{"id":"39b8d1bb-bffa-4997-9688-0530bb8ac22e","arxiv_id":"2507.13315","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Vertical bent-crystal collimation improves LHC halo cleaning by factors of 3 to 10 in the dispersion suppressor at 6.5 TeV, while horizontal crystals show little improvement.","lead":"This LHC study measures how well bent crystals clean stray beam protons at 6.5 TeV compared with standard collimators. The vertical crystal reduces losses in critical magnet regions by factors of 3 to 10, while the horizontal crystal gives only modest gains.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The vertical-plane claim includes a factor-3 improvement in Q7, yet the paper states Q7 losses are not directly comparable to simulations and hypothesizes they come from TCLA hadronic showers; the Q7 part of the central claim is unverified.","rationale":"The measured vertical-plane improvement in Q8-9 and Q10-11 is reasonably supported by Table 2 and by the rough agreement between measurement and simulation in those regions; that part of the claim is not the weak link. The weak link is Q7, where the simulation predicts zero losses and the paper itself attributes the measured signal to unmodeled hadronic showers. Because the central claim explicitly includes 'all DS regions,' the Q7 entry is load-bearing for the full statement. The paper deserves credit for flagging this limitation and for using a benchmarked SixTrack crystal routine, so the concern is not about the integrity of the measurements but about their interpretation. A FLUKA-based shower benchmark is the concrete, paper-suggested check that would settle whether the Q7 leakage ratio is a real cleaning improvement or an artifact of shower response. The reader's conditional verdict already captures this uncertainty, so no change is needed.","tokens_in":11591,"tokens_out":9807,"duration_ms":120050,"concrete_test":"Run a FLUKA (or equivalent) energy-deposition simulation using the SixTrack loss distributions for the standard and vertical-crystal configurations as inputs, including the full IR7 geometry with the TCLAs, and predict BLM signals in the Q7, Q8-9, and Q10-11 regions. Check whether the predicted Q7 BLM ratio reproduces the measured 3.49 ± 1.54. If the shower simulation cannot produce a Q7 ratio above unity, the 'all DS regions' claim fails; if it can, the Q7 hypothesis is confirmed and the measured claim is supported. This is exactly the benchmark the paper itself identifies as necessary in Sec. 6.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central vertical-plane claim (Sec. 6) asserts improvement in all IR7 dispersion-suppressor regions by a factor of about 3 to 10, citing Table 2 V-1: Q7 3.49 ± 1.54, Q8-9 16.43 ± 9.60, Q10-11 11.25 ± 2.99. The Q7 entry is load-bearing for the 'all DS regions' wording. However, Sec. 4 states that in SixTrack no particle is lost in Q7 and hadronic showers are not accounted for, so Q7 losses cannot be directly compared. Sec. 5 then hypothesizes that the measured Q7 losses are produced by direct hadronic showers from the upstream TCLAs. Sec. 6 concedes that 'the measured leakages are still not directly comparable with simulations' and calls for a full energy-deposition and shower-tracking benchmark. Without such a benchmark, the measured Q7 leakage ratio is not established as a local cleaning inefficiency: it could be a BLM response to shower background rather than a reduction in proton losses on the cold aperture. If the Q7 hypothesis is wrong, the claim of improvement in all DS regions is not supported; if it is correct, the Q7 factor still needs a quantitative benchmark before being presented as a measured cleaning improvement.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports measurements of crystal-assisted collimation cleaning performance at the LHC with 6.5 TeV protons, comparing normalized loss maps for standard collimation and for crystal-based configurations in the horizontal and vertical planes. The leakage ratios in Table 2 indicate that the vertical crystal improves cleaning in the IR7 dispersion suppressor and in IR3, while the horizontal crystal yields only slight improvements. The measurements are compared qualitatively with SixTrack simulations, with the paper explicitly acknowledging that measured BLM leakages are not directly comparable to simulated proton loss patterns