Re: [Wien] A few (more) elementary -so questions (with onsite -eece)
Sorry, the comment was confusing. The spin moments specified in the initial configuration of course point along the axis specified by the direction. What I wanted to point out is that for symmetry reasons this is the only meaningfull component you can specify. In this sense symmetry comes first. The direction is more than just the spin direction. Its the only direction along which any vector operator that Wien2k can represent in the given case can point. Someone should correct me if I am wrong, but in my understanding this is (one of) the reason(s) that Wien2k can only cope with collinear magnetic moments: The axial symmetry is introduced as a global symmetry. The direction of the axis is the same in all muffin-tin spheres. I expect the Wienncm code handles this on basis of the local symmetry. --- Dr. Martin Pieper Karl-Franzens University Institute of Physics Universitätsplatz 5 A-8010 Graz Austria Tel.: +43-(0)316-380-8564 Am 05.05.2015 19:31, schrieb Laurence Marks: I would be interested in clarification from others, but from what I can see in the code it appears that this is the spin direction that is used, not just the direction of breaking the symmetry. I may be wrong. On Tue, May 5, 2015 at 11:47 AM, pieper pie...@ifp.tuwien.ac.at wrote: I definitely am not an expert for -so, therefore I will not shoot down anything, only a comment: From my point of view from magnetism I would ask for some caution with identifying the direction given in .inorb and .inso with 'the spin direction'. As Gerhard pointed out earlier in this thread, it's all about symmetry: The specified direction only sets up the symmetry of the case to be compatibel with whatever has a rotational invariance with that (quantization) axis - be that a spin, or orbital moment, magnetization, a magnetic field, ... The symmetry of the basis has to allow for a magnetization otherwise it won't appear when you calculate expectation values. Personally I find Pavel's 'lecture on spin-orbit.ps [1]' here in the Wien documentation files (I hope it's still there) very illuminating. --- Dr. Martin Pieper Karl-Franzens University Institute of Physics Universitätsplatz 5 A-8010 Graz Austria Tel.: +43-(0)316-380-8564 Am 05.05.2015 14:51, schrieb Laurence Marks: Me a culpa, I should have checked the mailing list first for the answers. That said, this issue has come up enough times in the past that I think the UG should be tweaked so it is clearer. Let me try my interpretation, so I can be shot down if needed. Within Wien2k magnetic effects can be approximately included in a number of ways. Some such as the spin-orbit coupling assume a direction for the spin vector (for all electrons actively considered), others such as Bext in orb specify a direction for an applied magnetic field (in Tesla) and use the same direction for the spin vector. (The two spin states are then either parallel or anti parallel to the specified direction.) When a direction is specified in case.inso or case.inorb this fixes the spin vector and (if used) the external magnetic field direction. Via the output files from lapwdm (case.scfdmXX) one can monitor how the angular momentum changes [1]. By using different directions for the spin vector (and field) one can probe how the energy changes and/orbital occupancies with assumed directions for the spin/external magnetic field. To escape the assumption that the spin vectors all have one direction the Wienncm code has to be used. [1] My addendum. Changes in the occupancies can be a soft electronic mode, i.e. very small changes in the energy for quite large changes in the density. The older mixing algorithms (MSEC1 or MSEC3) are not so good for soft modes and can stagnate. MSR1 is better and the next release (7.0) is much better. With onsite -eece /or -orb it may help to push the mixer by either forcing a larger step (echo .2 .msec or echo .1 .pratt) or stopping, doing a force on the orbital potential (x orb -up; x orb -dn) then restarting with -NI. It is probably wise to check how the orbital momentum is converging (grep :ORB0 *.scf, perhaps other) and make sure that the mixing is not starving (grep GREED: *.scf and check the values are not small, e.g. 0.035). ___ Professor Laurence Marks Department of Materials Science and Engineering Northwestern University www.numis.northwestern.edu [2] [1] MURI4D.numis.northwestern.edu [3] [2] Co-Editor, Acta Cryst A Research is to see what everybody else has seen, and to think what nobody else has thought Albert Szent-Gyorgi On May 4, 2015 6:22 PM, Laurence Marks l-ma...@northwestern.edu wrote: Typo: although I remember don't symmetry operations being split into these two classes everywhere in the code On Mon, May 4, 2015 at 6:04 PM, Laurence Marks l-ma...@northwestern.edu wrote: I am a newbie at -so, so a few simple questions. a) What is the meaning of the orbital moment in case.scfdm*
Re: [Wien] A few (more) elementary -so questions (with onsite -eece)
