The effect of underfill material on reliability of flip chip on board (FCOB) assemblies is investigated in this study by using two-dimensional and three-dimensional finite element simulations under thermal cycling stresses from −55°C to 80°C. Accelerated testing of FCOB conducted by the authors reveals that the presence of underfill can increase the fatigue durability of solder interconnects by two orders of magnitude. Similar data has been extensively reported in the literature. It is the intent of this paper to develop a generic and fundamental predictive model that explains this trend. While empirical models have been reported by other investigators based on experimental data, the main drawback is that many of these empirical models are not truly predictive, and can not be applied to different flip chip architectures using different underfills. In the proposed model, the energy-partitioning (EP) damage model is enhanced in order to capture the underlying mechanisms so that a predictive capability can be developed. A two-dimensional finite element model is developed for stress analysis. This model accounts for underfill over regions of solder in an approximate manner by using overlay elements, and is calibrated using a three-dimensional finite element model. The model constant for the enhanced EP model is derived by fitting model predictions (combination of two-dimensional and three-dimensional model results) to experimental results for a given temperature history. The accuracy of the enhanced EP model is then verified for a different loading profile. The modeling not only reveals the influence of underfill material on solder joint durability, but also provides the acceleration factor to assess durability under life cycle environment, from accelerated test results. Experimental results are used to validate the trends predicted by the analytical model. The final goal is to define the optimum design and process parameters of the underfill material in FCOB assemblies in order to extend the fatigue endurance of the solder joints under cyclic thermal loading environments.
Recently, accelerated testing of surface mount interconnects under combined temperature and vibration environments has been recognized to be a necessary activity to ensure enhanced test-time compression. Successful use of vibration stresses requires a clear understanding of the correlation between vibrational damage and thermomechanical damage in surface mount solder joints. Hence, fatigue due to vibrational loading is important and accurate quantitative models are required to model effects due to vibrational fatigue. The proposed analysis in this paper contributes towards development of such quantitative models. This paper presents an approximate method to analyze stresses in surface mount solder joints subjected to vibration loading, using a generalized multidomain Rayleigh-Ritz approach (Ling and Dasgupta, 1995). The advantage of this approach is in its computational efficiency, compared to general-purpose finite element methods. Ling developed this approach in the context of thermomechanical stress analysis of solder joints. In this paper, the technique is modified and adapted for analyzing stresses caused by out-of-plane flexural dynamic modes of the printed wiring boards (PWBs). The analysis uses a two-step procedure where the local PWB curvatures are first estimated and the resulting deformations in the solder interconnect are then determined. The input boundary conditions for the first step are the bending moments in the PWB due to random vibrations. The stiffness of the interconnect assembly is then predicted using an energy method and curved-beam analysis. The bending moment and the computed stiffness of the interconnect assembly are then used to predict the local curvature of the PWB under any given surface-mount component by using an eigenfunction technique developed by Suhir (Suhir, 1988). In the second step of the analysis, the local curvature of the PWB is used as a boundary condition to predict the state of deformations, stresses, and strains in the solder joint using a modified version of the multidomain Rayleigh-Ritz approach. The overall method is applied to a specific example (J-lead solder joint) for illustrative purposes, and compared to finite element predictions for validation.
A generalized multi-domain Rayleigh-Ritz (MDRR) approach developed by Ling and Dasgupta (1995), is extended in this paper, to obtain the stress field in flip chip solder interconnects, under cyclic thermal loading. Elastic, Plastic and time-dependent visco-plastic analysis is demonstrated on flip chip solder interconnects. The method has been applied to other surface-mount interconnects in the past such as J-lead (Ling and Dasgupta, 1996a) and ball-grid joints (Ling and Dasgupta, 1997). The analysis results for the J-lead and ball grid joints have confirmed that the MDRR technique is capable of providing stress-strain hysteresis with adequate accuracy, at a fraction of the modeling effort required for finite element model generation and analyses. Nonlinear viscoplastic stress analysis results for flip chip interconnects without underfill are presented in this paper. The fatigue endurance of the solder joints is assessed by combining results from this stress analysis model with an energy-partitioning damage model (Dasgupta et al., 1992). The life predicted by the analytical damage model is compared with experimental results.
Parametric analytical studies are conducted to investigate whether the reliability of flip-chip solder interconnects are affected by inhomogeneities in the underfill, such as settling of the filler particles. The property gradation caused by filler settling is modeled with a micromechanics formulation. The predicted property gradients are then utilized in a finite element simulation of the flip-chip assembly, to assess the impact on solder interconnect reliability.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.