Search

Random-based learning paradigms exhibit efficient training algorithms and remarkable generalization performances. However, the computational cost of the training procedure scales with the cube of the number of hidden neurons. The paper presents a novel training procedure for random-based neural networks, which combines ensemble techniques and dropout regularization. This limits the computational complexity of the training phase without affecting classification performance significantly; the method best fits Internet of Things (IoT) applications. In the training algorithm, one first generates a pool of random neurons; then, an ensemble of independent sub-networks (each including a fraction of the original pool) is trained; finally, the sub-networks are integrated into one classifier. The experimental validation compared the proposed approach with state-of-the-art solutions, by taking into account both generalization performance and computational complexity. To verify the effectiveness in IoT applications, the training procedures were deployed on a pair of commercially available embedded devices. The results showed that the proposed approach overall improved accuracy, with a minor degradation in performance in a few cases. When considering embedded implementations as compared with conventional architectures, the speedup of the proposed method scored up to 20× in IoT devices.

Thermal, electrical and mechanical stresses age the electrical insulation systems of high voltage (HV) apparatuses until the breakdown. The monitoring of the partial discharges (PDs) effectively assesses the insulation condition. PDs are both the symptoms and the causes of insulation aging and—in the long term—can lead to a breakdown, with a burdensome economic loss. This paper proposes the convolutional neural networks (CNNs) to investigate and analyze the aging process of enameled wires, thus predicting the life status of the insulation systems. The CNNs training does not require any kind of assumption of how the factors (e.g., voltage, frequency and temperature) contribute to the life model. The experiments confirm that the proposal obtains better estimations of the life status of twisted pair specimens concerning existing solutions, which are based on strong hypotheses about the life model dependency on the factors.

Embedding the ability of sentiment analysis in smart devices is especially challenging because sentiment analysis relies on deep neural networks, in particular, convolutional neural networks. The paper presents a novel hardware-friendly detector of image polarity, enhanced with the ability of saliency detection. The approach stems from a hardware-oriented design process, which trades off prediction accuracy and computational resources. The eventual solution combines lightweight deep-learning architectures and post-training quantization. Experimental results on standard benchmarks confirmed that the design strategy can infer automatically the salient parts and the polarity of an image with high accuracy. Saliency-based solutions in the literature prove impractical due to their considerable computational costs; the paper shows that the novel design strategy can deploy and perform successfully on a variety of commercial smartphones, yielding real-time performances.

Video-based grasp classification can enhance robotics and prosthetics. However, its accuracy is low when compared to e-skin based systems. This paper improves video-based grasp classification systems by including an automatic annotation of the frames that highlights the joints of the hand. Experiments on real-world data prove that the proposed system obtains higher accuracy with respect to the previous solutions. In addition, the framework is implemented on a NVIDIA Jetson TX2, achieving real-time performances.

With the growth of pervasive electronics, the availability of compact digital circuitry for the support of data processing is becoming a key requirement. This paper tackles the design of a digital architecture supporting the $n -$mode tensormatrix product in fixed point representation. The design aims to minimize the resources occupancy, targeting low cost and low power devices. Tests on a Kintex-7 FPGA confirm that the architecture leads to an efficient digital implementation, which can afford real-time performances on benchmark applications with power consumption lower than 100mW.

Edge computing can take full advantage of data-driven models only if the eventual inference function can be deployed on low-power, resource-constrained digital devices. In this regard, single layer feedforward neural networks (SLFNs) based on the threshold activation function represent a suitable option, as they can suitably balance generalization performances and computational costs. Their robustness to perturbations at the node level, though, is an issue: a small perturbation affecting the input to the activation might result in a sign inversion at the output of the neuron. This in turn may severely affect the accuracy of the inference function when implemented on hardware. This paper shows that the robustness of this class of SLFNs can be improved by introducing in the cost function a regularization term specifically designed to limit the impacts of perturbations at the node level. The novel cost function indeed admits a closed-form solution. Experimental validation involved six real world benchmarks.

