Constant Stress Tests Research Articles

Bayesian Approach for Constant-Stress Accelerated Life Testing for Kumaraswamy Weibull Distribution with Censoring

The accelerated life tests provide quick information on the life time distributions by testing materials or products at higher than basic conditional levels of stress such as pressure, high temperature, vibration, voltage or load to induce failures. In this paper, the acceleration model assumed is log linear model. Constant stress tests are discussed based on Type I and Type II censoring. The Kumaraswmay Weibull distribution is used. The estimators of the parameters, reliability, hazard rate functions and p-th percentile at normal condition, low stress, and high stress are obtained. In addition, credible intervals for parameters of the models are constructed. Optimum test plan are designed. Some numerical studies are used to solve the complicated integrals such as Laplace and Markov Chain Monte Carlo methods.

Local Thinning Induced Less Oxide Breakdown in MOS Structures Due to Lateral Non-Uniformity Effect

The downscaling of gate oxide thickness leads to increasing importance of issues on oxide degradation characteristics due to larger operating oxide field and direct tunneling current. For gate oxide thinner than 4.0 nm, the direct tunneling becomes dominant mechanism for carrier transport in MOS capacitors. In this thickness regime, the voltage drop and electric field in oxide layer can be significantly altered by the magnitude of direct tunneling current, and the oxide field does not increase monotonically with decreasing oxide thickness. In addition, the lateral non-uniformity (LNU) effect is a great concern for MOS with ultrathin gate oxide, which may lead to unexpected electrical behavior. In this work, we demonstrate that in certain oxide thickness range (2.5－3.5 nm), the degradation behavior can be altered by introducing LNU in oxide thickness to reduce the effective oxide field in p-type MOSCAP. Two Al/SiO2/p-Si structured MOSCAPs are fabricated for constant voltage stress (CVS) tests. One of the samples (i.e., sample w/ LNU) has partially roughened silicon surface to form non-uniform SiO2 layer as shown in Fig. 1, while another sample (i.e., sample w/o LNU) has not. Fig. 2 and 3 show the I-V characteristics measured after various CVS ranging from +1 V to +4 V for 1000 seconds sequentially for samples with 2.9 and 3.2 nm oxide thickness, respectively. For 2.9 nm-oxide samples, slight increase in inversion current is observed with increasing stress voltage, but the accumulation current maintains even after +4 V CVS. For 3.2 nm-oxide samples, on the contrary, an apparent increase in accumulation current is observed in the sample w/o LNU after +1 V CVS as shown in Fig. 3 (a). Fig. 4 (a) and (b) plot the current-time (I-t) measurements for various stress voltages. It is clearly shown that the breakdown occurs during +1 V stress in the sample w/o LNU. The results are non-intuitive since that the oxide electric field is expected to decrease with thicker oxide, which results in smaller stress field. In addition, the sample w/ LNU appears to possess better oxide reliability compared to the sample w/o LNU in this experiment. For further analysis, the TCAD device simulation tools are applied to obtain the oxide electric field in MOSCAP with various oxide thicknesses as shown in Fig. 5 (a). Two different cases are considered in the simulations, the black square stands for the case including the direct tunneling model (DT) and the red circle stands for the case without DT model. For the case without DT, the voltage drop in oxide layer is approximately fixed for every oxide thickness, which results in inverse-proportionality of oxide electric field to oxide thickness since Eox = Vox/dox . In realistic device, the leakage current by direct tunneling significantly affects the electrical characteristics of the device. For thinner oxide, the high tunneling probability results in more expansion in depletion region and larger silicon band bending, and the oxide voltage drop varies accordingly. As shown in Fig. 5 (b), the depletion width expands in thinner oxide layer, and the expansion is most pronounced for oxide thickness between 3.5 nm and 2.5 nm, leading to the drastic reduction in oxide electric field as shown in Fig. 5 (a). This explains the breakdown behavior in 3.2 nm (w/o LNU) since there exists almost twofold increase in oxide electric field in 3.2 nm-oxide sample compared to 2.9 nm-oxide sample. The strong field transition region (2.5 nm－3.5 nm) also gives an explanation to the reduced degradation behavior in the sample w/ LNU. Fig. 6 illustrates MOSCAP with LNU in oxide layer. For non-uniform oxide, the depletion region expands non-uniformly due to various local tunneling probability. From macroscopic viewpoint, the device can be regarded as a sample with thinner effective oxide thickness. Therefore, the effective electric field shifts toward the lower field region [Fig. 5 (a)]. In summary, the field transition region gives rise to unique breakdown behavior for p-type MOSCAP with oxide thickness between 2.5 nm to 3.5 nm. It is demonstrated that by introducing LNU in oxide thickness, the degradation behavior can be reduced due to lower sustaining field in local region. It is worth noticing that the LNU brings up undesired interface traps and leakage current increase, which must also be carefully considered. Figure 1

