[en] Intelligent fault recognition has been a hot topic in the area of mechanical fault detection. However, it is difficult to collect sufficient monitoring data to represent the various kinds of faults and support network training. This paper proposes a convolutional adversarial auto-encoder (AE) for mechanical fault recognition with unseen classes via one-class classification. The generator is established based on the convolutional AE, while the discriminator is a multi-scale convolutional neural network. Through unsupervised adversarial training, the model can recognize unseen faults, which are not represented in the training data. The proposed method is verified by three bearing datasets, and some related research is also introduced for comparative analysis. Results show that the RR of the proposed method arrives at 100%, 100% and 97.3% in three cases, while the AC reaches 91.4%, 90.5% and 90.8% respectively. (paper)

[en] In situ continuous particle size measurement is desirable in a variety of industries. Acoustic emission (AE) is a particularly suitable technique to achieve on-line continuous sizing of particles in pneumatic conveying pipelines, which utilizes the AE signals due to the impact of particles with a waveguide protruding into the particle flow. Although early attempts have been made to reveal the relationship between the AE parameters and the particle size, the fundamental sensing mechanism of the AE-based technique for particle size measurement is still not established. For instance, the effect of particle size distribution on several AE parameters remains to be examined. This article aims to gain an in-depth understanding of the AE sensing mechanism for on-line particle sizing by quantifying the parameters of the AE signal, including peak amplitude, count, rise time, duration, energy and root-mean-square value. A theoretical model considering the energy dissipation during plastic impact is also developed to determine the particle size from the AE signal. The proposed method is verified through experimental tests with glass beads on a single-particle test rig. The experimental results obtained indicate that the proposed method is feasible to infer particle size information from the energy of an impact event. The particle size can be measured with a relative error mostly within $\pm <!--\; ??\; -->10\mathrm{\%<!--\; \%\; -->}$ over a range from 0.4 mm to 1.2 mm. (paper)

No abstract available

No abstract available

[en] A simplified quasilinear critical gradient model (QLCGM) for Alfven eigenmode (AE) driven energetic particle (EP) transport is used to treat the intermittency (burstiness) of the transport. Here intermittency is defined as the ratio of the root mean square deviation from mean flow to the mean flow. The model posits that the intermittency results from the micro-turbulent noise induced in the thermal plasma damping rate for the AE. The model is embedded in a time-dependent EP density transport code to generate the transport-flow time traces from typical critical gradient and slowing down density profiles. The QLCGM appears to be in reasonable agreement with the level and characteristics of the intermittency observed in the DIII-D fast particle loss detector flow time traces: high kurtosis [O(40)] of the burstiness with the intermittency O(1–2) increasing with the level of mean flow. (paper)

[en] 3D analysis of an arsenic-doped silicon fin sample is performed in a transmission electron microscope (TEM). High angle annular dark-field scanning TEM (STEM-HAADF) and energy-dispersive x-ray spectroscopy (STEM-EDX) modes are used simultaneously to extract 3D complementary multi-resolution information about the sample. The small pixel size and angular step chosen for the STEM-HAADF acquisition yield reliable information about the sidewall roughness and the arsenic clusters’ average volume. The chemical sensitivity of STEM-EDX tomography gives insights into the 3D conformality of the arsenic implantation and its depth distribution. Non-negative matrix factorization method is employed to identify the chemical phases present in the sample automatically. A total variation minimization algorithm, implemented in 3D, produces high-quality volumes from heavily undersampled datasets. The extension of this correlative approach to electron energy-loss spectroscopy STEM tomography and atom probe tomography is also discussed. (paper)

No abstract available

[en] Full text: Zr is added to the Ti-Nb system for controlling the mechanical properties and the microstructure, and this is due mainly to the suppression of the ω and the martensitic phases. Moreover, the addition of Zr decreases the critical stress for sliding deformation, which increases the deformation induced by the phase transformation, improving the superelastic behavior and the shape memory effect [1]. Zr is considered a neutral element for the stabilization of the β phase, but when added together with a β stabilizing element, such as Nb, it increases the stability of the β phase [2]. The addition of Zr in alloys with a Nb content greater than 20 at. % also produces a more homogeneous structure that improves the corrosion resistance and facilitates the recrystallization processes. Ternary alloys with a nominal chemical composition of Ti_80-x-Nb₂₀-Zr_x (x = 5, 10, 20, 30, and 40 at.%) were produced by electric arc casting. The structure, microstructure, and surface composition of the samples were characterized by X-ray diffraction (XRD), scanning electron microscopy (SEM), X-ray photoelectron spectroscopy (XPS); the elastic modulus was evaluated by an impulse excitation instrument, the hardness, by a microdurometer, and the corrosion resistance, by a potentiostat-galvanostat system coupled with electrochemical impedance. The results show that the addition of Zr causes a decrease in both the elastic modulus and the hardness of the alloys. The XPS analysis revealed that the environment has an effect on the Ti-Nb-Zr surfaces, showing the growth of oxide layers. All alloys with different Zr amounts showed wide passivation regions and no sign of pitting in the potential range that is characteristic of the human body. (author)

[en] Full text: Titanium (Ti) and its alloys have been the best choice for use in hard-tissue replacements, due to their promissory properties. Plasma electrolyte oxidation (PEO) is a surface modification treatment often applied to Ti surfaces, turning possible the surface coating with distinct biocompatible oxides. This study aims to characterize the crystalline structure, phase composition, average crystallite size, and microstrain of biocompatible oxides grown on CP-Ti grade 2 (α-type), Ti-6Al-4V ELI (α+β-type), and Ti-15Zr-15Mo (wt%; β-type) by PEO. PEO treatment was performed in a pulsed power source, at 300 V, 10 min, 40%, 1000 Hz. The electrolyte was composed of 10 g per liter of commercially pure powders (TiO2 , ZrO₂ , and TaOH) and 2 g per liter of KOH. XRD measurements were taken in a Panalytical X'Pert PRO, with monochromatic CuKα radiation (λ = 0.1544 nm), at 40 kV and 30 mA, in the Bragg-Brentano configuration, continuous scan mode, 3 degrees per minute, and step size of 0.02 degrees. The patterns were analyzed by the Rietveld method, using GSAS software with EXPGUI interface, with a standard Y₂O₃ sample for instrumental contribution, and crystallographic data sheets from the ICSD database. The results indicated the presence of a multicomponent coating composed of anatase and rutile TiO2 phases, tetragonal and monoclinic ZrO₂ phases, and orthorhombic Ta₂O₅ phase. The phase proportion, crystalline parameters, average crystalline size, and microstrain were affected by the bulk's chemical and phase compositions. Together with some preliminary characterizations (SEM/EDS, XPS, wettability, and profilometry), the results indicated that the coatings presented interesting properties to coat some metallic biomaterials. (author)

[en] Determination of the thickness of a nanometer thin film is an important topic in the nanostructure characterization field. Currently, the X-ray photoelectron spectroscopy and the ellipsometer have been used to analyze the film thickness. This work proposes another thickness measurement method based on secondary electron (SE) and backscattered electron (BE) yields which are produced by an electron beam with a certain energy. Through comprehensively analyzing a lot of simulated data of the ratio of SE to BE yields at different primary energies and overlayer thicknesses for a variety of overlayer/substrate structures, a semiempirical formula was obtained for the film thickness determination (up to 5 nm). According to the formula, the film thickness can be obtained based on the measured ratio of SE to BE yields. A specific example shows a good agreement is reached between the simulated and experimental measurements in the case of the Be/Pt systems. (author)