Providing user customized experience is one of the main goals for present-day electronic smart devices. Image polarity detection plays a crucial role in understanding users’ preferences due to the fact that information is massively represented by means of pictures. State-of-the-art frameworks are based on deep learning networks and continue evolving adding sophisticated structures to enhance generalization performances of the inference systems. Recent works proved that image analysis can be enhanced exploiting the information about salient regions. However, better performances are obtained at the cost of a higher computational load. This paper presents a hardware-friendly deep learning framework for image polarity detectors based on salient regions of an image. Experimental results show the reliable performances of the proposed solution on real-world data.

Image polarity detection opens new vistas in the area of pervasive computing. State-of-the-art frameworks for polarity detection often prove computationally demanding, as they rely on deep learning networks. Thus, one faces major issues when targeting their implementation on resource-constrained embedded devices. This article presents a design strategy for convolutional neural networks that can support image-polarity detection on edge devices. The outcomes of experimental sessions, involving standard benchmarks and a pair of commercial edge devices, confirm the approach suitability.

The online monitoring of a high voltage apparatus is a crucial aspect for a predictive maintenance program. The insulation system of an electrical machine is affected by partial discharges (PDs) phenomena that—in the long term—can lead to the breakdown. This in turn may bring about a significant economic loss; wind turbines provide an excellent example. Thus, it is necessary to adopt embedded solutions for monitoring the insulation status. This paper introduces an online system that exploit fully unsupervised methodologies to assess in real-time the condition of the monitored machine. Accordingly, the monitoring process does not rely on any prior knowledge about the apparatus. Nonetheless, the proposed system can identify the relevant drifts in the machine status. Notably, the system is designed to run on low-cost embedded devices.

The deployment of connectionist models on resource-constrained, low-power embedded systems brings about specific implementation issues. The paper presents a design strategy, aimed at low-end reconfigurable devices, for implementing the prediction operation supported by a single hidden-layer feedforward neural network (SLFN). The paper first shows that a considerable efficiency can be obtained when hard-limiter thresholding operators support the activation functions of the neurons. Secondly, the analysis highlights the advantages of using random basis networks, thanks to their limited memory requirements. Finally, the paper presents a pair of different architectural approaches to the effective support of SLFNs on CPLDs and low-end FPGAs. The alternatives differ in the specific trade-off strategy between area utilization and latency. Experiments confirm the effectiveness of both schemes, yielding a pair of viable implementation options for satisfying the respective constraints, namely effective area utilization or low latency.

The availability of compact digital circuitry for the support of neural networks is a key requirement for resource-constrained embedded systems. This brief tackles the implementation of single hidden-layer feedforward neural networks, based on hard-limit activation functions, on reconfigurable devices. The resulting design strategy relies on a novel learning procedure that inherits the approach adopted in the Extreme Learning Machine paradigm. The eventual training process balances accuracy and network complexity effectively, thus supporting a digital architecture that prioritizes area utilization over computational performance. Experimental tests confirm that the design approach leads to efficient digital implementations of the predictor on low-performance devices.

The development of automated tools capable of monitoring electric motors is important for industrial applications. The measure of partial discharges is one of the most prominent methodologies for the evaluation of electric motors operating conditions. This work proposes to apply tensor-based classification methods to discriminate between different partial discharges sources. The major advantage of tensor-based approaches relies on the fact that they can inherently handle partial discharges pattern using all the geometrical information embedded in the patterns. Secondly, noise filtering can be embedded in classification mechanism, without manual tuning. As a consequence, better performances are obtained in presence of additive sources of noise like 4G, GSM connections and electromagnetic sources that could be detected by conducted and irradiated sensors especially during an online monitoring. To demonstrate the ability of the tensor-based methods to overcome the noise presence, tests involved defects with different levels of addictive noises. Experimental results proved the effectiveness of the proposed approach.

Video-based hand grasp analysis can support both robotics and prosthetics. Indeed, computational aspects represent a major issue, as hand grasp analysis is expected to support grasping systems that are hosted on low-power embedded systems. This paper proposes a framework for video-based grasping classification that is designed for implementation on resource-constrained devices. The framework adopts a fully data-driven strategy and relies on deep learning to deal with advanced analysis of video signals. Nonetheless, the overall design takes advantage of CNN architectures that can cope with the constraints imposed by embedded systems. The experimental session involved a real-world dataset containing daily life activities collected using egocentric perspective. In addition, the complete inference system is implemented on a NVIDIA Jetson-TX2 obtaining real time performances. The results confirm that the proposed system can suitably balance the trade off between accuracy and computational costs.

