Small Number Of Failures Research Articles

Discussion of specifying prior distributions in reliability applications

AbstractTian et al. have reviewed and discussed various noninformative or weakly informative priors when reliability data are modeled using the log‐location‐scale family of distributions. They have compared these priors to partially informative priors, which are also called modified Jeffreys' priors in the literature. The authors recommend adaptations to Bayesian methods when analyzing reliability data with a small number of failures. Does a chosen Bayesian computation method play a role in the inference? Can a Bayesian model, without an informative prior, really solve the inference problem which is due to a lack of information provided by the data? How should a practitioner describe formal decisions based on the chosen Bayesian model (a likelihood as well as a prior)? Can one translate available information into an appropriate parameter space and a prior distribution on it, independent of the type of censoring? In this commentary, we put forward such questions and concerns that deserve a discussion, from a practitioner's point of view.

Інтервальна оцінка показників надійності за результатами випробувань компонентів складної системи

The goal of this work is to find the lower estimate of the no-failure probability (NFP) of a complex monotonic nonrecoverable system from the results of independent binomial tests of its components. Using the general-and-probabilistic method, the NFP is considered as a probability function polynomial, which is a linear homogeneous polynomial in each of the S variables where S is the number of system component types. Based on the method of confidence sets, the NFP lower estimate is found as the minimum of a function of an unknown multidimensional parameter at a probability of the aggregate test results (failure-free operation) of the system components equal to one minus the guaranteed confidence coefficient. The paper reports a system of equations, each of which for two component types relates the component reliability derivatives of the NFP (and one more equation relates the component reliability and the confidence coefficient). Conditions are found for the initial guess in a numerical solution of the above system of nonlinear equations (the number of the conditions is equal to the number of the component types minus one; each condition is a like sign for two functions each of which depends on the probability of the test results of a particular component type and the component reliability of this probability). In some specific cases, the program dimension can be reduced due to the simple structure of the probability function polynomial. The presented method gives a confidence reliability estimate with a guaranteed confidence coefficient for complex system that cannot be reduced to a serial-parallel or a parallel-serial structure and consist of components with an arbitrary type of failure time distribution. The method allows one to get an estimate at a small number of tests and a small number of failures or in their absence, which is of especial importance for high-reliability systems.

Specifying prior distributions in reliability applications

Especially when facing reliability data with limited information (e.g., a small number of failures), there are strong motivations for using Bayesian inference methods. These include the option to use information from physics‐of‐failure or previous experience with a failure mode in a particular material to specify an informative prior distribution. Another advantage is the ability to make statistical inferences without having to rely on specious (when the number of failures is small) asymptotic theory needed to justify non‐Bayesian methods. Users of non‐Bayesian methods are faced with multiple methods of constructing uncertainty intervals (Wald, likelihood, and various bootstrap methods) that can give substantially different answers when there is little information in the data. For Bayesian inference, there is only one method of constructing equal‐tail credible intervals—but it is necessary to provide a prior distribution to fully specify the model. Much work has been done to find default prior distributions that will provide inference methods with good (and in some cases exact) frequentist coverage properties. This paper reviews some of this work and provides, evaluates, and illustrates principled extensions and adaptations of these methods to the practical realities of reliability data (e.g., non‐trivial censoring).

Modern cladding systems for big sheds: The emerging state of the art

Stressed skin theory was first introduced in the 1960’s and intensive development continued until the mid-1980’s. This development was characterised by extensive research into the behaviour of low-rise portal frame structures clad with a single skin of profiled metal sheeting. The resulting state of the art was formalised in ECCS (1995). Since then, frames have become much taller and more slender while alternative cladding systems have proliferated. About 10 years ago, the construction industry in the UK was shaken by a small number of failures in the roofs of ‘Big Sheds’ clad with sandwich panels. Investigation by numbers of experts was inconclusive and non-disclosure orders (NDO’s) were issued by the companies concerned. In the UK, the Standing Committee on Structural Safety (SCOSS) was informed of this and, after discussion with the Institution of Structural Engineers issued an Alert entitled ‘Effects of scale’ (SCOSS, 2018). This drew attention to the fact that validation of our current design rules has been largely based on years of satisfactory building performance. However, such evidence may not be indicative of satisfactory performance as structures increase in scale. In deference to the NDO’s, a research program has been commenced based entirely on structural mechanics without making any direct reference to either forensic evidence from the sites of the failures or test results etc. that are not in the public domain. This is one of the papers arising from this research. It recognises significant deficiencies in the design procedures in the European Recommendations (ECCS, 1995) and the need for a full, non-linear, 3D numerical analysis of the complete clad structure. This, in turn, has resulted in the development of the ‘component method’ of analysis for structure cladding interaction problems that concentrates on the development of numerical models together with the performance of their ‘components’. The main alternative cladding systems are considered in turn and their characteristics and potential failure modes identified.

