Hadron Collider Experiments Research Articles

Triggering dark showers with conditional dual auto-encoders

Abstract We present a family of conditional dual auto-encoders (CoDAEs) for generic and model-independent new physics searches at colliders. New physics signals, which arise from new types of particles and interactions, are considered in our study as anomalies causing deviations in data with respect to expected background events. In this work, we perform a normal-only anomaly detection, which employs only background samples, to search for manifestations of a dark version of strong force applying (variational) auto-encoders on raw detector images, which are large and highly sparse, without leveraging any physics-based pre-processing or strong assumption on the signals. The proposed CoDAE has a dual-encoder design, which is general and can learn an auxiliary yet compact latent space through spatial conditioning, showing a neat improvement over competitive physics-based baselines and related approaches, therefore also reducing the gap with fully supervised models. It is the first time an unsupervised model is shown to exhibit excellent discrimination against multiple dark shower models, illustrating the suitability of this method as an accurate, fast, model-independent algorithm to deploy, e.g. in the real-time event triggering systems of large hadron collider experiments such as ATLAS and CMS.

Simultaneous energy and mass calibration of large-radius jets with the ATLAS detector using a deep neural network

Abstract The energy and mass measurements of jets are crucial tasks for the Large Hadron Collider experiments. This paper presents a new calibration method to simultaneously calibrate these quantities for large-radius jets measured with the ATLAS detector using a deep neural network (DNN). To address the specificities of the calibration problem, special loss functions and training procedures are employed, and a complex network architecture, which includes feature annotation and residual connection layers, is used. The DNN-based calibration is compared to the standard numerical approach in an extensive series of tests. The DNN approach is found to perform significantly better in almost all of the tests and over most of the relevant kinematic phase space. In particular, it consistently improves the energy and mass resolutions, with a 30% better energy resolution obtained for transverse momenta p T > 500 GeV.

Calculable neutrino Dirac mass matrix and one-loop θ¯ in the minimal left-right symmetric model

We revisit the contribution to the strong CP parameter θ¯ from leptonic CP violation at one-loop level in the minimal left-right symmetric model in the case of generalized parity as the left-right symmetry. The Hermitian neutrino Dirac mass matrix MD can be calculated using the light and heavy neutrino masses and mixings. We propose a parametrization of the right-handed neutrino mixing matrix VR and construct the heavy neutrino mass that maintains the Hermiticity of MD. We further apply it to evaluate the one-loop θ¯, denoted as θ¯loop, as a function of the sterile neutrino masses for explicit examples of VR. By requiring the magnitude of θ¯loop≲10−10, we derive the upper limits on the sterile neutrino masses, which are within reach of direct searches at the Large Hadron Collider and neutrinoless double beta decay experiments. Furthermore, our parametrization is pertinent for other phenomenological investigations. Published by the American Physical Society 2024

A Lightweight Algorithm to Model Radiation Damage Effects in Monte Carlo Events for High-Luminosity Large Hadron Collider Experiments.

Radiation damage significantly impacts the performance of silicon tracking detectors in Large Hadron Collider (LHC) experiments such as ATLAS and CMS, with signal reduction being the most critical effect; adjusting sensor bias voltage and detection thresholds can help mitigate these effects, generating simulated data that accurately mirror the performance evolution with the accumulation of luminosity, hence fluence, is crucial. The ATLAS and CMS collaborations have developed and implemented algorithms to correct simulated Monte Carlo (MC) events for radiation damage effects, achieving impressive agreement between collision data and simulated events. In preparation for the high-luminosity phase (HL-LHC), the demand for a faster ATLAS MC production algorithm becomes imperative due to escalating collision, events, tracks, and particle hit rates, imposing stringent constraints on available computing resources. This article outlines the philosophy behind the new algorithm, its implementation strategy, and the essential components involved. The results from closure tests indicate that the events simulated using the new algorithm agree with fully simulated events at the level of few %. The first tests on computing performance show that the new algorithm is as fast as it is when no radiation damage corrections are applied.

