Calculation of the interatomic forces fails

Running Abinit Response calculations
1

Hi,

Could someone please check this electro phonon interatomic forces calculation which throws an error saying:

message: |

  • Information is missing in the DDB.*
  • The dynamical matrix number 6 cannot be built,*
  • since no block with qpt: 5.000000E-01 0.000000E+00 5.000000E-01*
  • has been found.*
  • Action: add the required block in the DDB, or modify the q-mesh your input file.*

However, dataset 2,3 and 4 calculates for the q points (000) (0.5 0 0) and (0 0 0.5)! Why is it looking for a block with qpt (0.5 0 0.5)?
It is hexagonal lattice P6/mmm. I attach all the input files and log files here.

Best,
Rajesh
B2Mo_4.abi (1.1 KB)
B2Mo_3.abi (414 Bytes)
B2Mo_2.abi (87 Bytes)
log4.abi (4.1 KB)
B2Mo.abi (4.0 KB)

2

You changed the structure and symmetries, so you can not expect the list of irreducible q and perturbations to remain the same: your input copies those for Al FCC, but the number and coordinates of the q will change.

Please check the tutorial and see how to generate the list of irred q. The simplest is to run a dummy GS calculation with kptopt 1 and ngkpt = your qpt grid. This will give you the list of irred q in the kpt variable (no need to converge or anything).

Check the new eph tutorial as well: It explains you can use qptopt 1 and iqpt with the index of the q you want.

Finally, have a look at the abipy tutorials. There are workflows which do all of this indexing for you and avoid forgetting perturbations etc.

3

Hi mverstra,

Thanks for the reply. I have checked with your 1st suggestion, but there is a confusion about generated list of irred q. When I use different ngqpt from ngkpt I get same list. Seemingly, the list is depending only on ngkpt value.

For example, with ngqpt 222 and ngkpt 666 input it gives the same list if I use ngqpt 666 and ngkpt 666.

Here I attach my input and output files.

B2Mo_ngqpt222.abo (74.1 KB)
B2Mo.abi (2.2 KB)
B2Mo_ngqpt666.abo (74.1 KB)

regards,
Rajesh