[BTeam] startup actions when otherwise stalled

[Michael Tiefenback]

I sent this to those listed below, but thought I should send it more widely, and also note the potentially useful prospect of an "early" bunch length measurement using the 1A SLM.  That measurement is (at lowest level) independent of linac acceleration, and is possible at relatively low NL energy gain.  We might consider what this means in terms of injector qualification when doing a significantly "cold" start, as at present.  Being forewarned of bunch length anomalies can be a good thing.

The lattice I propose for direct measurement of body gradient is only one of many options.  I have a FODO layout with the singlets and center magnets of the alternating triplets with a K1 value of about 0.55.  Setting the positive lenses near this value to zero the dipersion at 1R01, and setting the negative lenses in the range from 0.8 to 1.2 times this strength, results in lattices for whcih the M12 and M34 from the beginning of 1A bound the body gradient fairly well.  This supposes, by the way, that the quadrupoles are doing what we think they are.  (It's always good to watch your back when hardware is involved.  And in other cases, too.)

The actions I suggested for phasing the NL if SRF is available and stable on swing or owl but before full energy operation is possible (chiller limitations, for instance) involve only normal operations procedures, but with <whatever> as the NL energy gain.  There is nothing critically different about it.  The other actions on the table are more "expert" activities.

Michael Tiefenback

________________________________
From: Michael Tiefenback <tiefen at jlab.org>
Sent: Wednesday, July 22, 2020 14:36
To: Will Oren <oren at jlab.org>; Paul Vasilauskis <vasilaus at jlab.org>; Brian Freeman <bfreeman at jlab.org>; Michael Aiken <aiken at jlab.org>; Yves Roblin <roblin at jlab.org>; Todd Satogata <satogata at jlab.org>; Jay Benesch <benesch at jlab.org>; Steve Suhring <suhring at jlab.org>
Subject: and another thing....

During our discussions of sequence for startup, I suggested a purely transverse beam optics measurement which might illuminate our body gradient question in Arc1A.  Given our uncertain bldg temp status, I thought it might be worth mentioning again, with a twist.

The injector is thought by many to be ironed out and ready to deliver ~120 MeV beam to the NL.  This is sufficiently relativistic not to suffer velocity modulation with a change in acceleration profile.  If the NL SRF can be operated at some stable intermediate energy gain, it becomes possible to phase the operating portion.  This would assist in restoration of beam, as such phasing consumes many hours of dedicated time.  This appears clearly possible to accomplish, but it is not clear to me that "early" operation of the linac, say at 1/2 gradient, is possible in the presence of bldg temp issues.

If the above can be done, and beam propagated to the 1R02 dumplet at some stable energy, it becomes possible to configure the arc optics in a straightforward pattern, leaving the outboard magnets of the triplets off, and alternating the sign of the remaining FODO singlet lattice.  With well-crested RF, the energy spread will be small enough to tolerate the 10-ish meter dispersion peaks.  We can configure the focusing (strength and bias) to bring the dispersion to zero at 1R01.  A set of differential optics measurements (such as rayTrace) can readily reveal the body gradient contribution of the dipoles.  Executing this in otherwise dead time provides a prequalification of the process prior to later execution at the nominal beam energy in a "beam studies" period.  It also would prospectively provide a data concerning the potential excitation dependence of body gradient.

Such activities will also engage operations in a relatively gentle startup.  There would be reduced schedule pressure at that time, as multipass operation at unexpected NL energy gain makes that unhelpful.  It also provides an early shot at identifying diagnostics which need attention but somehow have escaped notice.

I have lattice configurations for doing the optics, but the preliminaries are all basically standard ops processes.  It's just that some of the early bits do not require full multipass capability, and we might be able to exploit otherwise dead time.  And if we do it well, we may get some fun and interesting information.

Michael Tiefenback

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/bteam/attachments/20200723/87ab5347/attachment-0001.html>