abst_1001_05.pdf
This is the very beginning of the 1001_05_y-waga_kyushu_u channel
@ISHI, Yoshihiro created this channel at Sep 10th, 2021.
Friday, September 10th, 2021
ISHI, Yoshihiro 16:58 PM
joined #1001_05_y-waga_kyushu_u.
Y. Kuriyama (KURNS, LOC) 16:58 PM
was added to #1001_05_y-waga_kyushu_u by ISHI, Yoshihiro
Koki Kozuka 16:58 PM
was added to #1001_05_y-waga_kyushu_u by ISHI, Yoshihiro
Monday, September 13th, 2021
Tuesday, September 14th, 2021
yasutoshi test 14:16 PM
joined #1001_05_y-waga_kyushu_u.
Yoshiharu Mori 14:22 PM
joined #1001_05_y-waga_kyushu_u.
Thursday, September 16th, 2021
Alex Pertica 18:01 PM
joined #1001_05_y-waga_kyushu_u.
Adam Steinberg 18:38 PM
joined #1001_05_y-waga_kyushu_u.
Carl Jolly 19:19 PM
joined #1001_05_y-waga_kyushu_u.
Laurence Wroe 19:44 PM
joined #1001_05_y-waga_kyushu_u.
Ryan Bodenstein 20:52 PM
joined #1001_05_y-waga_kyushu_u.
Friday, September 17th, 2021
Shinji Machida 17:09 PM
joined #1001_05_y-waga_kyushu_u.
Frank Zhang 22:03 PM
joined #1001_05_y-waga_kyushu_u.
Saturday, September 18th, 2021
Dejan Trbojevic 01:36 AM
joined #1001_05_y-waga_kyushu_u.
Sunday, September 19th, 2021
Jacinta Yap 15:51 PM
joined #1001_05_y-waga_kyushu_u.
Tuesday, September 21th, 2021
Marion Vanwelde 17:09 PM
joined #1001_05_y-waga_kyushu_u.
Wednesday, September 22th, 2021
Yuhi Waga 17:14 PM
joined #1001_05_y-waga_kyushu_u.
Kyosuke Adachi 17:17 PM
joined #1001_05_y-waga_kyushu_u.
EMI YAMAKAWA 21:45 PM
joined #1001_05_y-waga_kyushu_u.
Thursday, September 23th, 2021
Hiroki Okuno 20:27 PM
joined #1001_05_y-waga_kyushu_u.
Wednesday, September 29th, 2021
Thursday, September 30th, 2021
Haj Tahar Malek 14:14 PM
Dear Yuhi, thank you for the nice talk. It is nice to see that the tune excursion can be reduced by remodeling the magnet. Nevertheless, I have few comments: 1) In slides 8 & 9, you quote the formula: qh^2=k+1 and qv^2=-k+F^2. Unfortunately those formula, although they hold for machines with low k-values i.e. most cyclotrons nowadays, they break down for FFAG with large k-values because the Alternating Gradient focusing in FFAG is missing there. For instance, you showed that, after remodeling the ring, k=8 and F^2=7.6. This implies that the theoretical tunes are qh^2=9 and qv^2=-0.4 which is not possible. I believe your calculations are based on tracking. Nevertheless, the remaining tune excursion is, to a large extent, driven by the changes of the Alternating Gradient focusing due to closed orbit distortion (RF cavity for instance is a source) and azimuthal variations of k. Could you comment on that?
14:16
2) In order to understand the origin of the imperfections, it is very useful to plot the map of the local k i.e. R/B*dB/dR everywhere in the ring which is what we did with Francois in 2015 3) Scaling in this concept should be valid locally i.e. at every point in the ring. A constant average k will not suffice. Clearly it is impossible to match a 2D map with a theoretical 2D function. So, why not fix the scaling in an average sense? The tune is the number of betatron oscillations per turn, so it is an average quantity just like isochronism is for cyclotrons. This means that only a few magnets shall be corrected to compensate for the already existing imperfections.
Yuhi Waga 15:27 PM
Thank you for your comment. I am not good at English, so my answer may be irrelevant.
1)The shape of the structure attached to the magnet was optimized based on the variation of betatron tune obtained using tracking. It was not optimized based on effective local k and flutter factor. Effective local k and flutter factor were intended to be used for qualitative explanation.
I agree with your comment about the cause of the remaining tune excursions. For the remodeled magnet, the distortion of the closed orbit has not yet been considered, so tune excursions are expected to increase when that is considered.
2)Although not completely local k, effective local k was calculated separately for the focusing and defocusing magnets. As shown in the figure on the upper right of page 1 of the additional material, the boundary is the place where the positive and negative magnetic fields are swapped. As a result, even in the cace of after remodeling, effective local k of defocusing magnet is not constant. Therefore, as shown in the page 2 of the additional material, the magnetic pole shapes of the focusing and defocusing magnets were changed so that effective local k, effective local k of focusing magnet, and effective local k of defocusing magnet are 7.5.
As shown in the page 3 of the additional material, the variation of horizontal betatron tune can be reduced by further optimization. But this is not actually planned in this study, as further optimization would result in almost another accelerator.
1)The shape of the structure attached to the magnet was optimized based on the variation of betatron tune obtained using tracking. It was not optimized based on effective local k and flutter factor. Effective local k and flutter factor were intended to be used for qualitative explanation.
