Bug #1460

Updated by Sanchez del Rio, Manuel almost 10 years ago

Submitted by

Dear Frank,

I has a deep loop to your problems and the mosaic model and the situation is problematic.

1) The model had always problems to calculate the output direction when the angle is far from the Bragg one. However, as the intensity was very small, it was not a fundamental problem.

2) I have seen that the current version using the new shadow3 kernel is buggy: the exit direction after diffration is only on the positove side (x') rather than in both. It is due to a problem calling the random generator. It is not difficult to fix, I did it in my version.

3) I remember that I did some modifications in the algorithms that avoided to create an unrealistic "hole" in the image, as seen in the attached plot. I had solved it in the past, but unfortunately I could not find the code, which was not implemented in the distributed versions, I think. The good news is that I found my hand notes, so I could redo it, but takes some time.

4) Curved crystals should work after doing the fixes.

So, the conclusion is that it is not very safe to use the mosaic model in the current version. I will look into this problem, but I cannot do it very quickly (may be before one month or two).

I can send you my notes. The modifications are in the "mosaic" routine. It is not difficult to introduce them, but requires compilation. If you want to try, I can send you the material. In any case, I put this problem in my "to do" list with high priority.



On 08/22/2012 05:00 PM, Filsinger, Frank wrote:

Dear Manuel,

I am currently trying to use the latest release of the XOP (2.3) and ShadowVUI (1.11) to simulate a beamline with mosaic crystals. In the course of these simulations I encountered two problems and I was wondering if you know what these problems might be related to:

1) If one uses a mosaic crystal, the software can only deal with a very narrow range of incidence angles around the Bragg angle (roughly +/- 3° for a fixed energy of 2830 eV, the crystal parameters as shown in the attached screen shot, and a mosaicity of 1°). When simulating a non-mosaic crystal the reflectivity simply approaches zero for angles that are much smaller / larger than the Bragg angle. However, there seems to be no min / max incidence angle for the ray tracing in this case. I was wondering if this behavior is intrinsic due to the way the mosaic crystal is implemented ? If yes, is there a workaround to deal with large /small incidence angles ?

2) When using surface geometries with mosaic crystals that are non-planar, the ray tracing does not seem to work properly. I attached a screenshot of the BLViewer for a spherical mosaic crystal as an example. Similar problems occurred with other non-planar surfaces. Is the software supposed to work for non-planar mosaic crystals or is it supposed to be restricted to planar geometries ?
I also tried a workaround by using the waviness preprocessor together with a superimposed surface error to define a spherical surface “by hand”. This approach seems to work.

Thanks in advance for your time and best regards


P.S.: Please find the two corresponding workspaces attached.