There are 3 VITESS modules to simulate collimators: collimator for a usual Soller collimator, collimator_radial for a radial collimator (around the sample) and the old collimator_soller, which can be used for both types of collimators, but only removes a certain fraction of the neutrons without checking their paths through the collimator.
The module collimator simulates a soller collimator of a certain length, width and height consisting of a number of channels separated by absorbing walls of a certain thickness.
Their walls are vertical planes parallel to the x-z-plane. There is no separation in vertical direction.
Exit width and height can be different from entrance width and height.
For each trajectory, it is checked whether the neutron is outside the exit or entrance area and if a channel wall is hit; in these cases the trajectory is removed.
Otherwise it is propagated to the collimator exit (without loss in intensity).
This is a realistic description of a Soller collimator and is therefore the recommended choice.
| Parameter Unit |
Description |
Range or Values |
Command Option |
| entrance width exit width [cm] |
Total width of the collimator entrance and exit (between the two walls defining beginning of first and end of last channel) (center of the entrance window at y=0) |
>= 0 | -w -W |
| entrance height exit height [cm] |
Entrance and exit height of the collimator (center of the entrance window at z=0) |
>= 0 | -h -H |
| length [cm] |
Length of the collimator | >= 0 | -l |
| number of channels | Number of collimator channels along the y-axis. There are no channels in vertical direction. |
>= 1 | -n |
| wall thickness [cm] |
Thickness of the blades dividing the collimator into channels Zero thickness is also possible and still gives 100% absorption. |
>= 0 | -s |
In contrast, the module collimator_soller simulates a collimator by allowing a certain divergence range (in the x-y-plane), which is described by a triangular distribution
characterising the probability that a neutron of a certain divergence passes through the collimator. The maximal possible divergence δφ then corresponds to the width
w and length l of a Soller collimator channel, i.e. tan δφ = w/l. The FWHM of the probability distribution is also δφ, (the peak is centered at divergence 0).
The module collimator_soller does neither change the neutron position or direction nor the frame of reference.
This method of simulating collimation is a good approximation. If this module is used, entrance and exit area should be simulated in addition using either the module slit
or spacewindow to allow for the spatial limitation and take care of the propagation.
This module can also be used to simulate a radial collimator. In this case, the direction of the first channel, the number of channels and the 'additional angular spacing' Δφ
have to be given:
Each collimation center is defined by an angle corresponding to zero divergence. The first centre is defined by the 'minimum of angle range' φmin.
The following centers (always at higher angles) are calculated by considering a gap of (2*δφ + Δφ) between two centers.
Therefore, the direction of the last center can be calculated as:
φmax = φmin + (Ncenters - 1)*(2*δφ + Δφ)
Note that the collimator does not have any size, if represented by this module. For the visualization, fixed sizes of 20 x 10 x 10 cm³ (L x W x H) for the Soller collimator and 10 cm height, 50 cm radius and 5 cm thickness for the radial collimator are assumed. It shows 3 channels in each collimation center and up to 20 channels in the Soller collimator.
| Parameter Unit |
Description |
Range or Values |
Command Option |
| angular collimation |
yes: radial collimation with several collimation centers (= directions regarded as zero-divergence directions) no : normal Soller collimator (only 1 zero-divergence direction) |
'yes', 'no' | -k |
| collimator divergence [deg] |
Maximal horizontal divergence δφ (= FWHM of the triangular shape) | > 0 | -d |
| peak transmission | Maximal probability for passing through the soller collimator (corresponds to the transmission of neutrons with zero divergence) | ]0, 1] | -e |
| minimum of angle range [deg] |
The lowest angle φmin defining a zero-divergence direction 0 deg: direction of the beam impinging on the sample (= x-axis) 90 deg: to the left (= y-axis) The following centers (always higher angles) are calculated by considering a gap of 2*δφ + Δφ between two centers. |
[-180°, 180°[ | -m |
| number of collimation centers | Number of orientations Ncenters corresponding to divergence 0 (or to peak transmission resp.). The difference between two collimation centers is: 2*δφ + Δφ | >= 1 |
-n |
| angle spacing [deg] |
Additional angular distance Δφ between the collimation centers due to the size of collimator spacers | >= 0 | -a |
A more appropriate way to simulate radial collimators is to use the module collimator_radial.
It is realized by a number of channels defining different directions in the horizontal plane.
Their walls are vertical planes throught the origin; a thickness of these blades can be considered, but zero thickness works as well. All walls are treated as 100% absorbing.
There is no separation in vertical direction, but it considers entrance and exit height of the collimator.
As the module collimator it calculates the neutron trajectory through the collimator and removes it from the simulation, if it hits one of the (absorbing) walls.
Otherwise, it propagates the neutrons to the end of the radial collimator.
Furthermore, an oscillation of the collimator can be treated. If this is used, the orientation of the collimator (at the time the neutron is passing)
is determined by a random choice within a range given. In this way, the shadowing effect of the blades can be simulated and the extent of this effect estimated.
| Parameter Unit |
Description |
Range or Values |
Command Option |
| theta (center) [deg] |
Hor. direction to the center of the collimator 0 deg: direction of the beam impinging on the sample (= x-axis) 90 deg: to the left (= y-axis) |
[-180°,180°] | -a |
| width [deg] |
Angular range covered by the radial collimator (between the two walls defining beginning of first and end of last channel) Channel centers and blades are pointing to the origin, i.e. the centre of the sample. |
[0°,360°] | -w |
| oscillation width [deg] |
Full width amplitude of oscillation of the radial collimator The orientation of the collimator for a certain trajectory is randomly chosen within this range Oscillation width = 0.0 means: no oscillation regarded |
>= 0 | -a |
| entrance height exit height [cm] |
Entrance and exit height of the radial collimator The center of entrance and exit is set to z=0, i.e. the collimator is symmetrical to the x-y-plane |
>= 0 | -h -H |
| distance [cm] |
Distance of the collimator entrance from the origin (usually the center of the sample) | >= 0 | -d |
| length [cm] |
Length of the collimator channels, (i.e. distance between entrance and exit of a channel) | >= 0 | -l |
| number of channels | Number of collimator channels in the x-y-plane (defining different theta-directions) There are no channels in vertical direction. |
>= 1 | -n |
| wall thickness [cm] |
Thickness of the blades dividing the collimator into channels zZero thickness is also possible and still gives 100% absorption. |
>= 0 | -s |
Last modified: 2021-07-22 KL