because hadronic showers are not modeled.","tokens_in":11879,"tokens_out":3840,"duration_ms":44406,"significance":"If confirmed, the vertical-plane result is practically relevant for the HL-LHC era, since it suggests that a crystal primary collimator can reduce cold-aperture losses in the IR7 dispersion suppressor by factors of 3-10. The paper's strengths are the unique beam-loss data from a collider at top energy, the systematic scan of collimator configurations, the absence of fitted parameters, and the explicit statement of the simulation-comparison limitation. However, the central vertical-plane claim is currently stronger than the evidence: the Q7 component is not established as a local cleaning improvement, and the simulated leakage ratios are presented without uncertainties despite the paper's own caveat that measurements and simulations are not directly comparable.","major_comments":[{"comment":"The vertical-plane statement that \"an improvement by a factor 3 is observed in all the regions\" rests on the Q7 leakage ratio of 3.49 ± 1.54, but Section 4 states that no particle is lost in Q7 in SixTrack and that hadronic showers are not accounted for. Section 5 itself hypothesizes that the measured Q7 losses are produced by hadronic showers from the upstream TCLAs. Without a quantitative shower/energy-deposition simulation, the Q7 ratio cannot be interpreted as a measured local cleaning inefficiency; it could reflect BLM response to shower background. Please either benchmark this channel or restrict the \"all DS regions\" claim to Q8-9 and Q10-11.","section":"Section 5, Table 2"},{"comment":"The simulated leakage ratios are listed without any uncertainties, and Section 6 concedes that \"the measured leakages are still not directly comparable with simulations.\" Nevertheless, Section 5 claims for V-1 that simulation expectations are within the measurement error bars except for IR6. Without statistical or parametric uncertainties for the SixTrack ratios, this quantitative agreement statement is not supported. Provide uncertainties for the simulated ratios, or rephrase the comparison as qualitative agreement.","section":"Table 2, Sections 5 and 6"},{"comment":"The uncertainties on several key vertical measurements are large: Q8-9 is 16.43 ± 9.60 and IR3 is 31.05 ± 21.53. These uncertainties should be propagated into the conclusions; for example, the Q8-9 factor of 10 is only marginally inconsistent with a factor of 2 at the 1σ level, and IR3 is consistent with a wide range of improvements. The paper's summary statements should be based on the confidence intervals, not solely on the central values.","section":"Table 2, vertical V-1 row"}],"minor_comments":[{"comment":"The sentence \"The simulated cleaning ratios, between standard and crystal collimation, are reported in Tab. 1\" should refer to Table 2, not Table 1.","section":"Section 5"},{"comment":"The caption contains a duplicated word: \"shown for for the full ring\" should read \"shown for the full ring.\"","section":"Figure 4 caption"},{"comment":"The phrase \"the necessary information to built LMs\" should read \"to build LMs.\"","section":"Section 2"},{"comment":"In the vertical-plane discussion, \"For the vertical crystal Cfg#1 an improvement by a factor 3 is observed in all the regions\" is missing the word \"of\" after \"factor.\"","section":"Section 5"},{"comment":"The sentence \"In the simulation is expected to be 10 times better than standard collimation, while it is observed to have the same performance\" is grammatically incomplete; the subject of the expectation should be made explicit.","section":"Section 5"}],"recommendation":"major_revision","confidential_remarks":"The paper is a measurement-focused contribution from the LHC collimation group; the self-citations are appropriate for the technical context. The main concern is that the headline vertical-plane claim overreaches the Q7 evidence, which the authors themselves flag as a hypothesis requiring full shower simulations. This is fixable by softening the claim or adding the needed benchmark, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis paper reports new measured cleaning leakage ratios for crystal collimation at 6.5 TeV in the LHC, comparing standard versus crystal primaries for both planes and several collimator configurations. The vertical crystal shows a clear improvement in the IR7 dispersion suppressor: measured ratios around 11–16 in Q8-9 and Q10-11, and about 3.5 in Q7. The horizontal crystal is close to standard or worse, with a plausible explanation tied to the crystal's bending radius.