4.13 what='mur_lc'

With this option the code runs several self-consistent calculations at different geometries. The runs can be done in parallel when several images are available. This option has two working modes controlled by the logical variable lmurn. When lmurn=.TRUE. the total energy as a function of the volume is interpolated by a Murnaghan equation and a plot of the energy as a function of the volume and of the pressure as a function of the volume is produced. The volume is changed by changing only celldm(1). celldm(2)...celldm(6) remain fixed at the values given as input of pw.x. When lmurn=.FALSE. the energy is calculated in a uniform grid of parameters composed of ngeo(1) x ngeo(2)... x ngeo(6) points. The energies are fitted with a quadratic or quartic polynomial of N_k variables, where N_k is the number of independent crystal parameters for the given crystal system. A plot of the energy as a function of the lattice constant is produced for cubic systems. For solids of the hexagonal, tetragonal, and trigonal systems contour plots of the energy as a function of the two crystal parameters (a and c/a or a and cos$\alpha$ ) are plotted. For orthorhombic systems contour plots of the energy as a function of a and b/a are plotted for each value of c/a . Presently no graphical tool is implemented to plot the energy for monoclinic and triclinic crystal systems. When lmurn=.FALSE. the bulk modulus is not calculated. To obtain it, you can calculate the elastic constants at the minimum geometry (see the option what='mur_lc_elastic_constants'). With this option the pressure control is active. You can specify a finite pressure and the enthalpy is minimized instead of the energy. Note however that if the minimum is distant from the starting configuration its associated error can be large, larger for the quadratic than for the Murnaghan interpolation. For this reason the present option should be used starting from the minimum found by pw.x using the vc-relax option and the pressure should not be too different from the pressure used for vc-relax.
This option can be controlled by the following variables:

ngeo(1),...,ngeo(6) : the number of geometries to use for each celldm parameter.
             The lattice constant of these geometries is calculated from the
             input of pw.x. celldm(1),...,celldm(6) of this input is used 
             for the central geometry. For the others celldm(1),...,celldm(6),
             are changed in steps of step_ngeo(1),...,step_ngeo(6). 
             ngeo(1) must be odd. Only the values of celldm relevant for
             each Bravais lattice are actually changed.
             Default: integer 1,1,1,1,1,1 for what=scf_*, 9,1,1,1,1,1 for 
             what=mur_lc_* and lmurn=.TRUE. or for cubic systems, 5 on all 
             the relevant celldm parameters when lmurn=.FALSE. and the system
             is not cubic.
step_ngeo(1),...,step_ngeo(6) : The step between the lattice constants at 
             different geometries. step_ngeo(1) is, in atomic units, the change
             of a, step_ngeo(2), step_ngeo(3) are dimensionless and are the
             changes of the ratios b/a, c/a, step_ngeo(4), step_ngeo(5), &
             step_ngeo(6) are the changes in degree of the angles alpha,
             beta, and gamma. The cosine of the angle is calculated by the
             program.
             Default: real 0.05 a.u., 0.02, 0.02, 0.5, 0.5, 0.5
lmurn       : if .TRUE. the Murnaghan fit is done. Only ngeo(1) values of
             the energy are fitted, the other values of ngeo are not used. 
             if .FALSE. use a quadratic or quartic function to interpolate 
             the energy as a function of all celldm parameters. The number of 
             self-consistent calculations is ngeo(1) x ngeo(2) x ngeo(3)
             x ngeo(4) x ngeo(5) x ngeo(6). In this case only the 
             minimum energy and the optimal celldm are given in output. 
             Default: .TRUE. 
show_fit   : if .TRUE. show the contour plot of the fitted energy instead
             of the energy. Used by default when reduced_grid is .TRUE.
             Default: logical .FALSE.
vmin_input : minimum volume for the plot of the energy as a function of volume.
             Default: real 0.98 times the volume of the first geometry.
vmax_input : maximum volume for the plot of the energy as a function of volume.
             Default: real 1.02 times the volume of the last geometry.
deltav     : distance between two volumes in the plot of the energy as a 
             function of the volume.
             Default: real calculated from nvol.
nvol       : number of volumes in Murnaghan plot
             Default : integer 51
lquartic   : if .TRUE. fit the energy with a quartic polynomial.
             Default : logical .TRUE.
lsolve     : choose the algorithm used to fit the quartic polynomial parameters.
             Allowed values:
             1 explicitly minimize chi^2
             2 Use the QR algorithm to minimize chi^2 (lapack routine dgels)
             3 Use the SVD algorithm to minimize chi^2 (lapack routine dgelss)
             Default: integer 2
flevdat    : file where the Murnaghan equation is written. The results of the
             Murnaghan fit are then written in flevdat.ev.out.
             Default: character(len=*) 'output_ev.dat'
flpsmur    : postscript file of the Murnaghan plot
             Default: character(len=*) 'output_mur'
ncontours  : the number of contours in the energy plot. These levels can
             be determined automatically by the code or defined by the user.
             The energy levels can be defined after the INPUT_THERMO namelist 
             but before the path, as a list
             energy_level(1)      color(1)
             ...
             energy_level(ncontours)   color(ncontours) 
             Color is a string of the type color_red, color_green, etc.
             The list of available colors is at the beginning of each gnuplot
             script.
             energy_level is in Ry units.
             Default: integer 9
do_scf_relax : if .TRUE. the code makes a self-consistent relax calculation at
             the equilibrium geometry to find the optimized atomic 
             coordinates. This step is needed only for solids that have 
             internal degrees of freedom in the unstrained configuration. 
             If .FALSE. the coordinates of the input geometry are strained 
             uniformly to the equilibrium geometry.
             Default: logical .FALSE. 
flenergy   : name of the file that contains the energy in a form that
             can be used by gnuplot to make contour plots.
             Defaults: character(len=*) 'output_energy'
flpsenergy : file with the contour plots of the energy as a function of the
             crystal parameters.
             Default: character(len=*) 'output_energy'

An example of the use of this option can be found in example05.
Number of tasks for this option:
ngeo(1) when lmurn=.TRUE., ngeo(1)x ngeo(2)x ngeo(3)x ngeo(4)x ngeo(5)x ngeo(6)
when lmurn=.FALSE. and reduced_grid=.FALSE.,
ngeo(1)+ ngeo(2)+ ngeo(3)+ ngeo(4)+ ngeo(5)+ ngeo(6)-5
when lmurn=.FALSE. and reduced_grid=.TRUE..