**(1) prototype the support tube:
** (transparencies, 6 pages of pdf,
880kB) or tiff, 880kB
Yamaoka explained a study plan by constructing miniatures of support tubes. The prototype
system consists of two support tubes which are connected by CFRP or its equivalent. The
overall scaling factor is about 1/10, so the total length is 1.6m and the diameter is
50-100mm. The tube is actually solid rod of Aluminum. Properties relevant to spectrum
analysis, such as natural frequencies, should be very similar to the real system.
He also introduced an oscillator and seismometer whose typical output is 1V/13 micro-m
at 1Hz ( 13 nanometer / 1 mV ). One possible problem of this seismometer is too big for
the prototype support tube. As another monitor, Sanuki showed an laser interferometer
which his group (Tokyo univ.) has and whose sensitivity is 10 nanometer.
pdf,85kB) or
tiff, 132kB
We discussed on how the prototype test can simulate the real one since two scales are
very different. The primary purpose is to verify our ANSYS calculation for oscillation
problems. The plan will be re-examined especially on the size of prototype.

**(3) Dump line:** (transparencies, 7 pages of pdf,930kB) or
tiff, 930kB
Miyamoto showed the magnetic field of the conventional QC magnets (QC1 and QC2) along
extracted beam line. Calculated results were passed to Kubo who expanded them into
multipoles for SAD analysis. We will see new calculation of dump line with the magnetic
field in next meeting.

**(4) Tail re-population:** Sanuki explained an idea for implementation of gas
scattering in SAD calculation, which has been developed in discussions with Oide-san.
The major issue is that a probability of tails is extremely small (<<10^{-7} ). Since
SAD can provide transfer matrix at arbitrary points in main linac, we must set
beam-profiles corresponding to tail generated by Coulomb scattering in the phase space
(sigma_x(y), sigma_x(y)', p ) at each point. Multiplying them by the matrix, one can
get tail-distributions at the exit of linac. In this calculation, the tail
distributions are assumed to be Gaussian too. The other method was also discussed.
The tails can be simulated by particles, that is, by using SAD-tracking method.
The latter method does not assume Gaussian distribution in tails and it is more relevant
for high energy physicists. It may be worth to compare these two methods.
Anyway, he got KEK-computer account for interactive SAD.

**(5) Discussions **
We had a general discussion on the beam delivery system including pre-linac collimation,
although the details depend on our further studies.
The major issue is to minimize the total length (BDS) for site- and cost-optimizations.
In the second bunch compressor after
the pre-linac (8GeV), there is a (pre-linac) colimation section where beams are collimated
in < +/- 5 sigma_z , <+/- 6 sigma_x, , <;/- 40 sigma_y. The (second) collimation system
down-stream of the main linac shall be dedicated to a machine protection (MP), so only
momentum collimation. Following final focus section shall be optimized at 250GeV of
beam energy by Oide/Yamamoto's or Pantaleo's optics. Roughly we expect the pre-linac
collimation of <50m, the MP collimation of ~200m and the FF of ~300m, which are
compared to the present JLC design of 1200m collimation and 1600m FF.
For energy upgrade, FF should be optimized at Pantaleo's optics whose length is expected
to be "independent" on beam energy, and the collimation section could expand into the
main linac, assuming the energy upgrade is done by higher acceleration-gradient. There
were many discussions and they will continue.

The next meeting will be on 7 June, 2000 10:00 - 12:00 at 3 gokan, 425.