Me a culpa, I should have checked the mailing list first for the answers. That said, this issue has come up enough times in the past that I think the UG should be tweaked so it is clearer. Let me try my interpretation, so I can be shot down if needed. Within Wien2k magnetic effects can be approximately included in a number of ways. Some such as the spin-orbit coupling assume a direction for the spin vector (for all electrons actively considered), others such as Bext in orb specify a direction for an applied magnetic field (in Tesla) and use the same direction for the spin vector. (The two spin states are then either parallel or anti parallel to the specified direction.) When a direction is specified in case.inso or case.inorb this fixes the spin vector and (if used) the external magnetic field direction. Via the output files from lapwdm (case.scfdmXX) one can monitor how the angular momentum changes [1]. By using different directions for the spin vector (and field) one can probe how the energy changes and/orbital occupancies with assumed directions for the spin/external magnetic field. To escape the assumption that the spin vectors all have one direction the Wienncm code has to be used. [1] My addendum. Changes in the occupancies can be a soft electronic mode, i.e. very small changes in the energy for quite large changes in the density. The older mixing algorithms (MSEC1 or MSEC3) are not so good for soft modes and can stagnate. MSR1 is better and the next release (7.0) is much better. With onsite -eece /or -orb it may help to push the mixer by either forcing a larger step (echo .2 .msec or echo .1 .pratt) or stopping, doing a force on the orbital potential (x orb -up; x orb -dn) then restarting with -NI. It is probably wise to check how the orbital momentum is converging (grep :ORB0 *.scf, perhaps other) and make sure that the mixing is not starving (grep GREED: *.scf and check the values are not small, e.g. 0.035). ___ Professor Laurence Marks Department of Materials Science and Engineering Northwestern University www.numis.northwestern.edu MURI4D.numis.northwestern.edu Co-Editor, Acta Cryst A Research is to see what everybody else has seen, and to think what nobody else has thought Albert Szent-Gyorgi On May 4, 2015 6:22 PM, Laurence Marks l-ma...@northwestern.edu wrote: Typo: although I remember don't symmetry operations being split into these two classes everywhere in the code On Mon, May 4, 2015 at 6:04 PM, Laurence Marks l-ma...@northwestern.edu wrote: I am a newbie at -so, so a few simple questions. a) What is the meaning of the orbital moment in case.scfdm* ? Is that the average direction projected to the global axis system? b) What is the physical significance of the orbital moment being parallel (or not quite parallel) to the direction used in case.inso? c) I understand that the results for different directions of B in case.inso reflect the magnetic anisotropy, but what are the units of field (if any)? d) What else is worth looking at? The partial orbital moment (:POM) seems relevant, but what exactly is it? e) I am blindly trusting that initso knows what it is doing, and have left the B symmetry operations in case.struct (although I remember symmetry operations being split into these two classes everywhere in the code). This seems to conflict with Pavel's notes, although those may be too old. Thanks. -- Professor Laurence Marks Department of Materials Science and Engineering Northwestern University www.numis.northwestern.edu Corrosion in 4D: MURI4D.numis.northwestern.edu Co-Editor, Acta Cryst A Research is to see what everybody else has seen, and to think what nobody else has thought Albert Szent-Gyorgi -- Professor Laurence Marks Department of Materials Science and Engineering Northwestern University www.numis.northwestern.edu Corrosion in 4D: MURI4D.numis.northwestern.edu Co-Editor, Acta Cryst A Research is to see what everybody else has seen, and to think what nobody else has thought Albert Szent-Gyorgi ___ Wien mailing list Wien@zeus.theochem.tuwien.ac.at http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien SEARCH the MAILING-LIST at: http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
Re: [Wien] A few (more) elementary -so questions (with onsite -eece)
I definitely am not an expert for -so, therefore I will not shoot down anything, only a comment: From my point of view from magnetism I would ask for some caution with identifying the direction given in .inorb and .inso with 'the spin direction'. As Gerhard pointed out earlier in this thread, it's all about symmetry: The specified direction only sets up the symmetry of the case to be compatibel with whatever has a rotational invariance with that (quantization) axis - be that a spin, or orbital moment, magnetization, a magnetic field, ... The symmetry of the basis has to allow for a magnetization otherwise it won't appear when you calculate expectation values. Personally I find Pavel's 'lecture on spin-orbit.ps' here in the Wien documentation files (I hope it's still there) very illuminating. --- Dr. Martin Pieper Karl-Franzens University Institute of Physics Universitätsplatz 5 A-8010 Graz Austria Tel.: +43-(0)316-380-8564 Am 05.05.2015 14:51, schrieb Laurence Marks: Me a culpa, I should have checked the mailing list first for the answers. That said, this issue has come up enough times in the past that I think the UG should be tweaked so it is clearer. Let me try my interpretation, so