Choose timezone
Your profile timezone:
SPS BWS Task Force
→
Europe/Zurich
865/1-D17 (CERN)
Videoconference
SPS BWS Task Force
- Zoom Meeting ID
- 63016739397
- Host
- Raymond Veness
- Useful links
- Join via phone
- Zoom URL
-
Notes from meeting of 17/5/23 by RV.
Apologies: Roncarolo, Velotti,
Present: Ray Veness, Tom Levens, Michael Sullivan, Giulia Papotti, Nabil El-Kassem, Christine Vollinger, Carlo Zannini, Elena De la Fuente, Thibaut Lefevre, Ivan Karpov, Giannis Papazoglou, Federico Carra, Kevin Li, Benoit Salvant, Jonathan Emery, Jose Somoza,
Operational beams
Introduction (Ray): Aims are to understand exactly what caused the failures in the scanners, so comparing 2022 and 2023, but the also being able to feed back to OP on how to operate with a potentially 'mitigated' wire scanner after TS1.
Giulia (see slides): Summary of instruments available for measuring bunch length in the SPS: WCM with MD scope, SPS BQM, ABWLM, SPS Tomoscope, PS2SPS.
One line (with the MD scope) has less divisions so can achieve a better band width. Can also adjust the dynamic range.
Bunch length measurement has two modes, OP and MD. OP mode publishes every 20ms MD is turn-by-turn. Reminder that the accuracy is ~ 10% on, for example, bunch length.
Diagnostic on the event of 22/4/23@5am gives a result that is compatible with a setting mismatch. This meant that the setting was republished until end of cycle so the bunch length data from this event is not recoverable.
Action (Future operations): Manual check on the settings when needed in the future to ensure this is correctly configured.
Thibaut: will need to measure and control bunch length to avoid damage to a broad range of BI (and other) equipment in the future. Differences in bunch length in range of 150 ps can be significant. May not be able to change hardware to accept large variation in bunch length. ACTION: Kevin - what are the LIU specs for bunch length (variation)?
Carlo (see slides): Data on beams during scrubbing run. Plots shown of measured heat load (scaling with beam parameters shown) for the different fills. Periods of BWS failure shown. Discussion on when exactly second scanner failed. Not clear that they failed on the first high heat load scan, which is important for making a thermo-mechanical model. ACTION Carlo: Add the precise cycles (or best knowledge) of when the scanners failed. Comparison between 2022 and 23 heat loads. 0.5 C/h max in 2022 and 2.5 C/h in 2023.Another factor: Repetition rate was higher in 2023 by factor of 2 when wire scanners failed. Discussion on this point. High repetition rate would be a desired/necessary operational scenario for HL. This could be a main contributor to over-heating on a longer timescale Thibaut comments that the 1.65 ns bunch length shown could be shorter due to measurement. Possibly we may need to gain a factor ~5 on heating compared with the cycle that failed the scanners.
Nabil (see slide): plot of of intensity over time showing wire breakage times. Would be interesting to know exactlywhen the wire broke, ie on the first cycle, or after some time (ACTION Nabil).
Benoit asks if this can be shown for the injected beams as this could be significant depending on which modes are excited (ACTION Nabil).
Update on Simulations and lab tests:
Michael (see slides). Ferrites. R/Q quantifies how beam couples to RF, with higher values giving more coupling. Warns that CST definition of impedance is used, so divide by 2 for accelerator definition.
Preliminary ferrite integration shown dissipation on wire drops to ~2% from 34%, so encouraging for a ferrite mitigation. Need to optimise further.
Can we test with ferrites in the lab? Jose says should not clean with alcohol as materials are absorbant. Need to discuss how to use in lab test (ACTION Ray/William to organise a meeting) - perhaps fire first? Ferrites are available in RF.
Jose says option to spot weld a screw in the tank and then attach a ferrite. This has been done in-situ by VSC and tooling is available to do in the BWS tank.
ACTION: Michael. Power on ferrite is to-be-calculated and given ASAP to Federico Carra.
Kevin: need a discussion with ABT and vacuum about when/how we can make an intervention in BA5. Should also discuss further vacuum sectorisation with VSC at same time. ACTION Ray for a meeting.
Could we consider also changing the spare tank in BA4 for a vertical? ACTION William - how long to make this change.
Christine: What cable can we use for feedthrough. Jonathan - could use one of the 4 photomultplier. CK50 with ~170m of cable. Tom - may also have spare BPM cable. ACTION: Tom will check*.
Benoit, Elena:
Beam spectrum measured during test to reproduce the beam that broke the scanner.
Spectrum on 80 mHz and 160 MHz with bunch length. Bunch length shown is from the MD scope in OP mode (see Giulia's talk) This is the average bunch length in operation mode every 20 ms for all bunches in the machine. 160 Mhz mode is less likely to be a problem as there was more intensity at injection there last year. However, probably also a lower duty factor.
Jonathan has some data on temperature on wires before the failure of the scanners. Important also to understand if there was heating at injection as this would imply that a different mode was responsible. Jonathan was not able to present this, as we ran out of time, but the slides have been uploaded to the INDICO page.
There was a discussion as to whether the long-term consolidation should target a 'worst-case' heating model, or some intermediate case.
The meeting ended here as we ran out of time. However, in discussions after the meeting it was agreed that MME (Carra, Papazoglou) will start to make a model for ferrite heating and also for the fork/wire assembly to couple with CST. BI (Mariet) will make a model of 1D wire heat load vs. temperature using the full heat transfer model of Sapinski et al.
Next meeting: Friday 26th May at 9h00 with agenda TBA.
*[Update from T.Levens after the meeting. "I checked with Thierry, there should be 4x unused CMC50 cables from BPCE.41705 to the rack RA1225=HCA442. This BPM is about 10m downstream of the WS. Based on GESMAR the cables should be around 120m long. The rack is underground but in an area which is accessible when there is beam."]
There are minutes attached to this event.
Show them.
