Predictive Reliability Model Research Articles

Review on accelerated life testing plan to develop predictive reliability models for electronic components based on design‐of‐experiments

AbstractAccelerated testing has been commonly used for the assessment of reliability of products or systems. In particular, it has been used to produce models for predicting the reliability of electronic components as a function of design and environmental parameters, or to qualify reliability. Extensive literature exists on different aspects ranging from defining type of stresses and type of censoring data, to optimizing test plans for efficient and relevant testing. On the other hand, design of experiments methodology is commonly used for studying the robustness of systems and for quality applications. This being said, combining both approaches, taking into account the system's physics of failure, is scarcely put into practice in a context of reliability prediction. Yet, this could significantly improve reliability prediction, especially in the case of electronic components which constantly go through technological progress with new parameters or properties to consider. After first presenting existing predictive reliability guides, models and parameters related to accelerated life testing, the purpose of this article is to provide a review of what has been done concerning the combination of such approaches.

A Predictive Reliability Model to Assess the Performance of Photovoltaic Systems

Clean energy is extremely important not only because of economic purposes but also for health considerations. The use of photovoltaic (PV) systems is growing, with the increased needs for electricity. This requires more attention to research of PV systems. In this study, a method to predict the expected lifetime based on the reliability of system performance is proposed. Geographical data were collected near two different locations: Cairo, Egypt and Riyadh, Saudi Arabia. The PV system was simulated with inputs from collected data to obtain the device factors and system responses. To study the significance of inputs and device parameters on the system responses, the Taguchi OA method was used. The probability density function (pdf) of the time of acceptable performance was estimated from the simulation data. A reliability analysis method was applied to the obtained pdf to estimate the reliability function, lifetime or mean life, reliable life, and rate of failure of the used PV system as assessment factors. The results showed that the system efficiency is highly dependent on the ambient temperature, while the performance ratio depends on many variables. The reliability analysis revealed that the field orientation of 30° tilt and 20° azimuth and of 30° tilt and 30° azimuth are best for near Cairo and near Riyadh, respectively. These orientations lead to the longest mean life of 772.25 and 688.36 months for Cairo and Riyadh, respectively. It also resulted in the lowest failure rates of 0.001295 and 0.001228 per month for both regions.

Development of Predictive Reliability Model of Solar Photovoltaic System Using Stochastic Diffusion Process for Distribution System

Reliability assessment of power distribution systems is an essential concern for electric utilities to maintain supply continuity. In the event of supply failure, distributed energy resources play a vital role in restoring the supply to customers. This paper presents the development of a predictive reliability model for distribution system with solar Photovoltaic (PV) system. At the initial stage, the hourly solar irradiance and temperature are forecasted using Stochastic Diffusion Process (SDP)-based Monte Carlo simulation from historical data. Later, the cloud transients are modeled using the Jump Diffusion Process integrated into SDP. The monthly failure rates of PV system components are evaluated from historical data using the Physics of Failure (PoF) reliability model. Finally, the predictive reliability evaluation of the practical distribution system integrated with the proposed solar PV model is carried out. Results demonstrate the capability of the developed model to handle continuous and discrete uncertainties with more accuracy.

Mathematical Modeling of Survivability Function for Thermoelectric Module

Developing an optimized reliability model for thermoelectric module at the stress where the probability of module to functions without abruptive failure is a challenging aspect. One of the major reasons is the mismatch of thermal expansion coefficient, which has severe effects on segmented moduli compared to unsegmented moduli. The likelihood of a thermoelectric module to survive at certain level of thermo-mechanical stresses varies by varying number of component (layers) in thermoelectric leg. On another hand, selection of an adequate distribution model to predict reliability and sustainability of the thermoelectric module requires development of new optimized stress-strength-based model. In this paper the predictive reliability model for high temperature segmented module is derived from parametric Lognormal mean residual life and nonparametric Lognormal-kernel survival function to measure probability of module to survive at certain thermo-mechanical stress. A comprehensive comparative discussion has been done to illustrate the maximum likelihood based on Bayesian nonparametric lognormal-Kernel inference method regarding to Monte Carlo simulation, Weibull’s distribution, and Lognormal mean residual life for various shapes for the survival function. It has been demonstrated that nonparametric lognormal-kernel survival function has high ratio of probability to predict the survival of module at higher discrete thermo-mechanical stress data.

