Activity for Robert D. Nielsen
-
Robert D. Nielsen posted a comment on discussion Elk Users
Hi Kay, Thanks for the help. I am trying to set the q-spiral in the y-direction with moment in the x-direction to calculate Heisenberg exchange. How does the ngridq work with spin-spirals? I looked at the example and manual but was a bit confused. What would be the actual q-vectors calculated with tasks 350 ngridq 0 2 0 Does it set equivalent points between the FM and AFM (q=0 and q=0.5)? Thanks you, Robert D. Nielsen
-
Hi Kay and Andrew, Thank you for the responses, they have helped exceptionally. I have managed to get the system to converge to the proper magnetic moments. Now I am using it as a collinear starting state for a spin-spiral run. I have tried with both highq and the other suggestions but the spin-spiral doesn't want to play nice. Here is my elk.in: ! EuO - LDA+U ! Surface - Eu Termination nwrite 1 tasks 0 !highq ! .true. gmaxvr 17 rgkmax 8.0 msmooth 4 isgkmax -2 lradstp 1 epspot 3.674932525e-8 epsengy...
-
Hi Kay and Andrew, Thank you for the responses, they have helped exceptionally. I have managed to get the system to converge to the proper magnetic moments. Now I am using it as a collinear starting state for a spin-spiral run. I have tried with both highq and the other suggestions but the spin-spiral doesn't want to play nice. Here is my elk.in: ! EuO - LDA+U ! Surface - Eu Termination nwrite 1 tasks 0 !highq ! .true. gmaxvr 17 rgkmax 8.0 msmooth 4 isgkmax -2 lradstp 1 epspot 3.674932525e-8 epsengy...
-
Hey Jon, That did the trick! Thank you. I was wondering if you could help with the parallelism a bit. I have an 80 core server and I am trying to maximize it's usage but am confused on the nesting aspect of elk. Is there a source for me to read more about this? -Robert
-
Hi Jon, Thank you for the suggestion. When I try to run 7.1.14 I run into a segmentation fault. When I run the code for bulk systems it works fine. Any ideas? Any help is much appreciated. -Robert
-
Hi Andrew, I have been trying various settings with your suggestions but nothing seems to converge. I seem to be observing charge sloshing and the moments of the ions don't converge properly. Here is my elk.in: ! EuO - LDA+U ! Surface - Eu Termination !maxthd ! 72 ! !maxthdmkl ! 18 tasks 0 nwrite 1 !highq ! .true. gmaxvr 15 rgkmax 8 msmooth 4 lradstp 2 lmaxapw 14 lmaxo 14 lmaxi 0 epspot 3.674932525e-8 epsengy 3.674932525e-7 maxscl 128 beta0 0.01 autolinengy .true. swidth 0.01 spinpol .true. spinorb...
-
Hi Andrew, I have been trying various settings with your suggestions but nothing seems to converge. I seem to be observing charge sloshing and the moments of the ions don't converge properly. Here is my elk.in: ! EuO - LDA+U ! Surface - Eu Termination !maxthd ! 72 ! !maxthdmkl ! 18 tasks 0 nwrite 1 !highq ! .true. gmaxvr 15 rgkmax 8 msmooth 4 lradstp 2 lmaxapw 14 lmaxo 14 lmaxi 0 epspot 3.674932525e-8 epsengy 3.674932525e-7 maxscl 128 beta0 0.01 autolinengy .true. swidth 0.01 spinpol .true. spinorb...
-
Hey Andrew, I've been attempting to converge a setup utilizing FSM both from scratch and and with a pre-converged result and am running into problems. At some point during convergence the energy ends up swinging wildly away from convergence. Here is the elk.in I am using currently. ! EuO - Geometry Optimization ! Surface - Eu Termination maxthd 72 tasks 0 nwrite 1 highq .true. !lmaxapw ! 14 ! !lmaxo ! 14 ! !lmaxi ! 0 ! !nempty ! 64 ! !epspot ! 1.0e-8 ! !epsengy ! 1.0e-7 maxscl 666 beta0 0.01 betamax...
-
Andrew, Thank you for your help. Would it be better to start with an FSM type calculation prior to convergence or with a pre-converged calculation and run with FSM with task=1? -Robert
-
Hi Andrew, Thank you so much for your response Andrew, it was much appreciated. I re-ran the elk.in with the original nempty given by highq and the linearisation energies did indeed go away and reach convergence. However I am unable to reach a consistent ferromagnetic state. I end with the following moments and charges: Charges : core : 380.0000000 valence : 172.0000000 interstitial : 17.87647215 muffin-tins (core leakage) species : 1 (Eu) atom 1 : 61.19379350 ( 0.1528482194E-05) atom 2 : 61.16187855...
-
Thank you Lan. What would cause the magnetic moments to not converge properly? My assumption was that this error was linked to this erroneous behavior. This setup worked well with EuO in the bulk setup. -Robert
-
Hello all, I am attempting to calculate a surface geometry for EuO (rock-salt) with surfaces on the (111) direction. While running it I keep recieving the error: Warning(linengy): could not find 8 linearisation energies in s.c. loop 1 I have tried a few things (increasing gmax, nempty, highq, vhighq, etc.) and the error persists. Here is me elk.in file: ! EuO - Geometry Optimization ! Surface - Eu Termination maxthd 72 tasks 0 nwrite 1 highq .true. !lmaxapw ! 14 ! !lmaxo ! 14 ! !lmaxi ! 0 nempty...