Embedded machine learning relies on inference functions that can fit resource-constrained, low-power computing devices. The literature proves that single layer neural networks using threshold functions can provide a suitable trade off between classification accuracy and computational cost. In this regard, the number of neurons directly impacts both on computational complexity and on resources allocation. Thus, the present research aims at designing an efficient pruning technique that can take into account the peculiarities of the threshold function. The paper shows that feature selection criteria based on filter models can effectively be applied to neuron selection. In particular, valuable outcomes can be obtained by designing ad-hoc objective functions for the selection process. An extensive experimental campaign confirms that the proposed objective function compares favourably with state-of-the-art pruning techniques.

The online monitoring of a high voltage apparatus is a crucial aspect for a predictive maintenance program. The insulation system of an electrical machine is affected by partial discharges (PDs) phenomena that—in the long term—can lead to the breakdown. This in turn may bring about a significant economic loss; wind turbines provide an excellent example. Thus, it is necessary to adopt embedded solutions for monitoring the insulation status. This paper introduces an online system that exploit fully unsupervised methodologies to assess in real-time the condition of the monitored machine. Accordingly, the monitoring process does not rely on any prior knowledge about the apparatus. Nonetheless, the proposed system can identify the relevant drifts in the machine status. Notably, the system is designed to run on low-cost embedded devices.

The purpose of this work is to analyze how the presence of combined AC-DC voltages can influence the partial discharge (PD) activity. A DC offset can be introduced by power electronics components affecting the partial discharge inception voltage (PDIV). If the PD phenomenon is the main cause of the electrical insulation degradation, the lifetime can be influenced by the presence of a DC component in the voltage waveform. In this work a simplified configuration has been considered in order to analyze the influence of the DC offset on the PD behavior. Twisted Pair samples have been taken into account representing surface PD activity conditions. The enameled wires have been subjected to PD activity after a conditioning in oven at 150 °C. The effect introduced by the DC continuous component offset is considerable. The number of discharges increases and the level of inception voltage decreases proportionally with the decrease of the continuous component, consequently the lifetime values vary since they are influenced by the PD activity. It has been possible to observe how the DC component influences the PD activity and the relevant degradation processes. This preliminary work suggests to perform exhaustive studies to better understand how the presence of a AC-DC voltage waveform can influence the PD behavior in case in different types of defects.

The main insulating system components of stator windings of rotating electrical machines are organic enamel and impregnating resin. When power electronic converters supply motors, the stator insulation system is subjected to voltage waveforms such as PWM composed by trains of pulses. Due to reflection phenomena, overvoltage can be associated to pulsed waveforms and partial discharge (PD) activity can incepts leading to the insulating system failure in short time. In this work two types of conventional and nanofilled impregnating resins, based on unsaturated polyester and solvented alkydic resin matrices, have been compared. Suitable twisted pair samples prepared by means of wires insulated with conventional enamel and by two impregnating resins have been tested. The twisted pair samples have been subjected to sinusoidal voltage waveforms, having amplitudes slightly higher than the Partial Discharge Inception Voltage. Aging tests have been performed until the insulating system breakdown, which is assumed as end-point criterion: during the tests, the PD activity has been monitored by means of a PRPDA system. For each sample, starting from the PD signals post processing, the apparent current of positive discharges has been considered: this derived quantity showed a linear behavior with the degradation time. In order to define a life model, the slopes of the straight lines have been put in relation with the related lifetimes by means of the inverse power law.

