Beam Dump Facility

• Beam dump facilities are fixed-target systems that absorb intense beams to enable rare process detection and safeguard accelerator components.
• They employ advanced multiphysics simulations—combining Monte Carlo, thermal, structural, and fluid analyses—to optimize energy deposition and structural integrity.
• Their integrated design uses effective shielding and absorber materials to transform beam termination into a platform for new physics searches such as dark matter and heavy neutral leptons.

A beam dump facility is a fixed-target arrangement in which an intense primary beam is directed into a thick target or absorber designed to stop essentially all primary particles and most charged secondaries, while allowing neutral particles and weakly interacting states to emerge toward downstream instrumentation. In modern accelerator practice, beam dumps occupy a dual role. They are beam-intercepting devices required for machine protection, but they are also “intensity-frontier colliders” in which very high beam power and a thick target are traded for sensitivity to rare processes, including hidden-sector production, decay-at-rest neutrino sources, and parasitic muon beams (Buonocore et al., 2018). The concept spans dedicated experimental installations such as the SPS Beam Dump Facility, near-detector dark-sector proposals such as SBN-BD and PIP2-BD, internal accelerator dumps such as the SPS TIDVG#5, and high-energy machine-protection systems such as the LHC external dumps (Aberle et al., 2022).

1. Defining characteristics

Beam dump operation differs from a conventional neutrino beamline in three design choices. A neutrino beamline uses a relatively thin target, focuses charged pions and kaons with a magnetic horn, and provides a long decay volume so that mesons decay in flight. A beam dump instead uses a thick, dense absorber that stops charged mesons quickly, avoids long decay pipes and focusing horns, and surrounds the interaction region with substantial shielding so that ordinary neutrino and muon fluxes are minimized while neutral secondaries can escape (Toups et al., 2022).

Within the broader taxonomy of beam-intercepting devices, beam dumps sit alongside targets, scrapers, collimators, and protection absorbers. The distinguishing feature is that their primary function is beam termination rather than beam shaping or secondary-particle optimization, although in practice many facilities are hybrids: the target/dump at CERN’s BDF must absorb the full SPS beam while maximizing charm and beauty production, and the graphite dump in PIP2-BD is chosen to favor pion production and reduce neutron backgrounds (Reggiani, 2 Jun 2026).

A common misconception is that a dump is only a passive absorber. Several recent studies treat the dump itself as an active production region. The Fermilab Booster Neutrino Beam iron dump has been analyzed as a source of heavy neutral leptons produced by neutrino upscattering, and the central claim of that work is that the dump must be treated on the same footing as the dirt and the detector in short-baseline new-physics searches (Dutta et al., 16 Jan 2025). This suggests that “beam dump facility” is best understood as a coupled system of target, absorber, shielding, transport geometry, and detector acceptance, rather than as a single block of material.

2. Architectures and engineering realizations

The generic beam dump layout consists of a primary beam and target, a shielding or absorber region, often a decay or drift volume, and a downstream detector. That abstract pattern is realized very differently across facilities, depending on beam energy, time structure, and physics goals (Buonocore et al., 2018).

At CERN’s SPS Beam Dump Facility, design studies are anchored to a 400 GeV proton beam with $4\times 10^{19}$ protons per year. The target/dump is a dense, high-$Z$ assembly of TZM and tungsten discs, cladded with Ta or Ta–2.5%W, with 5 mm water gaps between blocks, about 12 interaction lengths long, and embedded in a massive target complex with a hadron absorber, muon shield, and a $\sim 50$ m vacuum decay volume for SHiP (Kershaw et al., 2018). In one design phase, the average beam power is quoted as 356 kW with 305 kW deposited in the target; in prototype studies the target is treated as absorbing about 300 kW on average, with 355 kW total beam power and the remainder deposited in shielding (Sola et al., 2019).

At Fermilab, SBN-BD proposes a dedicated 8 GeV proton beam dump aligned with the Short Baseline Neutrino detectors. Its baseline absorber is an Fe target about 2 m long and about 1 m wide, placed directly at the end of the beam pipe with no air gap, and sized for up to 80 kW of deposited power. That geometry is chosen to suppress decay-in-flight neutrinos: relative to standard Be-target BNB running, the design reduces decay-in-flight neutrino-induced backgrounds by a factor of 1000, and relative to the MiniBooNE 50 m absorber configuration by about a factor of 20 (Toups et al., 2022).

PIP2-BD takes a different approach. The reference dump is graphite, 3.6 m long and 27 cm transverse, with an incident Gaussian beam profile of 4.5 cm sigma. The choice is explicitly physics-driven: graphite gives higher pion production per proton than heavy Hg or W spallation targets and yields fewer beam-related fast neutrons, which is advantageous for a nearby low-threshold detector (Toups et al., 2022).