Time-resolved synchrotron light source X-ray detection with Low-Gain Avalanche Diodes

Low Gain Avalanche Diodes (LGADs) represent the state-of-the-art in timing measurements and will instrument future timing detectors at the High Luminosity Large Hadron Collider (HL-LHC) experiments. While conceived as a sensor for charged particles, the intrinsic gain of LGADs makes it possible to detect low energy X-rays with good energy resolution and excellent timing (tens of picoseconds). Using the Stanford Synchrotron Radiation Lightsource (SSRL) at SLAC, several LGADs designs were characterized with energies from 5 to 35 keV. The SSRL provides 10 ps pulsed X-ray bunches separated by 2.1ns intervals, and with an energy dispersion (ΔE/E) of 1 × 10-4. LGADs fabricated by Hamamatsu Photonics (HPK) and Brookhaven National Laboratory (BNL) with different thicknesses ranging from 20µm to 50µm and different gain layer designs were read out a two stage fast amplification circuit and digitized with a high bandwidth, high sampling rate oscilloscope. PIN devices from HPK were characterized as well. A systematic and detailed characterization of the devices’ energy linearity, resolution and timing resolution as a function of X-ray energy was performed for different biasing voltages at room temperature.

High-rate tests on resistive plate chambers operated with eco-friendly gas mixtures

Results obtained by the RPC ECOgas@GIF++ Collaboration, using Resistive Plate Chambers operated with new, eco-friendly gas mixtures, based on tetrafluoropropene and carbon dioxide, are shown and discussed in this paper. Tests aimed to assess the performance of this kind of detectors in high-irradiation conditions, analogous to the ones foreseen for the coming years at the Large Hadron Collider experiments, were performed, and demonstrate a performance basically similar to the one obtained with the gas mixtures currently in use, based on tetrafluoroethane, which is being progressively phased out for its possible contribution to the greenhouse effect. Long term aging tests are also being carried out, with the goal to demonstrate the possibility of using these eco-friendly gas mixtures during the whole High Luminosity phase of the Large Hadron Collider.

Entropy and heat capacity of the transverse momentum distribution for pp collisions at RHIC and LHC energies

We investigate the transverse momentum distribution (TMD) statistics from three different theoretical approaches. In particular, we explore the framework used for string models, wherein the particle production is given by the Schwinger mechanism. The thermal distribution arises from the Gaussian fluctuations of the string tension. The hard part of the TMD can be reproduced by considering heavy tailed string tension fluctuations, for instance, the Tsallis q-Gaussian function, giving rise to a confluent hypergeometric function that fits the entire experimental TMD data. We also discuss the QCD-based Hagerdon function, another family of fitting functions frequently used to describe the spectrum. We analyze the experimental data of minimum bias pp collisions reported by the BNL Relativistic Heavy Ion Collider (RHIC) and the CERN Large Hadron Collider (LHC) experiments (from s=0.2 TeV to s=13 TeV). We extracted the corresponding temperature by studying the behavior of the spectra at low transverse momentum values. For the three approaches, we compute all moments, highlighting the average, variance, and kurtosis. Finally, we compute the Shannon entropy and the heat capacity through the entropy derivative with respect to the temperature. We found that the q-Gaussian string tension fluctuations lead to a monotonically increasing heat capacity as a function of the center-of-mass energy, which is also observed for the Hagedorn fitting function. This behavior is consistent with the experimental observation that the temperature slowly rises with increments of the collision energy. Published by the American Physical Society 2024