Ahstract-In order to recognize on-line different kind of defects where partial discharges take place it is necessary to employ an embedded acquisition system, possibly low-cost, near to the discharges site. Such acquisition system should be capable to analyze the acquired data in real time to recognize different defect sources. For this reason, a single hidden-layer feedforward neural network (SLFN) is presented in this work. Such neural network inherits the approach adopted by the Extreme Learning Machine (ELM), but its activation functions are based on hardware friendly step (hard-limit) function. This facilitates the hardware resources optimization. In order to show the effectiveness of this solution different kind of PDs signals have been acquired. Once the data have been available, they have been divided in two sets: training and test. The training data have been used off-line to construct the model. Therefore the test data have been used on-line to perform the neural network. Moreover a comparison, in term of generalization error, between the classic ELM (based on complex activation functions, e.g. sigmoid), the Support Vector Machine (SVM) and the digital realization has been analyzed and showed. The results demonstrate how a simple algorithm as a single layer feedforward neural network is able to distinguish different kind of PDs using limited resources. Eventually the occupation area is shown on different low-cost devices varying the number of neurons in the hidden layer. The new ELM digital implementation is able to achieve good results, in term of generalization error when involved in supervised problems recognizing different kind of PD activity. These results are comparable to the ones obtainable by mean of standard ELM and SVM.

In Type I insulating systems the presence of partial discharge (PD) leads to the fast degradation of the insulating system dielectric characteristics. This could result in the short-medium term, to a total breakdown discharge provoking the failure of the system. A predictive diagnostic system is advisable to prevent the failure of the motor. This kind of diagnostic system should be non-invasive and economically feasible. For these reasons different types of antenna sensors have been employed and evaluated. The captured irradiated partial discharge signals are then processed with different filters and amplifiers. This study has been conducted on suitable twisted pairs specimens subjected to PWM-like waveforms typical of electronic variable speed drive supplies. A comparison between measuring sytems based on the detection of conducted or irradiated PD signals have been performed. The peculiarities and potentiality of both systems have been assessed and evidenced.

In order to maintain reasonable costs and high measure speed, the PRPDA (Phase Resolved Partial Discharges Acquisition) acquisition systems usually employ 8 bits A/D converters. Normally the PDs activity shows signals having different amplitudes (few high amplitudes PD signals with many low amplitudes). Avoiding the A/D converter saturation that lead to measurements information loss, it has been necessary to set the reference voltage on the expected maximum PD signal magnitude. Due to the presence of high amplitude PD signals in real cases it could be difficult recognize bad quantized PD signals (low amplitude) from the noise. In fact, the LSB (Least Significant Bit) of the noise results similar to the ones of PDs signal. In this study, the main issues relevant to this identification problem have been analyzed and discussed. The main purpose is to recognize PD signals from noise. PDs measurements have been performed maximizing the resolution of the maximum acquired discharge and, with the same resolution, noise measurements have been collected. This way the comparison among the noise and the PD signals have been performed. The applied method to distinguish the PD signal is the Equivalent Time — Equivalent Bandwidth (TW). The amplitude of each PD has been varied, keeping the same LSB resolution, in order to study the algorithm behavior.

Nowadays the massive adoption of electronic converters to supply electrical motors leads to critical problems during the normal operation. In fact when the motors are supplied by electronic converters, the stator insulating system could be subjected to over voltages due to reflection and resonance phenomena. Frequently this over voltages exceed the partial discharge inception voltage and partial discharge activity incepts. The presence of partial discharge activity implies the degradation of the dielectric characteristics of the type I insulating system and in the short-medium time, leads to a total breakdown discharge. Three different types of type I insulating system used in rotating electrical machines fed from inverters have been considered and studied. The PDIV and RPDIV at different temperatures have been measured and compared. During this research study the stators of asyncronous motors have been subjected to sinusoidal waveform and unipolar positive voltage pulses at different temperatures. In order to perform the RPDIV measurements a modified surge tester has been used so to apply unipolar positive voltage pulses at the terminals of the windings. The correlation between the repetitive partial discharge inception voltage and the partial discharges inception voltage measured at different temperatures has been studied.

This research analyzes the affinities between two well-known learning schemes that apply randomization in the training process, namely, Extreme Learning Machines (ELMs) and the learning framework using similarity functions. These paradigms share a common approach to inductive learning, which combines an explicit data remapping with a linear separator; however, they seem to exploit different strategies in the design of the mapping layer. This paper shows that the theory of learning with similarity functions can stimulate a novel reinterpretation of ELM, thus leading to a common framework. This in turn allows one to improve the strategy applied by ELM for the setup of the neurons’ parameters. Experimental results confirm that the new approach may improve over the standard strategy in terms of the trade-off between classification accuracy and dimensionality of the remapped space.