I agree with your comment about the cause of the remaining tune excursions. For the remodeled magnet, the distortion of the closed orbit has not yet been considered, so tune excursions are expected to increase when that is considered.
2)Although not completely local k, effective local k was calculated separately for the focusing and defocusing magnets. As shown in the figure on the upper right of page 1 of the additional material, the boundary is the place where the positive and negative magnetic fields are swapped. As a result, even in the cace of after remodeling, effective local k of defocusing magnet is not constant. Therefore, as shown in the page 2 of the additional material, the magnetic pole shapes of the focusing and defocusing magnets were changed so that effective local k, effective local k of focusing magnet, and effective local k of defocusing magnet are 7.5.
As shown in the page 3 of the additional material, the variation of horizontal betatron tune can be reduced by further optimization. But this is not actually planned in this study, as further optimization would result in almost another accelerator.
Friday, October 1st, 2021
Stephen Brooks 01:43 AM
This is a very good improvement. Perhaps if you want to make the tune even more constant, it will require pole-face windings and experimental correction of the tune, based on measured tune variation.
Yuhi Waga Oct 1st, 2021 at 12:57:40 PM
thank you for your comment. I think so too.
J. Scott Berg 05:49 AM
I observed a couple things which you may wish to comment on:
1. In your remodeling, the vertical tune was reduced significantly. I can understand why this might happen based on your modifications to the D portion of the magnet. But was this the desired result? Had you thought much about your working point, or were your solely trying to reduce the tune variation. It appears as though a slightly higher tune in both horizontal and vertical might be favorable.
2. Your computed tunes only get up to 100 MeV. Is the machine not expected to be able to reach higher energies?
1. In your remodeling, the vertical tune was reduced significantly. I can understand why this might happen based on your modifications to the D portion of the magnet. But was this the desired result? Had you thought much about your working point, or were your solely trying to reduce the tune variation. It appears as though a slightly higher tune in both horizontal and vertical might be favorable.
2. Your computed tunes only get up to 100 MeV. Is the machine not expected to be able to reach higher energies?
Yuhi Waga Oct 1st, 2021 at 12:37:58 PM
I too think it would be better if the horizontal and vertical tunes were a bit higher, to avoid resonance lines. But I only focused on the variation of betatron tune.
The return yoke attached to the focusing magnet is the cause of the decrease in vertical tune. The improved magnetic independence of the F and D magnets has resulted in a smaller field for the D magnet. Therefore, the current value of the coil of the D magnet was increased. However, due to the performance of the available power supply, it was not possible to increase the vertical tune any further. For the same reason, the remodeled magnet can only accelerate up to 90 MeV. If a better power supply can be obtained or the number of turns in the coil can be increased, the vertical tune and energy can be increased.
To increase the horizontal tune, it is necessary to change the geometry of both the F and D poles. In this study, the aim was to reduce the variation of tune with as few modifications as possible. So, changing the geometry of both the F and D poles is not planned in this study.
The return yoke attached to the focusing magnet is the cause of the decrease in vertical tune. The improved magnetic independence of the F and D magnets has resulted in a smaller field for the D magnet. Therefore, the current value of the coil of the D magnet was increased. However, due to the performance of the available power supply, it was not possible to increase the vertical tune any further. For the same reason, the remodeled magnet can only accelerate up to 90 MeV. If a better power supply can be obtained or the number of turns in the coil can be increased, the vertical tune and energy can be increased.
To increase the horizontal tune, it is necessary to change the geometry of both the F and D poles. In this study, the aim was to reduce the variation of tune with as few modifications as possible. So, changing the geometry of both the F and D poles is not planned in this study.
Francois Meot 07:43 AM
In the RACCAM spiral FFAG magnet, the poles were chamfered with R-dependence of the chamfer width (see Planche et als., NIM). This contributed improving the flutter (in addition to field clamps) and thus vertical tune stability with R. Could this help here?
Yuhi Waga Oct 1st, 2021 at 12:55:21 PM
Thank you for the advice.
I think that improving the flutter factor by r-dependent chanfer is also valid for the 150 MeV FFA accelerator. However, the method of attaching additional poles to the sides of the poles is easier because it does not require cutting the poles. Changing the D-pole shape is also considered by adding iron instead of cutting the current pole surface.
I think that improving the flutter factor by r-dependent chanfer is also valid for the 150 MeV FFA accelerator. However, the method of attaching additional poles to the sides of the poles is easier because it does not require cutting the poles. Changing the D-pole shape is also considered by adding iron instead of cutting the current pole surface.
JB Lagrange 17:42 PM
Impressive! I think you could improve it further by playing with the clamps to change the integrated field as a function of radius. I could not see on the figure, did you connect the clamps and the D magnet, or are they separated?
Yuhi Waga Oct 1st, 2021 at 17:56:23 PM
Thank you for your comment.
Field clump and D magnet are separated.
I had previously examined the case where the field clamp shape was the same as the magnetic pole shape, or flat, but the effect on the tune was small. Probably this is because the edge field is small in the modded electromagnet.
Field clump and D magnet are separated.
I had previously examined the case where the field clamp shape was the same as the magnetic pole shape, or flat, but the effect on the tune was small. Probably this is because the edge field is small in the modded electromagnet.