Impact of implementing quality control systems in laboratories associated with teaching and research institutions − The case study of the laboratory for macromolecules and colloids in the petroleum industry

The competitiveness between institutions and companies generates the need to cheapen products and services and facilitate their production. Thus, several companies establish partnerships with specialized university laboratories in order to promote the optimization of their processes. These laboratories, in order to guarantee their reliability, often resort to the application of technical standards and the obtaining of international certifications. This paper aims to discuss the relevance of international certifications by research laboratories by report a case study from a laboratory of the Federal University of Rio de Janeiro. The results of the implementation of the quality management system of Laboratory for Macromolecules and Colloids in the Petroleum Industry (LMCP) are presented, where excellent satisfaction results were obtained, as well as a very small number of failures, demonstrating that the system is effective, which generates more and more customers and carry out increasingly sophisticated research. As perspectives it is expected that the quality management system can increase the number of collaborations and high-quality service delivery between industry and the LMCP.

Improved test for the scale parameter of the power-law process with incomplete failure data

The problem of testing the scale parameter (γ) of the power-law process for a single system is taken into account when dealing with incomplete failure data. In this study, the modified signed log-likelihood ratio test is introduced as a method for testing parameter γ. The accuracy of the proposed method is evaluated in the literature. For incomplete failure data, the modified signed log-likelihood ratio test is satisfactory and preferable to the signed log-likelihood ratio test for most situations in terms of controlling type I errors, and this method works well like the generalized variable test which is the exact method since it yields the empirical type I errors close to the nominal level 0.05. In terms of empirical powers, the modified signed log-likelihood ratio test performs satisfactorily and seems to be superior to the generalized variable test when the ratio of is small even for small number of failures and large missing ratios.

Intermediate to Long-Term Follow-up of Cementing Liners into Well-Fixed Acetabular Components.

The cementation of a new liner into a well-fixed acetabular component is common during revision total hip arthroplasty (THA) for many indications, but most commonly for lack of a modern, compatible, highly cross-linked polyethylene (HXLPE) liner. However, little is known about the intermediate-term to long-term durability of this strategy. The purpose of this study was to evaluate the implant survivorship, risk of complications, clinical outcomes, and radiographic results of cementing a new HXLPE liner into a well-fixed acetabular component. We retrospectively identified 323 revision THAs in which a nonconstrained HXLPE liner was cemented into a well-fixed acetabular component. The mean age at the time of the revision THA was 63 years, and 50% of patients were female. The most common indications for revision THA were polyethylene wear and osteolysis (48%), aseptic femoral loosening (35%), and hip instability (8%). The mean follow-up was 9 years. Polyethylene liner failure occurred in 11 cases (3%). In all cases, the cemented liner dissociated from the acetabular component. At 10 years, the survivorship free from any revision was 80% (95% confidence interval [CI], 75% to 84%) and the survivorship free from any reoperation was 77% (95% CI, 72% to 82%). The most common reason for re-revision was dislocation (45% of reoperations). A dislocation occurred in 17% of cases. Hips that underwent revision for instability were significantly more likely to dislocate compared with hips that underwent revision for liner wear (hazard ratio [HR], 2.3 [95% CI, 1.2 to 4.5]; p = 0.02). Elevated rim or face-changing liners were significantly more likely to dissociate than flat liners (HR, 9.0 [95% CI, 1.2 to 70.6]; p = 0.04). Cementation of a nonconstrained HXLPE liner into a well-fixed acetabular component during revision THA provided durable fixation with only a small number of failures at the cement interface (3%). Instability after this procedure remains a concern, but this is multifactorial in nature. These data support the continued use of this technique, when necessary, during revision THA. Therapeutic Level IV. See Instructions for Authors for a complete description of levels of evidence.

Improving Gnutella Query Search Algorithm with Jumps

The measurement of a search algorithm for unstructured P2P network centres on the number of nodes not receiving their requested files (number of failures) and the number of hops per query. Most current search algorithms are unable to guarantee the success of the query. This study involves a comparison of the strengths and weaknesses of three algorithms of Gnutella P2P protocol, namely Flood, Random Walk, and Random Walk with Neighbours Table. Based on this comparison, a new query search method—referred to as Random Walk with Jumps—is proposed. The experiment proves that the proposed algorithm can obtain a better result with a small number of failures and a minimum number of hops.