Local Thinning Induced Less Oxide Breakdown in MOS Structures Due to Lateral Non-Uniformity Effect

The reliability of MOSCAPs with ultrathin gate oxide is investigated by constant voltage stress tests. The results show that the degradation behavior in gate oxide can be reduced by introducing lateral non-uniformity in oxide thickness to obtain lower sustaining electric field in oxide layer. The strong oxide field transition region between 2.5 nm and 3.5 nm SiO2 thickness is believed to be the origin for the observed degradation behavior, as confirmed by TCAD simulations.

Online state of charge estimation of lithium-ion batteries: A moving horizon estimation approach

Online state of charge (SOC) estimation of lithium-ion batteries (LIBs) relies not only on accurate battery model but also on effective state estimation method. In this study, a nonlinear battery state-space model based moving horizon estimation (MHE) approach is proposed to estimate SOC within the full range. The relationship between SOC and circuit parameters in battery model is captured by polynomial functions. The essential arrival cost in the MHE problem formulation is approximated by the filtering scheme and its covariance matrix is updated by extended Kalman filter (EKF) method. Hybrid pulse power characterization test is first used to guide battery model construction and tuning parameters determination in MHE. The constant current discharge test and dynamic stress test are then used to validate the applicability of the MHE and investigate the performance comparisons between MHE and EKF. The results demonstrate that compared to the EKF, the MHE is less sensitive to the poor initial SOC guesses and has faster convergence to the true SOC. The results thus validate that the MHE provides a potential promising approach to perform accurate, reliable and robust SOC estimation of LIBs.

Adequacy of dual-variable Weibull failure distribution for oil-impregnated paper under pulsating DC voltage

This paper presents an experimental work investigating the failure data of oil-impregnated paper under pulsating DC voltage. The failure data of the oil-impregnated paper specimens were obtained through constant voltage stress tests; the Weibull distribution was used to estimate the statistical distribution of failure data. The exponential function (EF) and the inverse power function (IPF) were used to modify the Weibull distribution of failure data, so as to obtain two dual-variable Weibull failure distributions (WFDs), which were used to analyze failure data of oil-paper specimens obtained under pulsating DC and individual DC and AC voltages. Their adequacies were evaluated by using failure data obtained through the constant stress tests under various voltages. Both EF-modified and IPF-modified WFDs show more adequacy for failure data of oil-paper insulation under DC and pulsating DC voltages than AC voltages. The EF-modified WFD was more adequate than IPF-modified WFD in estimating the statistical distribution of failure data.

