Minutes of ATF2 weekly meeting, 27 July, 2011

July 27 15:00 - 16:30 (JST time), Webex and ATF Meeting Room, KEK

The new ATF wiki page (http://atf.kek.jp/twiki/bin/view) can be reached here, where the collaboration-page requests your login or registration.

ATF2 LATTICES SWAPPING THE MAGNETS, Edu Marin (CERN)

Edu reported the swapping effects in the alternative lattices (re-fitted with the MAD/KEK multipoles of QEA-magnets and the FD multipoles ) of the ATF2 nominal and ultra-low optics by proposing three swapping scenarios.

First, in order to select "good magnets" for the swapping, two lists of magnets were shown by sorting with the skew sextupole and octupole components, respectively, i.e. ordering from magnets with the least skew components. Next, in order to select magnets to be swapped, the magnets were listed in order of sensitivity of 2% vertical beam size increase by the skew sextupole and octupole components, respectively. Then, the three proposals were shown as;

IP vertical beam sizes with the 3 proposals were calculated as a function of the IP vertical beta function from 25 to 100μm together with ones without the swapping and the ideal ones, where the IP horizontal beta function is 10mm. The best solution is obtained when the magnets are sorted according to only the sextupole multipole component (1st swap). When the FD is swapped in addition, it corresponds to the worst scenario (2nd swap). When the magnets are sorted considering both the sextupole and octupole multipole component, an intermediate solution is obtained (3rd swap). Also, IP vertical beam sizes were calculated as a function of IP horizontal beta function from 4 to 10mm in the ATF2 nominal lattice.
Q : The results show that the skew sextupole components have dominant effect. Did you include a newly installed skew sextupole magnet in your analysis ?
A : Yes, but it is not important for it has small effect. The FD has major effect.
Q : Is the second conclusion for the ATF2 nominal optics ?
A : Yes. We will study in the case of ultra-low beta, too.
C : We thought that the skew sextupole magnet must contribute.
A : Yes, but it is true before the new multi-components.
C : Is the beam size (sigma) RMS ?
A : Yes.
C : There is non-Gaussian tail. So the sigma never goes to 35nm, if the optimization is done by RMS .
A : The optimization in MAPCLASS can be done by using Shintake sigma.
Q : If you can introduce a cut in the tail, you can get the same result as the Shintake sigma ?
A : There is no cut in MAPCLASS. I do not know how to implement cut. I will ask Rogelio, the author of MAPCLASS.
After the meeting, Rogelio sent a following message, i.e. update of results at the nominal optics.
"The lower β*x lattices (nominal β*y=100μm) were never re-optimized for the new multipoles. These lattices were optimized for the old multipoles and stored in the repository. Since the new multipoles have a lower impact in the lattice performance a better beam size was expected. Edu has worked hard these days to optimize all nominal lattices with the lower β*x using the rms beam size (as always done with MAPCLASS) and compute final rms and shintake beam sizes. Please find the results in the attached plot (uploaded as the sub-contribution). Certainly the rms vertical beam size remains far too large at low β*x indicating that indeed there are important aberrations and beam tails. However a quite interesting Shintake beam size is achieved at nominal β*x=4mm of ~42nm. Before the optimization the Shintake beam size was above 50nm!
In the light of these new results we understand that one can indeed leave the sorting for the ultra-low β* option."

The re-matched optics study (updated), Sha Bai(IHEP)

First, the MAD multipole definitions and the multipole rotation(tilt) angle are introduced, where the angle is defined with respect to the quadrupole angle. So, the rotation angles of sextupole and octupole components are expressed by T2=θ3/3-θ2/2 and T3=θ4/4-θ2/2, respectively.

She found following corrections to IHEP measured rotation angles to be good agreement with KEK ones which have been implemented in MAD, except for T3's for QEA25 to QEA34 ;

Above corrections are just to show that angles are actually rather consistent between KEK and IHEP, as long as we accept some polarity flips (presumably arising from different interpretations or confusions in the many signs involved in the measurement and interpretation...). Thus, we think that the amplitudes of skew components are rather consistent, while the signs are not. However, in the all the subsequent tracking results shown, such arbitrary sign flipping rotations were not applied. (Thanks Philip for this explanation)

The most sensitive magnets were selected by their significance on the beam size growth at IP. The significance is defined by a ratio of measured skew components and ones for the 5% beam size growth, where the skew multipoles are included up to the dodecapoles. They are QF9AFF, QF9BFF, QF5AFF, QF5BFF, QD4AFF and QD4BFF, which are candidates for the swapping.

The IP beam sizes (RMS and sigma by Gaussian fit) were calculated by tracking in MAD under various conditions. Results are summarized in tables in the slides. The beam optics is the nominal one without re-fiting, which is different from Edu's analysis. Also, the IP vertical beam sizes were calculated with enlarging the IP horizontal beta function up to 4cm, where the optics was re-matched by the matching quadrupole magnets ( QM12FF to QM16FF).

C Philip : There should be no miracle by the swapping, There might be cancellation in multipole components in the FD and other magnets. Also, there might be mistakes in the signs. Can we look for any robust solution ?
A Edu : The swapping has some improvement. If the major contribution comes from the FD, the swapping may not be relevant.
Q Philip : Is it enough for new QF1FF with small multipole components to reach 20nm beam size in the ultra low beta optics ?
A Edu : It is under study.
C Philip : Sha just uses the 6 matching quadrupole magnets for re-matching the IP beta functions in order to study differences between IHEP and KEK multipole components. Also, her study must be complementary with Edu's analysis.

ATF2 Laser-wire move to MFB2FF, S. T. Boogert(JAI at RHUL)

Stewart reported a move of the laser wire to a location of MFB2FF. Major motivation is to be able to run most of time in the nominal ATF2 optics, i.e. completely parasitic to the goal 1, low backgrounds and smaller vertical beam size for the virtual vertical IP. Expected beam sizes are 844nm and 376.2∼125.4um in vertical and horizontal directions, respectively.

Major effects of the movement are re-configuration of the straightness monitor system, removal of MFB2FF but replacement by a BINP cavity BPM with the same readout electronics. The detailed configurations are shown in the slides.

The LW system was actually moved in yesterday, and will be completed in today.

Q : Where is the BINP cavity BPM to replace MFB2FF ?
A : It is about 10cm or a few cm upstream to previous one.
Q : Where is the FFTB mover used for the MFB2FF?
A : It is moved to the upstream. So, it can be used for a spare mover.
Q : Are the beam sizes at the nominal optics ?
A : Yes, there is no optimization yet.
Q : The horizontal beam size seems to be too large ?
A : The laser spot size is about 50um. The beam size is about twice of it. There should be no problem in signal yields, at least.
C : The LW may contribute the ATF2 beam tuning for the virtual IP focal point. It is worse to study such possibility.
C : It is a good idea for the new LW location.

Next meeting will be in September, 2011.