PROJETO, FABRICAÇÃO E QUALIFICAÇÃO DE EQUIPAMENTO PARA FLEXÃO ROTATIVA DE FIOS SUPERELÁSTICOS DE NITI EM ENSAIOS DE FADIGA

In this work it was developed a rotating bending apparatus for fatigue tests of superelastic NiTi wires, and other materials with high elasticity. It was evaluated the performance, robustness, operability, and reliability through testing of 1 mm thick stainless steel wires. This device is mounted on a steel frame and features semiautomatic rotation speed control, time and testing bath temperature for sample immersion. The equipment qualification tests were performed controlling the following parameters: deformation of the wire, power level and ambient temperature. The results indicated lower discrepancies for the following parameters evaluated: number of cycles in fatigue life, rotation speed, the bath temperature and arc angle of rupture. Besides the reliability, the robustness and operability of the equipment also meet the purpose of the research as evidenced by the small number of failures in the qualification tests and calibration

Focus section on program debugging

Focus section on program debugging

Detection of knockdown resistance (kdr) mutations in Anopheles gambiae: a comparison of two new high-throughput assays with existing methods

BackgroundKnockdown resistance (kdr) is a well-characterized mechanism of resistance to pyrethroid insecticides in many insect species and is caused by point mutations of the pyrethroid target site the para-type sodium channel. The presence of kdr mutations in Anopheles gambiae, the most important malaria vector in Africa, has been monitored using a variety of molecular techniques. However, there are few reports comparing the performance of these different assays. In this study, two new high-throughput assays were developed and compared with four established techniques.MethodsFluorescence-based assays based on 1) TaqMan probes and 2) high resolution melt (HRM) analysis were developed to detect kdr alleles in An. gambiae. Four previously reported techniques for kdr detection, Allele Specific Polymerase Chain Reaction (AS-PCR), Heated Oligonucleotide Ligation Assay (HOLA), Sequence Specific Oligonucleotide Probe – Enzyme-Linked ImmunoSorbent Assay (SSOP-ELISA) and PCR-Dot Blot were also optimized. The sensitivity and specificity of all six assays was then compared in a blind genotyping trial of 96 single insect samples that included a variety of kdr genotypes and African Anopheline species. The relative merits of each assay was assessed based on the performance in the genotyping trial, the length/difficulty of each protocol, cost (both capital outlay and consumable cost), and safety (requirement for hazardous chemicals).ResultsThe real-time TaqMan assay was both the most sensitive (with the lowest number of failed reactions) and the most specific (with the lowest number of incorrect scores). Adapting the TaqMan assay to use a PCR machine and endpoint measurement with a fluorimeter showed a slight reduction in sensitivity and specificity. HRM initially gave promising results but was more sensitive to both DNA quality and quantity and consequently showed a higher rate of failure and incorrect scores. The sensitivity and specificity of AS-PCR, SSOP-ELISA, PCR Dot Blot and HOLA was fairly similar with a small number of failures and incorrect scores.ConclusionThe results of blind genotyping trials of each assay indicate that where maximum sensitivity and specificity are required the TaqMan real-time assay is the preferred method. However, the cost of this assay, particularly in terms of initial capital outlay, is higher than that of some of the other methods. TaqMan assays using a PCR machine and fluorimeter are nearly as sensitive as real-time assays and provide a cost saving in capital expenditure. If price is a primary factor in assay choice then the AS-PCR, SSOP-ELISA, and HOLA are all reasonable alternatives with the SSOP-ELISA approach having the highest throughput.

Control Charts for Monitoring Field Failure Data

One responsibility of the reliability engineer is to monitor failure trends for fielded units to confirm that pre-production life testing results remain valid. This research suggests an approach that is computationally simple and can be used with a small number of failures per observation period. The approach is based on converting failure time data from fielded units to normal distribution data, using simple logarithmic or power transformations. Appropriate normalizing transformations for the classic life distributions (exponential, lognormal, and Weibull) are identified from the literature. Samples of size 500 field failure times are generated for seven different lifetime distributions (normal, lognormal, exponential, and four Weibulls of various shapes). Various control charts are then tested under three sampling schemes (individual, fixed, and random) and three system reliability degradations (large step, small step, and linear decrease in mean time between failures (MTBF)). The results of these tests are converted to performance measures of time to first out-of-control signal and persistence of signal after out-of-control status begins. Three of the well-known Western Electric sensitizing rules are used to recognize the assignable cause signals. Based on this testing, the ―X-chart with fixed sample size is the best overall for field failure monitoring, although the individual chart was better for the transformed exponential and another highly-skewed Weibull. As expected, the linear decrease in MTBF is the most difficult change for any of the charts to detect. Copyright © 2005 John Wiley & Sons, Ltd.

A brief historical overview of the fatigue of railway axles

The history of railway axle design against fatigue is reviewed through a discussion of several notable accidents caused by broken axles. The earliest accidents prompted research into fatigue which resulted in a phenomenological understanding. In the last 50 years, largely through research directed towards the aircraft industry, fatigue has been recognized to be a process involving crack initiation and propagation. Fracture mechanics is now used to assist in the timing of crack inspection intervals, but a small number of failures still occur. The paper concludes with a short discussion of modern design methods, which have their origin in the nineteenth century. Some current pressures, particularly the need to reduce unsprung mass at high speeds, continue to challenge axle designers.

