Could a single reconfigurable dump geometry efficiently capture multiple exotic particle types simultaneously, or is optimization inherently particle-specific?

Your question touches on a real tension in experimental design: can a single detector geometry serve multiple particle types well, or does optimization inevitably sacrifice generality?

The short answer from the archive is optimization is inherently particle-specific, but clever detector placement can recover much of the loss.

Why particle-specificity matters

Different exotic particles have fundamentally different angular distributions when produced. [AG-2026.01-1312] directly addresses this: "the geometric setup is particularly relevant for the specific new-physics scenario under study, since different production mechanisms can generate different angular distributions of new particles." For instance, a particle produced in a beam dump via photon coupling will spray outward differently than one from neutrino upscattering in iron [AG-2025.01-1235].

This is physics, not engineering: the geometry that maximizes acceptance for one production channel can blind you to another. That said, [AG-2026.01-1312] shows that "even the most minimalistic reconfiguration of the existing NA62 experiment's detectors can already provide a very competitive sensitivity"—suggesting that modest geometric adjustments can capture multiple scenarios without a complete redesign.

The beam-dump "ceiling" complication

There's another wrinkle: [AG-2024.01-1042] identifies a fundamental trade-off called the "beam-dump ceiling" for prompt-decay mediators. Beyond a certain parameter space, sensitivity gains saturate regardless of geometry—the physics itself limits what any single configuration can achieve. This applies across different particle types, suggesting that multi-particle optimization hits a hard wall.

Practical compromise: complementarity

Rather than a single reconfigurable geometry serving all particles equally, [AG-2025.01-1235] advocates a different strategy: using multiple locations (dirt, detector, dump) that produce "distinct features in the signal events' kinematical properties" for separating signals. This trades single-detector versatility for multi-site complementarity.

The provided preprints don't directly benchmark a single reconfigurable dump against simultaneous optimization for, say, axion-like particles *and* heavy neutral leptons—so I can't give you quantitative guidance on efficiency loss across particle types.