Medical Undertakings Should Be Quantified to Have A Potential to Be Improved

The today’s efforts of system engineers to assure adequate and reliable performance of electronic devices, including medical devices, use, as a rule, qualitative, rather than quantitative, considerations. This is even truer about the performance of the healthcare and medical personnel. In the best case, the available pertinent aposteriori statistics is accumulated and taken into account. The author of this brief review developed, during his long professional lifetime, numerous predictive reliability models and accelerated test techniques in various areas of engineering and applied science and is convinced that the successful outcome of a medical mission or a more or less typical clinical situation cannot be expected, if it is not quantified. Every significant healthcare effort has to be quantified to be improved and because of the various inevitable intervening uncertainties, affecting its outcome, such a quantification should be done on the probabilistic basis.

Cycle life and statistical predictive reliability model for all-solid-state thin film microbatteries

This paper focuses on the development of a predictive cycle life model of all-solid-state thin film Cu/Li/LiPON/LiCoO2 microbatteries based on experimental accelerated aging tests. More than 120 samples were tested in 8 different conditions including several electrochemical cycling modes and different temperature levels. Hence, an original reliability model, based on an exponential function especially designed to describe the degradation mechanism occurring at the positive electrode, is proposed to describe the observed capacity loss evolution of microbatteries. An extended mathematical model, which includes a large set of accelerated factors, physicochemical considerations and statistical parameters is then built around such an exponential function. This model thus enables to simulate a wide variety of battery aging profiles from input parameters such as temperature, depth of charge (DoC), discharge rate, failure rate and capacity fade. We believe that the development of such a reliability model is of most importance not only from an industrial point of view, since it enables to determine precisely the expected behavior of microbatteries in various mission profile conditions, but it represents also a key tool that enables to further understand the observed degradation mechanisms, and to identify any technological optimization opportunities.

Evaluating and predicting molecular mechanisms of adhesive degradation during field and accelerated aging of photovoltaic modules

AbstractExtending photovoltaic (PV) module lifetimes beyond 30 years is a goal of significant priority. A challenge that must first be addressed, however, is the development of a predictive reliability model that captures the synergy of terrestrial stressors on module degradation, particularly at encapsulant interfaces. Using a metrology designed specifically for PV modules, a comprehensive study of the widely used ethylene vinyl acetate encapsulant was performed in which encapsulant adhesion was evaluated as a function of environmental stressors (UV exposure, temperature, and humidity) for modules aged both under accelerated lab and internationally located field conditions for months to nearly 3 decades. Mechanical and chemical characterization methods are combined with fundamental polymer reaction engineering to unravel the degradation processes active at the molecular scale that lead to encapsulant delamination. An analytical and modular model framework is put forward enabling the prediction of long‐term PV module durability, starting from fundamental principles at the molecular level and explicitly accounting for bond rupture events in the bulk encapsulant and at the encapsulant interfaces. Successful parameter tuning to adhesion data indicates a dominant occurrence of deacetylation, β‐scission, and hydrolytic depolymerization, respectively. The model contributes to the longstanding challenge of predicting module lifetimes in any geographic location while minimizing time‐consuming and costly aging studies.

A New Methodology for EMC Prediction of Integrated Circuits After Aging

Electromagnetic compatibility (EMC) of integrated circuits (ICs) represents a major constraint for the qualification of electronic circuits. Today, there exist standardized models such as integrated circuit emission model for conducted emissions (ICEM-CE) that allows the prediction of conducted emission generated by an IC at a printed circuit board (PCB) level. However, the EMC levels may change after a certain period of operation due to the aging of components. On the other hand, no existing model or tool can predict the long-term EMC levels. This paper presents a new methodology for modeling of ICs in order to construct an EMC model, which takes into account the aging based on a new reliability model called MultiphySics mulTistressOrs predictive Reliability Model (M-STORM).