This paper describes a digital circuit architecture which implements a recently proposed algorithm for the solution of the point location problem in the evaluation of piecewise affine functions. The circuit is suitable for FPGA implementation of explicit Model Predictive Control. The performances of the architecture are tested in a case study through hardware-in-the-loop simulation. Results show that the proposed circuit can be implemented on limited hardware resources also for quite complex, possibly discontinuous control functions, thus representing a good spare solution when other existing circuit architectures (generally faster but more resource-demanding) cannot be deployed.

A new analytical procedure for the processing of XRD patterns with conventional powder geometry was used to study phase composition and properties of a typical α–β Ti alloy. This approach was particularly effective in studying the Ti–6Al–4V alloy, since the small amount of fβ phase can not be easily identified bytraditional XRD methods. In addition to quantitative phase analysis, relevant structural information was produced by the refinement of cell parameters, thermal factors, and atomic positions. Further microstructural information, including texture, crystallite domain size, and distortion from lattice defects, was also obtained with the same procedure. The proposed method was particularly useful in studying the effect of surface polishing, which produced a compressive residual stress (also measured by non-destructive XRD. techniques) of –450 MPa in the surface region; changes in the microstructure and chemical composition of alloy elements in the two Ti polymorphs after stress relief heat tre...

X-ray microdiffraction (XRμD) was used to characterize the surface layers of a-β-Ti alloys (Ti-6Al-4V, Ti-4Al-4Mo-2Sn, Ti-5Al-2Fe and Ti-4Al-2Mn) ion nitrided at 1273 K for 12 h. On the alloys Ti-4Al-4Mo-2Sn, Ti-5Al-2Fe and Ti-4Al-2Mn the ion-nitrided layers consist of the phase δ (TiN x , ϵ (Tin 2 N) and a (nitrogen solid solution in a-Ti); the phases δ and α have already been identified by conventional surface diffractometry. The alloy Ti-6Al-4V presents only the phases δ and α. It is of particular interest to note that the ϵ phase was detected by means of both X-ray diffraction and XRμD in the surface layers of the same alloys after ion nitriding and annealing at 1073 K. The presence of the ϵ phase in alloys that were ion nitrided but not annealed seems to be correlated with the alloy chemical composition and with the differentiated alloy element distribution between the hardened layer and the metal matrix, which involves diffusion of the β-stabilizing elements (iron, molybdenum and vanadium) in the matrix with their maximum amounts in the interface zone.

A study has been made in order to determine the transition α (h.c.p.) ⇒ β (b.c.c.) temperatures in some two-phase α-β titanium alloys for industrial use: for this aim D.T.A. tests and for comparison, specific metallographic analyses on samples specially heat treated have been carried out. The correlation between β transition temperatures, treatment temperatures and the behaviour of the tested alloys in the ion-nitriding process has been examined: thus, it has been possible to define in which measure the selection of the treatment temperature, with reference to the B transition temperatures, determines the phases present in the nitrided surface layers and their mutual distribution. Finally, the morphological characterization of the hardened surface layers, obtained through the thermochemical nitriding treatment, has been defined: in particular, the nitrogen distribution in the phases present in the surface layers has been studied by means of back scattered electron analysis techniques.

The characterization of surface-hardened metallic materials can be carried out using X-ray microdiffractometry. For this research we examined a sample of ion-nitrided c.p. titanium, sectioned at an acute angle with respect to the original surface to make analysis easier. This process was used to ascertain the exact constitution of the nitrided layer, to evaluate (at increasing depths) the relative quantities of the constituent phases and to measure the reticular distortion induced in hexagonal titanium, mainly by the dissolution of nitrogen. The microdiffractometric conclusions are a useful addition to the microanalysis and microhardness results previously obtained, with which they closely agree.