Partial Discharge Characteristics in LLDPE-Natural Rubber Blends: Correlating Electrical Quantities with Surface Degradation

Partial discharges (PD) lead to the degradation of high voltage cables and accessories. PD activities occur due to the existence of impurities, voids, contaminants, defects and protrusions during the manufacture and installation of power cables. Commonly, insulation failures occur at cable joints and terminations, caused by inhomogeneous electric field distributions. In this work, a blend of natural rubber (NR) and linear low density polyethylene (LLDPE) was investigated, and the optimal formulation of the blend that could resist PD was discussed. The experiments were conducted under a constant high voltage stress test of 6.5 kV AC and the magnitude of partial discharge activities was recorded using the CIGRE method II. Pattern analysis of PD signals was performed along with the interpretation of morphological changes. The results showed that the addition of 10 wt% of NR and 5 wt% of Alumina Trihydrate (ATH) provided promising results in resisting PD activities. However, as the NR content increased, more micropores existed, thus resulting in increased PD activities within the samples.

Experimental Demonstration of Time-Dependent Breakdown in GaN-Based Light Emitting Diodes

This letter reports the first experimental demonstration of time-dependent breakdown in GaN-based light-emitting diodes (LEDs); based on a number of constant voltage stress tests, carried out below the breakdown limit of the devices, we show that: 1) when submitted to reverse-bias stress, the LEDs show a gradual increase in current, well correlated with an increase in the breakdown luminescence signal; 2) for sufficiently long stress times, the LEDs can reach a catastrophic (sudden) breakdown, which leads to the failure of the devices; 3) the breakdown process is time-dependent and the time to failure (TTF) has an exponential dependence on stress voltage; and 4) TTF is Weibull distributed. The results presented within this letter demonstrate that the GaN-based heterostructures can show a TDDB-like behavior, and can be useful for the interpretation of the degradation data of LEDs and HEMTs.

Fatigue Assessment of Cement-Treated Base for Roads: An Examination of Beam-Fatigue Tests

AbstractThis paper presents an examination of the fatigue characteristics of cement-treated base (CTB), a cement-stabilized material for use in road pavement construction based on CTB beam test specimens prepared and tested under laboratory conditions. The beam fatigue protocol for asphalt concrete, following the Austroads standard test method of AG:PT/T233, were adapted to examine the fatigue characteristics of the CTB test specimens. Standard road base (Granite/Diorite) from the local quarry in Western Australia was employed as the parent material of CTB. The beam-fatigue tests were conducted under the strain-controlled (constant strain) and stress-controlled (constant stress) testing conditions with varying cement contents (3–10%) of CTB test specimens. Two different loading waveforms of sinusoidal and haversine load curves were also investigated. The results show that for the strain-controlled testing condition, the fatigue characteristics of the CTB test specimens were not affected by the difference ...

Strength and post-peak response of Colorado shale at high pressure and temperature

The paper presents the thermo-mechanical strength characterization of Colorado shale which serves as a cap rock to underlying oil-rich formations in Cold Lake area, east-central Alberta, Canada. Samples from the Westgate and Joli Fou formations were tested at high temperatures along triaxial compression stress paths with effective consolidation pressures and temperatures consistent with in-situ stresses and operative conditions. These lab tests were complemented with indirect tensile strength, unconfined compression and triaxial constant effective mean stress tests at ambient temperature for a more comprehensive knowledge of the material strength, particularly, at low confining pressures. The lab experimental results revealed brittle behavior of the tested shales with considerable strain softening. Both peak and post-peak strength envelopes of the tested material were observed to be nonlinear with a pronounced curvature at low mean stresses and to decrease at high temperatures. A simple thermo-mechanical failure criterion is proposed to capture the aforementioned constitutive features. This provides a modeling framework for determining the material strength close to the zone of tensile failure at different temperatures which is of utmost relevance to cap rock integrity assessments. Finally, measured axial and radial deformations as well as X-ray tomography images were used to verify that the observed constant rate of dilation in the post-peak response of the samples was a reflection of the morphology of the virtually smooth shear plane, and not representative of the material behavior.

