Understanding the effect of corona discharge on ATH/LMGP filled hybrid silicone rubber composites adopting FEM and ML approach

Abstract Prolonged exposure of insulation structure to corona discharge leads to gradual deterioration of dielectric materials, resulting in the loss of hydrophobic properties and surface degradation. This study investigates the effects of highly hydrophobic fillers, including low‐melting glass powder (μ‐LMGP) and nano‐alumina trihydrate (n‐ATH), in a silicone rubber (SiR) matrix under corona exposure. The incorporation of ATH/LMGP fillers within the silicone rubber matrix significantly enhances the dielectric constant and hydrophobic performance while maintaining low tan δ values. Additionally, finite element method (FEM) simulations were conducted to examine the variations in the electric field concerning surface characteristics, hydrophobicity, and dielectric parameters under different levels of corona exposure. Furthermore, a quadratic support vector machine (Q‐SVM) regression model demonstrated superior predictive performance, achieving an R² value of 0.96 for accurate electric field estimation. Using electric field values at the triple junction, a predictive model was developed to estimate corona aging levels and the water droplet initiation corona inception voltage (CIV). This study provides a novel approach for evaluating corona aging levels and predicting water droplet‐initiated CIV values based on electric field measurements, offering valuable insights into the long‐term performance of silicone rubber‐based insulating materials in high‐voltage applications. Highlights SiR added with ATH/LMGP fillers improves resistance against corona discharge. ATH/LMGP fillers enhance the relative permittivity with lower tan δ values. Corona aged samples have been modeled with surface and dielectric parameters. Corona aging level and CIV values could be predicted from electric field values. The SVM based regression model is suitable for accurate electric field estimation.