Hi Azar, The relaxation information (energies/forces) may or may not be sensitive to k grid depending on your system.
If you are not sure, try doing the same calculation on different k grid sizes and look how the output changes. -------------------------------------------------------- Aaron Thong PhD research student Department of Materials B321 Desk 20 Imperial College London -------------------------------------------------------- On 25 Nov 2015, at 16:17, azar ostovan wrote: Dear Aaron Thong Thank you so much I try such kgrid just for siesta calculation after structural relaxation in transport calculation the z-direction will be extended to 100. So if I got correctly you mean the siesta relaxation information are not sensitive to the number of kgrid? Cheers azar On Wed, Nov 25, 2015 at 7:34 PM, Thong, Aaron <[email protected]<mailto:[email protected]>> wrote: Hi Azar, Would be helpful to think about what you are trying to calculate. The MP grid can differ across calculations (i.e. SIESTA will not complain. Advanced users please correct me if I'm wrong here). But this may not mean that the answer you obtain is correct/physical. Using 1 k point would save time but also you lose out information on electron densities/transmission/properties (depending on what job you are running) at other K points. This may or may not be important depending on your system. -------------------------------------------------------- Aaron Thong PhD research student Department of Materials B321 Desk 20 Imperial College London -------------------------------------------------------- On 25 Nov 2015, at 15:46, azar ostovan wrote: Dear transiesta users . Do the number of kgrid_Monkhorst_Pack block ( or periodicity in each direction) could differ in structural relaxation, transiesta and tbtrans calculations? Or it must be the same? I mean can I for time saving used the below block in the optimization and transiesta %block kgrid_Monkhorst_Pack 1 0 0 0.0 0 1 0 0.0 0 0 1 0.0 %endblock kgrid_Monkhorst_Pack but for tbtrans extend the structure in z-direction? Warm regards azar
