View source for Cold Chopper suggestion by Mads Bertelsen

This site contains details on the performance and specifications of the guide which I suggested could fill the requirements for the cold chopper instrument.
The concept is simple and can be seen in figure 1, an elliptical guide from as close to the moderator as possible to the sample with a beamstop in the middle of the guide to stop line of sight from moderator to sample. As far as I know, the Idear for this concept was by Uwe Filges.

[[File:mads-cold_chopper-sketch-no-cut.png|800px|center|thumb|Figure 1: A sketch of the guide concept which shows how the beamstop blocks line of sight.]]

The last 10 m of the guide was however removed because they were not needed as the divergency requirement was quite low and this gives a smoother spatial distribution which was a specific requirement as well. This is shown in figure 2. It  will still be possible to block line of sight, as shielding can be placed in the same manner as the guide which was removed without performance loss.

[[File:mads-cold_chopper-cut.png|800px|center|thumb|Figure 2: A sketch of the guide concept where the last 10 m of the guide have been removed.]]

As the figures show, the guide starts 2 m from the moderator and after 74.75 m there is a beamstop with dimensions just big enough to block line of sight. Then the remaining 64.75 m of the elliptical guide followed by 10m of shielding and 0.5 m from the guide end to the sample.


The performance of the instrument is shown through snapshots at specific wavelengths, 2Å,4Å,6Å and 10Å. The results are in brilliance transfer where relevant, and the integration box is always the same as the requirements: 0.4deg divergence, and 30 mm wide and 50 mm high sample.

'''Position monitors'''

[[File:Mads-cold_chopper-Type1-psd_lin-2AA.png|600px|thumb|center| Position monitor for 2Å snapshot.]]
[[File:Mads-cold_chopper-Type1-psd_lin-4AA.png|600px|thumb|center| Position monitor for 4Å snapshot.]]
[[File:Mads-cold_chopper-Type1-psd_lin-6AA.png|600px|thumb|center| Position monitor for 6Å snapshot.]]
[[File:Mads-cold_chopper-Type1-psd_lin-10AA.png|600px|thumb|center| Position monitor for 10Å snapshot.]]
From the above data it is clear that there is no big issues with uneven beam intensity on the sample. There is a small bump at 2Å which is about a 10% change between highest and lowest intensity.

'''Divergency monitors'''

[[File:Mads-cold_chopper-Type1-div_lin-2AA.png|600px|thumb|center| Divergency monitor for 2Å snapshot.]]
[[File:Mads-cold_chopper-Type1-div_lin-4AA.png|600px|thumb|center| Divergency monitor for 4Å snapshot.]]
[[File:Mads-cold_chopper-Type1-div_lin-6AA.png|600px|thumb|center| Divergency monitor for 6Å snapshot.]]
[[File:Mads-cold_chopper-Type1-div_lin-10AA.png|600px|thumb|center| Divergency monitor for 10Å snapshot.]]
The divergency monitors do show signs of the beamstop in the middle of the guide. There is a dip at the center for each horizontal and a dip in each vertical where the placement is wavelength dependent because of gravity. Other than this dip, there is no problems to report, and it should be mentioned that the brilliance transfer drops of very quickly for divergencies larger than the instrument requirement. This is a result of the cut in guide length.

'''2d monitors'''

[[File:Type2-samlet-2AA.png|600px|thumb|center| 2d monitors for 2Å snapshot.]]
[[File:Mads-cold_chopper-Type2-samlet-4AA.png|600px|thumb|center| 2d monitors for 4Å snapshot.]]
[[File:Mads-cold_chopper-Type2-samlet-6AA.png|600px|thumb|center| 2d monitors for 6Å snapshot.]]
[[File:Mads-cold_chopper-Type2-samlet-10AA.png|600px|thumb|center| 2d monitors for 10Å snapshot.]]
The 2d monitors should be as boring as possible within the black box that shows the figure of merit. Especially in the 2Å case there is some structures. For higher wavelengths these problems does not seem as serious. Notice that the hole in the divergency plot created by the beamspot have a placement which is wavelengt dependent because of gravity. For the 10Å this hole becomes mirrored in the guide and creates some strange effects, but these are generally outside of the figure of merit box and quite small.

'''Wavelength monitor'''
[[File:Type3-wavelength-all-AA.png|600px|thumb|center| Brilliance transfer as a function of wavelength.]]

The brilliance transfer as a function of wavelength looks very good, as it is about 95% in the entire useful region without line of sight to the moderator. There is also quite a sharp drop off for lower wavelengths which is desired and can be tuned by changing m-values of the mirrors.

'''Parameters'''

The focus points are different for the mirrors affecting the horizontal and vertical movement of the neutron. For the horizontal movement, meaning the vertical mirrors, the focus points of the ellipse is 2.25 m before the start of the guide and 1.03 m after the guide end. The guide is 40 cm wide at the widest point. For the vertical movement, meaning the horizontal mirrors, the focus points of the ellipse is 2.02 m before the guide start and 0.899 m after the guide end, and also 40 cm wide at the widest spot. The optimisation crept towards the largest possible width of the guide and thus the selected width is the maximum I would consider possible. The m values for the mirrors is 6 for the first 10% of the guide length (~15 m), and m=3 for the rest of the guide. This have not been minimized, it should be possible to lower these values a lot without sacrificing performance by introducing cost into the figure of merit. 


'''Conclusions'''

In conclusion this design have some problems with dips in divergency from the beamstop, but the guide does remarkably well when it comes to raw flux. At the meeting it was pointed out that the design transport a lot of unnecessary neutrons which are then absorbed close to the sample which is a definite problem. For the next meeting I will look at ways to overcome this problem.

- Mads Bertelsen