A study of trap-limited conduction influenced by plasma damage on the source/drain regions of amorphous InGaZnO TFTs

This study investigated electrical characteristics and stability variations of amorphous indium gallium zinc oxide thin film transistors (a-IGZO TFTs) with plasma damage on their source/drain (S/D) regions. The influence of the plasma damage on the TFT performance is absent as the channel length is 36–100 μm. When the channel length is decreased to 3–5 μm, the mobility () of the bottom gate TFT (BG TFT) with plasma damage is significantly degraded to 0.6 cm2 (V s)−1, which is much lower than 4.3 cm2 (V s)−1 of a damage-free BG TFT. We utilized the TFT passivation layer and the indium tin oxide (ITO), which was used as the pixel electrode material in the TFT backplane, to be the top gate insulator and top gate electrode of the defective BG TFT to obtain the defective dual-gate TFT. The mobility can be restored to 5.1 cm2 (V s)−1. Additional process steps are not required. Besides, this method is easily implemented and is fully compatible with TFT backplane fabrication process. The transfer curves, hysteresis characteristics, stabilities under constant voltage stress and constant current stress tests were measured to give evidences that the traps created by the plasma damage on the S/D regions indeed can affect electron transport. This trap-limited conduction can be improved by using the top gate. It was proven that the top gate was not for contributing an observably additional current. It was for inducing electrons to electrically passivate the plasma-induced defects near the back channel. Thus, the trapping/detrapping of the electrons transporting in the front channel can be reduced. The trap density near the Fermi level, hopping distance and hopping energy are 1.1 × 1018 cm−3 eV−1, 162 Å, and 52 meV for the BG TFT with plasma damage on the S/D regions.

Changing anisotropy of G0 in Hostun sand during drained monotonic and cyclic loading

The inherent and stress-induced anisotropy of saturated Hostun sand has been examined by means of shear wave velocity measurements under drained monotonic and cyclic triaxial conditions. Tests were carried out on samples prepared by moist tamping for a wide range of initial densities in a fully instrumented stress-path triaxial apparatus equipped with bender element instrumentation in three directions. Data have been normalised to account for the variations in void ratio and mean effective stress. The evolution of shear stiffness under anisotropic stress conditions has been explored by performing both constant mean effective stress tests and constant radial stress tests in compression and extension. Some samples were subsequently subjected to multiple drained deviatoric loading cycles, and the evolution of shear stiffness was examined. The tests showed that the initial inherent anisotropy was small, namely, lower than that previously observed in dry pluviated samples of the same sand in a cubical cell apparatus. Changing stress levels in compression or extension resulted in stress-induced anisotropy even when there were no changes in the values for p′, indicating that the widely-used empirical expressions for G0 under isotropic stress states are not strictly applicable under anisotropic stress conditions. The comparison between stress-induced anisotropy in compression and in extension is discussed, and the evolution of shear stiffness during drained cyclic loading is examined. It is concluded that the changes in inherent anisotropy during these tests were small, but that new constitutive models should account for such changes in state with appropriate normalization.

The Role of Dielectric Heating and Effects of Ambient Humidity in the Electrical Breakdown of Polymer Dielectrics

Polymer-based dielectric materials have potential applications in microelectronics, power electronics, photovoltaics, flexible electronics, MEMS, and sensing industries. The possibility of premature electrical breakdown due to high electric fields, particularly at high frequencies and in high ambient temperature and humidity conditions, has restricted its widespread adoption. In this paper, we generalize the thermochemical model of dielectric breakdown and establish dielectric heating as the primary ac degradation mechanism in polymers and develop an analytical dielectric breakdown model that satisfactorily explains measured trends in constant and ramp stress tests under both ac and dc electric fields applied to 5- $\mu\mbox{m}$ -thick PBO capacitors. We also study and quantify the effects of exposure to ambient relative humidity on the electrical breakdown lifetime of polymer dielectrics. Our study provides a fundamental physical understanding of the frequency, ambient humidity, and thickness dependencies of lifetime and breakdown strength for polymer dielectrics; the proposed breakdown model suggests far more optimistic dielectric lifetimes when accelerated test results are scaled to normal operating conditions.

