CALCE EPSC Graduate Student Theses and Dissertation Abstracts (2013)

Song, Bo (Ph.D)
Impact of Dust on the Reliability of Printed Circuit Assemblies

Dust is a ubiquitous component of the environments in which we live and work. It can deposit on printed circuit assembly to act as a source of ionic contamination. Two common consequences of dust contaminations in the printed circuit boards are loss of impedance (i.e., loss of surface insulation resistance) and electrochemical migration between traces and component leads. Both failure mechanisms involve the contamination forming a current leakage path on a printed circuit board. Based on studies on ionic contaminations, researchers have argued that the impact of dust in these two failure mechanisms is dependent on its pH, its hygroscopic compositions, and the critical relative humidity of the salts in it. However, due to the lack of experimental results and the complexity of dust compositions, the argument is not substantiated. Very few papers concerning the impact of different natural dust on these two failure mechanisms can be found in the literature. In practice, mixtures of Arizona dust and salts are used as a substitute for dust in experiments. In this research, natural dusts were collected from four locations: natural outdoor and indoor dust samples from Massachusetts, U.S., natural outdoor dust from Tianjin, China, and the ISO standard test dust (Arizona test dust). Loss of impedance in dust contaminated printed circuit boards was investigated under controlled temperature (20ºC to 60ºC) and relative humidity (50% to 95%) ranges. The impact of dust on electrochemical migration and corrosion was evaluated under temperature-humidity-bias tests (50ºC, 90% RH, and 10 VDC). In addition to the conventional DC measurement where only resistive data can be obtained, electrochemical impedance spectroscopy were adopted to obtain nonlinear equivalent circuit models of the electrochemical process, which helps to understand the underlying physics-of-failure.

The variation of impedance with relative humidity exhibited a transition range. Below the range, the impedance was constant, and above it, the impedance degraded by orders of magnitude. The value of the transition range decreased with an increase of dust deposition density. The equivalent circuit modeling showed that the dominant resistive path gradually shifted from the bulk to the interfacial with the increase of temperature from 20 ºC to 60 ºC. There were big variations among different dusts, which were quantified using the degradation factor introduced in the research, the critical transition range, and time-to-failure. This result demonstrated that a single salt or a mixture of compounds can not be representative of all dusts. It also indicated that using the ISO standard test dust in place of natural dust samples for reliability evaluation could lead to inaccurate results. Dust should be collected from the field in order to evaluate its impact. It is showed in this thesis that some critical characteristics of dust can be used to classify different dusts for the failure mechanisms of interest. Moisture sorption capability of dust can be used to classify different dusts regarding the loss of impedance failure. The dust with the highest moisture sorption capability had the highest degradation factor. Ion species/concentration or conductivity of dust aqueous solution can be used to classify dust regarding the electrochemical migration related failures. Dust with the highest ion concentration and conductivity had the lowest time-to-failure. The underlying principals behind those critical characteristics were described and discussed based on the physics-of-failure.


Chai, Fei (Ph.D)
Solder Interconnect Life Prediction under Complex Temperature Cycling with Varying Mean and Amplitude

Electronic devices are under concurrent loading of the power cycling of the devices and the temperature cycling from the surrounding environment. Temperature histories resultant from these concurrent loading would be a complex temperature cycling with varying cyclic temperature mean and amplitude, as well as spatial thermal gradient. This study developed modeling approaches and quantified accuracies for predicting solder interconnect life under complex temperature cycling. Three modeling approaches were presented in this study: 1) modeling the strain energy under the resultant complex temperature cycling and employing the energy based fatigue life models; 2) segmenting the resultant complex temperature cycle into multiple simple temperature cycles with a single temperature range for each first, then assessing the life expectancy of the solder interconnect under the segmented simple temperature cycles and at last applying Miner's rule to superpose the damage; 3) estimating solder damage under the resultant complex temperature cycling by a standard temperature cycling with a single temperature range. Two case studies were included in this thesis: 1) chamber controlled complex temperature cycling with mini cycles occurring at the upper excursion on ceramic leadless chip carriers assembled by Sn36Pb62Ag2 and SnAg3.0Cu0.5 solder (without spatial thermal gradient); 2) combined temperature and power cycling on plastic ball grid array assembled by Sn63Pb37 and SnAg3.0Cu0.5 solder (with spatial thermal gradient). Physical tests were also conducted to quantify the developed modeling approaches.


Xiaofei, He (Ph.D)
Evaluation and Modeling of Electrochemical Migration on Printed Circuit Boards

Electrochemical migration (ECM) is the growth of conductive metal filaments on a printed circuit board (PCB) through an electrolyte solution under a direct-current (DC) voltage bias. ECM can cause a reduction in surface insulation resistance (SIR) between adjacent conductors and lead to intermittent or catastrophic circuit failures.

