Search results

This article aims to develop procedures for estimation and prediction in case of Type-I hybrid censored samples drawn from a two-parameter generalized half-logistic distribution (GHLD).

The GHLD is a versatile model which is useful in lifetime modelling. Also, hybrid censoring is a time and cost-effective censoring scheme which is widely used in the literature. The authors derive the maximum likelihood estimates, the maximum product of spacing estimates and Bayes estimates with squared error loss function for the unknown parameters, reliability function and stress-strength reliability. The Bayesian estimation is performed under an informative prior set-up using the “importance sampling technique”. Afterwards, we discuss the Bayesian prediction problem under one and two-sample frameworks and obtain the predictive estimates and intervals with corresponding average interval lengths. Applications of the developed theory are illustrated with the help of two real data sets.

The performances of these estimates and prediction methods are examined under Type-I hybrid censoring scheme with different combinations of sample sizes and time points using Monte Carlo simulation techniques. The simulation results show that the developed estimates are quite satisfactory. Bayes estimates and predictive intervals estimate the reliability characteristics efficiently.

The proposed methodology may be used to estimate future observations when the available data are Type-I hybrid censored. This study would help in estimating and predicting the mission time as well as stress-strength reliability when the data are censored.

Inferences for multicomponent reliability is derived for a family of inverted exponentiated densities having common scale and different shape parameters.

Different estimates for multicomponent reliability is derived from frequentist viewpoint. Two bootstrap confidence intervals of this parametric function are also constructed.

Form a Monte-Carlo simulation study, the authors find that estimates obtained from maximum product spacing and Right-tail Anderson–Darling procedures provide better point and interval estimates of the reliability. Also the maximum likelihood estimate competes good with these estimates.

In literature several distributions are introduced and studied in lifetime analysis. Among others, exponentiated distributions have found wide applications in such studies. In this regard the authors obtain various frequentist estimates for the multicomponent reliability by considering inverted exponentiated distributions.

This paper adopts a modified Taguchi approach to develop empirical relationships to the performance characteristics (output responses) in terms of process variables and demonstrated their validity through comparison of test data. The method suggests a few tests as per the orthogonal array and provides complete information for all combinations of levels and process variables. This method also provides the estimated range of output responses so that the scatter in the repeated tests can be assessed prior to the tests.

In order to obtain defect-free products meeting the required specifications, researchers have conducted extensive experiments using powder bed fusion (PBF) process measuring the performance indicators (namely, relative density, surface roughness and hardness) to specify a set of printing parameters (namely, laser power, scanning speed and hatch spacing). A simple and reliable multi-objective optimization method is considered in this paper for specifying a set of optimal process parameters with SS316 L powder. It was reported that test samples printed even with optimal set of input variables revealed irregular shaped, microscopic porosities and improper melt pool formation.

Finally, based on detailed analysis, it is concluded that it is impossible to express the performance indicators, explicitly in terms of equivalent energy density (E_0ˆ*), which is a combination of multiple sets of selective laser melting (SLM) process parameters, with different performance indicators. Empirical relations for the performance indicators are developed in terms of SLM process parameters. Test data are within/close to the expected range.

Based on extensive analysis of the SS316 L data using modified Taguchi approach, the optimized process parameters are laser power = 298 W, scanning speed = 900 mm/s and hatch distance = 0.075 mm, for which the results of surface roughness = 2.77 Ra, relative density = 99.24%, hardness = 334 Hv and equivalent energy density is 4.062. The estimated data for the same are surface roughness is 3.733 Ra, relative density is 99.926%, hardness is 213.64 Hv and equivalent energy density is 3.677.

Even though equivalent energy density represents the energy input to the process, the findings of this paper conclude that energy density should no longer be considered as a dependent process parameter, as it provides multiple results for the specified energy density. This aspect has been successfully demonstrated in this paper using test data.

The purpose of this article is to derive inference for multicomponent reliability where stress-strength variables follow unit generalized Rayleigh (GR) distributions with common scale parameter.

The authors derive inference for the unknown parametric function using classical and Bayesian approaches. In sequel, (weighted) least square (LS) and maximum product of spacing methods are used to estimate the reliability. Bootstrapping is also considered for this purpose. Bayesian inference is derived under gamma prior distributions. In consequence credible intervals are constructed. For the known common scale, unbiased estimator is obtained and is compared with the corresponding exact Bayes estimate.

Different point and interval estimators of the reliability are examined using Monte Carlo simulations for different sample sizes. In summary, the authors observe that Bayes estimators obtained using gamma prior distributions perform well compared to the other studied estimators. The average length (AL) of highest posterior density (HPD) interval remains shorter than other proposed intervals. Further coverage probabilities of all the intervals are reasonably satisfactory. A data analysis is also presented in support of studied estimation methods. It is noted that proposed methods work good for the considered estimation problem.

In the literature various probability distributions which are often analyzed in life test studies are mostly unbounded in nature, that is, their support of positive probabilities lie in infinite interval. This class of distributions includes generalized exponential, Burr family, gamma, lognormal and Weibull models, among others. In many situations the authors need to analyze data which lie in bounded interval like average height of individual, survival time from a disease, income per-capita etc. Thus use of probability models with support on finite intervals becomes inevitable. The authors have investigated stress-strength reliability based on unit GR distribution. Useful comments are obtained based on the numerical study.

THE enlarged scale and more varied content of this year's Business Efficiency Exhibition recalls the importance of the office in modern business. So, too, do those vast, austere rectangular boxes whose modular construction makes them so depressingly alike. Those in London and other major cities can be matched exactly in Frankfurt or Brussels, Milan or Montreal. This is a natural consequence of the tendency for the number of manual workers to decline and of clerical ones to increase.

