Solder Joint Reliability – Prelude
DR. JENNIE S. HWANG, CEO, H-TECHNOLOGIES GROUP
Needs in advanced electronic hardware to enable today’s and emerging products (Smart devices, Internet of Things, Intelligent Things, autonomous vehicles, and augmented reality, 5G and AI gear) are mind-boggling. Embracing the fact that the transistor density of integrated circuits (IC) doubles every 18-months and arguably the IC density even surpasses the Moore’s law, the industry’s focus has been to meet the on-going challenges. Phenomenal semiconductors, IC packaging and circuit board-level manufacturing using surface mount technology (SMT) are facing ever-more-relentless thrusts and ever-stronger market pull. New IC package designs to connect the innovative “semiconductors” to the outside world are nothing short of astonishing.
Performing as the critical interconnections under increasingly demanding service conditions, particularly in the package- and board-level, solder alloys have been the de facto material to serve as physical, electrical and thermal interconnections. With the goal to produce reliable products while achieving high yield and low-cost manufacturing, the packaging-level and board-level manufacturing has to respond by delivering enhanced materials performance and changing production know-how to ensure the target performance and reliability of interconnections between semiconductor, package and circuit board, thus the solder joint reliability. Accordingly, these on-going changes and challenges prompted the thought of the solder joint reliability series.
Embracing the fact that the transistor density of integrated circuits (IC) doubles every 18-months, the industry’s focus has been to meet the on-going challenges.
First, I want to define the premise of the terminologies or just plain words used in discussing reliability. There is a distinction between reliability and test results. When a set of tests is conducted, raw data are generated. Data as a collection of numbers, images, graphs or other outputs produced from equipment or devices are viewed as the lowest level of information, from which knowledge is then derived or deduced. After the test results are analyzed and interpreted, the characteristics and behavior of a system become indicative. At this time, its performance under this given set of test conditions is concluded. Its reliability, however, is “tricky” and yet to be assessed.
Reliability is a relative term. The system’s operating conditions must be specified, and its functional period of time be also specified. In engineering, reliability is the probability that a system will perform a required function without failure under an anticipated set of conditions for an intended period of time. As examples, the reliability for a telephone switching equipment targets at not having more than X hours (e.g. 2 hours) of downtime in Y years (e.g. 20 – 30 years); for submarine light guide system, reliability requires less than X failures (3 failures) that need ship repair in Y years (e.g. 25 years). For laptop computers or smartphones, what the mean time between failures should be (e.g. 9000 hours or 27,000 hours)?
Weibull distribution, a continuous probability distribution invented by Waloddi Weibull in 1951, has been widely used in reliability engineering and failure analyses. As the Weibull plot is one convenient method of calculating the parameters of the Weibull distribution, almost all publications related to solder joint reliability include the Weibull plot, which plots empirical cumulative distribution of data: time-to-failure. In solder joint reliability studies, accelerated temperature cycling (ATC) is the most used test and a Weibull plot of failure rate vs. the number of cycles has been adopted as a de facto form of data.
One of the most informative parameters of Weibull plot is the Shape Parameter (β), which is the slope of the Weibull plot. Generally, the Shape Parameter relates to actual failure in three stages: early-life failure, useful life and wear-out failure, equivalent to the bathtub curve in reliability engineering. When β < 1, a failure rate decreases with time, also known as infantile or early-life failures. When β is equal to 1, a failure rate is nearly constant, indicative of useful life or random failures. When β > 1 a failure rate increases with time, also known as wear-out failures. Another informative parameter is the Characteristic Life, the time at which 63.2% of the units fail. However, for a target life expectancy, the longer Characteristic Life (e.g. 5,000 cycles) may not be more reliable than a lower Characteristic Life (e.g. 3,000 cycles). Rather, the lower failure rate at the target life expectancy (e.g. 2,500 cycles) tells the story of reliability, meaning more reliable.
It should be noted that although a Weibull plot is able to conveniently compare the relative time-to-failure for various systems, the comparison per se may not necessarily result in a conclusion in reliability. Equally important to note is that, for an unknow system, the Weibull plot of test results from one single set of the ATC test is far from being adequate to draw a reliability conclusion.
Anther practical premise to be refreshed is the intrinsic material properties. A solder joint cannot perform better than the intrinsic materials properties of a specific solder composition can offer, such as fatigue resistance, creep resistance. A material can only measure up to its intrinsic properties, and the intrinsic properties in solder materials are pre-determined by its composition, metallurgical phases, microstructure and the microstructural stability in response to service conditions.
To this end, this column will be organized as a series to address the important and relevant aspects related to solder joint reliability, including scientific fundamentals, state-of-the-art technologies, the challenges and prevalent issues. Our effort is to provide useful information and our hope is to spark interesting thoughts and intriguing ideas to ensure the delivery of reliable products and the sustainment of industry prosperity. Theme topics selected are listed below, and each of the topics is to be discussed in consecutive publication issues as a series. To the esteemed audience, your input in topics under this subject that are not listed below and you would like to be included in this column is very much welcomed.
Our planned theme topics of this series include the following:
Solder Joint Reliability – Prelude
Solder Joint Reliability – The Role in the Reliability of Electronic Products
Solder Joint in Electronic Assemblies – Likely Failure Modes and Fundamentals
Solder Joint Failure Mode – The Role of Fatigue
Solder Joint Failure Mode – The Role of Creep
Solder Joint Failure Mode – The Role of Intermetallics
Solder Joint Reliability – Microstructure
Solder Joint Reliability – Predictive Modeling
Solder Joint Materials – Strengthening Mechanisms
BTC, BGA Solder Joint Reliability
Solder Joint Reliability – in Harsh Environments
Solder Joint Reliability – Alloy Material System Ranking
Solder Joint Reliability vs. Cost of Ownership
Solder Joint Reliability – Testing and Assessment
Solder Joint Ultimate Reliability
Dr. Hwang will present a tutorial on “Preventing Manufacturing Defects and Product Failures” on June 6, 2018 at SMT Hybrid & Packaging Conference, Nuremburg, Germany.
Dr. Hwang, an international businesswoman, international speaker, and business and technology advisor, is a pioneer and long-standing contributor to SMT manufacturing since its inception as well as to the lead-free electronics implementation. Among her many awards and honors, she is inducted to the International Hall of Fame-Women in Technology, elected to the National Academy of Engineering, named an R&D-Stars-to-Watch and YWCA Achievement Award. Having held senior executive positions with Lockheed Martin Corp., Sherwin Williams Co., SCM Corp, IEM Corp., she is currently CEO of H-Technologies Group providing business, technology and manufacturing solutions. She serves as Chairman of Assessment Board of DoD Army Research Laboratory, Commerce Department’s Export Council, National Materials and Manufacturing Board, various national panels/committees, international leadership positions, and the board of Fortune-500 NYSE companies and civic and university boards. She is the author of 500+ publications and several books, and a speaker and author on trade, business, education, and social issues. Her formal education includes four academic degrees (Ph.D., M.S., M.A., B.S.) as well as Harvard Business School Executive Program and Columbia University Corporate Governance Program.