Simulations and measurements of beam loss patterns at the CERN Large Hadron Collider

The CERN Large Hadron Collider (LHC) is designed to collide proton beams of unprecedented energy, in order to extend the frontiers of high-energy particle physics. During the first very successful running period in 2010–2013, the LHC was routinely storing protons at 3.5–4 TeV with a total beam energy of up to 146 MJ, and even higher stored energies are foreseen in the future. This puts extraordinary demands on the control of beam losses. An uncontrolled loss of even a tiny fraction of the beam could cause a superconducting magnet to undergo a transition into a normal-conducting state, or in the worst case cause material damage. Hence a multistage collimation system has been installed in order to safely intercept high-amplitude beam protons before they are lost elsewhere. To guarantee adequate protection from the collimators, a detailed theoretical understanding is needed. This article presents results of numerical simulations of the distribution of beam losses around the LHC that have leaked out of the collimation system. The studies include tracking of protons through the fields of more than 5000 magnets in the 27 km LHC ring over hundreds of revolutions, and Monte Carlo simulations of particle-matter interactions both in collimators and machine elements being hit by escaping particles. The simulation results agree typically within a factor 2 with measurements of beam loss distributions from the previous LHC run. Considering the complex simulation, which must account for a very large number of unknown imperfections, and in view of the total losses around the ring spanning over 7 orders of magnitude, we consider this an excellent agreement. Our results give confidence in the simulation tools, which are used also for the design of future accelerators.

Figure 1

Stored beam energy, for the two beams called B1 and B2, at the beginning of each LHC physics fill in 2011 and 2012 with proton collisions. The operational energy was 3.5 TeV in 2011 and 4 TeV in 2012. At the beginning of each year, a gradual ramp up in intensity was performed for machine-protection reasons.

Figure 2

The schematic layout of the LHC (the separation of the two rings is not to scale). The two beams collide at the four experiments ATLAS, ALICE, CMS, and LHCb. Adapted from Ref. [27].

Figure 3

A secondary collimator jaw made of CFC (left) and two parallel jaws installed in a collimator tank seen from the top (right), where the beam should pass in the center between the jaws. The primary collimators are similar but with an effective length of only 60 cm.

Figure 4

Examples of the horizontal phase space of the starting distributions at the TCP for annular halo (left) and direct halo (right). Each blue point represents a single particle, and the red ellipse the matched 5.7 $\sigma$ envelope. The vertical red lines represent the cuts of the TCP jaws.

Figure 5

Examples of beam loss distribution around the LHC, measured with the BLMs using a 1.3 s integration time, from physics operation on August 17, 2011 (top) and from a qualification loss map of the betatron collimation system on March 11, 2011 (bottom). Here losses were provoked in B1 by crossing the third-order resonance in the horizontal plane. The physics loss distribution was measured 10 min after the start of stable beams in fill number 2031.

Figure 6

Beam loss distributions around the LHC as measured by BLMs during a qualification loss map on April 12, 2011 (top) and from a SixTrack simulation (bottom), with the results binned in 1 m intervals. Both simulation and measurement assume a beam energy of 3.5 TeV and ${\beta }^{*}=1.5\mathrm{m}$. They are both normalized to the highest loss, and the initial losses occur in the horizontal plane in B1.

Figure 7

Loss locations in IR7 (zoom of Fig. 6) from measurement (top) and SixTrack (bottom). The layout of the main magnetic elements (quadrupoles and dipoles) as well as the collimators is also shown, together with the LHC cell numbers at the cold loss locations.

Figure 8

The simulated distribution, at 3.5 TeV and ${\beta }^{*}=1.5\mathrm{m}$, of the most common histories of previously impacted collimators for the protons that are finally lost in a nuclear inelastic interaction on the TCTH in IR1 (55: TCTH IR1; 10–12: vertical, horizontal, and skew TCP in IR7; 16-44: different TCSs in IR7). For readability multiple entries of the same collimator have been neglected, e.g. particles with the history (11,11,44,55), where the two hits on collimator 11 occur on different turns, are counted in the same bin as (11,44,55).

Figure 9

The simulated distribution of histories of previously undergone pointlike interactions in other collimators for the protons that are finally lost in a nuclear inelastic interaction on the TCTH in IR1 (NE designates nuclear elastic scattering; SD designates single-diffractive scattering; RU designates Rutherford scattering). The particles that have not undergone any pointlike interactions have acquired small offsets in angle and energy in repeated passages in collimators through multiple scattering and ionization. Multiple entries of the same physical processes have been grouped together, e.g. the left-most bin contains protons that have undergone nuclear elastic scattering one or several times but no other pointlike interaction.

Figure 10

The dependence of various losses on the TCP impact parameter $b$, as simulated with SixTrack. All simulations were carried out for a horizontal halo in B1.

Figure 11

The measured distribution of horizontal orbit drifts in B1 at the beam position monitors over six cells left and right of IR7. The orbit was sampled every minute during all physics fills in 2011.

Figure 12

The influence, as simulated with SixTrack, of different machine imperfections (T designates tilt error; C designates center error; G designates gap error; O designates optics errors introduced in the lattice in MAD-X; F designates jaw flatness error; A designates aperture misalignments) on different losses.

Figure 13

The ratio of BLM signal, or particles lost, at the BLMs with the highest signal in the IR7 DS, to the signal at the horizontal TCP, in simulations and measurements for the 2011 machine for horizontal losses in B1. The errors on the SixTrack simulations indicate the standard deviation over different random seeds with collimator imperfections. The error bars on the measurements are taken as the standard deviation over the data set.

Figure 14

The 3D geometry surrounding the three IR7 TCPs, as implemented in FLUKA. The collimator jaws are contained in the green metal tanks and the BLMs are the upright cylinders below each collimator.

Figure 15

The ratio of BLM signal, or particles lost, on horizontal and vertical TCTs to the TCPs in simulations and measurements in the 2011 machine. Simulation results are shown both from counting primary losses in SixTrack, as well as with a two-step simulation where FLUKA simulates the shower to the BLMs, starting from the SixTrack impacts in the simulations including imperfections. The errors on the SixTrack simulations indicate the standard deviation over different random seeds with imperfections.