Reliability-Driven Hybrid Modeling for Warpage Prediction and Optimization in FOWLP Decarrier

In Fan-Out Wafer-Level Packaging (FOWLP), the decarrier process is a critical manufacturing stage where severe coefficient of thermal expansion (CTE) mismatches induce significant wafer warpage, posing serious risks to structural integrity and long-term reliability. To address this challenge, this study proposes a reliability-driven hybrid modeling framework that integrates high-fidelity Finite Element Method (FEM) simulations with an Artificial Neural Network (ANN) surrogate model. To overcome the limitations of conventional data-hungry AI approaches, we employed a "Smart Data" strategy: an experimentally anchored FEM (error96% reduction from the baseline. Ultimately, this work provides a fully transparent, scalable Digital Twin methodology that empowers engineers to quantitatively balance thermo-mechanical reliability with mass-production manufacturability.