Previous studies reported that music therapy (MT) exerts a positive effect on many medical and neuropsychiatric disorders. The use of MT has been proposed also for patients with severe mental illnesse (SMI), altrough further studies are still needed. The aim of the present study was to evaluate the effects on a structured MT program on clinical and social functioning indices of patient with SMI, hospitalized in an psychiatric emergency ward. The MT intervention followed the Benenzon model of MT and was delivered biweekly to 61 patients consecutively admittted to the psychiatric emergency ward. Subjects who did not complete the two-week MT intervention (N=45) were considered as the control group. all subjects were administred the Brief Psychiatric Rating Scale (BPRS) to evaluate the general psychopatology, the Hospital Anxiety And Depression Scale (HADS) for affective symptomatology, the Clinical Global Impression Scale (CGI-S) for severity of symptoms and the Global Assestment of Functioning (GAF) for psychosocial functioning. A repeated measures analysis of variance revealed that patients who unerwent the MT intervention had a statistically significant reduction of general and affective psychopatology scores and of symptoms severity with respect to the control group, after observation period. Our result are in line with previous studies confirming that MT may exert positive effects on psychopatology (in particular, on affective symptomatology) of patient with SMI, and extend this observation to an emergency setting, with short period of hospital stay.

Samples of titanium with a commercial purity grade 4 and of alloys Ti-6Al-4V and Ti-4Al-2Sn-4Mo were surface hardened by laser alloying with boron. This treatment was carried out by contemporaneously laser melting of the titanium alloy surface and powders containing boron which were injected on the surface of the specimens. The alloying powders were deposited on the metal surface either before or during laser treatment. Boronizing powders with different chemical compositions (B, B 4 C, TiB 2 , MoB, MoB 2 and W 2 B 5 ) were used to prepare surface coatings. Treatments were carried out with CO 2 laser equipment (1.2 kW) under a protective argon atmosphere using specific power ranging between 10 and 14 kW/cm 2 and interaction times between 0.3 and 0.6 s in order to assess the best processing conditions. Morphology, composition and microhardness of the layers were studied by microscopy, X-ray diffraction and Vickers microhardness analyses. Hardened layers showed a content of boride crystals and a progressively decreasing hardness according to the distance from the surface. These coatings proved to be well adherent to the substrate and showed a hardness up to 14 × 10 3 MPa. The results obtained on different alloys showed that the matrix composition can affect the coating features. Borided samples submitted to annealing treatment displayed a satisfactory thermal stability.

The chemical and physical characteristics of ion-nitrided surface layers, obtained on α-β titanium alloys, are examined and correlated both with the working conditions adopted in the ion-nitriding process and with the alloy chemical composition. Besides the influence of the working parameters on the morphology and on the microstructures of the ion-nitrided surface layers, mainly the alloy element distributions both in surface coatings and in the substrate are analysed for five α-β titanium alloys of industrial use, and for titanium c,p. as reference, ionnitrided at various treatment temperatures. The nitriding process forms, on titanium alloy parts, high-hardness surface layers consisting of TiN (δ phase) and Ti2N (ɛ phase) nitrides and an interstitial solid solution of nitrogen in the close-packed hexagonal lattice of titanium (α phase). The presence and the extent of these phases as well as the ion-nitrided layer morphology are essentially determined by the alloy chemical composition and the working parameters. In particular a low-temperature treatment produces an extended nitrogen diffusion in the matrix beneath a thin continuous nitrided layer, while a high-temperature treatment produces prevalently a continuous nitrided surface layer. The alloy element distribution appears differentiated in the various phases and may be correlated with the different affinity of these elements with nitrogen.

The crystallographic characteristics and the equilibrium ratios for solid solutions containing 33 and 50 at% boron which belong to the Fe-Ni-B system are described. The study was carried out on samples obtained by diffusion at 1073 K, of mixtures of the elements in powder form. It was possible to confirm the existence of a complete solid solution between the compounds Fe2B and Ni2B. On the other hand, there are two solid solutions containing 50 at% boron: one is derived from the compound FeB in which up to 70% of the iron atoms can be substituted by nickel, the other derives from the boride NiB and extends as far as the component (Ni0.85Fe0.15)B. Curves were obtained indicating the variation of lattice parameters and volume of the respective elementary cells with composition for these solid solutions. The different insertion of nickel in the borides was also quantified and the relative limit distribution coefficient was evaluated. Its value is consistent with those calculated on analogous borides of transition elements at the same temperature.

