In this core task, the methods and conclusions of the SMA study are investigated to validate newly developed CALCE failure models for PWBs and CCAs. In addition, existing models are enhanced with guidance from the SMA program.
Approach
Long-term reliability versus wearout failure mechanisms is investigated using the highly accelerated SMA method, and simulated with physics-of-failure models. The reduction in reliability due to manufacturing and material defects is investigated through error-seeding techniques. Material failure models are developed and combined with electrical performance models to examine the nominal influence of wearout failures such as fatigue and corrosion on electrical parameters at the bare-board and CCA levels. Relevant models are incorporated into software for easy implementation.This is a multi-unit task: (1) augment the BCAG extended core project for stimulating failures under the SMA program; (2) develop material failure models at the PWB and CCA packaging levels; (3) develop predictive models for relevant electrical failure modes, and (4) incorporate the results in appropriate software for members. During the 1992-93, year several important failure modes have been addressed. The importance of these failure modes was inferred from field data, from feedback from the IAB, and through SMA testing.
Work Accomplished
Task 1: Augmentation of SMA methodology. Extended core funds from BCAG (Task C93- 06) and internal funds have been utilized as seed money to set up experimental facilities for highly accelerated stress applications on test articles. A Qualmark OVS-1 test chamber has been installed and calibrated extensively to establish a procedure for understanding the stresses caused by random 6-axis vibrational loads and highly accelerated thermal loads. Error-seeded surface-mount PWB samples from Honeywell have been tested under combined loads and failures have been precipitated in highly accelerated environments. The PWB dynamic response has been used to understand the stresses acting at the failure sites. Results of extensive metal migration tests, under highly accelerated stress conducted under separate funding are also available in this program to determine appropriate failure models for conductive filament growth.Task 2: Development of Material Failure Models. Two categories of solder-joint stress analysis models are being developed to overcome the well-known limitations of Engelmaier's models in CALCE software. The first category of model is semi-empirical, and is based on extensive finite element modeling and design of experiments. These models have been initiated with the current year's funding and with associated funding from a related project, and will be enhanced in future years. The second kind of solder-joint stress analysis model involves actual mechanics-based analysis. Elastic stress analysis has already been accomplished. Plastic and creep stress analysis models will be developed in future years. Lead fatigue and lead-seal fatigue models have already been developed in previous years. Fully elastic-plastic models for PTH fatigue have also been developed, and were augmented this year to increase their applicability to Aramid reinforced boards and to include crucial manufacturing variabilities, such as plating waviness. Models for corrosion have been obtained from the literature and models for metal migration have been developed empirically in a related project. These failure models and sites were picked based on preliminary surveys of failure data available in the literature. In future years, the new generation of solder joint fatigue models will be completed, incremental enhancements will be made to other existing models, and new failure mechanism models will be developed for aging and degradation at interfaces of dissimilar materials, and for electromigration.
Task 3: Development of Electrical Failure Models. Electrical failure modes due to the material degradation predicted in TASK 2 have been modeled through physics-of-failure techniques. Predictive models for resistance or impedance changes due to corrosion in metal traces have been developed. These models are used in conjunction with lossy coupled transmission-line models in SPICE to predict degradation of electrical performance over time. Thus, reliability of packages can now be predicted based on loss of electrical performance. Similar models for electrical opens due to fatigue-crack propagation in solder joints and PTHs have been developed. Empirical models for electrical shorts due to metal migration are also available from other related projects. In future years, additional models will be included for electrical failure modes due to material failure mechanisms.
Task 4: CCA Level Software. The failure models developed in Tasks 2 and 3 are being incorporated in stand-alone software, to be ultimately integrated into the software. The design inputs and I/O formats are obtained from board layout data in Gerber files and from other databases, such as material properties, geometry (component footprint, lead style, and so on), and electrical functionality. The input and output formats for these files have been determined based on meetings with DoD. This software will be enhanced in future years, with the addition of new failure models and better user interfaces. Eventually, this software will be integrated within the CALCE software.