To evaluate current leakage and electrochemical migration behavior on printed circuit boards, IPC B-24 comb structures were exposed to 65.C and 88% relative humidity conditions under DC bias for over 1000 hours to determine the effects of solder alloy (eutectic tin-lead and lead-free), board finish (organic solderability preservative versus lead-free hot air solder leveling), spacing (25 mil versus 12.5 mil), voltage (40 V versus 5 V bias), solder mask (using and not using), and flux solid content on ECM. In-situ measurements of SIR, energy-dispersive spectroscopy after testing, and optical inspection before and after test were used. The relationships between the electrical and electrochemical behaviors of solder alloys were established. The long term electrochemical behaviors of tin-lead and lead-free solders were obtained. The morphology and distribution of migrated species, including Sn, Pb, Cu, and Ag were investigated. Compared with solder alloy, board finishes played a secondary role in affecting SIR due to their complexation with or dissolution into the solder. The competing effect between electric field and spacing was also investigated. Solder mask was found to stabilize SIR and reduce the chances for ECM to occur due to its .walling effect.. Compared to low solids flux, medium solids flux increased the characteristic lives of PCBs due to their encapsulation effects.

The prolonged SIR decline of Sn-3.0Ag-0.5Cu soldered boards was simulated by three-dimensional progressive and instantaneous nucleation models, whose predictions were compared with experimental data. The kinetics of the electrochemical migration process between copper traces in deionized water were investigated using electrochemical impedance spectroscopy and cyclic voltammetry. The rate limiting step was identified before and during dendritic growth. The time to generate an embryonic dendrite was measured experimentally on copper traces and modeled using Nernst Planck equations, which matched the experimental results well.


Sutrisno, Edwin (M.S)
Fault Detection and Prognostics of Insulated Gate Bipolar Transistor (IGBT) Using a K-Nearest Neighbor Classification Algorithm

Over the past 20 years, power electronic systems have been increasingly required to operate in harsh environments including automotive, deep-well drilling and aerospace applications. In parallel, the higher power densities and miniaturization of the power switching module result in elevated stress levels on the control circuitry. The objective of this study was to develop methods and models for assessing the interconnect reliability of components used in the control circuitry for power electronic systems. Physics-of-Failure modeling and a series of thermal and reliability simulations were conducted on a 2.2 kW variable-frequency drive to evaluate the susceptibility of system level and component level failure mechanisms. Assessment methods consisted of developing CalcePWA simulation models of the primary sub-assemblies and constructing a power cycling apparatus to perform accelerated testing of the drive.


Lillie, Edwin (M.S)
Assessing the Cost of Risk for New Technology and Process Insertion

Adoption and insertion of new technologies and processes into systems is inherently risky, especially for critical systems. The projected cost of failure consequences (PCFC) is defined as the cost of all failure events (of varying severity) that are expected to occur over the service life of the system. This paper uses cost-based FMEA to determine the PCFC for a technology insertion control plan for the adoption of lead-free solder for the assembly of electronic systems that previously used tin-lead solder. The study demonstrates an assessment of the cost-effectiveness of the control plan for the same product in various risk scenarios.


Boettcher, Robert (M.S)
Assessing the Cost of Risk for New Technology and Process Insertion

Make operation studies have been conducted on Ag/SnO2 contact materials under high power DC conditions. The circuit allows the relay contacts to be tested at different power and energy levels in order to determine the relationship to cycles to failure. Relays from various manufacturers were subjected to capacitor discharge pulses of 250 V at 10-80 .F to show vendor dependent differences in relay reliability. Failure analysis was then conducted on the welded contacts using scanning electron microscopy (SEM) and wavelength-dispersive spectroscopy (WDS) in order to address material properties and design variations that affect the welding susceptibility of relays. The incidence and extent of degradation is correlated to material characteristics including contact material, oxide content, hardness, contact geometry, and surface roughness using a physics of failure approach. The relays with a higher percent content of indium oxide exhibited a greater reliability than those without. Both power and energy were then varied to further investigate the one cycle to failure boundary region. Results in the form of a cycle to failure map are presented.


