Event
CALCE Webinar- Quantitative Prediction of Warpage after Molding Processes: Is It a Myth?
Thursday, June 5, 2025
11:00 a.m.-12:00 p.m.
https://web.calce.umd.edu/seminars/cws20250605.htm
Molding processes produce residual stresses in encapsulated components. They are combined with the stresses caused by the coefficient of thermal expansion (CTE) mismatch to dictate the final warpage at room and reflow temperatures, which must be controlled for fabrication of redistribution layers (RDL) as well as yield during assembly.
With advanced computational mechanics techniques and computing hardware, one can model/simulate any kind of semiconductor packages for complex loading and boundary conditions. Then, why is it still difficult to make a quantitative prediction of warpage after molding processes?
The answer is rather simple. Some critical properties are missing during prediction. To compensate for the missing properties, commercially available warpage measurement tools have been employed, through which the simplified material properties are adjusted until numerical predictions have reasonably good agreement with measured warpage. Unfortunately, the adjusted models often do not produce accurate warpage predictions for different package architectures using the same epoxy molding compounds (EMCs) - "architecture-dependent" modeling, which is defined as "a model that works for a specific packaging design but does not work for different designs using the same material sets". If accurate material properties are employed, however, the prediction accuracy should remain the same regardless of the package architectures - “authentic material-based” modeling.
This seminar discusses two critical properties missing in the current practice of warpage prediction: (1) effective cure shrinkage and (2) temperature-dependent viscoelastic bulk modulus. Advanced measurement techniques to measure the two properties and related modeling approaches are presented with measurement results obtained from advanced epoxy molding compounds (EMCs).
About the Presenter
Dr. Bongtae Han is currently a Keystone Professor, the Mechanical Engineering Department of the University of Maryland. Dr. Han started his career at IBM Microelectronics in 1991, where he was involved in the first development of flip-chip, build-up, and ball grid array technologies. He returned to academia in 1996. Since then, his research interest has been the design and analysis of semiconductor and photonics systems for optimum reliability – "design for reliability". The current LOMSS projects address: (1) mechanical design and reliability assessment of microelectronics devices; (2) advanced characterizations of semiconductor packaging materials; (3) accelerated test and life prediction of packages and package assemblies; (4) adhesion analysis of critical interfaces; (5) experimental characterization of package stresses and warpages; and (6) moisture ingress, behavior and prediction inside semiconductor packaging.
Dr. Han co-authored a textbook entitled "High Sensitivity Moiré: Experimental Analysis for Mechanics and Materials" (Springer-Verlag, 1994). He edited two books and has published 13 book chapters and over 300 journal and conference papers in the fields of microelectronics and experimental mechanics (https://scholar.google.com/citations?hl=en&user=-giPkhcAAAAJ). He holds 2 US patents and 4 invention disclosures.
He served as an Associate Technical Editor for Experimental Mechanics, from 1999 to 2001; for Journal of Electronic Packaging, Transaction of the ASME from 2003 to 2012; for Microelectronics Reliability from 2017 to 2020. He is currently serving as Editor-in-Chief for Microelectronics Reliability.
He was elected a Fellow of the SEM and the ASME in 2006 and 2007, respectively.