SSUH model: A small-strain extension of the unified hardening model

A small strain unified hardening (SSUH) model is proposed in the present study to tackle the small strain behavior of clay. The model is an extension of the unified hardening (UH) model for overconsolidated (OC) clays accounting for the small strain stiffness. The new features of the SSUH model over the UH model include: (a) a new elastic hysteretic stress-strain relationship to evaluate the stiffness degradation at small strains and to generate the hysteresis loop under cyclic loading condition; (b) a revised unified hardening parameter to enhance the plastic stiffness at small strains; and (c) a new overconsolidation parameter, which is crucial to make the UH model working with the elastic hysteretic stress-strain relationship effectively. With these enhancements, the SSUH model can describe a high initial stiffness and the highly nonlinear stress-strain relationship at small strains, in addition to the shear dilatancy and strain hardening/softening behaviors of OC clays at large strains. In comparison with the Modified Cam-clay (MCC) model, the proposed model needs two more small strain related parameters, which can be easily obtained from laboratory tests. Finally, some drained triaxial compression tests at large strains, drained triaxial compression/extension tests at small strains, an undrained compression test at small strains and a drained cyclic constant radial stress test are employed to validate the new model.

Bayesian Estimation of Exponentiated Weibull Distribution Under Partially Acceleration Life Tests

This paper presents statistical methods for analyzing partially accelerated life test from constant-stress test. The lifetime of items under use condition follows the two-parameter exponentiated Weibull distribution. Based on progressively Type-II censored sample, the classical maximum likelihood method as well as a fully Bayesian method based on Lindley (Trabajos de Estadistica 31:223–237, 1980) approximation form and the Markov chain Monte Carlo technique are developed for inference about model parameters and acceleration factor. Furthermore, approximate confidence intervals and credible intervals are presented. A Monte Carlo simulation is given to study the precision of different estimators.

Extensive Investigation of Time-Dependent Breakdown of GaN-HEMTs Submitted to OFF-State Stress

This paper reports the experimental demonstration of time-dependent dielectric breakdown in GaN-based high-electron mobility transistors (HEMTs) submitted to OFF-state stress. Based on combined breakdown measurements, constant voltage stress tests, and 2-D simulations, we demonstrate the following relevant results. First, GaN-based HEMTs with a breakdown voltage higher than 1000 V (evaluated by dc measurements) may show time-dependent failure when exposed to OFF-state stress with $V_{\mathrm {DS}}$ in the range 600–700 V. Second, time-to-failure (TTF) is Weibull-distributed, and has an exponential dependence on the stress voltage level. Third, time-dependent breakdown is ascribed to the failure of the SiN dielectric at the edge of the gate overhang, on the drain side. Fourth, 2-D simulations confirm that—in this region—the electric field exceeds 6 MV/cm, i.e., the dielectric strength of SiN. Finally, we demonstrate that by limiting the electric field in the nitride through epitaxy and process improvements, it is possible to increase the TTF by three orders of magnitude.

Improving the performance of Ground Coupled Heat Exchangers in unsaturated soils

Ground heat exchangers (GHE) have serious limitations imposed upon them when installed in soils having low thermal conductivity and diffusivity. The restrictions to heat flow are: pipe surface area, the formation low thermal diffusivity and unsaturated soils moisture migration due to elevated temperatures. A comparative theoretical analysis has been conducted to ascertain the feasibility of a ‘Membrane Conduction Augmentation System’ (MCAS) to improve performance relative to a conventional Horizontal Ground Coupled Heat Exchanger (HGCHE). Results show that significant reductions in HGCHE length for equivalent cooling could be obtained in analyses both cyclic (simulating intermittent air-conditioning usage) and 10 day constant heat transfer stress test. Most significantly, the lower the soil's thermal conductivity and diffusivity the higher the efficiency of the proposed system. The MCAS simulated in the same size trench (300mm) as that of a conventionally constructed ‘direct burial’ HGCHE yielded efficiency 46% higher for the 10 day stress test, while the 1200mm MCAS improved by 150%. Both simulations used soil thermal conductivity of 0.54W/m·K. Reductions in interface soil temperatures for the MCAS corresponded inversely to the size of the MCAS.

