Failure-mechanism models for conductive-filament formation

This tutorial illustrates design and qualification situations where conductive filament formation, a failure mechanism that can occur in laminates, can affect product performance. Empirical models to design against this failure mechanism are presented in this paper. Examples illustrate the use of these models for design and qualification tests in electronic packaging     

Failure-mechanism models for creep and creep rupture

This tutorial illustrates design situations where creep and creep rupture of components can compromise system performance over time, thereby acting as a wearout failure mechanism. Polycrystalline materials, such as metals and ceramics, and polymers are treated. Analytic microstructural creep mechanisms leading to failure of these materials are presented. Continuum microscale models for predicting long-term creep are explained for practical design purposes. Two examples illustrate these models for mechanical engineering and electronic packaging situations     

Failure mechanism models for ductile fracture

This tutorial illustrates design situations where ductile fracture of some components can compromise system performance, thereby acting as an overstress failure mechanism. Analytic (physics-of-failure) methods, based on continuum fracture mechanics principles, are presented to design against such failures. Examples illustrate the use of these models in practical design situations in mechanical engineering and electronic packaging. The design equations are based on continuum mechanics rather than on molecular micromechanics, and can be implemented in an engineering design environment. The associated stress analysis of ten requires numerical finite-element techniques. The methods for material-property characterizations have matured appreciably over the past 40 years and are specified in engineering handbooks     

Failure mechanism models for electromigration

This tutorial illustrates situations where electromigration (a wearout failure mechanism) in electronic devices can degrade performance. Electromigration and its relation to microstructure is discussed. Temperature considerations are treated. A practical model for electromigration and two application examples of it are given. Qualitative design procedures for avoiding solid-state electromigration failure are briefly discussed     

Failure mechanism models for material aging due to interdiffusion

This tutorial illustrates design situations where material aging due to interdiffusion in some components can compromise system performance over time, thereby acting as a wearout failure mechanism. Microstructural diffusion mechanisms, continuum diffusion models, and interdiffusion analysis techniques are presented to design against such failures. An example illustrates the application of the mechanisms, models, and techniques in electronic packaging     

Failure mechanism models for cyclic fatigue

This work illustrates design situations where mechanical fatigue under cyclic loading, of one or more components, can compromise system performance. In this failure mechanism, damage accumulates with each load cycle, thereby causing a physical wearout failure mechanism. Phenomenological continuum length-scale models, based on micromechanical considerations, are presented to predict the onset (or initiation) of fatigue cracking in ductile materials. Fatigue crack propagation is modeled with continuum fracture mechanics principles. The number of load cycles required to cause failure is predicted based on these models. Approaches for modeling creep fatigue interactions are briefly discussed. Analytic physics-of-failure method and examples are presented for designing against wearout failure due to cyclic fatigue. These models can be implemented in an engineering design environment. The associated stress analysis requires numerical finite element techniques in many cases. The associated material property characterization techniques have matured since the 1950s and are specified in engineering handbooks     

Failure-mechanism models for excessive elastic deformation

Design situations where excessive elastic deformations can compromise system performance, thereby acting as a failure mechanism, are illustrated. Models, based on continuum mechanics, to design against such failures are presented. Two examples illustrate the use of these models in practical design situations in electronic packaging and in mechanical systems. The examples use deterministic models, for simplicity of presentation. In reality, each of the variables in the design equation is uncertain, e.g. geometric dimensions, material properties, and applied loads. If these uncertainties are important then they must be accounted for by the usual statistical methods     

Asynchronous circuits as alternative for mitigation of long-duration transient faults in deep-submicron technologies

The use of deeper-submicron technologies in integrated circuits worsens the effects of transient faults. In fact, the transient-fault durations become as important as the clock periods of synchronous circuits. Electronic systems are thus more vulnerable to failure situations. Nevertheless, this paper shows innovatively that such a worse scenario does not happen in asynchronous circuits. This additional novel benefit pushes on the asynchronous design as a better alternative to mitigate transient faults in deep-submicron technology-based circuits.     

基于嵌入式DSP系统的数据记录与分析

嵌入式DSP系统上电后启动工作,断电后停止,数据丢失。而在许多场合,如系统故障时,希望系统能对其工作过程中产生的状态和数据信息进行记录,以便于事后分析。为解决该问题,文中从硬件和软件两方面详细叙述了在DSP系统中利用电可擦除存储器AT28C010进行数据记录与提取分析的设计方案。为系统试验的事后分析与故障定位提供了数据支持,且该记录方式灵活简便、电路规模小,具有较好的工程应用价值。     