A study has been made of the preferential distribution between the phases (M’, M″)2B and (M M″)B present in the two-phase fields of the M’-M″-B systems for the transition metals with atomic numbers between 24 and 28. It has been ascertained that the metal with the smaller atomic number is systematically concentrated in the phase richer in boron. The extreme values of the distribution ratio α°, calculated according to Hagg and Kiessling1 in the temperature range 1073 to 1473 K, diminish progressively in relation to the increase in the difference between the atomic numbers of the metallic atoms. Further, α° increases gradually as the temperature increases. The preferential distribution has been justified considering the different metal-nonmetal affinity, and recalling analogous considerations expressed in relation to nitrides, carbides, and hydrides of transition metals.

Phase equilibria in the (Ce, Te) O system, originating from interaction between Ce(NO3)3 · 6 H2O and H6TeO6 followed by calcination up to 700°C, have been studied by methods of physico-chemical analysis (DTA, TGA, x-ray diffraction, XPS). Evidence is given for the formation of non-stoichiometric fluorite-type mixed-crystals (Ce, Te)O2 in the 450–500°C range, up to a tellurium solubility limit corresponding to a lattice parameter of 5.666(3) A. At higher temperatures (500–600°C) Ce2(TeO4)3 is stable and Ce(TeO3)2 is formed above about 550°C. Ce(TeO3)2 is also obtained by solid-state interaction between CeO2 and TeO2 and is identical with the previously reported CeTe3O8 phase. Preparative conditions of the various compounds are described and x-ray diffraction data are reported. The TeO2CeO2 phase diagram was established in the temperature interval from about 650 to 1100°C. Eutectic temperatures are 689°C and 794°C at the TeO2 and CeO2-rich sides, respectively. Tellur-Cer-Oxide Die Phasengleichgewichte im (Ce, Te)O-System, erhalten aus Ce(NO3)3 · 6 H2O und H6TeO6 und Temperung bis 700°C, wurden durch physikalisch-chemische Analysen (DTA, TGA, Rontgenbeugung, XPS) untersucht. Offensichtlich bilden sich nicht-stochiometrische Mischkristalle (Ce, Te)O2 vom Fluorittyp im Temperaturbereich von 450–500°C bis zu einer Tellur-Loslichkeitsgrenze entsprechend einem Gitterparameter von 566,6(3) pm. Bei hoheren Temperaturen (500–600°C) ist Ce2(TeO4)3 stabil und Ce(TeO3)2 wird uber etwa 550°C gebildet. Ce(TeO3)2 wird auch durch Festkorperreaktion zwischen CeO2 und TeO2 erhalten und ist identisch mit der bereits mitgeteilten CeTe3O8-Phase. Praparative Bedingungen fur die verschiedenen Verbindungen und Rontgenbeugungsdaten werden mitgeteilt. Das TeO2CeO2-Phasendiagramm im Temperaturbereich von 650–1100°C wird aufgestellt. Eutektische Temperaturen sind 689°C bzw. 794°C fur die TeO2- bzw. CeO2-reiche Seite.

The variation in hardness of the phases (Fe, M)B and (Fe, M)2B (M ≡ Cr or Ni), which are the predominant components of the borided layer obtained on iron alloys, was defined and related to increase in chromium, nickel and carbon contents. It was found that chromium increases the hardness both of the borided layer as a whole and of the boride components, even though these values are systematically lower than those measured on pure borides. Carbon, which is insoluble in this type of phase, accumulates at the boride-matrix interface and, because of its modification of the boron diffusion mechanism, it indirectly increases the hardness of the borided surface. Nickel reduces slightly but systematically the hardness of the borides, in particular of the (Fe, Ni)2B phase in which it has its highest concentration.

Synthetic ferrous alloys containing chromium and/or nickel were prepared and borided at 1173 K with powders containing B4C, KBF4 and SiC for times varying from 20 to 60 h. The surface layers composed of borides of type (Fe, M)B and (Fe, M)2B were characterized by means of X-ray diffraction, microscopic observations, analysis with the microprobe and microhardness measurements. A quantitative study was carried out on the differentiated distribution of chromium and nickel in the phases constituting the borided layer. The chromium spreads from the matrix towards the borided layer, where it is concentrated in the phase richest in boron (Fe, M)B. Nickel behaves in the opposite manner. Alloys without carbon were used in order to avoid the formation of an area containing carbides and boron carbides, situated between the borided layer and the matrix, which could influence chromium distribution. The distribution phenomenon is in accordance with what was found in the study of Fe-M-B ternary systems (M=Cr, Ni). The influence of chromium and nickel on the thickness, the morphology and the microhardness of the borided layer are discussed.