Building bridges to education

Since it started in May 1998, InterAct's Bridges to Education programme has helped over 1000 clients with mental health issues to access lifelong learning. The underlying purpose of the programme is to help clients involve themselves in their communities. Client feedback about the benefits is overwhelmingly positive. A few clients have not been helped. This article looks at factors that have led to the successes and factors that may have contributed to the small number of failures.

Rheumatic Fever—A Half-Century Perspective

Rheumatic Fever—A Half-Century Perspective

Improved reliability estimates for samples containing suspensions

A method is presented for obtaining better estimates of reliability levels used in Weibull analysis and related techniques. The method, which is applicable when there are suspended items, overcomes an inherent pessimism which occurs in existing plotting‐based methods, and provides reliability estimates which are continuously sensitive to any changes in the ages of suspended items. Together, these features provide a more realistic estimate of the reliability function than the existing methods provide, particularly in samples with a small number of failures.

Election in asynchronous complete networks with intermittent link failures

Considers the problem of fault-tolerant leader election in asynchronous complete (fully-connected) distributed networks. The processors are reliable, but some of the communication channels may fail intermittently before or during the execution of the algorithm. Channel failures are undetectable due to the asynchronous nature of the network. Let n be the number of processors in the network and f be the maximum number of faulty channels incident on each processor, where f/spl les/ 1/2 [n-1]. Our algorithm uses at most O(n/sup 2/+nf/sup 2/) messages to elect a unique leader of the network. Each message consists of at most O(log|T|) bits, where |T| is the cardinality of the set of processor identifiers. All previous algorithms either tolerated only benign failures such as fail-stop failures, assumed that the network is synchronous, tolerated only a small number of failures, or assumed that the faults are detectable. Our algorithm is the first election algorithm that is designed specifically for asynchronous intermittently faulty complete networks in which up to 1/4 n[n-1] channels may be faulty, where each processor is adjacent to no more than 1/2 [n-1] faulty channels, and where the faults are undetectable. >

Personality changes after anti-incontinence operations in women

Seventy-three multiparous women with stress incontinence were evaluated before and 1 year after anti-incontinence procedures to assess the effects of successful surgical therapy on the personality of patients with genuine stress urinary incontinence. Urodynamic studies were used to assess the urinary incontinence; Taylor-Johnson temperament analysis was used to evaluate the personality characteristics of the subjects; and paired t-tests and t-tests were performed for statistical analysis. Nine of the 73 patients were excluded because of invalid answers on the questionnaire. Of the remaining 64, 57 (89%) were subjectively cured and 51 (80%) were objectively cured. The preoperative temperament evaluation revealed, in all the patients, traits of depressiveness and quietness which were significantly different from the normal range of the average population. Postoperatively, the authors found significantly (P<0.05) reduced depressiveness and increased social involvement, demonstrativeness and self-discipline only in the women subjectively cured. A trend was noted in terms of psychosocial adaptation between the patients cured and the ones with persistent incontinence. However, due to the small number of failures (n=7) no statistical differences could be demonstrated. It was concluded that women with stress urinary incontinence have increased depressiveness and decreased psychosocial involvement compared to the average population. Successful anti-incontinence procedures significantly improve the psychosocial status of these individuals.

The use of a microtensile strength bench for testing the strength of growth rings of softwoods and hardwoods parallel to the grain

A self-developed microtensile strength bench was used to detect possible differences in tensile strength between early wood and late wood of Scots pine and beech. The difference in tensile strength parallel to the grain between early wood and late wood in Scots pine is more obvious than that of beech. The use of a gelglue (ethylcyanoacrylate) for fastening the small specimens to the testing bench was very promising. The tests showed that the method is very useful, because of a very small number of failures.

Water‐Induced Corrosion of Materials Used for Semiconductor Passivation

We show for the first time that both chemical corrosion and electrochemical corrosion of silicon‐based films does occur, and that this corrosion leads to device failure. Chemical corrosion is the dominate failure mechanism when devices are aged between 500 and 1500 h. Typically, a small number of failures are found as a result of chemical corrosion. However, if the devices are improperly packaged, then this mechanism can be responsible for most if not all failures. The mechanism is characterized by an activation energy between 0.6 and 1.0 eV depending on the passivation stoichiometry. Electrochemical corrosion dominates long term failure and is strongly dependent upon the resistivity of the passivation. We found that the median lifetime could vary by two orders of magnitude depending on the film resistivity. Because of this large variation in median lifetimes, it is important to characterize the resistivity of the passivation films used for given technologies, remembering, of course, that the resistivity is strongly dependent on film stoichiometry, applied field, and local geometry.