Approaches for reliability modeling of continuous-state devices

Three approaches for reliability modelling of continuous state devices are presented in this paper. One uses the random process to fit model parameters of a statistical distribution as functions of time. This approach allows the data set to be from any general distribution. The second approach uses the general path model to fit parameters of the model as functions of time. The relationship between the random process model and the general path model is illustrated. The third approach uses multiple linear regression to fit the distribution of lifetime directly. This approach has less restriction on the degradation data to be analyzed. All three approaches are illustrated with examples. Finally a mixture model is proposed which can be used to model both catastrophic failures and degradation failures. This mixture model also shows engineers how to design experiments to collect both hard failure data and soft failure data. Topics for further investigation in continuous device reliability modelling include further investigation of the mixture model, application of these models to practical situations, and using complex statistical distributions to fit degradation data     

Integrated VP-Based Control Strategies for ATM Survivable Networks

ＩｎｔｅｇｒａｔｅｄＶＰ－ＢａｓｅｄＣｏｎｔｒｏｌＳｔｒａｔｅｇｉｅｓｆｏｒＡＴＭＳｕｒｖｉｖａｂｌｅＮｅｔｗｏｒｋｓ￥ＣｈｅｎＳｈａｎｚｈｉ；ＣｈｅｎｇＳｈｉｄｕａｎ；ａｎｄＣｈｅｎＪｕｎｌｉａｎｇ（ＳｔａｔｅＫｅｙＬａｂｏｒａｔｏｒｙｏｆＳｗｉｔ...     

Hypothesis testing of equality between exponential distributionswith matched sets

In matched design, if the unit cost from one comparison group is higher than the unit cost from the other group, then one can consider matching each unit randomly selected from the former with more than one unit from the latter, to increase power of the test. This paper extends the discussion on testing equality between exponential distributions for one-to-one paired design to that for K-to-one matched design, where K can be any finite positive integer. This paper considers the asymptotic test procedure using the central limit and Fieller's theorems (CLFT), the asymptotic test procedure using the marginal likelihood ratio test (MLRT), an exact parametric test (EXPT) and applies Monte Carlo simulation to evaluate the performance of these procedures. When the number of matched sets, n, is as small as 10, the estimated type-I error for the two asymptotic procedures can still agree well with the nominal level. When the number of matched units, K, exceeds 4, the effect due to an increase in K on power generally becomes minimal. When the intra-class correlation between failure times within matched sets is small, using the CLFT generally has larger power than using either the MLRT or EXPT in one-to-one paired design. On the other hand, when the intra-class correlation between failure times within matched sets is large, the power for the MLRT is higher than the power for both the CLFT and EXPT in almost all the situations considered in this paper. Hence the author recommends the MLRT     

An exponential reliability-growth model in multicopy testingprogram

An exponential reliability-growth model, which incorporates step changes in the failure intensity, is proposed as an alternative to other reliability-growth models commonly used to analyze the failure process of repairable equipment in a development program. A multicopy testing-modification scenario is used in which several s-identical copies are put on test, and design modifications are introduced into all copies at each failure occurrence. Maximum likelihood estimation of quantities which index the reliability of the equipment as it goes into production, is investigated under the hypothesis that the unconstrained estimators are inside the parameter space. Exact and approximate procedures for obtaining interval estimates of the current failure intensity and interval prediction of the current lifetime are provided. A goodness-of-fit test is developed, and exact critical values of the test statistic are given. The situations in which unconstrained results can be correctly used are investigated and a numerical application drawn from real data is presented     

Optimal replacement policy for induced draft fans

This paper derives the optimal block replacement policies for four different operating configurations of induced draft fans. Under the usual assumption of higher cost of repair or replacement on failure compared to preventive replacement, the optimal preventive replacement interval is found by minimising the total relevant cost per unit time. Specifically, this paper finds optimal preventive maintenance strategies for the following two situations.

1. (i)|Both the time to failure and time to carry out minimal repair or replacement are exponentially distributed.

2. (ii)|The time to failure follows the Weibull distribution and there is no possibility of on-line repair or replacement.

For both situations closed form expressions are derived whose solutions give optimum preventive maintenance intervals.     

On the design of a hierarchical SS7 network: a graph theoreticalapproach

Discusses the design of Signaling System No. 7 networks based on graph theoretical methods. A hierarchical network topology is derived by combining the advantage of the hierarchical network structure with the realization of node disjoint routes between nodes of the network. By using specific features of this topology, the authors develop an algorithm to construct circle-free routing data and to assure bidirectionality also in case of failure situations. The methods described are based on the requirements that the network topology, as well as the routing data, may be easily changed     

Shared-Risk Logical Span Groups in Span-Restorable Optical Networks: Analysis and Capacity Planning Model

