First, he studied luminosities as a function of vertical beam-offset at IP by CAIN, where the beam parameters are those of the roadmap report, i.e. 1.4nsec bunch separation, 192 bunches/pulse, intensity of 0.75 x 10^{10}/bunch, sigma^{*}_{x}/sigma^{*}_{y}=243/3.0 nm, beta^{*}_{x}/beta^{*}_{y}=8/0.11mm and the nominal luminosity of 2.5 x 10^{34}/cm^{2}/s. At the offset of 4 (10) sigma^{*}_{y} the luminosity drops to 50% (17%) of the nominal one. Second, deflection angles were calculated after collisions as a function of the offset, where the angle is the average one. Values of the angle are 40, 120 and 190 micro radian at the offset of 1, 3 and 6 sigma^{*}_{y}, respectively. Then, he made a simple algorithm of the feedback with vertical beam offset only for this moment, so called "linear model", in reference to the Schulte's paper ( LCC-0026 20/09/99, CLIC Note 415). In his simulation, a key parameter is a gain ( g ) which is defined as delta/sigma^{*}_{y}= g def.angle/sigma^{*}_{y'}, where delta is the correction factor in unit of sigma^{*}_{y}, and def.angle is the deflection angle measured by BPM, and sigma^{*}_{y'} is a sigma of vertical beam divergence at IP. Trailing bunches will be kicked (or corrected) bunch by bunch with a constant gain of g. Generally, the "corrected" offset would start to oscillate if g is large. He found that such oscillation could disappear with g=0.05 and the luminosity recovered at 40th bunch after the first bunch passed through BPM.

Spread of synchrotron radiations was estimated at the final focus quadrupole magnet(QD0) across the IP, i.e. s=L^{*}=3.5m ( s=0 at IP) . The maximum divergent angles are theta_{x}=112 micro-radian and theta_{y}=25 micro-radian in horizontal and vertical directions, respectively, which are produced at the final doublet(QD0+QF1). The spread can be calculated by linear optics simply as sigma(SR)_{x}=(s+2.9m) x theta_{x} and sigma(SR)_{y}=s x theta_{y} with s=3.5m. Putting s=3.5m, we obtained sigma(SR)_{x}=0.72mm and sigma(SR)_{y}=87.6micro-m. Therefore, the spread corresponding to 11 sigma_{x} x 46 sigma_{y} is calculated to be sigma(SR)_{x}=7.93mm and sigma(SR)_{y}=4.03m. The spreads are small enough for passing through the final magnet without interaction.
He also calculated energy loss by radiations along the beam line by using SAD for two types of beam (Gaussian and flat of 20 sigma_{x} x 60 sigma_{y}) with/without octupoles (OCT on/off). The results are summarized in the table, where first, second and third numbers are transmission rate, average energy loss and number of photons per electron emitted.

OCT | Gaussian | flat (20sigma_{x}, 60 sigma_{y}) |
---|---|---|

on | 100%,9.25MeV, 44.6photons/e | 994%,21.1MeV,51.9photons/e |

off | 100%,9.10MeV, 43.8photons/e | 92.2%,17.7MeV,44.7photons/e |

There was a suggestion that SAD should be modified in order to calculate energy of photons and emitted angles for comparison with simulation results of GEANT4.

He demonstrated an effect of the tail-folding octupoles which can allow the use of wider spoiler jaw settings, say more than double. He installed MUCARLO-program in jlc.cc.kek,jp in order to estimate muon background in cooperation with Namito-san. In present BDS, there seems to be no space for 120m long muon-attenuator while only 70m long attenuator can be used. Effect of the shorter attenuator shall be studied. If necessary, it might be desirable to modify the BDS for the 120m long attenuator.