Capacity Fade Estimation in Electric Vehicle Li-Ion Batteries Using Artificial Neural Networks

In this paper, an artificial neural network (ANN) based approach is proposed to estimate the capacity fade in lithium-ion (Li-ion) batteries for electric vehicles (EVs). Besides its robustness, stability, and high accuracy, the proposed technique can significantly improve the state-of-charge (SOC) estimation accuracy over the lifespan of the battery, which leads to more reliable battery operation and prolonged lifetime. In addition, the proposed technique allows accurate prediction of the battery remaining service time. Two identical 3.6-V/16.5-Ah Li-ion battery cells were repeatedly cycled with constant current and dynamic stress test current profiles at room temperature, and their discharge capacities were recorded. The proposed technique shows that very accurate SOC estimation results can be obtained provided enough training data are used to train the ANN models. Model derivation and experimental verification are presented in this paper.

Parameter Estimation Under Failure-Censored Constant-Stress Life Testing Model: A Bayesian Approach

This article compares likelihood and Bayesian estimations for partially accelerated constant-stress life test model under type II censoring assuming Pareto distribution of the second kind. Both maximum likelihood and Bayesian estimators of the model parameters are derived. The posterior means and posterior variances are obtained under the squared error loss function using Lindley's approximation procedure. The advantages of this proposed procedure are shown. Monte Carlo simulations are conducted under different samples sizes and different parameter values to assess and compare the proposed methods of estimation. A noninformative prior on the model parameters is used to make the comparison more meaningful. It has been observed that Lindley's method usually provides posterior variances and mean squared errors smaller than those of the maximum likelihood estimators. That is, Lindley's method produces improved estimates, which is an advantage of this method.

Thickness and temperature dependences of the degradation and the breakdown for MgO-based magnetic tunnel junctions

The reliability of a magnetic tunnel junction (MTJ) with an MgO tunnel barrier was evaluated. In particular, various voltage tests were used to investigate the effects of the barrier thickness and the temperature on the resistance drift. We compared the resistance change during a constant voltage stress (CVS) test and confirmed a trap/detrap phenomenon during the interval stress for different barrier thicknesses and temperatures. The resistance drift representing degradation and the time to breakdown (TBD) representing the breakdown characteristic were better for a thicker barrier and lower temperature, but were worse for a thinner barrier and a higher temperature. The results suggest that breakdown and degradation due to trap generation strongly depend on both the barrier thickness and the temperature. Furthermore, as the TBD varies at steady rates with changing barrier thickness, temperature, and electric field, we assume that a MTJ with an adnormal thin layer of MgO can be screened effectively based on the predicted TBD. As a result, the barrier thickness and the temperature are very important in determining the reliability of a MTJ, and this study is expected to be helpful in understanding the degradation and the breakdown of a MTJ.

Positive-Bias Stress Test on Amorphous In–Ga–Zn–O Thin Film Transistor: Annealing-Temperature Dependence

We performed constant positive gate bias stress tests on sputter-deposited amorphous In-Ga-Zn-O (a-IGZO) thin-film transistors (TFTs) fabricated with various annealing conditions. We found that the time evolution of threshold voltage shift (ΔV <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">th</sub> ) during the stress test exhibits two different behaviors; a power function type and a logarithmic function type. Combining with thermal desorption spectroscopy (TDS), we found that the ΔV <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">th</sub> type changes from the log-function type to the power-function type as H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> desorption increases. The power-function type is attributed to a H-diffusion limited process. As for the log-function type, the magnitude of ΔV <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">th</sub> has correlation with the hysteresis width in a first transfer curve and the desorption amount of H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> O and O <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> species between 200 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">°</sup> C-300 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">°</sup> C. Hence, it is considered that the traps causing the log-function type ΔV <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">th</sub> are meta-stable states related to O- and/or OH-related weak bonds.