LL-GNN: Low Latency Graph Neural Networks on FPGAs for High Energy Physics

This work presents a novel reconfigurable architecture for Low Latency Graph Neural Network (LL-GNN) designs for particle detectors, delivering unprecedented low latency performance. Incorporating FPGA-based GNNs into particle detectors presents a unique challenge since it requires sub-microsecond latency to deploy the networks for online event selection with a data rate of hundreds of terabytes per second in the Level-1 triggers at the CERN Large Hadron Collider experiments. This article proposes a novel outer-product based matrix multiplication approach, which is enhanced by exploiting the structured adjacency matrix and a column-major data layout. In addition, we propose a custom code transformation for the matrix multiplication operations, which leverages the structured sparsity patterns and binary features of adjacency matrices to reduce latency and improve hardware efficiency. Moreover, a fusion step is introduced to further reduce the end-to-end design latency by eliminating unnecessary boundaries. Furthermore, a GNN-specific algorithm-hardware co-design approach is presented which not only finds a design with a much better latency but also finds a high accuracy design under given latency constraints. To facilitate this, a customizable template for this low latency GNN hardware architecture has been designed and open-sourced, which enables the generation of low-latency FPGA designs with efficient resource utilization using a high-level synthesis tool. Evaluation results show that our FPGA implementation is up to 9.0 times faster and achieves up to 13.1 times higher power efficiency than a GPU implementation. Compared to the previous FPGA implementations, this work achieves 6.51 to 16.7 times lower latency. Moreover, the latency of our FPGA design is sufficiently low to enable deployment of GNNs in a sub-microsecond, real-time collider trigger system, enabling it to benefit from improved accuracy. The proposed LL-GNN design advances the next generation of trigger systems by enabling sophisticated algorithms to process experimental data efficiently.

Generative machine learning for detector response modeling with a conditional normalizing flow

In this paper, we explore the potential of generative machine learning models as an alternative to the computationally expensive Monte Carlo (MC) simulations commonly used by the Large Hadron Collider (LHC) experiments. Our objective is to develop a generative model capable of efficiently simulating detector responses for specific particle observables, focusing on the correlations between detector responses of different particles in the same event and accommodating asymmetric detector responses. We present a conditional normalizing flow model (𝒞𝒩ℱ) based on a chain of Masked Autoregressive Flows, which effectively incorporates conditional variables and models high-dimensional density distributions. We assess the performance of the 𝒞𝒩ℱ model using a simulated sample of Higgs boson decaying to diphoton events at the LHC. We create reconstruction-level observables using a smearing technique. We show that conditional normalizing flows can accurately model complex detector responses and their correlation. This method can potentially reduce the computational burden associated with generating large numbers of simulated events while ensuring that the generated events meet the requirements for data analyses. We make our code available at https://github.com/allixu/normalizing_flow_for_detector_response.

Experimental overview: Quarkonium production in relativistic particle collisions

Quarkonium provides a golden probe of the formation of the Quark Gluon Plasma (QGP). Its production in heavy-ion collisions can be affected by an interplay of different phenomena such as medium-induced dissociation and heavy-quark (re)combination in the QGP, and cold nuclear matter effects due to the presence of the nucleus. The measurement of quarkonium states in different collision systems and beam energies allows to gain further insights on these effects and probe the properties of the QGP. In these proceedings, an overview of the main experimental quarkonium measurements in proton-proton, protonnucleus and nucleus-nucleus collisions is presented, giving emphasis on the most recent results from the Large Hadron Collider and Relativistic Heavy Ion Collider experiments shown at the 2023 Quark Matter conference.

A holistic study of the WLCG energy needs for the LHC scientific program

A holistic study of the WLCG energy needs for the LHC scientific program

Automated Network Services for Exascale Data Movement

The Large Hadron Collider (LHC) experiments distribute data by leveraging a diverse array of National Research and Education Networks (NRENs), where experiment data management systems treat networks as a “blackbox” resource. After the High Luminosity upgrade, the Compact Muon Solenoid (CMS) experiment alone will produce roughly 0.5 exabytes of data per year. NREN Networks are a critical part of the success of CMS and other LHC experiments. However, during data movement, NRENs are unaware of data priorities, importance, or need for quality of service, and this poses a challenge for operators to coordinate the movement of data and have predictable data flows across multi-domain networks. The overarching goal of SENSE (The Software-defined network for End-to-end Networked Science at Exascale) is to enable National Labs and universities to request and provision end-to-end intelligent network services for their application workflows leveraging SDN (Software-Defined Networking) capabilities. This work aims to allow LHC Experiments and Rucio, the data management software used by CMS Experiment, to allocate and prioritize certain data transfers over the wide area network. In this paper, we will present the current progress of the integration of SENSE, Multi-domain end-to-end SDN Orchestration with QoS (Quality of Service) capabilities, with Rucio, the data management software used by CMS Experiment.