(Invited) Toward an Understanding of Defects and Device Reliability in GaN Hemts

GaN device reliability has become a major factor in technology acceptance for various electronics applications including high radiation and space applications. GaN’s properties make it ideal for a variety of harsh conditions including high energy particle radiation, high power RF biasing, high temperature, and simultaneous combinations of these stressors. Understanding how trapping behavior impacts the electrical characteristics and how defects form or are activated with these various stressors is important to predicting lifetime and end of life performance and approaches to reduce the degradation rate, improve the radiation hardness, etc. This work explores different modes of GaN high electron mobility transistors (HEMT) degradation when subjected to proton irradiation and electrical stressing. Specific trap states associated with different stress conditions are revealed using conventional and HEMT-based deep level optical spectroscopy (DLOS) and deep level transient spectroscopy (DLTS) methods. These trap characterization techniques provide quantitative information about degradation-limiting traps that can exist anywhere within the device bandgap profile and can be distributed within specific regions of the device structure. In addition, using the newly developed scanning-DLTS, we will show the ability to quantitatively map specific defects on the nanometer scale to identify how defects are related to specific microstructure. Multiple types of AlGaN/GaN HEMTs were characterized in a systematic fashion using conventional, HEMT-based, and scanning-DLTS and DLOS before and after the application of RF switching, high voltage switching and irradiation. Overall, radiation is found to form defects that primarily impact the threshold voltage where shallower traps cause dynamic threshold voltage VT instabilities and deeper traps cause static VT shifts. In particular, proton irradiation is found to increase the concentration of traps at EC-1.3 eV and EC-3.28 eV in the GaN buffer, which appear to be responsible for large and persistent VT shifts in GaN HEMTs, and a state at EC-0.72 eV in the GaN buffer that separately correlates with an increase in dynamic VT instability. The RF accelerated life testing (ALT) has led to different degradation pathways in different generations of HEMTs and different vendors but has consistently been in the drain access region where the electric fields are maximized, which is quite different than the radiation-induced degradation. Early generations consistently showed correlation of the RF output power with the concentration of the EC-0.57 eV level, which we have recently mapped with scanning-DLTS to directly correlate the EC-0.57 eV level with edge dislocations and Fe doping. Newer generations have degraded due to deeper, near mid-gap defects indicating that several degradation mechanisms are possible in GaN HEMTs and they evolve as the materials and device designs are improved. Semi-on DC ALT exhibits degradation somewhat different to RF ALT where the degradation is still due to defects in the drain access region but the traps responsible are different indicating that different testing mechanisms may not be predictive of actual degradation in some cases. The use of DLOS and DLTS to track the activation of these defects will be shown as essential to accurately determine exact degradation mechanisms to establish accurate predictive reliability models and guidelines for improved end of life material and device designs.

Predictive Reliability Model of 10G/25G Mesa-Type Avalanche Photodiode Degradation

<p class="1Body">Avalanche photodiodes (APDs) are important building blocks for high-sensivity, low<strong>-</strong>noise receivers deployed in the datacenter, wireless and cloud computing networks. Maintaining stable dark current is a crucial task for overall robust sysem reliability. To achieve design-in low dark current stability, good knowledge of reliability physics is indispensable. In this work, we study the physical mechanisms of 10G/25G mesa-type APD degradation. We institute a predictive reliability model to account for the degradation processes. A comprehensive comparison of APD and IC transistor is also illustrated in terms of dielectric breakdown, mobile ion migration and hot carrier injection. The model suggests that surface leakage current is the dominant factor for the mesa-type APD degradation. Based on the model, it is predicted that highly reliable 10G/25G APD can be achieved with the suppression of weak links at the surface/interface states.</p>

Diseño de Esquemas de Autorestauración mediante la Ubicación Estratégica de Reconectadores utilizando un Modelo Predictivo de Confiabilidad

El principal objetivo de este trabajo es el diseño de esquemas de autorestauración del servicio, mediante la ubicación estratégica de reconectadores en los alimentadores del sistema de distribución de la Empresa Eléctrica Regional Centro Sur C.A. (CENTROSUR), aplicando filosofías de protección de sobrecorriente tales como “Fuse Saving” y criterios de automatización de la distribución “Fault Location, Isolation and Service Restoration – FLISR”. Se calibra un modelo predictivo de confiabilidad para evaluar oportunidades de mejora y luego se prioriza el portafolio de proyectos en función de la reducción del tiempo de interrupciones y de un análisis incremental beneficio/costo.