-
Hello all, I am attempting to calculate a surface geometry for EuO (rock-salt) with surfaces on the (111) direction. While running it I keep recieving the error: Warning(linengy): could not find 8 linearisation energies in s.c. loop 1 I have tried a few things (increasing gmax, nempty, highq, vhighq, etc.) and the error persists. Here is me elk.in file: ! EuO - Geometry Optimization ! Surface - Eu Termination maxthd 72 tasks 0 nwrite 1 highq .true. !lmaxapw ! 14 ! !lmaxo ! 14 ! !lmaxi ! 0 nempty...
-
Hello, I am curious if there is a way to perform an NSCF calculation in elk similar to the ICHARG=11 tag in VASP. Thank you, Robert D. Nielsen
-
Just tried and you're right. Even without the dmaxbnd is occurs. Any ideas? I know the system is large, but since I have plenty of memory, should it not handle it? Perhaps I need to recompile the program with a larger value somewhere? -Robert
-
Could this issue arise from the high demaxbnd value? -Robert
-
I will try this. Thank you very much Michael. -Robert
-
I have Intel's MPI installed and am have tried launching it with both 1 MPI processes and 72 OpenMP threads and 72 MPI processes with 1 OpenMP thread each. I have also tried with serial ifort with only OpenMP active and given it 72 threads. All have resulted in the segmentation fault. The system is indeed large, which is why I am utilizing a large memory system. The system is a SLURM managed cluster which would report in the memory is overloaded, which it does not. It seems to come from a miss managed...
-
Hello, I am attempting to calculate the reconstruction on the polar 111 surface of EuO and have a question regarding the setup. I have been able to calculate the ground state of the surface but am now undergoing the relaxtion of the geometry. My question is, as the surface layers of either Eu or O ( I am testing both cases) relax, it is my understanding that the local electronic structure will alter for these atoms in relation to the bulk like atoms within the surface model and as such, would it...
-
I have been trying to run this program for a few days now and though it seems the linearisation energy problem has been fixed, I am running into a segmentation fault issue. When I run the program the set runs and it enters into loop 1 without any linearisation energy error output, but then crashes due to a segmentation fault issue. The systems administrator has informed me that the issue is due to the program attempting to call a memory register beyond the scope of the allocated memory. I am sure...
-
I've enable autolinengy, autoswidth, and cranked demaxbnd to a high 15.0 and it seems to have fixed the issue. I will continue monitoring the execution to see that all goes well. Thank you so much for your help thus far. -Robert
-
The error persisted. -Robert
-
I have submitted the job. The LDA+U is indeed on Fe. -Robert
-
Michael, The lattice constants are indeed in bohr radii. You are correct regarding the LDA+U correction. I had re-done the geometry and the order of the ions were changed. I did not catch that error. Thank you. However, after changing this I still recieve the error. Any suggestions? -Robert
-
I am trying to do a simple magnon dispersion calculation but am having troubles with the base collinear calculation. I am calculating a conjugated microporous material which resembles a heme complex surrounded by large pores. The issue I am having is that when I try to run the base collinear calculation I am recieving the following error: Warning(linengy): could not find XXX linearisation energies in s.c. loop 1 From previous posts I double checked my geometry and found that it is fine and should...
-
I am trying to do a simple magnon dispersion calculation but am having troubles with the base collinear calculation. I am calculating a conjugated microporous material which resembles a heme complex surrounded by large pores. The issue I am having is that when I try to run the base collinear calculation I am recieving the following error: Warning(linengy): could not find XXX linearisation energies in s.c. loop 1 From previous posts I double checked my geometry and found that it is fine and should...
-
Hello, I am working with an MCA calculation and am trying to go about it using the Andersen force therom but am having a bit of confusion. When i set tasks to 1 and set the maxscl to 1, I get the same results back as if they were not modified. So I then set the maxscl to 2 and saw that the moment was changing from the original the described moment by the mommtfix tag, but not entirely equal to it. It seems that the magnetic moments aren't being totally constrained into the direction of mommtfix....
-
Hello, I am working with an MCA calculation and am trying to go about it using the Andersen force therom but am having a bit of confusion. When i set tasks to 1 and set the maxscl to 1, I get the same results back as if they were not modified. So I then set the maxscl to 2 and saw that the moment was changing from the original the described moment by the mommtfix tag, but not entirely equal to it. It seems that the magnetic moments aren't being totally constrained into the direction of mommtfix....
-
Markus, Thanks for the quick response, its very much appreciated. I have looked into the species file and am still confused. I see at the bottom where the local orbitals are listed: 7 : nlorb 0 2 : lorbl, lorbord 0.1500 0 F : lorbe0, lorbdm, lorbve 0.1500 1 F 1 2 : lorbl, lorbord 0.1500 0 F : lorbe0, lorbdm, lorbve 0.1500 1 F 2 2 : lorbl, lorbord 0.1500 0 F : lorbe0, lorbdm, lorbve 0.1500 1 F 1 3 : lorbl, lorbord 0.1500 0 F : lorbe0, lorbdm, lorbve 0.1500 1 F -26.0000 0 T 0 3 : lorbl, lorbord 0.1500...
-
Hello, I am working on a calculation with Eu and am running into troubles with the inclusion of local orbitals. Specifically, I cannot seem to find anywhere which allows me to specify the inclusion of the 5s and 5p local orbitals. I have checked the Elk Code Manual and found the lorbcnd tag and set it to true, but I cannot find anything that suggests that additional local obitals have been included. Where can I find the output responsible for listing which orbitals were set to local and how would...
1