Dear All - we're doing a Castep phonon calculation on a small organic crystal structure, a param file is below. A DFPT calculation appears to run fine when started from this file. The odd thing is that after the phonons have been calculated a new Thermodynamics job is generated, and the program seems to start the phonon calculation all over again.

The contents of the cell and param files for the original phonon and thermodynamics jobs are identical except for the lines added to do the thermodyamics calculation (e.g. task : thermodynamics and the thermo_t_start stop and npoints entries). 

Any idea why it would repeat the calculation?

I assume we can just kill the current thermodynamics job. I'm concerned about the integrity of the checkpoint file if I do this, but I notice there's a file with a .castep_bin_2 extension which looks like it might be a copy of the castep_bin file at the end of the first phonon calculation.  

All advice gratefully received. 

Simon Parsons (University of Edinburgh)

task : Phonon
continuation : default
CHECKPOINT: CPCODE_PhonDOS.castep_bin
comment : CASTEP calculation from Materials Studio
xc_functional : PBE
sedc_apply : true
sedc_scheme : TS
spin_polarized : false
spin_orbit_coupling : false
opt_strategy : Default
page_wvfns : 0
cut_off_energy : 990.000000000000000
grid_scale : 2.000000000000000
fine_grid_scale : 3.000000000000000
finite_basis_corr : 2
finite_basis_npoints : 3
elec_energy_tol : 1.000000000000000e-10
elec_eigenvalue_tol : 1.000000000000000e-09
max_scf_cycles : 100
fix_occupancy : true
metals_method : dm
mixing_scheme : Pulay
mix_charge_amp : 0.500000000000000
mix_charge_gmax : 1.500000000000000
mix_history_length : 20
nextra_bands : 0
num_dump_cycles : 0
calculate_stress : false
calculate_ELF : false
popn_calculate : false
calculate_hirshfeld : false
calculate_densdiff : false
phonon_sum_rule_method : reciprocal
phonon_calc_lo_to_splitting : false
born_charge_sum_rule : true
backup_interval : 900
phonon_energy_tol : 1.000000000000000e-05
phonon_fine_method : interpolate
phonon_force_constant_cutoff : 7.257734973167010 ang
phonon_max_cycles : 100
bs_max_iter : 250
bs_max_cg_steps : 25
bs_eigenvalue_tol : 1.0e-9
pdos_calculate_weights : false
spectral_write_eigenvalues : false