Hybrid Inertial–Centrifugal Gas–Liquid Separator: CFD‐Driven Optimization Using Coupled Discrete Phase and Wall‐Film Modeling With Response Surface Methodology

Abstract

Efficient removal of liquid aerosols from compressed gas streams is essential in chemical and energy processes. Conventional separators relying on a single mechanism (inertial or centrifugal) struggle to capture both fine and coarse droplets simultaneously. This work presents a compact gas–liquid separator that synergistically combines inertial impaction with stationary‐blade‐induced centrifugal guidance in a uniaxial configuration. A high‐fidelity computational framework coupling the discrete phase model (DPM) with an Eulerian wall‐film model is developed in ANSYS Fluent to simulate droplet dynamics, breakup, and liquid film formation. Thirteen geometric parameters are screened using a Taguchi L27 orthogonal array, followed by response surface methodology (RSM) optimization focused on the two most influential parameters: number of stationary blades (x 7 ) and blade‐fixture‐to‐outlet distance (x 13 ). The optimized design achieves 98.86% separation efficiency at a pressure drop of 5005 Pa—substantially outperforming conventional cyclones (85%–92% at 3–6 kPa). Grade efficiency exceeds 99% for droplets > 5 μm and reaches 95% for droplets as small as 2 μm. Off‐design simulations confirm robust performance across 40%–160% flow rate and 0.3–0.6 MPa inlet pressure. The framework is validated by grid independence (GCI < 0.12%), model robustness, and RSM prediction errors below 1.25%. This work provides both a high‐performance separator and a generalizable optimization methodology for multiphase separation processes.