Page 1 of 1

Bad termination error during epw1 calculation

Posted: Tue Jul 27, 2021 2:20 am
by jhosamelly
I am doing mobility calculations for pristine MoS2. Was able to do scf>ph>pp.py>nscf
but I encountered a problem when I did the epw1 calculations. There is a bad termination error.

Here is the last few lines before the error and the error itself.

Code: Select all

  Reading data about k-point neighbours 
 
      - All neighbours are found 
 
     AMN
      k points =    64 in   44 pools
            1 of    2 on ionode
            2 of    2 on ionode

===================================================================================
=   BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
=   RANK 0 PID 32331 RUNNING AT XXXXXX-cpu-31
=   KILLED BY SIGNAL: 9 (Killed)
===================================================================================
Here is my epw1.in code

Code: Select all

&inputepw
  prefix      = 'mos2'			  ! prefix used in pw.x calculations			
  amass(1)    = 95.94000,		  ! mass of Mo
  amass(2)    = 32.06600,		  ! mass of S
  outdir      = './'			  ! same outdir used in pw.x calculations

  elph        = .true.			  ! If .true. calculate e-ph coefficients. default .false.
  epbwrite    = .true.			  ! If epbwrite = .true., the electron-phonon matrix elements in the coarse Bloch 
					  ! representation and relevant data (dyn matrices) are written to disk. default .false.
  epbread     = .false.			  ! If epbread = .true. the above quantities are read from the ‘prefix.epb’ files. Pool dependent files.	
  epwwrite    = .true.			  ! If epwwrite = .true., the electron-phonon matrix elements in the coarse 
  					  ! Wannier representation and relevant data (dyn matrices) are written to disk. default .true.
  epwread     = .false.			  ! If epwread = .true., the electron-phonon matrix elements in the coarse
  					  ! Wannier representation are read from the ‘epwdata.fmt’ and ‘XX.epmatwpX’ files. default .false.
  etf_mem     =  1			  ! When etf_mem = 1, more IO (slower) but less memory is required. default 1
  lpolar      = .true.   		  ! polar material default .false.
  vme         = 'wannier'  		  ! If .true. then calculate the velocity matrix elements beyond the local approximation. default .false.
  use_ws      = .false.

  lifc        = .false.			  ! default .false. If .true. uses the real-space inter-atomic force constant generated by q2r.x.
  asr_typ     = 'simple'		  ! default 'simple'. Kind of acoustic sum rule that can be imposed in real space.
  					  ! Possible ASR are ‘simple’, ‘crystal’, ‘one-dim’ and ‘zero-dim’.
!  lphase      = .true.			  ! If .true. then fix the gauge for the interpolated dynamical matrix and electronic Hamiltonian. default .false.

  nbndsub     =  3			  ! Number of wannier functions to utilize.
  bands_skipped = 'exclude_bands = 1'     ! List of bands to exclude from the wannierization.

  wannierize  = .true.			  ! default .false. Calculate the Wannier functions using W90 library calls
 					  ! and write rotation matrix to file ‘filukk’. If .false., filukk is read from disk.
  num_iter    = 1500			  ! number of iteration in the minimization of wavefunction??
  iprint      = 3			  ! Output verbosity level default 1
  dis_win_max = 3.0			  ! Top of the outer energy window
  dis_win_min = -3.0			  ! Bottom of the outer energy window
  dis_froz_max= 2.5			  ! Top of the inner (frozen) energy window
  proj(1)  = 'Mo:d'
  proj(2)  = 'S:p'
  wdata(1) = 'bands_plot = .true.'
  wdata(2) = 'begin kpoint_path'
  wdata(3) = 'G 0.00  0.00 0.00 M 0.50 -0.50 0.00'
  wdata(4) = 'M 0.50 -0.50 0.00 K 0.67 -0.33 0.00'
  wdata(5) = 'K 0.67 -0.33 0.00 G 0.00  0.00 0.00'
  wdata(6) = 'end kpoint_path'
  wdata(7) = 'bands_plot_format = gnuplot'
  wdata(8) = 'guiding_centres = .true.'
  wdata(9) = 'dis_num_iter      = 1000'	  ! Number of iterations for the minimisation of wavefunction??
  wdata(10) = 'num_print_cycles  = 10'	  ! Control frequency of printing 
  wdata(11) = 'dis_mix_ratio     = 1.0'   ! Mixing ratio during the minimisation of wavefunction??
  wdata(12) = 'conv_tol = 1E-8'		  ! The convergence tolerance for finding wavefunction
  wdata(13) = 'conv_window = 4'		  ! The number of iterations over which convergence of wavefunction?? is assessed
  wdata(14) = 'use_ws_distance = T'
  elecselfen  = .false.			  ! Calculate the electron self-energy from the el-ph interaction. default .false.
  phonselfen  = .false.			  ! Calculate the phonon self-energy from the el-ph interaction. default .false.
  a2f         = .false.			  ! Calculate Eliashberg spectral function, α2F(ω), transport Eliashberg spectral function α2Ftr(ω),
  					  ! and phonon density of states F(ω). Only allowed in the case of phonselfen = .true. default .false.

  fsthick     = 100.0 			  ! Width of the Fermi surface window to take into account states in the self-energy delta functions in [eV].
  					  ! Narrowing this value reduces the number of bands included in the selfenergy calculations. default 1.d10
  temps       = 1 ! K 			  ! Temperature values used in superconductivitiy, transport, indabs, etc..
  					  ! If no temps are provided, temps=300 and nstemp =1.
  nstemp      = 1
  degaussw    = 0.01 			  ! Smearing in the energy-conserving delta functions in [eV] default 0.025

  dvscf_dir   = './save/'		  ! Directory where ‘prefix.[dvscf|dyn]_q??’ files are located.

  efermi_read = .true.			  ! If .true. the Fermi energy is read from the input file.
  band_plot   = .true.			  ! Writes files for band structure and phonon dispersion plots.
  					  ! The k-path and q-path is provided using filkf and filqf. default .false.


  filkf       = './LGXKG.txt'		  ! File which contains the fine k-mesh or the k-path of electronic states to be calculated for elinterp.
  filqf       = './LGXKG.txt'		  ! File which contains the fine q-mesh or the q-path of phonon states to be calculated for phinterp.

   nk1        = 4			  ! Dimensions of the coarse electronic grid, corresponds to the nscf calculation and wfs in the outdir.
   nk2        = 4
   nk3        = 4
   nq1        = 4			  ! Dimensions of the coarse phonon grid, corresponds to the nqs list.
   nq2        = 4
   nq3        = 4
 /

What can I do to resolve this? If other relevant files are needed please let me know. Thanks.

Re: Bad termination error during epw1 calculation

Posted: Tue Jul 27, 2021 5:42 pm
by hlee
Dear jhosamelly:

Code: Select all

  nbndsub     =  3			  ! Number of wannier functions to utilize.

Code: Select all

  proj(1)  = 'Mo:d'
  proj(2)  = 'S:p'
The above two blocks are not consistent; you are considering 11 projections (5+2*3) (I assume the case without using spin-orbit coupling).
Please familarize yourself with Wannierization before running EPW.

Sincerely,

H. Lee