I can be shot down if needed. Within Wien2k magnetic effects can be approximately included in a number of ways. Some such as the spin-orbit coupling assume a direction for the spin vector (for all electrons actively considered), others such as Bext in orb specify a direction for an applied magnetic field (in Tesla) and use the same direction for the spin vector. (The two spin states are then either parallel or anti parallel to the specified direction.) When a direction is specified in case.inso or case.inorb this fixes the spin vector and (if used) the external magnetic field direction. Via the output files from lapwdm (case.scfdmXX) one can monitor how the angular momentum changes [1]. By using different directions for the spin vector (and field) one can probe how the energy changes and/orbital occupancies with assumed directions for the spin/external magnetic field. To escape the assumption that the spin vectors all have one direction the Wienncm code has to be used. [1] My addendum. Changes in the occupancies can be a soft electronic mode, i.e. very small changes in the energy for quite large changes in the density. The older mixing algorithms (MSEC1 or MSEC3) are not so good for soft modes and can stagnate. MSR1 is better and the next release (7.0) is much better. With onsite -eece /or -orb it may help to push the mixer by either forcing a larger step (echo .2 .msec or echo .1 .pratt) or stopping, doing a force on the orbital potential (x orb -up; x orb -dn) then restarting with -NI. It is probably wise to check how the orbital momentum is converging (grep :ORB0 *.scf, perhaps other) and make sure that the mixing is not starving (grep GREED: *.scf and check the values are not small, e.g. 0.035). ___ Professor Laurence Marks Department of Materials Science and Engineering Northwestern University www.numis.northwestern.edu [1] MURI4D.numis.northwestern.edu [2] Co-Editor, Acta Cryst A Research is to see what everybody else has seen, and to think what nobody else has thought Albert Szent-Gyorgi On May 4, 2015 6:22 PM, Laurence Marks l-ma...@northwestern.edu wrote: Typo: although I remember don't symmetry operations being split into these two classes everywhere in the code On Mon, May 4, 2015 at 6:04 PM, Laurence Marks l-ma...@northwestern.edu wrote: I am a newbie at -so, so a few simple questions. a) What is the meaning of the orbital moment in case.scfdm* ? Is that the average direction projected to the global axis system? b) What is the physical significance of the orbital moment being parallel (or not quite parallel) to the direction used in case.inso? c) I understand that the results for different directions of B in case.inso reflect the magnetic anisotropy, but what are the units of field (if any)? d) What else is worth looking at? The partial orbital moment (:POM) seems relevant, but what exactly is it? e) I am blindly trusting that initso knows what it is doing, and have left the B symmetry operations in case.struct (although I remember symmetry operations being split into these two classes everywhere in the code). This seems to conflict with Pavel's notes, although those may be too old. Thanks. -- Professor Laurence Marks Department of Materials Science and Engineering Northwestern University www.numis.northwestern.edu [1] Corrosion in 4D: MURI4D.numis.northwestern.edu [2] Co-Editor, Acta Cryst A Research is to see what everybody else has seen, and to think what nobody else has thought Albert Szent-Gyorgi -- Professor Laurence Marks Department of Materials Science and Engineering Northwestern University www.numis.northwestern.edu [1] Corrosion in 4D: MURI4D.numis.northwestern.edu [2] Co-Editor, Acta Cryst A Research is to see what everybody
Re: [Wien] A few (more) elementary -so questions (with onsite -eece)
I would be interested in clarification from others, but from what I can see in the code it appears that this is the spin direction that is used, not just the direction of breaking the symmetry. I may be wrong. On Tue, May 5, 2015 at 11:47 AM, pieper pie...@ifp.tuwien.ac.at wrote: I definitely am not an expert for -so, therefore I will not shoot down anything, only a comment: From my point of view from magnetism I would ask for some caution with identifying the direction given in .inorb and .inso with 'the spin direction'. As Gerhard pointed out earlier in this thread, it's all about symmetry: The specified direction only sets up the symmetry of the case to be compatibel with whatever has a rotational invariance with that (quantization) axis - be that a spin, or orbital moment, magnetization, a magnetic field, ... The symmetry of the basis has to allow for a magnetization otherwise it won't appear when you calculate expectation values. Personally I find Pavel's 'lecture on spin-orbit.ps' here in the Wien documentation files (I hope it's still there) very illuminating. --- Dr. Martin Pieper Karl-Franzens University Institute of Physics Universitätsplatz 5 A-8010 Graz Austria Tel.: +43-(0)316-380-8564 Am 05.05.2015 14:51, schrieb Laurence Marks: Me a culpa, I should have checked the mailing list first for the answers. That said, this issue has come up enough times in the past that I think the UG should be tweaked so it is clearer. Let