Remaining useful life prediction for degrading systems with random shocks considering measurement uncertainty

Degradation data have been widely used for the remaining useful life (RUL) prediction of systems. Most existing works apply a preset model to capture the degradation process and focus on the degradation process without shocks or constant shock effects. More generally, the actual degradation path is unobservable due to the existence of measurement uncertainty, which interferes with the determination of the degradation model. Besides, the effect of random shocks is usually fluctuating. Given these problems, a general degradation model with the random shock fluctuant effects considering the measurement uncertainty is first developed to describe the degradation process, and a two-step approach combining the arithmetic average filter and the Bayesian information criterion is adopted to identify the degradation path. Subsequently, the transfer processes of the actual degradation state and the abrupt change caused by shocks are depicted using a two-dimensional state-space model, and an expectation-maximization algorithm combined with the particle filtering is developed for parameter estimation. Furthermore, the explicit solution of RUL distribution is obtained when only considering harmful shocks, while a simulation method of RUL distribution is provided when both harmful and beneficial shocks exist. Finally, the effectiveness of the proposed method is verified by a numerical example and two practical case studies.     

Diffusion behaviors and atomic mobilities in Mg–Sc hcp and bcc alloys: Investigation via single-phase and multi-phase diffusion couples

Diffusion behaviors in Mg–Sc hcp and bcc solid solutions between 773 and 873 K were investigated using both single-phase and multi-phase diffusion couple techniques. The EPMA detected composition-distance profiles were smoothed and fitted using the error function expansion (ERFEX). The interdiffusion coefficients were extracted using Sauer–Freise integral. The interdiffusion coefficients in hcp phase showed a slightly parabolic composition dependence at the Mg-rich part and the maximum value was around 2–3 at. % Sc. However, the interdiffusion coefficients in the bcc phase monotonously decreased with the increase of solubility of Sc. The determined inter- and impurity diffusion coefficients in the hcp Mg–Sc alloys were assessed to develop the atomic mobility database, and their validity was justified by reproducing the composition profiles and diffusion fluxes obtained in this diffusion couple experiment. Meanwhile, the development of bcc atomic mobility was realized via the Maclaurin approximation, extrapolation, and optimization. The results make up for the missing data of Mg–Sc diffusion kinetics.     

Systematic study of short circuit activation on the performance of PEM fuel cell

During the operation of proton exchange membrane fuel cell (PEMFC), it always suffers from reversible performance loss caused by the oxidation of platinum catalyst on its electrode, which reduces the electrochemical active surface area. Short circuit method has been found to improve the performance of fuel cells by stripping of oxides and other adsorbed species from platinum, which needs systematical understanding the effective parameters of short circuit method on fuel cell performance. In this paper, the effects of different short circuit activation parameters (duration, interval, cycles, cut-off voltage, operating current) are carefully studied and analyzed during short circuit operations. In addition, the mechanism revealing how relevant parameters influence short circuit activations is deeply analyzed. The results show that five groups of activation parameters have obvious influence on the activation of fuel cell, indicating that the short-circuit activation effect can be optimized. Among these parameters, the short-circuit duration parameter have the greatest impact on activation, because the platinum hydroxides and oxides is gradually removed during short-circuit duration and results in a larger effective surface area of the platinum catalyst for the electrochemical reaction. However, the smallest impact is short-circuit interval. Another finding is that the five activation parameters are not independent, so the optimal activation parameter value needs to be analyzed in combination with the operating conditions. Finally, according to the activation principle, selection of appropriate short circuit activation parameters for application are proposed to further improve performance and fuel utilization by considering the safety of the stack.     

40SiO2–40P2O5–20ZrO2 sol-gel infiltrated sSEBS membranes with improved methanol crossover and cell performance for direct methanol fuel cell applications

Membranes commonly used in direct methanol fuel cell (DMFC) are expensive and show a great permeability to methanol which reduces fuel utilization and leads to mixed potential at the cathode. In this work, sulfonated styrene-ethylene-butylene-styrene (sSEBS) modified membranes with zirconia silica phosphate sol-gel phase are developed and studied in order to evaluate their potential use in DMFC applications. The synthesized hybrid membranes and sSEBS are subjected to an exhaustive physicochemical characterization by liquid uptake, ion exchange capacity, atomic force microscopy, X-ray photoelectron spectroscopy and dynamic mechanical and thermogravimetric analyses. Likewise, the potential use of the prepared membranes in DMFC is evaluated by means of electrochemical characterizations in single cell, determining the limiting methanol crossover current densities, proton conductivities and DMFC performances. The hybrid membranes show lower water and methanol uptakes, higher stiffness, water retention capability, upper power density and lower methanol crossover than sSEBS and Nafion 112.     

