You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: MDhelp/docs/phasedata.md
+27-14Lines changed: 27 additions & 14 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -26,27 +26,40 @@ For single crystal data, the only parameters are scale, extinction and disordere
26
26
27
27
***Scale factor** - Used for single crystal data: relates \(F^2_{obs}\) to \(F^2_{calc}\).
28
28
29
-
***Crystallite size peak broadening** – This is computed from size factor(s) in microns (1 μm = \(10^{-6}\) m), with the Scherrer constant assumed as unity. Sizes can be computed in three ways: isotropic, uniaxial and ellipsoidal.
29
+
***Crystallite size peak broadening** –
30
+
Peaks can be broadened due to the finite size of crystallites or due to microstrain (see below). Microstrain is more common other than in nanoparticles.
31
+
The broadening is computed from size factor(s) in microns (1 μm = \(10^{-6}\) m), with the Scherrer constant assumed as unity. Sizes can be computed with a choice of three models: isotropic, uniaxial and ellipsoidal.
30
32
Typical sensitivity for crystallite size is to no more than 4 μm (less for lower resolution instruments); beyond that the particles are effectively infinite for a diffraction experiment.
31
33
32
-
* In isotropic broadening, crystallites are assumed to average as uniform in all directions and a single size value is supplied;
33
-
* with uniaxial broadening, a preferred direction (as a crystallographic axis, such as (001) is supplied) -- note that for most crystal systems only one axis makes sense -- and two size parameters are defined, one for along the axis and one for in the perpendicular plane;
34
-
* with ellipsoidal, six terms are used to define a broadening tensor that has arbitrary orientation -- this model may require constraints and is seldom needed.
34
+
* In **isotropic** broadening, crystallites are assumed to average as uniform in all directions and a single size value is supplied;
35
+
* with **uniaxial** broadening, a preferred direction (as a crystallographic axis, such as (001) is supplied) -- note that for most crystal systems only one axis makes sense -- and two size parameters are defined, one for along the axis and one for in the perpendicular plane;
36
+
* with **ellipsoidal**, six terms are used to define a broadening tensor that has arbitrary orientation -- this model may require constraints and is seldom needed.
35
37
36
-
Note that size broadening is usually Lorentzian, which corresponds to a LGmix value of 1.0; if this value is between 0. and 1., both Gaussian and Lorentz size broadening is modeled and a value of 0.0 is pure Gaussian. Values less than 0. or greater than 1. make no physical sense. LGmix is not commonly refined.
38
+
Note that size broadening is usually Lorentzian, which corresponds to a LGmix value of 1.0; if this value is between 0 and 1, both Gaussian and Lorentz size broadening is modeled and a value of 0.0 is pure Gaussian. Values less than 0 or greater than 1 make no physical sense. LGmix is not commonly refined.
37
39
38
-
***Microstrain peak broadening** - This is computed as unitless fraction of Δd/d (or equivalently ΔQ/Q) times \(10^6\). Microstrain can be represented in three ways: isotropic, uniaxial and generalized. Typical microstrain is ~1000, but may be significantly higher.
39
-
* In isotropic broadening, microstrain broadening assumed to be the same in all crystallographic directions and a single value is supplied;
40
-
* with uniaxial broadening, a preferred direction (as a crystallographic axis, such as 0,0,1) is supplied -- note that for most crystal systems only one axis makes sense -- and two microstrain parameters are defined, one for along the axis and one for in the perpendicular plane;
41
-
* with generalized, the [Nicolae Popa](https://journals.iucr.org/j/issues/2020/06/00/es5029/index.html)/[Peter Stephens](https://journals.iucr.org/paper?hn0085) second-order expansion model is used and the number of terms will depend on the crystal system. It is typically possible to refine all terms when significant anisotropic strain broadening is present.
40
+
***Microstrain peak broadening** - This is computed as unitless fraction of Δd/d (or equivalently ΔQ/Q) times \(10^6\). Typical microstrain is ~1000, but may be significantly higher in physically processed materials. Note that the term residual stress is sometimes used for microstrain, but residual stress can be computed from microstrain when the elastic strain constants are known. Microstrain was can be computed in GSAS-II via a choice of three models: isotropic, uniaxial and generalized:
41
+
* In **isotropic** broadening, microstrain broadening assumed to be the same in all crystallographic directions and a single value is supplied;
42
+
* with **uniaxial** broadening, a preferred direction (as a crystallographic axis, such as 0,0,1) is supplied -- note that for most crystal systems only one axis makes sense -- and two microstrain parameters are defined, one for along the axis and one for in the perpendicular plane;
43
+
* with **generalized**, the [Nicolae Popa](https://journals.iucr.org/j/issues/2020/06/00/es5029/index.html)/[Peter Stephens](https://journals.iucr.org/paper?hn0085) second-order expansion model is used and the number of terms will depend on the crystal system. It is typically possible to refine all terms when significant anisotropic strain broadening is present.
42
44
43
-
Note that microstrain broadening is usually Lorentzian, which corresponds to a LGmix value of 1.0; if this value is between 0.0 and 1.0, both Gaussian and Lorentz broadening is applied and a value of 0.0 is pure Gaussian. Values less than 0 or greater than 1 make no physical sense. LGmix is not commonly refined.