\n\nWhat is genuinely new here is the systematic comparison across configurations, with loss maps normalized by beam flux and a clean leakage-ratio definition. The paper also handles limitations honestly: it explicitly says measured leakages are not directly comparable to simulations, and that Q7 losses are not reproduced in SixTrack because hadronic showers are not modeled. The Q7 shower hypothesis is reasonable and testable with FLUKA.\n\nThe soft spots are real but not fatal. The conclusion's phrase 'improvement in all DS regions by a factor ~3 to ~10' leans on the Q7 ratio of 3.49 ± 1.54, which is the least certain entry and rests on the unverified shower hypothesis. If Q7 turns out to be BLM response to TCLA showers rather than local proton losses, the 'all DS regions' wording is too strong. That is a wording issue, not a reason to reject: the Q8-9 and Q10-11 improvements are large and independent of that hypothesis. Two other weaknesses: simulated ratios come without uncertainties, so 'within error bars' claims cannot be checked; and the horizontal/vertical asymmetry explanation is qualitative. Also, the paper's own statement that measured and simulated leakages are 'still not directly comparable' sits oddly with earlier comparisons, though the overall tone is appropriately cautious.\n\nBottom line: a useful operational result for the LHC collimation community, and the data will matter for HL-LHC planning. It deserves a serious referee. I would recommend conditional acceptance, asking the authors to temper the Q7 claim, add uncertainties to the simulated ratios, and either provide a shower benchmark or clearly label Q7 as provisional.","headline":"Solid measured comparison of LHC crystal versus standard collimation, but the vertical improvement claim including Q7 is softer than the conclusion suggests.","tokens_in":12366,"tokens_out":3706,"would_cite":true,"duration_ms":39316,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper claims that replacing the amorphous primary collimator with a bent silicon crystal in the vertical plane improves LHC halo cleaning by a factor of about 3 to 10 in the dispersion suppressor regions at 6.5 TeV, and that the…","keywords":["crystal collimation","LHC","beam halo cleaning","bent silicon crystal","planar channeling","loss maps","leakage ratio","dispersion suppressor"],"falsifier":"A full Monte Carlo simulation that tracks hadronic showers and energy deposition from the upstream TCLAs to the Q7 beam-loss monitors, seeded by the SixTrack loss patterns, would settle it: if the predicted Q7 leakage does not reach the measured values, the shower hypothesis and the quantitative gain would need revision; if it does, the measured leakage ratios become directly interpretable.","tokens_in":11428,"feed_emoji":"💎","tokens_out":8349,"duration_ms":86072,"temperature":0.7,"pith_summary":"The paper sets out to show that inserting a short bent silicon crystal as the first stage of the LHC's betatron collimation system improves the disposal of beam halo at 6.5 TeV, and to quantify that improvement with loss maps. It reports that in the vertical plane the crystal outperforms standard amorphous-jaw collimation by a factor of about 3 to 10 in the dispersion suppressor regions, where off-momentum halo particles most easily hit superconducting magnets, and by a large margin at the momentum-cleaning collimator. In the horizontal plane the improvement is smaller and mainly visible in the dispersion suppressor, with a persistent leakage region, Q7, that the paper attributes to hadronic showers from upstream absorbers rather than to off-momentum losses. The measured leakage ratios are compared with SixTrack simulations, which reproduce the main loss patterns qualitatively but cannot yet be compared directly with the measured leakages.","feed_headline":"Bent crystals cut LHC halo losses 3-10x in vertical plane","feed_subtitle":"LHC loss maps show a crystal primary stage cuts dispersion-suppressor losses, easing cold-magnet heat load.","key_machinery":"The load-bearing mechanism is planar channeling in a short bent silicon crystal: protons entering nearly parallel to the