In an optical transport network distinct logical groups of lightwave channels between neighboring OXC nodes (called spans) may sometimes be realized over a common physical resource such as a duct or conduit, and hence share a common cause of failure. This is closely related to the concept of shared risk on individual channels or links, called SRLGs, which is relevant to pre-planned path protection schemes with shared capacity on backup paths. But when considering span-restorable networks, shared risk over logical spans (not individual channels) is the corresponding issue of concern. This work considers several aspect of how such shared-risk span groups (SRSG) affect the protection capacity design and other aspects of span-restorable mesh networks. We provide a model for capacity planning any span-restorable network in the presence of a known set of such shared-risk spans and study the relationship between capacity requirements and the number and placement of such situations. This provides guidelines as to how many SRSGs can be sustained before the capacity penalty becomes severe and methods to diagnose which of them are the most limiting to overall protection efficiency. One finding of interest is that if a given percentage of all possible dual-failure combinations incident to a common node are allowed for in the design, then nearly the same percentage of other dual-span failure combinations (any two spans in the network) will also be restorable. We also show that designing a network to withstand even a small number of multi-span co-incident spared-risk span groups will yield a significant improvement in overall dual-failure restorability and hence also in network availability.Presented at Optical Networking and Communications Conference (Opti Comm 2002), Boston, MA, USA, July-August 2002.     

Design for performance of broadband signalling and services

As the telecommunications industry evolves, and broadband services start to arrive, customers are increasingly coming to expect the perception of instantaneous access to service providers, together with transparency to network failures. System performance dictates that response times need to be minimised, sufficient redundant capacity be installed in case of failure, and controls be embedded within the design to manage the exceptional situations (such as media-stimulated events) that continually threaten network integrity. A vital part of this system is the broadband signalling network, which underpins the dialogue with the customer, and which enables the delivery of the service. Network design based on a 'top-down', 'end-to-end' methodology plays a fundamental role in delivering solutions that meet customers' performance needs. Simple, approximate performance models at an early stage in this life cycle are valuable to uncover major performance problems which affect the design of the architecture. Performance issues identified can then be fed back into the design process in an iterative way, to ensure that the design solution will conform to performance requirements.This paper outlines performance studies of a number of design scenarios for providing broadband services. These studies consider interactive multimedia services, looking at service demand, physical network topology, signalling message flows, the mapping of functional entities to physical components, and routeing, as part of the network design process. The most significant performance issues identified relate to bottle-necks, capacity requirements, and load-dependent response times as perceived by the customer.     

Copper trace fatigue models for mechanical cycling, vibration and shock/drop of high-density PWAs

When assessing the mechanical durability of electronic assemblies, the focus is generally on the solder interconnects. However, in many package styles, such as ball grid arrays (BGAs), land grid arrays (LGAs), micro lead frame (MLFs) and quad flat no-lead (QFNs), the Cu trace emanating from the solder pad may be the weakest failure site, especially if the solder joint is copper-defined rather than mask defined. This is particularly true in situations with cyclic mechanical loading, such as cyclic quasi-static bending, vibration and repetitive drop/shock.This study focuses on quasi-static mechanical cycling durability of LGA assemblies with copper-defined pads. Specimens were cycled to failure under zero-to-max, three-point bending and failure statistics were collected. The failure mode was confirmed to be fatigue cracks in copper traces emanating from the corner solder pads, just at the edge of the solder mask where the solder joint ends. The cracks were identified by lateral polishing after desoldering the component.The cyclic bending of this assembly was modeled with 3D, elastic-plastic, large deformation finite element analysis. Due to the complexity of the geometry, a global-local approach was used to identify the strain history and the mean stress at the failure site.A generalized strain-based fatigue model was used to characterize these failures, and preliminary model constants were iteratively estimated by ensuring that they were simultaneously compatible with both the durability test data and the copper stress-strain curves used in the FEA (finite element analysis).These preliminary model constants were then refined by separately modeling the initiation and progression history of fatigue damage in the cyclic bend tests. Damage progression was modeled in this study by using a technique of ‘successive initiation’ developed earlier by this research group. In this method, finite element simulations are used to progressively ‘kill’ elements that have accumulated sufficient fatigue damage to lose their load-carrying capacity. When the killed elements span the entire cross-section of the copper trace, the trace is assumed to electrically fail.This calibrated model is then used to demonstrate its ability to predict copper trace failures in other situations such as quasi-static four-point bending of LGA assemblies and vibration of BGA assemblies. More important, we demonstrate its use for re-designing of BGA assemblies to prevent copper trace failures under drop/shock loading.The important impact of this study includes insight into copper trace failures in Printed Wiring Assemblies (PWAs) under mechanical cycling, a quantitative model to predict its occurrence, and validated guidelines to prevent it by design.     

Stochastic reliability functions for failure rates derived from Gauss - Markov processes (Corresp.)

An extension of the well-known Cameron-Martin formula can be interpreted as the expectation of a stochastic reliability function applicable in those situations where nondecreasing failure rates are desired. This follows ff the failure rate is modeled as the square of a Gauss-Markov process. We describe the methodology for the general vector case, and then specialize the results to the one-dimensional case so as to obtain an exact closed-form expression for the reliability function. Using the theory of recurrent and transient processes, we then show how the choice of a model parameter and the initial state influence reliability.