Transient shrinkage behavior of wustite-centered binary/ternary/quaternary/quinary-component oxide packed beds in a reducing atmosphere up to 1773 K

Wustite (FeO)-centered multicomponent oxides play an important role in the ironmaking process, and a complete understanding of their high temperature behaviors is of great importance for process optimization to achieve high efficiency and low emissions during industrial production. In this work, the transient shrinkages of FeO-centered multicomponent oxide packed beds are quantitatively determined in a reducing atmosphere up to 1773 K, and the effects of the interactions between the oxides on the shrinkage rate (SR) are qualitatively evaluated. The results show that although mixing CaO with FeO increases the SR to 0.42%/K below 1173 K, further mixing with SiO₂ or Al₂O₃ significantly limits this enhancement effect due to the formation of an olivine or spinel phase. However, in the subsequent stage, the SR increases to as high as 0.44%/K after CO is injected. The interaction between FeO and MgO leads to an SR of greater than 0.20%/K at lower temperatures, but it causes a decrease in the SR from 0.33%/K to 0.16%/K between 1173 K and 1273 K. Meanwhile, adding SiO₂ slows the reduction reaction, and the SR correspondingly decreases further to 0.04%/K. On the other hand, the interaction between CaO and Al₂O₃ takes precedence over the interaction between SiO₂ and MgO and dominates the shrinkage process in the quinary-component case, and the preferentially formed CaAl₂O₄ spinel phase hinders the formation of the Mg₂SiO₄ olivine phase.     

A novel predictive model for mechanical behavior of single-lap GFRP composite bolted joint under static and dynamic loading

Mechanical connection of composite is critical due to its complicated meso-structure and failure mode, which has become a bottleneck on reliability of composite material and structure. Although many researches on composite bolted joints have been carried out, the theory and experiment on mechanical behavior of such a joint structure under dynamic loading were rarely reported. Here, we propose a novel predictive model for quasi-static and dynamic stiffness of composite bolted joint by introducing the strain rate dependent elastic modulus into the mass spring model. Combined with the composite laminate theory and Tsai-Hill theory, the present model was capable of predicting the strain rate dependent stiffness and strength of the composite bolted joint. Quasi-static and impact loading experiments were carried out by Zwick universal hydraulic testing machine and split Hopkinson tension bar, respectively. The stiffness and strength predicted by our model showed good accordance with the experiment data with errors below 12% under quasi-static loading and below 30% under impact loading. The results indicated that under impact loading, stiffness and strength of the composite bolted joint were significantly higher than their quasi-static counterparts, while the failure mode of the joint structure trended towards localization which was mainly bearing failure. Among various lay-up ratios studied, the optimal lay-up ratio for quasi-static and dynamic stiffness was 0:±45:90 = 3:1:1.     

Fiber-continuous panel–pillar structure in insect cuticle and biomimetic research

Scanning electron microscope (SEM) observation shows that the cuticle of Dorcus titanus is a kind of natural sandwich structure consisting of upper and lower panels and middle pillars. The observation also shows that the material of the sandwich structure is a biocomposite consisting of chitin-fiber layers and sclerous-protein matrix. More careful observation shows that the fiber layers in the sandwich structure continuously join the panels and the pillars to form a fiber-continuous panel–pillar sandwich structure. The strength of the fiber-continuous panel–pillar structure is investigated and compared with that of the non-fiber-continuous panel–pillar structure based on their representative models. It is shown that the fiber-continuous panel–pillar structure has higher ultimate strength compared to that of the non-fiber-continuous panel–pillar structure. Based on the observations and analyses, the fiber-continuous panel–pillar structure is biomimetically fabricated with a special mould and process. The ultimate strength of the structure is tested and compared with that of the non-fiber-continuous panel–pillar structure. It is indicated that the ultimate strength of the fiber-continuous panel–pillar structure is distinctly larger than that of the non-fiber-continuous panel–pillar structure.     

Effects of annealing on copper substrate surface morphology and graphene growth by chemical vapor deposition

Understanding the mechanism of graphene synthesis by chemical vapor deposition and the effect of process parameters is critical for production of high-quality graphene. In the present work, we investigated the effect of H₂ concentration during annealing on evolution of Cu surface morphology, and on deposited graphene characteristics. Our results revealed that H₂ had a smoothening effect on Cu surface as its surface roughness was reduced significantly at high H₂ concentration along with the formation of surface facets, dents and nanometer-sized particles. Furthermore, H₂ content influenced the graphene morphology and its quality. A low H₂ concentration (0% and 2.5%) during annealing promoted uniform and good quality bilayer graphene. In contrast, a high concentration of H₂ (20% and 50%) resulted in multilayer, non-uniform and defective graphene. Interestingly, the annealed Cu surface morphology differed considerably from that obtained after deposition of graphene, indicating that graphene deposition has its own impact on Cu surface.     

Amorphous carbon nitride thin films for electrochemical electrode: Effect of molecular structure and substrate materials

We report on the effect of molecular structure and substrate material on amorphous carbon nitride (a-CN:H) electrode properties including film adhesion to the substrate and electrochemical properties. Films were prepared by neutral beam enhanced chemical vapor deposition on different substrate materials (p-type Si, Cu, Ti, and Pt) below room temperature. When depositing on Si, doping nitrogen into carbon improved the electrochemical properties despite weak adhesion to the substrate. Nitrogen in a-CN:H formed two different bonding configurations: incorporation into aromatic carbon rings and hydrogen nitride by infrared (IR) spectroscopy. Therefore, delocalization of π bonds by incorporation of nitrogen affected the electrochemical improvement of the a-CN:H electrode. For samples deposited on a different metal substrate, the adhesion to substrate increased as a function of decreasing oxygen concentration on the metal substrate surface; the Pt substrate performed well with no delamination in our evaluation. The electrochemical properties were improved only in the case of deposition on Pt. Moreover, Pt surface modification by hydrogen beam was also effective; consequently, the electrochemical property of the a-CN:H electrode was superior to the graphite electrode with high temperature annealing. The observed increases in IR spectra of aromatic clusters were in line with the electrochemical improvements of a-CN:H.