Participants : Brett Parker, Andrew Marone (BNL), Nobuhiro Kimura, Glen White, Nobuhiro Terunuma,Toshiaki Tauchi, Tatsuya Kume (KEK), Philip Bambade (Tokyo), Andrei Seryi (SLAC), Eduardo Marin (CERN) etc.
Keeping the total current limit for correction coils, there are a few options. If there is no Q corrector in the main sextupole magnet, skew octupole or dodecapole corrector can be added.
There are questions on strength of skew octupole and dodecapole, where the both are limited by maximum current of 20A. Are they appropriate?
There is possibility to have both correctors by adding one more corrector, i.e. 6 in total, if the maximum current is decreased to 18A.
We must decide the configuration soon.
Next, he presented an analytical estimation of the corrector performance running at a few Hz for possible slow feedback.
His direct explanation is given below;
--------
For the ATF2 magnet, with its thick 6.mm copper shield, there is no chance to get a few Hertz magnetic field to the beam. Doing this estimation made me realize that this amount of low temperature copper will block AC fields; only slowly changing magnetic fields can reach the beam. Note that this also applies to changes in field gradients, i.e. quad, sextupole and correctors. While it may be nice that power supply ripple won't reach the beam, when we make run plans we need to keep in mind that we cannot insantaniously change these fields.
Note: this is dramatically different from the ILC's QD0 situation. For QD0 there is only a single 2 mm thick stainless steel tube and the resistivity of steel does not drop as precipitously at low temperature as copper does. If we only had to worry about this tube we could probably use frequencies up to a couple of tens of Hertz.
Unfortunately the above neglects the infuence of the supercondcutor. With the baseline configuration having six layers of superconducting cable between the dipole corrector and the beam, my simple simulation shows that all this material does have a perceptible impact on the applied dipole field; however, since this would be making a small change to an already small correction field, it is possible that this could be acceptable.
Since my simple AC steady state calulation did not include the superconducting filaments (a bit tricky to model) it is hard to say with assurance (without doing direct experiments... which we could do) what the final level of field distortion would be.
If there really is interest in doing such few Hertz, low level, feedback at ILC/CLIC, we could certainly investigate further. As I indicated in my presentation it is possible that moving the dipole coils inside the quadrupole coils (which I have not simulated) could make things better. But even in this case I would expect some effect due to the quad conductor interacting with the (albeit weaker) dipole external field.
--------
In addition, there is a distortion of the dipole corrector field (Bo=14.4Gauss), e.g. +/-0.7% at 5mm. Is this level OK ?
So, we would like to discuss the interface part between the magnet cryostat and the connection box with BNL for the detailed design work.
In addition we have following questions .
Next webext meeting will be in May after the golden week in Japan. Meantime, we will exchange emails concerning the interface issues.