Radiation resistant optical components for high energy physics detectors

New detectors for future high-energy physics experiments will operate under unprecedented radiation dose rates. This condition requires improved radiation resistance on detector equipment. The consequent development of new materials, particularly optical materials, becomes crucial. In this work, optical components mean reflectors, light absorbers, or light transmitters. These materials, reflectors or light transmitters, are needed in detectors primarily to collect and transmit light from the scintillator to the PMT. When it comes to light absorbers, they are required to protect the detector from light from the environment. This work aims at studying selected optical materials with improved properties (functional and optical) and radiation resistance that can be used in new detectors at the Large Hadron Collider (LHC) experiments. Light transmittance, optical reflection, thermal characteristics, and radiation resistance were investigated to evaluate the proposed materials.We have developed optical systems based on siloxanes to continue our previous developments of radiation-resistant materials for radiation detectors. We also report a study of several reflective materials and light absorber introduced into the siloxane. Investigations have shown that these systems are radiation resistant to doses of at least 1 MGy.Tested samples were irradiated at linear electron accelerator LUE-40 in the National Science Center Kharkiv Institute of Physics and Technology (KIPT). The accelerated electrons were sent to the heavy complex targets to deliver the irradiation with gamma or neutron fluxes. The consequent gamma and neutron fluxes and doses were estimated with the GEANT4 simulation.

Discovering New Science in the Noise from LHC Experiments

Steven Schramm and his team of researchers in the DISCOVERHEP project are searching for evidence of new physics in previously ignored ‘noise’ data of Large Hadron Collider experiments.

Effect of ionizing radiation on thermal and mechanical properties of filled-epoxy adhesives

Stave assemblies at the Large Hadron Collider and other high-energy physics experiments require radiation-resistant adhesives composed of light elements with high thermal conductivity and low interfacial resistance. Porous carbon foams are currently bonded to carbon fiber faceplates with graphite-filled epoxy. Low-viscosity epoxies are used to increase the amount of filler in order to double the thermal conductivity without impacting the properties of the material under irradiation. Promising new adhesive formulations showed no degradation in thermal conductivity after exposure to a fluence of 1016 n/cm2 (>0.1 MeV neutrons) or 10 Gy gamma radiation, but showed variable degradation, compared to the currently used adhesive, after exposure to a fluence of 1015 p/cm2 (1 MGy, 67.5 MeV protons). Adhesives with more than double the irradiated thermal conductivity of the baseline material showed marked improvement in thermal transport in graphite/epoxy/foam/epoxy/graphite structures based on thermal diffusivity measurements. Silicon carbide particles can be added to graphite to stiffen the modulus of the filled epoxy. When spherical AlN was used as a filler, the degradation in thermal conductivity under proton irradiation varied based on the size of the ceramic filler. Options for improving the thermal performance of irradiated adhesives for high-energy physics experiments are discussed.

MetaHEP: Meta learning for fast shower simulation of high energy physics experiments

For High Energy Physics (HEP) experiments, such as the Large Hadron Collider (LHC) experiments, the calorimeter is a key detector to measure the energy of particles. Incident particles interact with the material of the calorimeter, creating cascades of secondary particles, so-called showers. A detailed description of the showering process relies on simulation methods that precisely describe all particle interactions with matter. Constrained by the need for precision, the simulation of calorimeters is inherently slow and constitutes a bottleneck for HEP analysis. Furthermore, with the upcoming high luminosity upgrade of the LHC and a much-increased data production rate, the amount of required simulated events will increase. Several research directions have recently investigated the use of Machine Learning based models to accelerate particular calorimeter response simulation. These models typically require a large amount of data and time for training, and the result is a simulation tuned specifically to this configuration. Meta-learning has emerged recently as a fast learning algorithm using small training datasets. In this paper, we use a meta-learning approach that “learns to learn” to generate showers from multiple calorimeter geometries, using a first-order gradient-based algorithm. We present MetaHEP, the first application of the meta-learning approach to accelerate shower simulation using very high granular data and using one of the calorimeters proposed for the Future Circular Collider (FCC), a next-generation of high-performance particle colliders.

