For find groupvelocity

GULP
1

Dear. Julian

In the keywords, i’d like to get groupvelocity

So I include groupvelocity keyword in input.

But, I could not find the info about groupvelocity at the calculation output.

Where could i find the infor about groupvelocity in the output?

or Could I get the groupvelocity concurrently with calculation of thermal conductivity?

Thanks

2

The output for the group velocities should be after the frequencies for each
k point in the GULP output:

K point 10 = 0.062500 0.187500 0.187500 Weight = 0.004

Frequencies (cm-1) :

35.48 36.40 55.11 65.00 65.35 69.57 76.47 80.02 81.49
90.39 92.30 105.13 108.42 117.60 120.22 121.90 126.78 129.64
131.65 136.17 140.87 142.78 146.15 150.66 153.42 156.77 161.99
162.30 168.54 174.83 185.29 191.95 196.24 196.84 210.00 217.95
522.69 523.77 527.95 534.62 548.82 550.07 552.20 555.57 560.18
563.76 567.80 568.72 571.30 576.21 579.03 579.69 611.04 611.96
616.38 618.67 623.27 631.00 635.63 645.70 856.06 857.68 864.77
866.60 869.39 870.27 878.00 883.53 896.97 903.78 917.22 920.06

Group velocities (cm-1.Ang) :

Mode X Y Z

 1      10.143314     89.060616     90.753465
 2      29.048987     87.030304     96.541942
 3      22.785549    117.812017     75.555719
 4      -6.002210     83.839838     23.017065
 5      10.194852     17.701837    -19.132623
 6       4.171938     17.434621     30.391648
 7      14.302651     59.521725      8.909531
 8       1.668237     59.952394    -17.005749
 9     -14.394475     22.969855    -13.108723
10     -22.593768     44.701118    -30.223168
11      19.527123     14.810086    -36.115895
12       7.481936     34.740610    -36.160432

…

This is just a quick partial grab from an output to show you what should
appear. I’d recommend running separately from the thermal conductivity
calculation where you are trying to get the third order force constants.
It’s cheap in comparison to thermal conductivity and so it’s not a big overhead
to run it as a separate calculation.

1 Like
3

Dear. Julian

Thanks so much for your reply.

I confirmed the group velocity thanks to your comment

I’d better to calculate this value separately like your comment. It is a good tip to me. thanks.

And Group velocities (cm-1.Ang) look like that the unit phase/cm or phase/Angstrom? or cm/s?

also, to get Group velocity near gamma point , almost limit to gamma point (kpoints 0 0 0) ,

Is the below input setting reasonable?
kpoints
0.001 0.001 0.001

Sincerely,

4

The group velocities are in cm-1/Ang as they are the frequency (cm-1) derivatives with respect to reciprocal space (Ang-1).
I haven’t tested, but a value of 0.001 should be pretty close to gamma. It’s always best to run a series of points to see how to extrapolate the trend to the limit.

1 Like
5

Dear. Julian

I appreciate your reply

as your comment, 0.001 fine to me

Best regard.

6

Dear Julian

in the group velocity

Mode X Y Z

 1      10.143314     89.060616     90.753465
 2      29.048987     87.030304     96.541942

Are 10.143314 , 89.060616 and 90.753465 the velocity of each directions of x,y,z respectively?

Thanks

7

The units are output as “Group velocities (cm-1.Ang)” and so that would be a reasonable conclusion.

1 Like