Machine-protection dumps show another engineering branch of the concept. The SPS internal beam dump TIDVG#5 operates inside the accelerator vacuum and was designed for proton beams from 14 to 450 GeV/$c$ and average deposited power up to $\sim 270$ kW. Its 5.0 m core uses 4.4 m of graphite, followed by 0.2 m of TZM and 0.389 m of tungsten, all surrounded by CuCr1Zr heat sinks and extensive shielding (Francia et al., 2023). The LHC external dumps push the scale further: each 6.4-tonne unit receives up to 539 MJ per beam in Run 3, with about 397 MJ deposited in graphite and about 23 MJ in the stainless-steel vessel during a full dump (Maestre et al., 2021).

3. Simulation, multiphysics design, and event generation

Beam dump design is intrinsically multiphysics. A pragmatic workflow begins with energy-deposition assessment, usually with FLUKA or MCNP, then couples that result to thermal, structural, and fluid-dynamic analyses, typically in ANSYS or equivalent tools. The central design drivers are energy deposition, temperature gradients, thermal stress, fatigue, radiation damage, and activation (Reggiani, 2 Jun 2026).

This workflow is visible across facilities. The CERN BDF target prototype program used FLUKA for 3D energy deposition, ANSYS CFX for cooling and heat-transfer coefficients, and ANSYS Mechanical for transient thermal and structural calculations. In the prototype, power density per spill was explicitly matched to the final target at 38 MW/m$^3$, and measured temperatures were generally within 10% of simulation while strain variations were generally within 25% (Sola et al., 2019). The PSB dump used FLUKA for the 2 GeV proton beam, CFD to optimize forced-air cooling around a finned CuCr1Zr core, and thermo-mechanical analysis to show maximum dynamic von Mises stress of about 51.2 MPa in the most loaded upstream block (Perillo-Marcone et al., 2019).

For new-physics interpretation, end-to-end event generation is equally important. MadDump extends MadGraph5_aMC@NLO to beam dump geometries by factorizing production and detection. The key object is a multidifferential hidden-particle flux,

$\phi(E,\vec{x})=\frac{dn_{\rm HP}}{dE\,d\vec{x}},$

which is convolved with geometry and detection cross sections. MadDump defines an effective detection cross section

$$\sigma_D=\int dE \int d^n x\, \phi(E,\vec{x})\, W(E,\vec{x})\, \hat{\sigma}_D(E),$$

and uses adaptive fits to preserve energy-angle correlations and detector geometry (Buonocore et al., 2018). This matters because beam dump sensitivities are usually controlled by forward kinematics, path length through the detector, delayed decays, and off-axis acceptance, not merely by inclusive rates.

The same logic appears in facility-specific simulation stacks. PIP2-BD combines Geant4 for dump hadroproduction with BdNMC for dark-sector event generation, while the LHC dump program coupled FLUKA power-density maps to implicit and explicit finite-element simulations in ANSYS and LS-DYNA to resolve both slow thermal expansion and fast shock-driven motion (Toups et al., 2022, Maestre et al., 2021).

4. Physics programme

The modern beam dump facility is optimized for weakly coupled states: heavy neutral leptons, vectors, scalars, axion-like particles, and light dark matter. Production mechanisms fall into three broad classes: prompt production in primary interactions, rare meson decays, and bremsstrahlung-like emission in the target (Buonocore et al., 2018).

At $\mathcal{O}(10\,\mathrm{GeV})$ proton energies, neutral meson decays are central. In SBN-BD, the dump produces large numbers of $\pi^0$, $Z$0, and heavier mesons, enabling channels such as $Z$1 followed by $Z$2. With $Z$3 POT in five years, SBND at 100 m is projected to achieve greater than an order-of-magnitude improvement over the MiniBooNE dark matter search, with sensitivity presented in both NC$Z$4-like and NC electron scattering channels (Toups et al., 2022).

At lower proton energies, stopped-pion and light-meson production become the organizing principle. PIP2-BD uses 800 MeV to 1.2 GeV protons to generate a nearly pure decay-at-rest neutrino flux and prompt dark-sector signals. Its reference detector is a 100-ton LAr scintillation-only experiment, and the facility is explicitly framed as having world-leading sensitivities to low-mass dark matter, axion-like particles, sterile neutrinos, and non-standard interactions, with staged beam options ranging from 0.09 MW to 1.3 MW (Toups et al., 2022).

Electron-dump facilities probe a complementary regime. The Jefferson Lab proposal builds on the BDX underground vault to use a 12 GeV continuous-wave electron dump as a source of secondary muons, neutrinos, and hypothetical light dark matter. There the dump is treated as a compact source of decay-at-rest neutrinos, forward muons, and dark bremsstrahlung production, extending the beam dump concept beyond hadronic facilities (Achenbach et al., 6 Oct 2025).

