Inelastic Intensity Inconsistency

[GavinPhD]

I have two plots (shown below) that seem to have a discrepancy for what the intensity should be at a given reciprocal lattice point (0,0,0). If you look at the [-HH0] slice below it is much higher intensity at H=0 then the [H00] slice below. Is there something in the way the spectra is being calculated that would cause this?

[wardsimon]

Hey Gavin, Is there any chance of getting the color bar to go along with the plots? Also, in the spectra structure returned from obj.spinwave can you find the corresponding H=0 index in index = all(spectra.hkl == [0 ; 0; 0], 1) and check the difference in the spectra.Sab(:,:,:,index). There is no reason why the intensities should be different as point calculation is non-contiguous, so I'm guessing something has happened with plotting.

[GavinPhD]

Hey Simon, Thanks for the quick response. I checked Sab for H=0 on both and they seem to have the same values (clips 1 and 2 below). After posting the plots above, I realized that I hadn't checked the color scale. Turns out they were different, however, remedying that did not fix the intensity discrepancy. I have included the plots with colorbars below.

H00 Sab and hkl values:

H-H0 Sab and hkl values:

H00 Spectra with colorbar:

H-H0 Spectra with colorbar:

[wardsimon]

Can you try:

spectra = obj.spinwave({[-1.5 0 0], [0 0 0], [1.5 -1.5 0], 301})

This will be traversing all Q vectors at once. Then:

spec = sw_neutron(spec);
spec = sw_egrid(spec,'component','Sperp')
figure;
sw_plotspec(spec)

This will give the correct output (I hope!). If it does, then the difference comes from either different components or convolution doing something with normalization.

[GavinPhD]

I ran the code you gave and got the left plot below, which looks fine. However, I found that when I enabled the g-tensor in obj.spinwave, I suddenly get the intensity discrepancy (right plot below), so it seems it may be deeper issue.

[wardsimon]

Hey Gavin, I've been looking into this and can see theoretically where it arises. Is it possible that you could email me the script so that I can do a bit more testing of the specific problem? Then I'll report back :-)

[wardsimon]

I've played around with the code and think it is a peculiarity with the system, where there is a singularity at [0, 0, 0]. I cant say why it occurs, but if you look at the structure factor maps the same thing happens at [0, 0, 0] but along [0, K, 0] with out the g-tensor albeit a far lesser extent.

Horace code

You can use the Horace interface and play around with this and see if it makes sense.

horaceObj = d4d(obj.abc,[1 0 0 0],[-0.5,0.05,0.5],[0 1 0 0],[-0.5,0.05,0.5],[0 0 1 0],[-0.5,0.05,0.5],[0 0 0 1],[0,0.01,10]);
horaceObj = disp2sqw_eval(horaceObj,@obj.horace,{'component','Sperp','gtensor',true}, 0.09);
horaceObj2 = d4d(obj.abc,[1 0 0 0],[-0.5,0.05,0.5],[0 1 0 0],[-0.5,0.05,0.5],[0 0 1 0],[-0.5,0.05,0.5],[0 0 0 1],[0,0.01,10]);
horaceObj2 = disp2sqw_eval(horaceObj2,@obj.horace,{'component','Sperp','gtensor',false}, 0.09);

[H, K, 0] plane

With g-tensor	Without g-tensor

[H, 0, L] plane

With g-tensor	Without g-tensor

[0, K, L] plane

With g-tensor	Without g-tensor

[GavinPhD]

Thanks for looking into that, Simon. I am not sure why there would be a singularity at (0, 0, 0) for this system (as we definitely don't see one experimentally). If I happen to discover why that exists at some point I will let you know.