arXiv subjects

Mikio Sakai

Publications and source records attributed to Mikio Sakai.

3 recordsLinked to original sources

arXiv2026

Development of a single-parameter spring-dashpot rolling friction model for coarse-grained DEM

Simulating granular materials composed of non-spherical particles remains a major challenge in discrete element method (DEM) simulations due to the complexity of contact detection and rotational dynamics, rendering large-scale simulations computationally prohibitive. To address this limitation, rolling friction is commonly introduced as an approximation to account for particle shape effects by applying a resistive torque to spherical particles. Among existing rolling friction formulations, the spring-dashpot (S-D) type model is widely recognized for its numerical stability and realistic representation of rolling resistance. However, conventional S-D models require multiple empirical parameters that must be calibrated in an interdependent manner, leading to increased experimental effort, parameter ambiguity, and uncertainty in practical applications. To overcome these issues, this study proposes a new S-D type rolling friction model that reduces the parameter set to a single physically meaningful quantity: the critical rolling angle. Derived from theoretical considerations, this parameter characterizes the transition from static to rolling motion at particle contacts. The use of a single parameter simplifies implementation and eliminates the need for extensive calibration. Stability analysis demonstrates that the proposed model allows particles to reach a physically consistent equilibrium state without spurious rotational oscillations. For large-scale applications, the model is further integrated into a coarse-grained DEM framework. Validation using DEM-CFD simulations of an incinerator system confirms that the proposed approach successfully reproduces the macroscopic behavior of the original particle system. Overall, this study enhances the applicability of DEM for industrial-scale simulations involving non-spherical particles.

physics.comp-ph↗

arXiv2026

Mechanisms of particle entrainment in confined gas-particle systems under moving boundaries

Particle entrainment in confined gas-particle systems driven by moving boundaries is central to many industrial and natural processes, including pharmaceutical manufacturing, food processing, and chemical engineering. Although often termed a "suction effect," its physical origin remains unclear, especially under unsteady flow, strong particle interactions, and transient force networks. Here we study suction-induced entrainment in a prototypical confined system using high-fidelity coupled CFD-DEM simulations resolving unsteady gas flow and discrete particle motion with moving boundaries. By decomposing the forces on individual particles, we show that suction is not purely pressure-driven, but results from the combined action of pressure-gradient and unsteady drag forces generated by boundary-accelerated flow. Despite the heterogeneous and transient force fields, the final entrained mass is found to be governed primarily by a single energetic measure: the mechanical work performed on the particle assembly in the entrainment direction during boundary motion, rather than by peak instantaneous forces. Varying boundary kinematics demonstrates that changes in displacement or velocity history control entrainment mainly by modifying the duration over which fluid-particle forces perform work. These results reveal an organizing principle for suction-driven entrainment and establish a work-based framework for boundary-induced particle transport in confined gas-particle systems.

physics.flu-dyn↗

arXiv2025

An advanced heat transfer model for Eulerian-Lagrangian simulations of industrial gas-solid flow systems

The discrete element method (DEM) coupled with computational fluid dynamics (CFD), has been developed to simulate complex solid-fluid flow systems. Today, DEM is regarded as an established approach, with extensive applications in industrial systems. Heat transfer modeling might be essential to the DEM as the industrial applications. However, existing DEM heat transfer models have fundamental limitations. These issues arise from the soft spring model inherent in DEM, where heat conduction is mathematically influenced by the spring constant. Consequently, complex modeling, considering contact state such as contact area and duration, is typically required to estimate heat conduction accurately. Moreover, the current heat transfer models exhibit poor compatibility with scaling laws, such as the coarse-grained DEM, leading to amplified temperature errors relative to motion errors. To address these challenges, we develop a novel heat transfer model based on an Eulerian framework within DEM simulations. In our approach, the Eulerian description is applied to the heat transfer calculation, while particle motion remains treated by the DEM. Notably, the heat conduction in the solid phase is captured through a simple setup by specifying the void fraction, rather than relying on complex contact modeling. The adequacy of the proposed model is demonstrated through validation tests in gas-solid flow systems, showing that the temperature distribution is independent of the particle contact state. Furthermore, the model exhibits strong compatibility with coarse-grained DEM, maintaining accuracy even at reduced computational costs. These results establish the new model's reliability and universality, positioning it as a promising standard for DEM-CFD simulations in industrial applications.

physics.flu-dyn↗