Authored By:
Caiying He, Ph.D., Zuyao Liu, Hongxia Wang, Lei Wang, Feng Lu, Ph.D., Hongfeng Ran
Kaifa Technology
Shenzhen, Guangdong
P. R. China
Summary
In SMT assembly process, the thermal warpage becomes one of the root causes of solder joint reliability failures for PBGA. The accurate prediction for warpage of PBGA and PCB is an important and challenging task of numerical simulation of PBGA solder joint defect. In this paper, the numerical analysis of warpage in PBGA and PCBA is carried out in consideration of the residual stresses produced during SMT reflow process. The analysis methodology using a viscoelastic based material model is adopted to account the time and temperature dependent behavior of an epoxy molding compound. In order to get a series of material constant for Generalized Maxwell Viscoelastic Model, DMA (Dynamic Mechanical Analyzer) tests for mold compound material sample is performed to generate viscoelastic material relaxation function through curve fitting.
The experimental research on thermo warpage at different temperature for both PBGA and assembled PCBA is carried out by using the Shadow Morie method. The experimental data is compared with the simulation results as verification for FEA modeling. It is found that the predictions are in agreement with experimental data, especially in qualitative analysis. Optimization analysis is also performed to study the effect of PBGA material properties, molding curing temperature and SMT process parameters on warpage and solder joint failures.
Conclusions
The accurate prediction for warpage and residue stress of PBGA needs a huge amount of testing, analysis modeling, and fine-tuning. Material properties parameters selection and determination are the most important design step. Specifically, the accurate viscoelastic material behavior needs to be acquired through DMA experiment, which requires certain expertise in polymer material and the complicated curve-fitting techniques. Necessary measurements on either warpage or stress are much needed to validate the effectiveness of the FEA model.
Viscoelastic calculation considering thermal conduction using the 1/2 solid element model is the most appropriate method for predicting both the warpage and the stress of PBGA during SMT reflow process. However, due to large computing time and convergence concerns, a strip model can also be used to predict accurate warpage profile. The stress prediction using strip viscoelastic model in this paper still needs to be validated through consideration of thermal conduction in future work.
The warpage and solder joint defects optimization results show that the CTE mismatch between the components causes warpage and internal stress in the package. Low CTE of key components can significantly reduce the residue stress around the die area, which can effectively eliminate cracks and delaminations between silicon die and adjacent components. Hence, it is meaningful to optimize the coincident CTEs and Moduli through materials' selection for a good quality package design.
Initially Published in the SMTA Proceedings
|