45
+
Note that microstrain broadening is usually Lorentzian, which corresponds to a LGmix value of 1.0; if this value is between 0 and 1, both Gaussian and Lorentz size broadening is modeled and a value of 0.0 is pure Gaussian. Values less than 0 or greater than 1 make no physical sense. LGmix is not commonly refined.
44
46
45
-
***Hydrostatic/elastic strain** – This shifts the lattice constants for the contribution of a phase into a histogram. The values are added to the [reciprocal lattice parameter tensor terms](http://gsas-ii.readthedocs.io/en/latest/GSASIIutil.html#gsasiilattice-unit-cell-computations). They must be refined in sequential refinements or where the lattice constants are slightly different in different histograms (as an example see the [Combined X-ray/CW-neutron refinement of \(\rm PbSO_4\) tutorial](https://advancedphotonsource.github.io/GSAS-II-tutorials/CWCombined/Combined%20refinement.htm). But these values and the phase's lattice parameters (on the General tab) should not be refined at the same time. When the values are non-zero, the lattice constants after application of these strain tensor terms is shown.
47
+
***Hydrostatic/elastic strain** – This shifts the lattice constants for the contribution of a phase into a histogram. These $D_{ij}$ values are added to the [reciprocal lattice parameter tensor terms](http://gsas-ii.readthedocs.io/en/latest/GSASIIutil.html#gsasiilattice-unit-cell-computations). They must be refined in sequential refinements or where the lattice constants are slightly different in different histograms (as an example, see the [Combined X-ray/CW-neutron refinement of \(\rm PbSO_4\) tutorial](https://advancedphotonsource.github.io/GSAS-II-tutorials/CWCombined/Combined%20refinement.htm)) or may account for changes to the lattice constants due to external stress (as occurs in a high-pressure measurement.) Note that *these values and the phase's lattice parameters (on the General tab) should not be refined at the same time*. When the values are non-zero, the lattice constants after application of these strain tensor terms is shown with these $D_{ij}$ values.
46
48
47
-
<aname="Phase-Preferred_orientation"></a>
48
-
<aname="Preferred_orientation"></a>
49
-
<aname="preferred_orientation"></a>
49
+
For cubic material, an extra term, eA ($\epsilon_A$) with units of Å${}^{-2}$, is included that is particularly useful in high pressure work. This accounts for the shift of peaks due to macroscopic stress along the (111) directions:
***Preferred Orientation** – Preferred orientation (texture) can be treated in one of two different sections of GSAS-II, either the Preferred Orientation correction here in the Data tab, or the "[Texture](phasetexture.md)" tab, depending on what is desired.
52
65
The Preferred Orientation correction here is typically used for crystallographic studies, where intensity corrections are desired to repair for undesired texture in the sample, while the Texture tab is used for studies where the goal is to characterize preferred orientation in a sample.
* Installs or updates the GSAS-II files from the GitHub repo
407
-
* Creates the and installs the appropriate binary files from the
408
-
Fortran, C and Cython sources.
409
-
* Does a byte-compile on all ``.py`` files
410
-
* Creates shortcuts/icons for starting GSAS-II (OS specific)
411
-
412
-
Note that there are a number of options that can be used with the
413
-
script, for example ``python gitcompile.py --reset`` overwrites any
414
-
changes that have been made to GSAS-II files locally with the original
415
-
versions of the files. The other options are not likely to be needed,
416
-
but can be seen with ``python gitcompile.py --help``
417
-
418
-
Compiling with Scons
419
-
---------------------------
420
-
421
-
Compilation with scons (as opposed to meson, as discussed above) is not
422
-
recommended and will be removed from GSAS-II in the future. It will
423
-
work only with Python 3.11 or older and only with the ``master``
424
-
branch. Note that GSAS-II will fail with Python earlier than 3.7 and may have
425
-
some errors even with Python 3.8-3.10 as it is no longer tested on
426
-
those environments.
427
-
428
-
The compilation process requires installation of the gcc and GFortran
429
-
compilers. Others will probablu not work. Also, the Python Scons
430
-
package must be installed into Python. Compilation is then done with
431
-
commands::
432
-
433
-
cd fsource
434
-
scons
435
-
436
-
The scons file captures the compilation options needed for the supported platforms, but to compile on other platforms, it may be necessary to modify the ``Sconstruct`` file to configure for the new platform.
437
-
438
-
Installation of compilers is highly depend on the computer system being used, but in many cases they can be installed as a conda package, with a command such as::
439
-
440
-
conda install gfortran_osx-64 scons
441
-
442
-
Use the ``conda search gfortran`` command to find the name for the package.
443
-
On most linux systems, one can use a command such as ``sudo apt-get gfortran`` or ``yum install gcc-gfortran``. Also see https://gcc.gnu.org/wiki/GFortranBinaries for more information.
444
-
445
-
Note that the intent is that this Scons-based process is no longer in
446
-
use, but for reference some older web pages discussing compiling GSAS-II may be of use:
0 commit comments