of the basis
- > 0
- In addition,
print:
- The basis species after shift of origin
- The basis species after sorting into internal program order
- < 0
- Suppress printing of the basis species
- MPRINT is a 5
digit integer, with each digit controlling a different section of the
perfect lattice calculation. Numbering the digits from the right (units
digit) to left the digits have the following values and meanings.
- Digit 1
-
- 0
- No output
- 1
- List the contributions to the lattice energy and the atom forces before the first iteration and after the last iteration.
- 2
- List the contributions to the lattice energy after each iteration.
- Digit 2
- As for digit 1 but listing the elastic constant tensors.
- Digit 3
-
- 0
- No output
- Digit 4
-
- 0
- No output
- 1
- List the force constant matrix and gradient vector before the first cycle and after the last cycle.
- 2
- List the force constant matrix and gradient vector on every cycle of the perfect lattice iteration.
- 3
- List the gradient vector only before the first cycle and after the last cycle
- 4
- List the gradient vector only on every cycle
they must be multiplied by
14.3997584/RLAT3, where RLAT is the lattice constant.
The gradients must be scaled by
14.3997584/RLAT2. - Digit 5
-
- 0
- No output
- 1
- Currently not used, formerly used to print the Madelung and short range energies of each particle. (Use CCLS instead)
PLUT
MINI
-
- 0
- No output
- 1
- Print information about progress of the minimisation.
BOND
- This APRINT option
is retained for future use, it is only relevant for a system with three-body
force constants.
- 0
- No output
- 1
- Print the three-body bonds found in the unit cell.
TORS
- This APRINT option
is retained for future use, it is only relevant for a system with torsion
force constants.
- 0
- Do not print the torsion angles found in the unit cell.
- 1
- Print the torsion angle information
MOLE
- This APRINT option
is used in DMAREL to control output from the MOLE
directive.
- 0
- Do not print molecular bonding information.
- 1
- Print molecular bonding information
GEOM
-
- 0
- No output
- 1
- Print molecular local axis information
Program Control
| Operand type | I |
| Operand name | IRECL |
The input processor reads 80 characters per card image by default. IRECL should be less than 80, and gives the number of characters read in per record.
| Operand type | F |
| Operand name | ACCLAT |
ACCLAT is the limit below which numbers are taken as being equal. The default set in the program is 1.0E-06. ACCLAT is also used at various other points in the program for testing small quantities.
| Operand type | F |
| Operand Name | ACCMAD |
This directive changes the accuracy to which the Ewald sum is carried out. The default is 1 part in 10**6 (Default value of ACCMAD = 1000000.0). Setting a larger value of ACCMAD may change the perfect lattice properties slightly. It should also be noted that a larger value of ACCMAD will set a larger lattice sum cutoff (See BASI directive [BASIS ATOM INPUT]).
| Operands | None |
The programs checks that the unit cell is neutral and will stop if this is not so. Including the directive CHGC forces the program to continue with a non-neutral cell. DMAs generated by SCF programs often have a small residual charge, so this directive should be included.
SCAL
| Operand type | F |
| Operand name | SCALQQ |
Scaling factor for DMA charges, when it can be overestimated if DMA derived at SCF level. The scaling is used to approximate the effect of electron correlation on the charge density.
CLUS
| Operands | None |
Cluster calculation - has not been used for a long time. Email M. Leslie for details.
SYMM
| Operands | None |
Switch on symmetry adapted calculation - default from NEIGHBOURS is on.
FDAT
| Operands | None |
Read FDAT input file from fort.20 if SYMM is switched off.
ZVAL
| Operand type | F |
| Operand name | None |
Read in number of molecules per cell if needed.
Additional Data File Input/Output
This table gives a list of the additional FORTRAN unit numbers used by DMAREL.|
Filename
|
INPUT/OUTPUT
|
USE
|
|
filename.dmain
|
input
|
Input
file for setting up minimisation, directives and printing options,
created by NEIGHBOURS
|
|
fort.20
|
input
|
Created
by NEIGHBOURS, must exist since it is a symmetry information file
for the minimisation
|
|
fort.8 |
output
|
Lattice
vectors and molecule coordinates of minimised structure
|
|
fort.12
|
output
|
summary
file of lattice parameters and energies
|
|
fort.13
|
output
|
an
spf file for minimised crystal structure
|
|
fort.14
|
output
|
empty
file, please ignore
|
|
fort.15
|
output
|
an
FDAT file for minimised crystal structure
|
|
fort.16 |
output
|
an
SHELX file for minimised crystal structure
|
|
fort.30
|
output
|
Crystal
information, similar to those created in fort.20
|
|
graphic
|
output
|
a
graphical output file, as requested using VIDEO
directive
|
| Operand Type | I |
| Operand Name | IDBAS |
This directive controls whether a complete restart dataset is written after the perfect lattice section has completed. This is very useful if it is necessary to restart the perfect lattice calculation. A complete copy of the input dataset is written to FORTRAN unit 15 with the following changes. The text between TITLE and ENDS is changed to reflect that the dataset is a restart. The lattice vectors and basis positions are updated to the new values. The dataset written to unit 15 will contain all of the input dataset as far as it had been processed and down to the next following START or STOP directive. A STOP directive is then inserted.
- 0
- Do not print a restart dataset.
- 1
- Print a restart dataset.