06h-BN with finite-size extrapolation¶
00 Introduction¶
From this tutorial, you can learn how to calculate h-BN (finite-size extrapolation) with JDFT ansatz. You can download all the input and output files from here
.
01 DFT¶
# for all s_X_X_X
cd s_X_X_X/01trial_wavefunction/00makefort10/
cp ../../../9008997.cif .
turbogenius makefort10 -g -str 9008997.cif -s X X X -detbasis cc-pVTZ -jasbasis cc-pVDZ -detcutbasis -jascutbasis -pp ccECP
turbogenius makefort10 -r -post
cp fort.10 fort.10_in
turbogenius convertfort10mol -g -r -post
cd ../01DFT
cp ../00makefort10/fort.10 ./
cp ../00makefort10/pseudo.dat .
turbogenius prep -g -grid 0.10 0.10 0.10
# on a local machine (parallel version)
mpirun -np XX prep-mpi.x < prep.input > out_prep
# on a cluster machine (PBS)
qsub submit.sh
# on a cluster machine (Slurm)
sbatch submit.sh
turbogenius prep -post
02 Jastrow optimization¶
cd ../../02optimization/
cp ../01trial_wavefunction/01DFT/fort.10_new fort.10
cp ../01trial_wavefunction/01DFT/pseudo.dat ./
cp fort.10 fort.10_dft
turbogenius vmcopt -g -opt_onebody -opt_twobody -opt_jas_mat -optimizer lr -steps 500 -nw 480
# on a local machine (parallel version)
mpirun -np XX turborvb-mpi.x < datasmin.input > out_min
# on a cluster machine (PBS)
qsub submit.sh
# on a cluster machine (Slurm)
sbatch submit.sh
turbogenius vmcopt -post -optwarmup 450 -plot
03 JDFT ansatz - VMC¶
cd ../03vmc/
cp ../02optimization/fort.10 fort.10
cp ../02optimization/pseudo.dat .
turbogenius vmc -g -steps 500 -nw 480
# on a local machine (parallel version)
mpirun -np XX turborvb-mpi.x < datasvmc.input > out_vmc
# on a cluster machine (PBS)
qsub submit.sh
# on a cluster machine (Slurm)
sbatch submit.sh
turbogenius vmc -post -bin 10 -warmup 5
04 Summary¶
Try to plot the obtained energies per formula unit. v.s. 1/N, where N is the number of atoms in the simulation cell.