다중 물리 시뮬레이션 기반 이방성 열-기계적 특성을 고려한 등가 모델링 활용 대면적 반도체 패키지의 워피지 예측

초록

With the rise of new industrial paradigms such as AI and autonomous driving, semiconductors have become essential components across various platforms, from data centers to personal devices. As demands for high performance, high integration and lightweight packaging grow, challenges such as warpage arise due to large-area structures and heterogeneous materials. Although simulation-based warpage prediction methods have been developed, fully reflecting high-resolution circuit patterns in finite element model leads to excessive mesh counts, increasing computational cost and reducing numerical stability. To overcome these issues, in this research, proposal of the modeling approach that specifies Cu pattern orientation and density in each region of high-resolution circuit images. These analyzed results are used to construct an equivalent model with effective anisotropic thermo-mechanical properties. This method enables efficient thermo-mechanical analysis while preserving key circuit characteristics, allowing accurate warpage prediction in semiconductor packages.

키워드