PARTICLES 2025

Evaluation of a Gap Model for Improving Gas-Phase Calculations in MPS Simulations of a Liquid Ring Pump

  • Yang, Zhiqiang (The University of Tokyo)
  • Shibata, Kazuya (The University of Tokyo)
  • Takanashi, Takeshi (Tsurumi Manufacturing co., LTD.)
  • Sapkota, Sabin (Tsurumi Manufacturing co., LTD.)
  • Yasuda, Naoyuki (Tsurumi Manufacturing co., LTD.)

Please login to view abstract download link

Liquid ring pumps are widely used in industries that require efficient gas handling, such as the steel and resin molding sectors. The internal flow in these pumps is characterized by complex gas–liquid two-phase behavior, which poses significant challenges for optimizing hydraulic performance [1]. The Moving Particle Semi-Implicit (MPS) method [2], a well-established mesh-free particle method, is particularly well-suited for simulating such flows. In a previous study [3], the liquid ring vacuum pump was simulated using the MPS method, where the inflow and outflow of both liquid and gas phases were modeled. The governing equations adopted for incompressible flow were the continuity equation and the Navier–Stokes equations [4]. The pressure Poisson equation was discretized using a Laplacian model, and the improved pressure calculation method developed by Shibata et al. [5] was employed. In the present study, gas-phase calculations were further improved through the introduction of a gap model, and simulations were conducted under various conditions. The gap model was proposed to address the limitation in the previous two-dimensional simulations, where the axial clearance between the impeller end wall and the pump casing was neglected, potentially leading to gas backflow between the gas chambers and the inlets [1]. Comparisons were conducted among the simulation results obtained with the gap model, those without the gap model, and the experimental results of the liquid ring vacuum pump to evaluate the performance of the proposed gap model. These comparisons focused on the liquid ring shape and the circumferential pressure distribution along the casing.