Fritzler, Thomas (M.S)
Scintillation Conditioning of Tantalum Capacitors with Manganese Dioxide Cathodes

Scintillation testing is a method that activates the self-healing mechanism in tantalum capacitors. In preliminary experiments, the deliberate activation of self-healing yielded up to 27% higher breakdown voltages in weak parts that had an increased risk of ignition failure. This improvement results in a better performance under surge current conditions. This paper demonstrates that scintillation conditioning reduces surge current failures in tantalum capacitors with manganese dioxide cathodes. Tantalum capacitors with MnO2 cathodes from two manufacturers are subjected to scintillation conditioning and compared to non-conditioned populations in a surge current test. To ensure that the activation of the self-healing mechanism has no detrimental effect on the reliability of the parts, a life test is conducted. The results show that the conditioning method increases the breakdown voltage of self-healed tantalum capacitors by up to 25% under surge current conditions, which mitigates the risk of ignition failures. No detrimental effect on the life of the conditioned samples was observed. Additional tests to assess the reliability of tantalum capacitors with manganese dioxide cathodes under simultaneous thermo-mechanical and voltage stresses were performed. Even though these tests are not directly related to scintillation conditioning the study was included as an additional chapter, since it pertains to the general subject of tantalum capacitor reliability.


Bakhshi, Roozbeh (M.S)
Reliability Assessment of Voided Microvias in High Density Interconnect Printed Circuit Boards under Thermo-Mechanical Stresses

Printed circuit boards (PCBs) are made of several dielectric layers stacked on top of each other. These layers could be standard PCB core board or high density interconnect (HDI) layers. Microvias allow signal and power transmission between the HDI layers of the PCBs. The presence of voiding in filled microvias has raised concerns in industry about how they affect the degradation of microvias during the life cycle of the product. Voids can vary widely in shape and size and have been observed in both stacked and single-level microvias. This paper examines whether the presence of voids alone is responsible for the degradation of microvias, or if parameters such as void size and void shape have an influence as well. Both voided and nonvoided microvias were tested together using liquid-to-liquid thermal shock as the accelerated testing method to understand their different behaviors under thermomechanical stresses.


Kunche, Surya Tej (M.S)
Classifier Fusion Technique for Fault Diagnostics

Classification algorithms have been widely used to solve data-driven fault diagnostics problems. The number of classification algorithms used has been increasing in recent years. Each classification algorithm has its own strengths and weaknesses, and the accuracy of classifiers changes with the different features used for training. As a result, traditional methods of selecting an appropriate classification algorithm, including domain expertise and trial and error, are becoming complex and difficult to employ. Classifier fusion has been used to solve this problem of selecting an appropriate diagnostic algorithm, and it also improves the generalizability of an algorithm. The performance of a classifier fusion algorithm is governed by the combination rule adopted for fusing multiple classifiers and how the bias and variance are balanced by the combination rule. However, research still needs to determine which combination rule optimally balances the bias and variance during classifier fusion. Therefore, this research develops a fusion methodology that combines the classifiers by balancing the bias and variance. This methodology reduces the number of false negatives and positives, thereby improving the overall accuracy of the algorithm for fault detection. A cost function that considers bias and variance errors was developed to evaluate the performance of the algorithm. Sequential quadratic programming-based optimization was employed to find the optimal combination of classifiers and balance of bias and variance. The developed algorithm was used for fault diagnosis of analog circuits, and the results indicate that the developed fusion approach improved diagnostic accuracy over existing classifier fusion techniques.


Squiller, David (M.S)
Methods and Models for Assessing Solder Interconnect Reliability of Control Boards in Power Electronic Systems

Over the past 20 years, power electronic systems have been increasingly required to operate in harsh environments including automotive, deep-well drilling and aerospace applications. In parallel, the higher power densities and miniaturization of the power switching module result in elevated stress levels on the control circuitry. The objective of this study was to develop methods and models for assessing the interconnect reliability of components used in the control circuitry for power electronic systems. Physics-of-Failure modeling and a series of thermal and reliability simulations were conducted on a 2.2 kW variable-frequency drive to evaluate the susceptibility of system level and component level failure mechanisms. Assessment methods consisted of developing CalcePWA simulation models of the primary sub-assemblies and constructing a power cycling apparatus to perform accelerated testing of the drive.


Ernst, Matthew Ross (M.S)

Response and Durability of Large Radius of Gyration Structures Subjected to Biaxial Vibration

Multiaxial vibration tests were conducted using an electrodynamic shaker capable of controlled vibration in six degrees of freedom. The test specimen consisted of six large inductors insertion mounted on a printed wiring board. Average damage accumulation rate was measured for random excitation in-plane, out-of-plane, and both directions simultaneously. Under simultaneous biaxial excitation, the damage rate was found to be 2.2 times larger than the sum of the in-plane and out-of-plane rates. The conclusion was that multiple-step single-degree-of-freedom testing can significantly overestimate the durability of some structures in a multiaxial environment. To examine the mechanics behind this phenomenon, the response of a simple rod structure was analyzed with the finite element method. Axial vibrations, which produce negligible stress on their own, were found to contribute significant additional stress when combined with transverse vibration. This additional stress contribution was found to be highly dependent on the frequency ratio and phase relationship between the two participating axes.



Copyright © 2013 by CALCE and the University of Maryland, All Rights Reserved