crystal planes are captured in the interplanar electrostatic potential and coherently deflected by the crystal's full bending angle, so the crystal intercepts halo weakly while a downstream secondary collimator absorbs the deflected flux. The comparison machinery is the leakage ratio, built from loss maps in which every beam-loss-monitor signal is normalized by the instantaneous beam-loss flux rather than by the primary-collimator signal, because the crystal itself no longer produces a loss signal proportional to intercepted particles. The crystal's orientation is set by an interferometer-based goniometer, and the expected loss patterns are generated with the SixTrack tracking code plus a Monte-Carlo crystal routine.","core_discovery":"At collision energy with 6.5 TeV protons in Beam 1, replacing the amorphous primary collimator with a channeling-oriented bent crystal changes where and how much halo is lost. The paper's central quantitative result is a set of leakage ratios, defined as the standard-system leakage factor divided by the crystal-system leakage factor in each machine region; a ratio above 1 means the crystal cleans better. For the vertical crystal configuration V-1, the ratios are 3.49 at Q7, 16.43 at Q8-9, 11.25 at Q10-11 and 31.05 at IR3, corresponding to the quoted factor-3-to-10 improvement in the dispersion suppressor and roughly a factor-30 reduction at the momentum-cleaning collimator. The horizontal crystal gives a slight improvement only in some leakage clusters, and the paper explains the difference by the horizontal crystal's bending radius being closer to the critical radius, which increases dechanneling of particles at small deflection angles. A control run with the crystal oriented as an amorphous material is worse than standard collimation in every region, confirming that the gain comes from coherent channeling rather than from the mere presence of the crystal.","pith_inferences":["If the shower hypothesis for Q7 is right, adding local shielding or repositioning the upstream TCLAs could recover most of the missing horizontal-plane gain, a step the paper does not itself propose.","The measured vertical-plane advantage comes from loss maps taken under fast white-noise excitation; a test with natural halo diffusion or with slower excitation would show whether the factor 3 to 10 persists in steady-state operation.","Because the horizontal crystal's performance is tied to its bending radius being near the critical value, crystal curvature can be treated as a design parameter: a crystal with the vertical crystal's larger radius should show smaller dechanneling and better horizontal cleaning.","The flux-normalization method uncouples loss-map comparison from the primary-device signal, so it could be reused to benchmark any future non-intercepting or low-interception primary element against the standard collimation system."],"forward_implications":["If the vertical-plane result holds in operation, crystal collimation can reduce the thermal load on the LHC's superconducting magnets in the IR7 dispersion suppressor by a factor of 3 to 10 relative to the standard system.","The Q7 leakage, if it is caused by hadronic showers from upstream TCLAs, would be a separate inefficiency that standard off-momentum arguments miss, and it sets a floor on how much cleaning improvement is visible until those showers are intercepted.","Configurations with fewer secondary collimators downstream of the crystal degrade dispersion-suppressor cleaning, so the full TCSG family is a required part of a working crystal-collimation scheme.","A simulation chain that models energy deposition and hadronic showers, using the SixTrack loss patterns as input, is needed before measured leakage ratios can be compared quantitatively with predictions."],"supporting_citations":[{"why":"Defines the multistage LHC collimation system whose performance is the baseline in every comparison.","marker":"[1]"},{"why":"Defines the LHC crystal-collimation layout, goniometer requirements, and the secondary-absorber design the measurements use.","marker":"[8]"},{"why":"Reports the first observation of channeling at 6.5 TeV and introduces the flux-normalization procedure that makes crystal and standard loss maps comparable.","marker":"[9]"},{"why":"Provides the experimental configurations and earlier crystal-collimation results that the present