Board-level Ball Grid Array Research Articles

A Time Frequency Domain Based Approach for Ball Grid Array Solder Joint Fatigue Analysis Using Global Local Modeling Technique

AbstractThe fatigue life prediction of the electronic packages under dynamic loading conditions is an increasingly important area of research, with direct application in packaging industries. Current life prediction methodologies are, in general, developed through a finite element (FE) model that is correlated using an experimental data measured through sweep sine testing. The frequency response function (FRF) generated by using a sweep sine testing may suffer from leakage and windowing of the signal may not work correctly, which results in the shift in the amplitude and the resonance frequencies of the package. In consequence, there will be a significant deviation between the actual and the predicted natural frequencies and the amplitude of vibration response in the given excitation range, resulting in the longer time to fail the package during the laboratory based/virtual durability testing. Thus, it is necessary to develop a suitable validation technique in time/frequency domain to address this issue. In this paper, the step sine testing procedure is utilized to validate the FE model of a test vehicle consisting of a board level ball grid array chip package and the resonance-based fatigue testing is performed in the FE-based simulation. The global–local modeling approach is utilized to model the test vehicle and the volume average von Mises stress is used to predict the life of the solder joint. Following the numerical simulations, fatigue test is carried out in the test vehicle at the first resonance frequency obtained from the step sine test. Experimental results show that there are full openings of the corner balls in a very short interval of time. The results of the life prediction from the FE model and from experiments are comparable to each other thus validating the proposed methodology.

Machine learning for board-level drop response of BGA packaging structure

Board-level drop responses are critical to evaluate the mechanical reliability of solder joints to serve as electrical and mechanical connections in electronic devices to resist failure due to drop impact. A machine learning (ML) model based on the back propagation (BP) method is proposed for predicting three-dimensional board-level drop responses of the ball grid array (BGA) packaging structures. Compared with conventional finite element (FE) simulations with solid elements, the proposed approach with a deep neural network is more than three orders of magnitude faster in terms of computational efficiency. More importantly, detailed stress and strain of all solder joints and also the warpage of the printed circuit board (PCB) can be accurately predicted throughout the drop process. According to the contact type between BGA packaging structure and rigid ground, the drop conditions are divided into three types: surface contact, line contact and point contact. After training the ML model, the obtained nonlinear mapping relation can well predict the mechanical response of solder joints in BGA packaging structures. To significantly extend the application scope, the trained ML model could reasonably predict the dynamic responses of BGA structures with the drop angles which are beyond the training samples. Compared with FE results, the prediction accuracy of the proposed ML model is objectively measured by the Pearson correlation coefficient which is found to be above 0.95 after estimating the stress, strain and energy density of solder joints and PCB warpage. Except for those stress, train and energy density values close to 0, the prediction errors are mostly less than 10% compared with the finite element simulation values. Therefore, it is affirmed that the computational efficiency and accuracy of the proposed ML method are satisfactory to replace the traditional time-consuming FE modeling for predicting the dynamic responses of board-level BGA packaging structures even under more extreme and complicated loading conditions.

The Melting Characteristics and Interfacial Reactions of Sn-ball/Sn-3.0Ag-0.5Cu-paste/Cu Joints During Reflow Soldering

In this work, the melting characteristics and interfacial reactions of Sn-ball/Sn-3.0Ag-0.5Cu-paste/Cu (Sn/SAC305-paste/Cu) structure joints were studied using differential scanning calorimetry, in order to gain a deeper and broader understanding of the interfacial behavior and metallurgical combination among the substrate (under-bump metallization), solder ball and solder paste in a board-level ball grid array (BGA) assembly process, which is often seen as a mixed assembly using solder balls and solder pastes. Results show that at the SAC305 melting temperature of 217°C, neither the SAC305-paste nor the Sn-ball coalesce, while an interfacial reaction occurs between the SAC305-paste and Cu. A slight increase in reflow temperature (from 217°C to 218°C) results in the coalescence of the SAC305-paste with the Sn-ball. The Sn-ball exhibits premelting behavior at reflow temperatures below its melting temperature, and the premelting direction is from the bottom to the top of the Sn-ball. Remarkably, at 227°C, which is nearly 5°C lower than the melting point of pure Sn, the Sn-ball melts completely, resulting from two eutectic reactions, i.e., the reaction between Sn and Cu and that between Sn and Ag. Furthermore, a large amount of bulk Cu6Sn5 phase forms in the solder due to the quick dissolution of Cu substrate when the reflow temperature is increased to 245°C. In addition, the growth of the interfacial Cu6Sn5 layer at the SAC305-paste/Cu interface is controlled mainly by grain boundary diffusion, while the growth of the interfacial Cu3Sn layer is controlled mainly by bulk diffusion.

Simulation Investigations on Fluid/Structure Interaction in the Reflow Soldering Process of Board-Level BGA Packaging

The objective of the present study is to develop a Fluid/Structure Interaction model of a board-level Ball Grid Array (BGA) assembly for an infrared-convection reflow oven. The infrared-convection reflow oven is modeled in Computational Fluid Dynamic (CFD) software while the structural heating BGA package simulation is done using Finite Element Method (FEM) software. Both software applications are coupled bidirectional using the Multi-physics Code Coupling Interface (MpCCI). The simulation thermal profile is compared with the experiment thermal profile, and they were found to be in good conformity. The simulated flow fields show that the convection mode in an infrared-convection reflow oven played minor effect on heat transfer to the printed circuit board (PCB). The dominant heat transfer mode in an infrared-convection reflow oven is the radiation mode from a quartz heating tube. From the simulation results, the PCB near the edges or corners tended to heat up first at preheating, soaking and reflow stages. The PCB and component experience larges temperature difference in preheating stage. This situation runs the risk of an excessive board warpage. In addition, the maximum von-Mises stress is trapped in the interfaces between solder joint and die, which intend to form the nucleation of initial solder joint crack. This guideline is very useful for the accurate control of temperature and thermal stress distributions within components and PCB, which is one of major requirements to achieve high reliability of electronic assemblies.