Another misconception is that only dedicated dump mode matters for new physics. In short-baseline neutrino facilities, BSM production may occur in the dump, the dirt, or the detector. For dipole-portal heavy neutral leptons at the BNB, the dump, dirt, and detector produce distinct timing, energy, and angular signatures, and dump-origin events can be especially valuable because they are highly forward and delayed relative to prompt neutrino backgrounds (Dutta et al., 16 Jan 2025).

5. Shielding, backgrounds, atmosphere control, and lifecycle

Beam dump facilities are built around aggressive background suppression. Shielding is used not only for prompt radiation protection but also to suppress beam-related muons, neutrons, and neutrinos in downstream detectors. In SBN-BD, the dense Fe dump, no-air-gap beamline termination, and surrounding shielding are specifically intended to minimize decay-in-flight neutrinos and penetrating secondaries (Toups et al., 2022). In the SPS BDF studies, the public-dose design target is $Z$5, and the facility must remain underground so that the intense muon flux ranges out safely in the surrounding ground (Aberle et al., 2022).

Atmosphere control is another facility-level design choice. The CERN BDF target and shielding sit inside a helium vessel with a free gas volume of about 75 m$Z$6, maintained at +50 Pa relative to the surrounding underground volume. The passivation system is designed to keep the vessel at 99.9% helium purity, circulate about 770 m$Z$7/h for cooling, remove up to 3 kW of heat, and purify helium at about 75 Nm$Z$8/h via low-temperature adsorption around 77 K (Avigni et al., 2019). The rationale is explicit: helium reduces gas activation, limits corrosion, and extends component lifetime.

High activation makes remote handling intrinsic to the facility concept. The BDF target complex was developed with both crane-based and trolley-based remote handling, shielded storage pits, remote manipulators, and dedicated hot-cell functions. The BDF prototype target tests at CERN also validated fully metallic remote plug-in systems for water and electrical services under 22 bar and high-radiation conditions (Kershaw et al., 2018, Sola et al., 2019).

Lifecycle evidence from the LHC beam dumps shows why this matters. Post-irradiation examination of the Run 1–2 dumps found that extruded graphite plates exhibited a cracking pattern, likely due to the dynamic response of the device upon beam impact, and their retaining rings were displaced. By contrast, expanded graphite sheets were intact and the isostatic graphite blocks showed no evidence of material degradation, despite deposited energy densities up to 1.5 kJ/g (Solieri et al., 7 May 2025). This suggests that for very high-energy dumps the limiting factor may be dynamic structural details rather than the absorber bulk alone.

6. Representative facilities and technological directions

The term now covers a family of implementations rather than a single canonical machine.

Facility	Beam	Characteristic feature
CERN SPS BDF / SHiP	400 GeV protons	Thick high-$Z$9 target/dump, muon shield, long decay volume
Fermilab SBN-BD	8 GeV protons	Fe dump on-axis with SBN, factor-1000 DIF neutrino reduction
Fermilab PIP2-BD	800 MeV–1.2 GeV protons	Graphite dump with 100-ton LAr detector
Jefferson Lab BDX-based facility	12 GeV electrons	Underground vault for secondary muons, neutrinos, and LDM
CERN SPS TIDVG#5	14–450 GeV/$\sim 50$0 protons	Internal dump in UHV with graphite–TZM–W graded core
LHC external dumps	6.8 TeV protons	External dumps absorbing up to 539 MJ per beam

Across these systems, the engineering trend is clear. Facilities converge on detailed Monte Carlo plus thermo-mechanical plus CFD design, on radiation-tolerant materials and instrumentation, and on remote handling as a baseline requirement (Francia et al., 2023). They diverge where the physics diverges. High-$\sim 50$1 refractory targets are chosen when charm, beauty, and compact hadronic absorption dominate, as in the SPS BDF. Graphite is favored when pion yield and neutron suppression matter, as in PIP2-BD. Dense Fe absorbers are favored when one wants to suppress decay-in-flight backgrounds while preserving neutral-meson production, as in SBN-BD (Sola et al., 2019, Toups et al., 2022, Toups et al., 2022).

A further extension is the use of parasitic beams behind dumps. At the Fermilab Test Beam Area, operation in low-energy pion mode produces a muon beam behind the concrete dump, with momenta in the range 10–50 GeV/$\sim 50$2, beam spot FWHM between 10 and 20 cm, and peak fluxes of 1.1, 2.2, and 2.6 cm$\sim 50$3 per spill for 16, 24, and 28 GeV/$\sim 50$4 pion settings, normalized to $\sim 50$5 pion counts per spill in MT6SC1 (Denisov et al., 2016). This illustrates that even when a dump is not itself the primary experiment, it can become a controlled secondary-beam source.

Taken together, these developments indicate that the contemporary beam dump facility is an integrated accelerator-and-detector system. Its essential variables are beam power, duty factor, absorber material, geometric suppression of ordinary secondaries, and the ability to model, instrument, service, and eventually dissect highly activated hardware. The scientific value of the facility follows directly from how well those engineering choices transform a beam terminus into a source of measurable rare processes.