Update on possibilities to reach *=40cm (work in progress) R. - - PowerPoint PPT Presentation
Update on possibilities to reach *=40cm (work in progress) R. Bruce, S. Redaelli Acknowledgement: R. de Maria, S. Fartoukh, M. Giovannozzi, M. Huhtinen R. Bruce, 2015.01.19 R. Bruce, 2015.01.19 1 1 Introduction LMC 3/9/2014:
1
1
– Based on optimistic assumptions: Can go to tighter collimator settings, decrease emittance and beam-beam separation, use full theoretical gain from BPM buttons in collimators
2
Chamonix 2014
3
– In 2012: Triplet aperture was ~4.5 σ behind IR7 TCSGs – With 6.5 TeV, mm kept settings, 11 σ BB, 40 cm: aperture 1.5 σ behind TCSGs inIR7 – With 2 σ retraction settings: 40 cm aperture is 2 σ behind TCSGs inIR7
– Remember: IR7 hierarchy also larger than nominal to keep down impedance
4
– Assumes 90 deg phase advance from dump kicker MKD to any TCT or bottleneck – in Run 1, we knew that real phase advance was better (hidden margin). Pessimistic!
– If we have a good phase advance, we might be unnecessarily pessimistic
– Ongoing work since long time: see earlier studies in e.g. MPP review 2013
5
6
7
– Improved phase advance – further away from odd multiples of 90 deg – in IR1 B1 (previously most critical case for nominal optics), IR1 B2 and IR5 B1 – Worse phase advance at IR5 B2 – Worst phase advance at β*=40cm is about 60 deg away from 90 or 270.
8
9
– Losses at IR1 are factor 20 below plastic deformation limit even with TCTs inside IR7 secondary collimators – Losses from secondary halo at IR5 B2 are very similar to 55cm
– Much larger impact parameters – see talk E. Quaranta in ColUSM 19/9/2014
– Have to deal with them even if there are no errors in the machine. Increasing margins not likely to help
10
– Different approach required: step away from the 90 deg assumption – Not meaningful to base the calculation of margins on this scenario
11
– Taking the worst case out of 1000 random imperfect optics (phase advance 33 deg -> 45 deg , dynamic error), TCT losses increase by factor ~25 – Extremely unlikely case – Should probably not be a concern, but to be verified with beam
– On paper much larger aperture margin there. – However: In IR5 B2, we have ~90 deg phase advance to the experiment already at higher β* – SixTrack results: no losses in detectors even in extremely pessimistic conditions (backup slides) => should not be a concern
12
– Triplet BLMs should dump the beam (and possibly other interlocks). – No risk of damage
– If too tight: increased number of beam dumps. Can step back in β*
13
– Largest allowed setting given by triplet protection from cleaning losses – Smallest allowed setting given by background constraints and cleaning hierarchy
– With 40cm optics – 6.5 TeV – Collimator settings: 2σ retraction – Simulating B1 and B2 – Simulating horizontal and vertical halo
14
– What losses are acceptable?
Run 1 configuration 7 TeV LHC design configuration
15
– Limit not fully clear – Nominal LHC configuration studied in detail in the past – hopefully OK
– Discussion M. Huhtinen: Even a factor 10 increase compared to Run 1 could possibly be tolerated, but experimental verification needed
– Observed background will also depend on the beam lifetime – many uncertainties – Open question if we can go further
16
– Triplet aperture artificially reduced to expected machine aperture
17
– To be noted: no imperfections included
– Do not risk serious machine damage if violated – Protected by interlocks: TCT BPMs and triplet BLMs
– probably OK for both background and triplet protection – Corresponding to about 0.7 σ margin TCT – triplet – Possibly to be followed up by imperfection studies and MD tests
18
Collimator Setting (2 σ retr) TCP IR7 5.5 TCSG IR7 7.5 TCSG IR6 8.3 TCDQ IR6 8.8 TCT IR1/5 (preliminary) 8.8 Aperture 9.5
[σ with ε=3.5μm] Possibly to be put at 8.3 σ – under discussion with LBDS (reduces secondary halo leakage during asynch dumps). See backup slide Assumes 11 σ beam-beam separation with ε=3.75 µm. Gives 205 µrad half crossing angle
19
– Inject just enough intensity to get the orbit right (1 nominal + many pilots) – Squeeze to 40 cm, stay separated (most critical for aperture) – Introduce 2 σ retraction in IR7. Possibly measure tune shift / impedance – Align TCTs (should be fast with buttons), – Do betatron loss maps and scan TCT setting – 1 pilot per setting and plane. Observe TCT losses, triplet losses and maybe experimental background – Possibly measure IR1/5 triplet aperture at 40 cm using standard method – Do asynchronous dump test
20
– Assuming TCTs limited to a setting about 1 σ outside IR7 secondaries and about 0.5 σ inside the triplet aperture (preliminary numbers!)
– With more optimism, we could imagine 10 σ beam-beam separation for 2.5 µm emittance – With Run I aperture assumption, this brings limit to β* = 31 cm, assuming good phase persists, and cleaning is similar as at 40cm – Only considering aperture here – optics constraints should also be studied. Need ATS? – Latest news: optics provided by R. de Maria down to 20 cm – checks
– For the future: low-impedance collimators could allow reducing IR7 margin
21
22
– Can squeeze further in separation plane… Some example configurations.
β* (sep/cross) Half angle Aperture L (1034 cm-2 s-1) 80/80 cm 145 µrad 13.8 σ 0.65 40/40 cm 205 µrad 9.5 σ 1.0 30/40 cm 205 µrad 9.5 σ 1.2 β* (sep/cross) Half angle Aperture L (1034 cm-2 s-1) 40/40 cm 150 µrad 11.0 σ 1.6 31/31 cm 170 µrad 9.0 σ 1.8 25/40 cm 150 µrad 9.0 σ 2.0
2700 bunches, 1.15e11 p/bunch, 7.55 cm bunch length 11 σ BB sep, ε =3.75 µm 10 σ BB sep, ε =2.5 µm
Obviously, aperture numbers to be taken with some uncertainty…
– Possibly some TCSGs closed only at end of squeeze (CWG 2014.12.08)
– Are there other loss scenarios to include in the margins? Orbit bumps … ?
23
– Nominal β*=40 cm optics: this is the case! Easier to reach than thought – Next limitation: cleaning constrains outer and inner TCT position
– Only change w.r.t. startup : 2 σ retraction in IR7. β*<40cm might be within reach. New optics under study. – Further validations necessary, in particular with beam
24
25
26
27
– On paper much larger aperture margin there. – However: In IR5 B2, we have ~90 deg phase advance to the experiment already at higher β*
– At the end of the scan, with the TCDQ at 16σ, still no losses in experiments
28
29
30
– Risk that it becomes primary – Probably not acceptable
– Factor ~10 lower TCTV losses – Factor ~15 higher TCLA losses
31
– Account for shape of bump: take the crossing bump – Account for the (small) phase advance TCT->triplet – Account for machine drifts (as before)
– If accidentally increasing crossing angle, we eat up margin!
32
33