The purpose of this paper is to present the implementation in a finite element (FE) code of a recently developed material model for the analysis of cracked reinforced concrete (RC) panels. The model aims for the efficient nonlinear analysis of large‐scale structural elements that can be considered as an assembly of membrane elements, such as bridge girders, shear walls, transfer beams or containment structures.

In the proposed constitutive model, the equilibrium equations of the cracked membrane element are established directly at the cracks while the compatibility conditions are expressed in terms of spatially averaged strains. This allows the well‐known mechanical phenomena governing the behaviour of cracked concrete elements – such as aggregate interlock (including crack dilatancy effects), tensile fracture and bond shear stress transfer – to be taken into account in a transparent manner using detailed phenomenological models. The spatially averaged stress and strain fields are obtained as a by‐product of the local behaviour at the cracks and of the bond stress transfer mechanisms, allowing the crack spacing and crack widths to be obtained directly from first principles. The model is implemented in an FE code following a total formulation.

The fact that the updated stresses at the cracks are calculated explicitly from the current spatially averaged total strains and from the updated values of the state variables that are used to monitor damage evolution contributes to the robustness and efficiency of the implementation. Some application examples are presented illustrating the model capabilities and good estimates of the failure modes, failure loads, deformation capacity, cracking patterns and crack widths were achieved.

While being computationally efficient, the model describes the complex stress and strain fields developing in the membrane element, and retrieves useful information for the structural engineer, such as concrete and reinforcement failures as well as the crack spacing and crack widths.

Fat‐tailed distributions have been found in many financial and economic variables ranging from forecasting returns on financial assets to modeling recovery distributions in bankruptcies. They have also been found in numerous insurance applications such as catastrophic insurance claims and in value‐at‐risk measures employed by risk managers. Financial applications include:

In adaptive slicing, the number of layers is drastically reduced by using sloping layer walls. For both vertical (2.5D slices) and sloping (ruled slices) outer walls, the strategies for determining slice height generally consider a number of vertical sections along the contour of a slice. Surface deviation error is calculated at these sections and slice height subsequently determined. Instead, a method is proposed which calculates error at every part of the surface. This method approximates the outer wall between two successive contours by a series of taut cubic spline patches. It is proposed that the deviation between such a patch and the actual surface is a better and more exhaustive estimate of surface error. Results show that the predicted number of slices is slightly higher than that predicted by existing methods for sloping layer walls.

The purpose of this paper is to analyze the thermal and fluid dynamic behaviors of mixed convection in air because of the interaction between a buoyancy flow and a moving plate induced flow in a horizontal no parallel-plates channel to investigate the effects of the minimum channel spacing, wall heat flux, moving plate velocity and converging angle.

The horizontal channel is made up of an upper inclined plate heated at uniform wall heat flux and a lower adiabatic moving surface (belt). The belt moves from the minimum channel spacing section to the maximum channel spacing section at a constant velocity so that its effect interferes with the buoyancy effect. The numerical analysis is accomplished by means of the finite volume method, using the commercial code Fluent.

Results in terms of heated upper plate and moving lower plate temperatures and stream function fields are presented. The paper underlines the thermal and fluid dynamic differences when natural convection or mixed convection takes place, varying minimum channel spacing, wall heat flux, moving plate velocity and converging angle.

The hypotheses on which the present analysis is based are two-dimensional, laminar and steady state flow and constant thermo physical properties with the Boussinesq approximation. The minimum distance between the upper heated plate of the channel and its lower adiabatic moving plate is 10 and 20 mm. The moving plate velocity varies in the range 0-1 m/s; the belt moves from the right reservoir to the left one. Three values of the uniform wall heat flux are considered, 30, 60 and 120 W/m², whereas the inclination angle of the upper plate θ is 2° and 10°.

Mixed convection because of moving surfaces in channels is present in many industrial applications; examples of processes include continuous casting, extrusion of plastics and other polymeric materials, bonding, annealing and tempering, cooling and/or drying of paper and textiles, chemical catalytic reactors, nuclear waste repositories, petroleum reservoirs, composite materials manufacturing and many others. The investigated configuration is used in applications such as re-heating of billets in furnaces for hot rolling process, continuous extrusion of materials and chemical vapor deposition, and it could also be used in thermal control of electronic systems.

This paper evaluates the thermal and velocity fields to detect the maximum temperature location and the presence of fluid recirculation. The paper is useful to thermal designers.

This paper aims to present a methodology for estimating the risk of component level electrical bridging failures from unattached conductive (tin) whiskers.

Based on experimental data an algorithm was developed and assessed by further experiments. The risk estimate is based on whisker parameters, generated from experiments over a period of time. A bridging failure risk is defined as the probability of a conductive whisker landing between two isolated electrical conductors. A probabilistic estimate for electrical bridging failure risk is achieved by randomly sampling distributions of conductive whisker length, deposition angle, and density for a defined electrical structure. A fine pitch quad flat package attached to a printed wiring board is used as test vehicle to verify the risk estimate.

The estimated risk is found to be higher than planned in the experimental test. The lower experimentally determined risk was found to be the result of high contact resistance between the conductive whisker and the electrical conductors that form the unintended circuit. Contact resistance between the whisker and electrical conductors was found to mitigate the whisker shorting risk.

This is the first attempt to quantify the risk failure due to unattached conductive whiskers in electronic products. A methodology for estimating electrical bridging risk due to unattached conductive whiskers is provided. Contact resistance of conductive whiskers is found to be a critical issue that may be mitigate failure risks.