Optimization strategies for the greenhouse gas consumption of the resistive plate chamber detectors at the CERN LHC experiments

Resistive Plate Chamber (RPC) detectors at the Large Hadron Collider (LHC) experiments are operated with a gas mixture containing two greenhouse gases (GHG): C2H2F4 (R134a) and SF6. These gases are used because they allow to achieve specific detector performance. However, due to their high Global Warming Potential (GWP) they are currently subject to a phase down policy that affects price and availability. To reduce the GHG usage, different strategies have been identified. As immediate actions, during the LHC Long Shutdown 2 (LS2) the RPC gas systems was upgraded to cope with new detector requirements and, in parallel, extensive campaigns for fixing leaks at detector level were performed. Since R134a dominates the GHG consumption, the development of a R134a gas recuperation plant is ongoing. A first prototype system was tested with encouraging results. A second prototype is under construction and it will be ready by beginning of 2023. For future long-term detector operation, R&D studies are ongoing for finding “green” alternatives to the currently used gases. Unfortunately, new gases developed by industry as refrigerant fluids are not behaving as the R134a in particle detectors which makes the replacement for present experiments difficult.

Operational radiation protection challenges for the LHC experiments.

Radiation protection physicists at CERN are often required to assess residual activation for the Large Hadron Collider (LHC) experiments during stop periods in order to ensure adequate optimization during planned exposure situations and to establish proper procedures for the radiological control of materials. Given the complexity of the facilities and of the high-energy and mixed fields inducing the activation, Monte Carlo transport codes are an essential tool to simulate both prompt and residual radiation. The present work highlights the challenges of assessing residual dose rates for the LHC experiments in shutdown configurations and of establishing residual activation zonings. For the latter, a method based on fluence conversion coefficients was developed and is efficiently employed. The practical example of the assessment of the activation of the 600 tons of austenitic stainless steel in the future Compact Muon Solenoid (CMS) High Granularity Calorimeter will be used to demonstrate how these challenges are dealt with and the capabilities of the method developed.

Analysis of interstrip capacitance of p+n MCz Si microstrip detector using TCAD simulation combining surface and bulk damage effects by mixed irradiation

Interstrip capacitance plays a crucial role on the detector electronic noise performance at the readout electrode of the irradiated detectors in the harsh radiation environment of the high luminosity Large Hadron Collider experiments. The major challenge is to optimize the design of the detector to get the low interstrip capacitance in the real ‘mixed’ irradiation environment of the experiment. A few of experimental studies have been performed within CERN RDD0 collaboration on the n-MCz Si detectors to extract the microscopic parameters for the surface and bulk damage effects especially by mixed irradiations in the detectors. The results on the microscopic parameters are fed into ATLAS TCAD commercial simulation program to compare the experimental data and simulation result. A very good agreement has been recorded between the data and simulation result. In this paper, the microscopic model parameters are used to investigate the effects of mixed irradiation up to a fluence of 8.82× 1014 neq./cm2 on the interstrip capacitance in n-MCz 〈100〉 silicon microstrip detectors. The results are discussed in detail using electric field distribution, electron concentration profile in the irradiated detectors for the different metal overhang width, and an optimized p+n-MCz Si microstrip design is proposed for the future collider experiment.

Improvement of q 2 Resolution in Semileptonic Decays Based on Machine Learning

The neutrino closure method is often used to obtain kinematics of semileptonic decays with one unreconstructed particle in hadron collider experiments. The kinematics of decays can be deducted by a twofold ambiguity with a quadratic equation. To resolve the twofold ambiguity, a novel method based on machine learning (ML) is proposed. We study the effect of different sets of features and regressors on the improvement of reconstructed invariant mass squared of ℓ ν system ( q 2 ). The result shows that the best performance is obtained by using the flight vector as the features and the multilayer perceptron (MLP) model as the regressor. Compared with the random choice, the MLP model improves the resolution of reconstructed q 2 by ~40%. Furthermore, the possibility of using this method on various semileptonic decays is shown.