Reliability Performances of BST Capacitors for Impedance Tuning Applications

Over the past few years, ferroelectric thin films have been widely considered for the development of tunable microwave device [1]. One of the best candidates identified so far for such applications is BaxSr1-xTiO3 (BST) material, which exhibits a significant variation of permittivity as the electrical field increases, with quite low dielectric losses, as shown in figure 1.STMicroelectronics recently developed a solution of integrated tunable capacitor based on BST material, which effectively offers excellent RF performances, low power consumption and high linearity required in adaptive radiofrequency tuning applications. These tunable capacitances are controlled through a bias voltage, typically ranging from 2 to 20 V, which lead to a capacitance tuning ratio up to 4. When embedded in mobile phone circuits, these advanced tunable chips enable to maintain an efficient transfer of energy from the handset amplifier to the antenna under various operating conditions, making the performance almost insensitive to the external environment.During the qualification of BST thin film capacitor process, the evaluation of reliability performances has to be carefully carried on, to ensure that BST material effectively fulfills the reliability expectations required in application mission profiles. This paper focuses on the development of a predictive reliability model, established from highly accelerated tests performed directly at wafer level. This reliability model is representative of Time-Dependent Dielectric Breakdown (TDDB) failure mechanism, which leads to an irreversible loss of the dielectric insulating properties.To develop such a model, the time-to-breakdown (Tbd) evolution of BST capacitors were recorded under various stress conditions, applying voltage stresses from 22V to 30V, at temperature levels ranging from 160°C to 200°C. In such conditions, failures are observed to occur before a few hours, which effectively represent pretty high accelerated test conditions compared to the product lifetime expected to exceed five years. Each test was performed on at least 30 capacitors. The statistical distributions of time-to-breakdown data were fitted by the Weibull law. Indeed, the Weibull approach is known to be the most appropriate one to describe the statistical behavior of breakdown data, not only for silicon oxide dielectrics, but also for high-k dielectrics such as ferroelectric materials, as recently demonstrated [2]. To describe the capacitors time-to-breakdown evolution as a function of stress level, a mathematical model structure can be proposed, associating an Arrhenius law to account for the temperature acceleration, with a power law for the voltage acceleration effect [3]: Tbd = C . exp(Ea/kT) . V-n The lifetime simulation results provided by the model fit perfectly with the experimental data obtained at wafer level, but it also turns out to be in very good agreement with additional reliability results obtained at package level, in moderate acceleration conditions, as shown in figure 2. We believe that such agreement validates the relevance of the reliability approach adopted in this study, and the consistency of the mathematical structure of the model.To conclude, this study presents the development of a predictive model that enables to simulate the reliability performances of a new generation of tunable capacitors, based on BST dielectric material. The applications of such reliability model are numerous; it represents a key asset for any manufacturer. The lifetime predictions provided by the model turn out to be perfectly compliant with RF tunable applications, as demonstrated on three different qualification lots. Besides, by implementing a wafer level monitoring test in-line, this reliability tool offers the opportunity to monitor the lifetime performance of any sample, and to guaranty the reliability of the whole production line.[1] A.K.Tagantsev et al., Journal of Electroceramics, 2003.[2] M-T.Chentir et al., Microelectronics reliability, 2009.[3] S-C.Huang et al., Journal of Applied Physics, 1998.

Reliability Performances of BST Capacitors for Impedance Tuning Applications

STMicroelectronics recently developed a solution of integrated tunable capacitor based on BST material, which offers excellent RF performances, low power consumption and high linearity required in adaptive radiofrequency tuning applications. This paper focuses more particularly on the development of a predictive reliability model established from highly accelerated tests performed directly at wafer level on dedicated capacitor test structures. The validity of such model, representative of Time-Dependent Dielectric Breakdown (TDDB) failure mechanism, was confirmed by additional package level reliability tests, carried out on final product samples. The lifetime predictions provided by the model turn out to be perfectly compliant with RF tunable applications, considering typical mission profiles for such products.

Equilibrium and dynamic studies for adsorption of boron on calcium alginate gel beads using principal component analysis (PCA) and partial least squares (PLS)

AbstractIn this paper, boron removal from aqueous solutions by adsorption on calcium alginate gel beads was investigated. Based on the batch tests, calcium alginate gel beads demonstrated a good capacity of adsorption; the maximum capacity obtained was 121 mg/g at pH 6. The aim of this paper is the study of analytical tools for extracting more analytical information from a few experiments. Principal component analysis (PCA) was employed to perform a comparative analysis between different isotherms obtained by varying the percentage of alginate on the adsorbent beads. The calculated fraction of the cumulative sum of squares for all the components (R2X(cum)) and the predictive reliability model based on the value of the correlation coefficient of cross validation (Q2 (cum)) obtained were 0.9639 and 0.8523, respectively. Dynamic system study using 20 different columns was carried out. Partial least squares (PLS) modeling are applied to characterize the effect of the experimental variables on the adsorption process. Some of the descriptors studied were the initial boron concentration (Co), adsorbent mass (m), pH, column diameter (dc), column depth (Z) and superficial flow velocity (U0). Experimental adsorptions were compared against predicted adsorptions and a correlation coefficient (r2) of 0.9927 was obtained.

