Hidden valley scenario sensitivity in the CMS muon endcap detector

TL;DR

This study reinterprets the CMS displaced shower search to test Hidden Valley scenarios with long-lived dark-sector hadrons decaying in the muon endcap, parameterizing HV theory space by quantities such as the confinement scale $\Lambda$, the ratio $m_\pi/\Lambda$, and the flavor content ${N_F}/{N_C}$ while treating the dark-sector lifetime as a free parameter to obtain model-independent cross-section upper limits. By simulating HV/DS showers with a PYTHIA-based hadronization model and applying CMS CSC-cluster selection and MET-based triggers, the authors quantify how the signal efficiency depends on theory parameters through LLP multiplicity, geometry, reconstruction, and selection factors, and they present both current and HL-LHC projections for cross-section limits. They show that the CMS displaced shower search has sensitivity to regions of HV/DS parameter space, with stronger constraints at smaller $\Lambda$ and $m_\pi/\Lambda$, and weaker dependence on ${N_F}/{N_C}$; however, substantial hadronization uncertainties prevent a precise translation to fundamental theory parameters. The paper also provides a single model-dependent exclusion on the mediator coupling $g_D^{tot}$ and outlines how future HL-LHC analyses could substantially improve constraints, while emphasizing the importance of presenting results in HV/DS parameter terms rather than simplified-model abstractions. Overall, this work highlights a principled, theory-guided reinterpretation strategy for LLP searches in strongly interacting hidden sectors, balancing informative sensitivity studies with transparent caveats about non-perturbative modeling.

Abstract

We study the sensitivity of the CMS search to displaced showers arising from the decays of long-lived particles in the muon system, within the framework of Hidden Valley scenarios. To establish our simulation setup, we employ a parameterization of Hidden Valley theory space and adopt a hybrid strategy where the lifetime is treated as a free parameter to provide model-independent Hidden Valley quark production cross-section upper limits. Our results indicate that the CMS search is broadly sensitive to variations in Hidden Valley parameters such as the overall scale and relevant mass ratios, while showing comparatively weak dependence on the number of Hidden Valley colors or flavors. The exact quantitative results we derive depend on the underlying hadronization model employed and thus, we urge caution in interpreting the results. Within these limitations, we also establish model-dependent limits on the maximum strength of the mediator coupling to Hidden Valley quarks. Our approach illustrates how one can move beyond simplified model strategies by employing a first-principles-inspired theory setup, while highlighting the theoretical uncertainties inherent in modeling Hidden Valley phenomenology.

Figures (34)

• Figure 1: A sketch of the production and decay mechanisms for HV/DS quarks and subsequently HV/DS mesons considered in this work.
• Figure 2: (left panel) Final state pion multiplicity as a function of $m_{\pi}/\Lambda$ and ${{N_F}/{N_C}}$ for ${N_C} = 5, \Lambda = 5\,\rm{GeV}$, (right panel) Ratio of the final state pion multiplicity to total final state meson multiplicity as a function of $m_{\pi}/\Lambda$ and ${{N_F}/{N_C}}$ for ${N_C} = 5, \Lambda = 5\,\rm{GeV}$.
• Figure 3: The LLP multiplicity (left panel) and boost (right panel) as a function of $\Lambda$ with $m_{Z_D} = 3.5\,\rm{TeV}, {{N_F}/{N_C}} = 1, m_{\pi}/\Lambda = 0.6, {N_C} = 5$.
• Figure 4: Average hadronic energy deposited in the muon endcap detector per HV/DS pion (left panel), and fraction of events containing at least one cluster (right panel) for $c \tau_{\rm LLP}$ = 100 and 1000 mm (blue and orange lines) as a function of $\Lambda$. $m_{Z_D} = 3.5\,\rm{TeV}, {{N_F}/{N_C}} = 1, m_{\pi}/\Lambda = 0.6, {N_C} = 5$ are kept fixed.
• Figure 5: The multiplicity (left panel) and boost (right panel) of the diagonal pions in the events a function of ${{N_F}/{N_C}}$, with $m_{Z_D} = 3.5\,\rm{TeV}, \Lambda = 5\,\rm{GeV}, m_{\pi}/\Lambda = 0.6, {N_C} = 5$ kept fixed.