Sagittally bent Double Laue Mono (horizontally focusing) - problem with exit vertical divergence

[bernd73meyer]

I am new user of XRT (used to SHADOW). Trying to replicate the sagittally bent Laue mono as in the paper attached. I have used the BentLaue2D class for it (see xrt-script attached). Rsag is convex for both crystals, Rmer is concave. The Bragg incidence is lower case. As the results show, there is an offset of ~ -4mrad in vertical divergence after the first crystal. The horizontal focus size is right. In general, I am still a bit struggles with the rotation direction (clockwise or counterclockwise) of the local coordinates/rotations and inside the local_n, local_z methods. I tried to change the asymmetric angle, pitch etc., but then it doesn't focus in the vertical...
About to get a reasonable flux estimate I plan to overwrite the get_amplitude method by using the rocking curves obtained by XOPPY crystal routine and assign an estimated reflectivity related to the incidence angle of each ray. Does it sounds viable?

model from Glow as reference:

SBDL_XRT.zip

[kklmn]

I recommend setting materials to None initially. When you know your geometry is correct, add the right materials.

[bernd73meyer]

Update: the problem is the diffraction order. In this case, incidence is below the planes, therefore it should be -1. In 'oes_base.py' the order for Laue case always enters as 1 ('_grating_deflection.'-function). This _grating_deflection function is a very beautiful general solution for most of the optics. But, in this case, lower diffraction order happens just in one direction, in sagittal direction order should be 1. Now I don't know how to implement this, and it is in the main code....

[kklmn]

You still have the freedom to define (a) the sign of your asymmetry, (b) add π to your pitch if you see the crystal plane normal has the wrong sign, (c) define your OE class with whatever orientation of both normals - of surface and crystal planes. I haven't tried your code but I would blame the sign of asymmetry.

[bernd73meyer]

After checking all options, I discovered that the incident beam direction in z (lb.c) is rotated by thetaBragg in the wrong direction. This actually causes the small divergence offset of 4mrad in pitch (because of the wrong lb.c input in _grating_deflection). Beam focuses as expected in sagittal direction. I don't know how the directions of the incidence beam are calculated. Maybe as a temporary solution I could rotate the beam back? But how exactly? with raycing.rotate_beam()?

[kklmn]

Then give your thetaBragg the right sign if you see it is wrong. What is the problem?

[bernd73meyer]

Attached you find a sketch of the configuration and how it is used in the code. Also the plots of the beam size and divergence after the Crystal1 and 2, and the beam size at focus position. You see, that using the correct sign of theta, there is this ~4mrad vertical offset and focus in sagittal is OK. If you change sign of theta (pitch), there is no vertical angular offset, but sagittal focus is wrong. I wast a week already trying to define the local OE and asymmetric angle + incidence angle (pitch ) in all possible ways, but always get the exact results. Because it is related to the sign of theta added to pitch. It affects not only pitch (lb.b), but also yaw (lb.c). This is the core problem of this case. And as I said before, there is somehow the wrong sign used in the rotation of yaw (lb.c).
I also thought, that this offset could be correct and the second crystal corrects for this. But if changing the sign of asymmetric angle, the result keeps absolutely the same.

xrt_sbdl.pdf

XRT_XPD_BNL_testBL2D.py