Reliability Model of Motorway Toll Collection System

Reliability Model of Motorway Toll Collection System In the paper, a methodology and a concept for establishment of a predictive reliability model of toll collection system at motorway section are presented. Markov processes and methods for projecting exploitation reliability have been used for modelling reliability. Elaboration of the model will allow for defining a projectable availability index of the system and its individual elements. The indices will allow for determining to what extent the system meets specific requirements and for preparing relevant exploitation strategy.

Science-based MEMS reliability methodology

In cases where device numbers are limited, large statistical studies to verify reliability are impractical. Instead, an approach incorporating a solid base of modelling, simulation, and material science into a standard reliability methodology makes more sense and leads to a science-based reliability methodology. The basic reliability method is (a) design, model and fabricate, (b) test structures and devices, (c) identify failure modes and mechanisms, (d) develop predictive reliability models (accelerated aging), and (e) develop qualification methods. At various points in these steps technical data is required on MEMS material properties (residual stress, fracture strength, fatigue, etc.), MEMS surface characterization (stiction, friction, adhesion, role of coatings, etc.) or MEMS modelling and simulation (finite element, analysis, uncertainty analysis, etc.). This methodology is discussed as it relates to reliability testing of a micro-mirror array consisting of 144-piston mirrors. In this case, 140 mirrors were cycled full stroke (1.5 μm) 26 billion times with no failure. Using our technical science base, fatigue of the springs was eliminated as a mechanism of concern. Eliminating this wear-out mechanism allowed use of the exponential statistical model to predict lower bound confidence levels for failure rate in a “no-fail” condition.

Software reliability and cost models: Perspectives, comparison, and practice

Computer software has gradually become an indispensable element in many aspects of our daily life and an important factor in numerous critical applications such as nuclear plants, medical monitoring control, real-time military and air traffic control. In this paper, we present recent studies in software reliability engineering that include nonhomogeneous Poisson process (NHPP) software reliability models, NHPP models with environmental factors, and cost models. Several applications are presented to illustrate the models. More applications are needed to fully validate the descriptive and predictive software reliability models in a general industrial setting. Future research directions in software reliability are also discussed.

Predictive reliability models through validated correlation between power cycling and thermal cycling accelerated life tests

Aims to show that with careful modelling, the fatigue life of solder joints of identical geometry and microstructure can be predicted very accurately (through empirical correlations) under different environmental test or field use conditions. Here, on the TI 144 chip ‐scale package, the empirical correlation for fatigue life developed under thermal cycling conditions is used to predict the life under power cycling. This accurate model has served as the physical basis which in to demonstrate quantitatively the equivalence of thermal cycling and power cycling as valid accelerated life tests. Describes the great importance of spatial refinement, temporal refinement, and accurate boundary conditions, including the often ignored natural convection boundary conditions, and their effect on predicted life.

A Bayesian predictive software reliability model with pseudo-failures

In our previous paper (2000), a Bayesian software reliability model with stochastically decreasing hazard rate was presented. Within any given failure time interval, the hazard rate is a function of both total testing time as well as number of encountered encountered failures. In this paper, to improve the predictive performance of our previously proposed model, a pseudo-failure is inserted whenever there is a period of failure-free execution equals (1-/spl alpha/)th percentile of the predictive distribution for time until the next failure has passed. We apply the enhanced model with pseudo-failures inserted to actual software failure data and show it gives better results under the sum of square errors criteria compared to previous Bayesian models and other existing times between failures models.

Architectural factors influencing the reliability of fault-tolerant VLSI arrays

Current trend in VLSI technology is toward highly modular structures in which identical Processing Elements (PE) are connected in a regular lattice. Enhancement of the operational reliability of these VLSI devices is obtained by means of fault-tolerance. Fault-tolerance is achieved incorporating spare PE's into the array and designing a flexible interconnection network which is able to support the reconfiguration of the array in the presence of a fault. This paper discusses how architectural related factors influence the configuration and the technology of the device, and how these factors can be accounted for in a predictive reliability model.