[Wannier] problem about "kmesh_get_bvector: Not enough bvectors found"

Jonathan Yates jonathan.yates at materials.ox.ac.uk
Wed Jul 27 11:14:27 CEST 2011


Dear Yun Yung,

 The solution is to either increase or decrease kmesh_tol. Do not set search_shells or shell_list.

Your unit cell causes some problems for the automatic shell finder routine in W90. It has a very slight distortion away from orthorhombic. The 2nd shell should contain b-vectors in the direction of a* and b* - the distortion breaks their symmetry. When 
kmesh_tol=0.00001
 W90 treats them as degenerate (ie multiplicity of 4)
When
kmesh_tol : 0.0000001
 W90 notices the degeneracy and gives two shells each with multiplicity 2.
I've copied the full b_vector information below.

 Unfortunately it seems that the default value causes one check to pass and one to fail. I'll need to work on fixing that.

 Note that these are small differences, and I would not expect the two choices to give very different MLWF. 

Jonathan





kmesh_tol : 0.0000001

 |                  b_k Vectors (Ang^-1) and Weights (Ang^2)                  |
 |                  ----------------------------------------                  |
 |            No.         b_k(x)      b_k(y)      b_k(z)        w_b           |
 |            ---        --------------------------------     --------        |
 |             1         0.000000    0.000000    0.269302     6.884154        |
 |             2         0.000000    0.000000   -0.269302     6.884154        |
 |             3         0.378480    0.000000    0.000000     3.490474        |
 |             4        -0.378480    0.000000    0.000000     3.490474        |
 |             5         0.000000    0.378480    0.000785     3.480299        |
 |             6         0.000000   -0.378480   -0.000785     3.480299        |
 |             7         0.000000    0.378480   -0.268517     0.010178        |
 |             8         0.000000   -0.378480    0.268517     0.010178        |
 |             9         0.378480    0.000000   -0.269302     0.000002        |
 |            10        -0.378480    0.000000    0.269302     0.000002        |
 +----------------------------------------------------------------------------+

kmesh_tol : 0.00001
+----------------------------------------------------------------------------+
 |                  b_k Vectors (Ang^-1) and Weights (Ang^2)                  |
 |                  ----------------------------------------                  |
 |            No.         b_k(x)      b_k(y)      b_k(z)        w_b           |
 |            ---        --------------------------------     --------        |
 |             1         0.000000    0.000000    0.269302     6.873978        |
 |             2         0.000000    0.000000   -0.269302     6.873978        |
 |             3         0.000000    0.378480    0.000785     3.480299        |
 |             4         0.378480    0.000000    0.000000     3.480299        |
 |             5        -0.378480    0.000000    0.000000     3.480299        |
 |             6         0.000000   -0.378480   -0.000785     3.480299        |
 |             7         0.000000    0.378480   -0.268517     0.010178        |
 |             8         0.000000   -0.378480    0.268517     0.010178        |
 |             9         0.378480    0.000000    0.269302     0.005089        |
 |            10         0.378480    0.000000   -0.269302     0.005089        |
 |            11        -0.378480    0.000000    0.269302     0.005089        |
 |            12        -0.378480    0.000000   -0.269302     0.005089        |



-- 
Department of Materials, University of Oxford, Parks Road, Oxford, OX1 3PH, UK
tel: +44 (0)1865 612797                http://users.ox.ac.uk/~oums0549/




More information about the Wannier mailing list