The first result is expected to be produced until the 6th ACFA-LC workshop on 15-17 December, 2003, India. Comprehensive results are expected to be presented at next International LC physics/detector workshop (LCWS04) on 19-23 April 2004, Paris.
(2)LINAC simulation (T. Yamamura)
(transparencies, 14 pages, pdf,16.2MB)
Yamamura continued to investigate the effect of non-linear optics at the main LINAC for maximum(injection) beam energy of 500 (8) GeV. In terms of previous observation of less acceleration with large amplitude, he found a mistake in the initial phase conditions. Right condition is 90 degree behind the previous one, that is, tail of a bunch shall be accelerated more than head of the bunch. With this correction, particles with larger amplitude are accelerated more. For an example, when xi=+/-8mm, dp/p and dz become to +2.5% and +/-0.75mm, respectively. He investigated validity of this result by phase-change caused by dz; i.e. dphi=2 pi dz/lamda , where dphi is phase-change and lamda=2.63cm of x-band, in dividing the LINAC into 2 sections at s=8km and Ebeam=300GeV. Therefore, dphi = 0.188 for dz=0.75mm, where zi=0.0. Assuming phi=-1.61 and -2.09 at s<8km and s>8km, respectively, beam energy with xi=8mm was simply estimated to be 516.5GeV. So, dp/p=2.5% must be relevant.
He also confirmed that the non-linear effects would appear visible only at B3/B2=0.1. Rotating the sextupole fields randomly or fixed at several angles, non-linear effects of final positions were calculated to be less than 4nm .
He would like to investigate scattering effects with the non-linear optics.
(3) Progress report of support tube R&D (H.Yamaoka)
(transparencies, 12 pages, pdf, 1MB and ppt(9.1MB))
Yamaoka previously reported the results of 1/10 prototype with taper flanges (57th FFIR/BDSIM meeting on 7/30/2003). In this time, he reported the case of flat flanges. He also reported spectrum analyses of the support tube with ground motions measured at ATF.
Test procedures of prototype with flat flanges were the same as that with the taper ones. He observed four major resonances at 81.5, 97.7, 266 and 541Hz, which should be compared with 76, 256 and 489Hz of the FEM analysis. So, the first two are "hyper fine" split ones due to imperfect fixed-points. Effects of joint strength were also measured in resonant frequencies by changing a number of bolts (M6). The effects were apparent, that is, about 3 % less resonance frequencies with half number of bolts, while there is no significant effect for that with taper flanges. Next, he will simulate heavy mask and quadrupole magnet by dummy weight of 1kg.
Inside the detector, additional support can be available at the edge of iron yoke as already discussed. He updated the spectrum analyses of the support tube inside the detector with and without the additional supports. Input ground motion was the measured one at ATF. The first and second resonances moved from 18 and 89Hz to 76 and 225Hz, respectively, by the additional supports. Power densities were 4 x 10-19m2/Hz and 1 x 10-23m2/Hz at 76 and 225Hz, respectively.
There was a suggestion to input measured ground motions directly to the ANSYS calculations rather than the power density spectrum. In principle, it is possible. So, he will try it.
(4) FEATHER status (N. Delerue )
First, a graduate student(M1) of Tokyo Metropolitan university, Hiroyuki Fujimoto was introduced as a new collaborator to the FEATHER project. His supervisor is Prof. Takayuki Sumiyoshi.
Nicolas briefly reported present status of FEATHER. For the beam test, a kicker and BPM were installed at the ATF extraction line. Positions of these movable electrodes have been calibrated in the atmosphere before the installation. However, the positions of the kicker-electrodes were found to be changed in vacuum because of bellows for moving. Magnitude of this movement was 0.6mm both for up/down electrodes. Fortunately, their levelness was kept even in vacuum.
Impedance(Z0) of the electrodes was also measured by the TDR(time domain reflectometry ) method based on a measurement of reflected pulse (stepwise signal with rise time of 20~50ps) from electrodes with unknown impedance ; where Z0 =((ƒ¢V1/(2ƒ¢V0-ƒ¢V1))x50[ƒ¶], ƒ¢V1 is potential in a 50ƒ¶ cable and ƒ¢V0 is a reflected potential from electrodes with unknown impedance. After examining several termination methods of connectors on the kicker-electrodes, the impedance could be successfully measured with one of electrodes grounded at the connectors. Results seems to be consistent with the calculations. We also found that velocity of RF pulse transmitting on the electrodes is almost same as light velocity in vacuum. ( This part of minutes is added by T.Tauchi, including informations got after the meeting. At this meeting, a few problem were reported in terms of impedance measurements which were resolved as mentioned above.)
The beam test will begin on 24 October.
(5)GM effects at LINAC by SLEPT (K. Kubo)
(transparencies, 6 pages, pdf, 32KB)
Kubo has tried to reproduce some of Linda Hendrickson's results presented at the video conference on 14 August.
Present calculation assumed as follows;
Emittance growth was calculated with no feedback correction. Result was ƒ¢ƒÃ/ƒÃo = 13.7 while Linda's was 3.7 . The reason is unknown at present.
Based on optics calculated by quadrupole magnets, so-called "calculated model", the growth was calculated for 8, 10, 12 sections/linac as a function of iteration number by the NLC type steering feedback with the GM. No error of the BPMs was assumed. In order to recover the growth, at least 10 sectors and two iterations are needed.
Once the BPM responses are well calibrated by steering beams, so called "measured steering-BPM response", no iteration shall be needed. The calibration must include effects of wakefield, beam energy spread. The emittance growth was calculated both for the NLC and SLC type feedbacks as a function of sector-number. No error of the BPMs was assumed, too. As Linda's result, the NLC type has better performance than the SLC. However, the gowth was larger than Linda's (LH); i.e. ƒ¢ƒÃ/ƒÃo =44% vs LH=9% for 10 sectors/linac.
Finally, for a case of 12sectors/linac and the NLC type feedback, the emittance growth was calculated as a function of BPM resolution with or without an iteration. Result shows that at least one iteration and BPM resolution of 0.2um is needed.
He will investigate the discrepancies in details. Especially, it is noted that there is a discrepancy even without feedback.