MULTIPHYSICS OPTIMIZATION OF AN ELECTRO-DYNAMIC FILTER FOR GAS TURBINE

Fouling, corrosion, and erosion are the significant particle ingestion degradation phenomena that can occur in a gas turbine. The solid contaminants carried by the airflow enter the gas turbine, and depending on their chemical properties and particle size, they can adhere to the internal surfaces (fouling), remove some material (erosion), or react with the internal part (corrosion). For this reason, multi-stage filtration systems are commonly employed to preserve the life and the efficiency of the entire system. Mechanical filters are used to separate small-diameter particles by trapping the contaminant in cartridges of porous material. These devices are characterized by high filtration efficiency, but also high pressure drop, which increases with the exposure time. Inertial filters are used to separate large-diameter particles. However, due to their lower collection efficiency and pressure drop than porous filters, they are usually used in the prefiltration stages. In recent years, electrostatic fields induced on dust-laden flows proved effective in trapping conductive particles and showing good potential for industrial applications. This work shows a geometrical multi-physics optimization of a combined electrostatic-inertial filtration system. The combination of an electrostatic filter with an inertial one aims to take advantage of both the filtration mechanisms, i.e., capture efficiency comparable to a high-filtration system and a pressure drop comparable to an inertial one. A numerical campaign of the innovative type of filter is performed using an in-house OpenFOAM lagrangian solver that can simulate dispersed flows in the presence of an electrostatic field. Regarding geometrical optimization, a simplicial homology optimizer is used to find the geometry characterized by the highest capture efficiency and the lowest possible pressure drop. Copyright © 2025 by Baker Hughes.