Multiple order reflections about xrt

[yxrmz]

Hi, have a look a the CrystalHarmonics class here. Definition is pretty straightforward, make sure you have pyopencl set up properly.
Same class is used in our of our xrtQook examples, BioXAS_Main.xml, you could try generate the script and play with harmonics.

[ts8guest]

Thanks a lot.

[ts8guest]

Dear developers,
We are interested in the calculation of several order reflections from a mosaic crystal in one simulation run.
Initially, we had calculated the first and the second order reflections separately.
We summed results and compared with the experiment. Some misalignment is observed (see figure).

The first question is: Is it correct to sum the results of two orders?

Afterward, we installed properly pyopencl, calculared with CrystalHarmonics class (Nmax=2) and also compared with the experiment. Unfortunately, the result is unsatisfactoty.

Could it be the problem in incorrectly defined parameters in crystalHarmonics class or parameters in OE class?
Thank you in advance.

[kklmn]

In the present implementation, CrystalHarmonics doesn't work with the mosaic model. We should probably rename it to SingleCrystalHarmonics or PerfectCrystalHarmonics.

Some misalignment is observed (see figure).

What do you mean?

The first question is: Is it correct to sum the results of two orders?

This depends on what and how you do. What are "the results" that you sum? Do you correctly model the polarization properties of your experiment? You may want to attach a script.

[ts8guest]

Thanks for your reply.
Now, that's clear about CrystalHarmonics.

Conserning the sum of first and secnd order reflections, I have to clarify. I've calculated the first order 002 reflection and the second order 004 reflection of HOPG crystal (please see figure) for the same source, OE and screen position in both cases.

After that, I added 002 and 004 reflections in order to compare with the experiment.

The souce definition is the next:

    beamLine.geometricSource01 = rsources.GeometricSource(
    bl=beamLine,
    center=[0, 0, 0],
    dx=4,
    dz=4,
    dxprime=Rm/D, #Rm - OE radius; D - source-OE distance
    dzprime=Rm/D,
    distE=r"lines",
    energies=energy,
    energyWeights=weight)

And OE definition:

beamLine.cylinder = Cylinder(
    bl=beamLine,
    center=[0, D, 0],
    Rm=Rm,
    limPhysY=[0, L],
    material=hopg)

[ts8guest]

I forgot to mention that an X-ray tube is a source in the experiment.

[kklmn]

To me, the calculations beautifully reproduce the experiment, except a slight deviation at the left of the main sharp peak.

Yes, you can sum together the total flux from the two reflections. I would do it with the same beam from the source. Like this:

def run_process(beamLine):
    source = beamLine.source.shine()

    beamOrder1global, beamOrfder1local = beamLine.cylinderOrder1.reflect(source)
    beamOrder2global, beamOrfder2local = beamLine.cylinderOrder2.reflect(source)
    # beamLine.cylinderOrder1 and 2 are OEs that differ by the used material:
    # crystals with different hkl's.
    
    # add components of the intensity matrix (aka coherency matrix):
    beamOrder2global.Jss += beamOrder1global.Jss
    beamOrder2global.Jpp += beamOrder1global.Jpp
    # Jsp = 0 in your case
    # same for the local beams if you use them
    
    # work with beamOrder2global for further propagation

Add to your definition of the source:

    GeometricSource(....    polarization=None, .....)

By default, polarization is horizontal, but your source is unpolarized.

[ts8guest]

Ok, now I see. Thank you very much for your help!