Influence of PCB vibration modal on peeling stress of solder joints in board-level package under drop impact

Peeling stress of solder joints is critical to assess the reliability of packages under drop/impact loadings. Numerical simulation is one important tool to tackle this issue. Usually an implicit dynamic transient finite element (FE) is used to conduct the drop/impact analysis of PCB board level packages under a specific drop condition. However, even using the most update computer this kind of simulation is time consuming. Therefore some researchers and engineers proposed an alternative static analysis as first order estimation of peeling stress. The alternative static models can be classified into two types. One is based on PCB bending moment equivalent, i.e, let the bending moment of PCB in the static model equal to the moment in a prior dynamic analysis, and then evaluate peeling stress by the static model. Another is based on PCB deflection equivalent in which the PCB deflection in the static model identifies to the greatest deflection of PCB in dynamic analysis. From the view of that the peeling stress is caused by the bending stiffness difference between the package and the PCB the package attached, it seems that the deflection equivalent scheme is more reasonable. However, there is no research reported in which a through investigation was conducted to verify the effectiveness of the static model. In this paper, a very detailed finite element model of PCB board and BGA package was constructed. A dynamic drop/impact simulation of the board-level package was conducted by LS-DYNA and Input-G approach under the drop Condition H recommended by JEDEC standard JESD22-B 111. A PCB deflection equivalent static analysis was conducted also in order to investigate the difference of peeling stress between the dynamic and the static model. The results show that although the maximum peeling stress occurs at the moment the PCB reaches its maximum deflection, in the deflection-equivalent static model the peeling stress is unexpectedly much greater than that in the dynamic model. And more, a sudden inverse stress change is observed in the dynamic model, which cannot be explained by the PCB deflection model. To explain the difference in peeling stress between the dynamic and the static model, a spectrum analysis to the dynamic stress and a modal analysis of the PCB-package system were carried out. The results indicate that not only the 1st modal of the PCB-package system, which causes the PCB deflection in the length direction, but also the . 5th modal, which causes the PCB deflection in the width direction, play dominant role in the evolution of the peeling stress in the solder joints. The Ist modal causes peeling stress by the PCB deflection at the same time the 5th modal produces inverse direction peeling stress in the solder joints. As a result, the combined effect of the 1st modal and the 5th modal leads to dynamic peeling stress much less than in the static modal, in which only the length direction deflection that analogizes to the 1st modal is considered. Conclusions: 1) the dynamic deflection of PCB-package system under drop/impact is governed by the 1st and 5th vibration modal; 2). the current equivalent static model always overstates the peeling stress comparing with a dynamic model because in the static model the 5th modal deflection of the PCB is not taken into account. 3) the drop/impact reliability design of the package based on a static peeling stress analysis might yield a conservative or oversafe design.