me try my interpretation, so I can be shot down if needed. Within Wien2k magnetic effects can be approximately included in a number of ways. Some such as the spin-orbit coupling assume a direction for the spin vector (for all electrons actively considered), others such as Bext in orb specify a direction for an applied magnetic field (in Tesla) and use the same direction for the spin vector. (The two spin states are then either parallel or anti parallel to the specified direction.) When a direction is specified in case.inso or case.inorb this fixes the spin vector and (if used) the external magnetic field direction. Via the output files from lapwdm (case.scfdmXX) one can monitor how the angular momentum changes [1]. By using different directions for the spin vector (and field) one can probe how the energy changes and/orbital occupancies with assumed directions for the spin/external magnetic field. To escape the assumption that the spin vectors all have one direction the Wienncm code has to be used. [1] My addendum. Changes in the occupancies can be a soft electronic mode, i.e. very small changes in the energy for quite large changes in the density. The older mixing algorithms (MSEC1 or MSEC3) are not so good for soft modes and can stagnate. MSR1 is better and the next release (7.0) is much better. With onsite -eece /or -orb it may help to push the mixer by either forcing a larger step (echo .2 .msec or echo .1 .pratt) or stopping, doing a force on the orbital potential (x orb -up; x orb -dn) then restarting with -NI. It is probably wise to check how the orbital momentum is converging (grep :ORB0 *.scf, perhaps other) and make sure that the mixing is not starving (grep GREED: *.scf and check the values are not small, e.g. 0.035). ___ Professor Laurence Marks Department of Materials Science and Engineering Northwestern University www.numis.northwestern.edu [1] MURI4D.numis.northwestern.edu [2] Co-Editor, Acta Cryst A Research is to see what everybody else has seen, and to think what nobody else has thought Albert Szent-Gyorgi On May 4, 2015 6:22 PM, Laurence Marks l-ma...@northwestern.edu wrote: Typo: although I remember don't symmetry operations being split into these two classes everywhere in the code On Mon, May 4, 2015 at 6:04 PM, Laurence Marks l-ma...@northwestern.edu wrote: I am a newbie at -so, so a few simple questions. a) What is the meaning of the orbital moment in case.scfdm* ? Is that the average direction projected to the global axis system? b) What is the physical significance of the orbital moment being parallel (or not quite parallel) to the direction used in case.inso? c) I understand that the results for different directions of B in case.inso reflect the magnetic anisotropy, but what are the units of field (if any)? d) What else is worth looking at? The partial orbital moment (:POM) seems relevant, but what exactly is it? e) I am blindly trusting that initso knows what it is doing, and have left the B symmetry operations in case.struct (although I remember symmetry operations being split into these two classes everywhere in the code). This seems to conflict with Pavel's notes, although those may be too old. Thanks. -- Professor Laurence Marks Department of Materials Science and Engineering Northwestern University
Re: [Wien] A few (more) elementary -so questions (with onsite -eece)
Typo: although I remember don't symmetry operations being split into these two classes everywhere in the code On Mon, May 4, 2015 at 6:04 PM, Laurence Marks l-ma...@northwestern.edu wrote: I am a newbie at -so, so a few simple questions. a) What is the meaning of the orbital moment in case.scfdm* ? Is that the average direction projected to the global axis system? b) What is the physical significance of the orbital moment being parallel (or not quite parallel) to the direction used in case.inso? c) I understand that the results for different directions of B in case.inso reflect the magnetic anisotropy, but what are the units of field (if any)? d) What else is worth looking at? The partial orbital moment (:POM) seems relevant, but what exactly is it? e) I am blindly trusting that initso knows what it is doing, and have left the B symmetry operations in case.struct (although I remember symmetry operations being split into these two classes everywhere in the code). This seems to conflict with Pavel's notes, although those may be too old. Thanks. -- Professor Laurence Marks Department of Materials Science and Engineering Northwestern University www.numis.northwestern.edu Corrosion in 4D: MURI4D.numis.northwestern.edu Co-Editor, Acta Cryst A Research is to see what everybody else has seen, and to think what nobody else has thought Albert Szent-Gyorgi -- Professor Laurence Marks Department of Materials Science and Engineering Northwestern University www.numis.northwestern.edu Corrosion in 4D: MURI4D.numis.northwestern.edu Co-Editor, Acta Cryst A Research is to see what everybody else has seen, and to think what nobody else has thought Albert Szent-Gyorgi ___ Wien mailing list Wien@zeus.theochem.tuwien.ac.at http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien SEARCH the MAILING-LIST at: http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html