Simulation of particle systems with retQSS: selected case studies and applications
This dataset contains the underlying experimental data used in a research article on retQSS, a novel methodology for efficient modeling and simulation of particle systems in reticulated geometries. We tackled four case studies of very different nature and analyzed the efficiency of our approach by establishing modeling and performance comparisons against other related simulation toolkits. Here, we provide not only the raw experimental data found during our research but also the definition and implementation of the corresponding particle models (including the geometry files in VTK format). Below we describe briefly the contents of each file attached: * Bouncing balls: a simple motivational example consisting of balls bouncing in a chamber with obstacles, used to drive the discussion of the most relevant features and capabilities of retQSS. This case study does not include experimental data. * Bird flocking: proposed retQSS models for bird flocking, following the classic approach introduced by Reynolds where birds are characterized as an agent system with emergent behavior. We also provide two alternative geometries (tesstellated sphere and torus). * High energy physics: a model of a charged subatomic particle describing helical trajectories in a mesh of cuboids. We compared the performance of retQSS against an equivalent model in Geant4, the reference simulation toolkit in HEP experiments. * Molecular interactions: a system of molecules interacting via an exponentially decaying potential. This case study showcases the ability of retQSS to implement particle-based numerical methods (in particular, Molecular Dynamics). We also established a performance comparison against Aboria, a general-purpose software library for particle-based methods. * Plasma flow: a model for the flow of plasma in a two-dimensional grid. This case study is devoted to show custom approaches to the Particle-In-Cell numerical method using retQSS. We studied the accuracy and efficiency of our proposed models against a reference implementation in Octave. The experimental data is usually given by unmodified raw outputs from the simulators (in some cases, though, it was conveniently post-processed). The experimentation was conducted in the TUPAC computer cluster (https://www.tupac.gob.ar/).