loss-map campaign extends.","marker":"[10]"},{"why":"Describes the beam-loss-monitor system that records the ionization-chamber signals used to build loss maps.","marker":"[13]"},{"why":"Motivates the white-noise excitation used to generate the controlled losses in each loss map.","marker":"[14]"},{"why":"Documents the horizontal crystal's curvature being close to the critical radius, the dechanneling effect invoked to explain the horizontal-vertical difference.","marker":"[15]"},{"why":"Supplies the bunch-by-bunch beam-current measurements used to evaluate the beam-loss flux for normalizing each BLM signal.","marker":"[18]"},{"why":"The tracking code whose aperture and loss-pattern model produces the simulated loss maps.","marker":"[19]"},{"why":"The Monte-Carlo crystal routine added to SixTrack to simulate channeling and dechanneling in the simulations.","marker":"[22, 23, 24, 25]"}],"fun_headline_variants":["Crystal collimation cuts LHC halo losses up to 30x","Bent crystal improves LHC cleaning up to 30x at 6.5 TeV","Vertical crystal slashes LHC beam halo losses 3-10x","LHC halo cleaning boosted by bent crystal primary stage","Crystal primary collimator reduces LHC losses to 1/30"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The comparison assumes that loss maps taken under forced white-noise excitation and normalized by beam flux reflect real operational halo cleaning, and that simulated proton-loss locations can be compared with beam-loss-monitor signals before hadronic showers and energy deposition are folded in.","fun_headline_variants_meta":{"raw":{"variants":["Crystal collimation cuts LHC halo losses up to 30x","Bent crystal improves LHC cleaning up to 30x at 6.5 TeV","Vertical crystal slashes LHC beam halo losses 3-10x","LHC halo cleaning boosted by bent crystal primary stage","Crystal primary collimator reduces LHC losses to 1/30"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000257,"raw_usage":{"total_tokens":1576,"prompt_tokens":938,"completion_tokens":638,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":554,"completion_tokens_details":{"reasoning_tokens":540}},"tokens_in":554,"tokens_out":638,"duration_ms":6970,"temperature":1.0,"reasoning_tokens":540,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T16:25:28.834214+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A full Monte Carlo simulation that tracks hadronic showers and energy deposition from the upstream TCLAs to the Q7 beam-loss monitors, seeded by the SixTrack loss patterns, would settle it: if the predicted Q7 leakage does not reach the measured values, the shower hypothesis and the quantitative gain would need revision; if it does, the measured leakage ratios become directly interpretable.","supporting_citations":[{"cited_title":"Assmann, O","cited_arxiv_id":null,"evidence_quote":"Defines the multistage LHC collimation system whose performance is the baseline in every comparison."},{"cited_title":"Scandale, G","cited_arxiv_id":null,"evidence_quote":"Reports the first observation of channeling at 6.5 TeV and introduces the flux-normalization procedure that makes crystal and standard loss maps comparable."},{"cited_title":"Rossi, Experimental Assessment of Crystal Collimation at the Large Hadron Collider","cited_arxiv_id":null,"evidence_quote":"Provides the experimental configurations and earlier crystal-collimation results that the present loss-map campaign extends."},{"cited_title":"Holzer, B","cited_arxiv_id":null,"evidence_quote":"Describes the beam-loss-monitor system that records the ionization-chamber signals used to build loss maps."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Motivates the white-noise excitation used to generate the controlled losses in each loss map."},{"cited_title":"Rossi, D","cited_arxiv_id":null,"evidence_quote":"Documents the horizontal crystal's curvature being close to the critical radius, the dechanneling effect invoked to explain the horizontal-vertical difference."},{"cited_title":"Belohrad, J","cited_arxiv_id":null,"evidence_quote":"Supplies the bunch-by-bunch beam-current measurements used to evaluate the beam-loss flux for normalizing each BLM signal."},{"cited_title":"Schmidt, CERN, Geneva, Switzerland, Rep","cited_arxiv_id":null,"evidence_quote":"The tracking code whose aperture and loss-pattern model produces the simulated loss maps."}],"review_version":1}