Surface layers having elevated hardness were produced by ion nitriding of titanium and α-β alloys of Ti-6wt.%Al-4wt.%V, Ti-4wt.%Al-2wt.%Sn4wt.%Mo and Ti-4wt.%Al-2wt.%Mn. The treatment was conducted at temperatures between 1073 and 1273 K for times of 4–32 h, using gaseous mixtures of nitrogen and hydrogen containing 20–80 vol.% nitrogen. X-ray diffraction, optical microscopy and microhardness were used to characterize the hardened surfaces. In most of the samples, this surface was composed of the phases δ (TiN type), ϵ (Ti 2 N type) and α (solid solution of nitrogen in titanium); in other cases only the δ and α phases were present. The operating parameters which influence the composition of the layer were identified and the particular crystalline texture of the ϵ phase is discussed.

Surface layers having elevated hardness were produced by ion nitriding of titanium and α-β alloys of Ti-6wt.%Al-4wt.%V, Ti-4wt.%Al-2wt.%Sn4wt.%Mo and Ti-4wt.%Al-2wt.%Mn. The treatment was conducted at temperatures between 1073 and 1273 K for times of 4–32 h, using gaseous mixtures of nitrogen and hydrogen containing 20–80 vol.% nitrogen. X-ray diffraction, optical microscopy and microhardness were used to characterize the hardened surfaces. In most of the samples, this surface was composed of the phases δ (TiN type), ϵ (Ti 2 N type) and α (solid solution of nitrogen in titanium); in other cases only the δ and α phases were present. The operating parameters which influence the composition of the layer were identified and the particular crystalline texture of the ϵ phase is discussed.

Phase equilibria in the (Ce, Te) O system, originating from interaction between Ce(NO3)3 · 6 H2O and H6TeO6 followed by calcination up to 700°C, have been studied by methods of physico-chemical analysis (DTA, TGA, x-ray diffraction, XPS). Evidence is given for the formation of non-stoichiometric fluorite-type mixed-crystals (Ce, Te)O2 in the 450–500°C range, up to a tellurium solubility limit corresponding to a lattice parameter of 5.666(3) A. At higher temperatures (500–600°C) Ce2(TeO4)3 is stable and Ce(TeO3)2 is formed above about 550°C. Ce(TeO3)2 is also obtained by solid-state interaction between CeO2 and TeO2 and is identical with the previously reported CeTe3O8 phase. Preparative conditions of the various compounds are described and x-ray diffraction data are reported. The TeO2CeO2 phase diagram was established in the temperature interval from about 650 to 1100°C. Eutectic temperatures are 689°C and 794°C at the TeO2 and CeO2-rich sides, respectively. Tellur-Cer-Oxide Die Phasengleichgewichte im (Ce, Te)O-System, erhalten aus Ce(NO3)3 · 6 H2O und H6TeO6 und Temperung bis 700°C, wurden durch physikalisch-chemische Analysen (DTA, TGA, Rontgenbeugung, XPS) untersucht. Offensichtlich bilden sich nicht-stochiometrische Mischkristalle (Ce, Te)O2 vom Fluorittyp im Temperaturbereich von 450–500°C bis zu einer Tellur-Loslichkeitsgrenze entsprechend einem Gitterparameter von 566,6(3) pm. Bei hoheren Temperaturen (500–600°C) ist Ce2(TeO4)3 stabil und Ce(TeO3)2 wird uber etwa 550°C gebildet. Ce(TeO3)2 wird auch durch Festkorperreaktion zwischen CeO2 und TeO2 erhalten und ist identisch mit der bereits mitgeteilten CeTe3O8-Phase. Praparative Bedingungen fur die verschiedenen Verbindungen und Rontgenbeugungsdaten werden mitgeteilt. Das TeO2CeO2-Phasendiagramm im Temperaturbereich von 650–1100°C wird aufgestellt. Eutektische Temperaturen sind 689°C bzw. 794°C fur die TeO2- bzw. CeO2-reiche Seite.