Simulation of SiGe:C HBTs using neural network and adaptive neuro-fuzzy inference system for RF applications

In this article, the small-signal equivalent circuit model of SiGe:C heterojunction bipolar transistors (HBTs) has directly been extracted from S-parameter data. Moreover, in this article, we present a new modelling approach using ANFIS (adaptive neuro-fuzzy inference system), which in general has a high degree of accuracy, simplicity and novelty (independent approach). Then measured and model-calculated data show an excellent agreement with less than 1.68?×?10^?5% discrepancy in the frequency range of higher than 300 GHz over a wide range of bias points in ANFIS. The results show ANFIS model is better than ANN (artificial neural network) for redeveloping the model and increasing the input parameters.     

ESTIMATION OF CONFIGURATION FACTOR FOR RADIATION FROM A RECTANGLE TO A PARALLEL SMALL ELEMENT OF SURFACE LYING ON THE PERPENDICULAR TO A CORNER OF THE RADIATOR

To calculate the radiant intensity at a point distant from a radiator, a geometrical or configuration factor must be used. If the temperature is known, then the emissive power can be calculated using the mean equivalent beam length, but before the radiant heat flux at a distance can be estimated, a configuration factor must be calculated. In this article a simple predictive tool is developed to estimate the configuration factor for a geometry including a receiving element lying on the perpendicular to one corner of a radiant rectangle. The results are found to be in excellent agreement with reported data in the literature with average absolute deviation being around 2%. The results can be used in follow up calculations to estimate heat fluxes on surfaces exposed to radiation. The tool developed in this study can be of immense practical value for engineers as a quick check on the configuration factor for a geometry including a receiving element lying on the perpendicular to one corner of a radiant rectangle without opting for any experimental trials. In particular, engineers would find the approach to be user-friendly with transparent calculations involving no complex expressions.     

Nash bargaining game model for two parallel stages process evaluation with shared inputs

Data envelopment analysis is a non-parametric technique for evaluating peer decision making units (DMUs) with using multiple inputs to produce multiple outputs. In the real world, DMUs usually have complex structures. One of these structures is a two-stage process with intermediate measures. In this structure, there are two stages and each stage uses inputs to produce outputs, separately where the outputs of the first stage are the inputs for the second stage. Cooperative model such as centralized model and non-cooperative model are game theoretic approaches to evaluate two-stage processes. Non-cooperative model supposes that one of the stages is the leader and another stage is the follower, whereas in the centralized model, both stages are evaluated simultaneously. In this paper, we propose a game theoretic model based on the Nash bargaining game to calculate weights when parallel stages with shared inputs compete to reach a high efficiency in the competitive strategy. Two data sets including the bank branches and thermal power plants in Iran are used to show the abilities of proposed model. This model can be applied in other processes such as supply chain, manufacturing and public service units.     

Improving the surface quality in wire electrical discharge machined specimens by removing the recast layer using magnetic abrasive finishing method

This study explores the feasibility of removing the recast layer formed on aluminum alloy cylindrical specimens machined by wire electrical discharge machining (WEDM) by using magnetic abrasive finishing (MAF). The WEDM is a thermal machining process capable of accurately machining parts with high hardness or complex shapes. The sparks produced during the WEDM process melt the metal’s surface. The molten material undergoes ultra-rapid quenching and forms a layer on the surface defined as recast layer. The recast layer may be full of craters and microcracks which reduce service life of materials tremendously, especially under fatigue loads in corrosive environments. This investigation demonstrates that MAF process, can improve the quality of WEDM machined surfaces effectively by removing the recast layer. The present work studies the effect of some parameters, i.e., linear speed, working gap, abrasive particle size, and finishing time on surface roughness and recast layer thickness using full factorial analysis. Three-level full factorial technique is used as design of experiments for studying the selected factors. In order to indicate the significant factors, the analysis of variance has been used. In addition, an equation based on regression analysis is presented to indicate the relationship between surface roughness and recast layer thickness of cylindrical specimens and finishing parameters. Experimental results show the influence of MAF process on recast layer removal and surface roughness improvement.     

Using AC field measurement data at an arbitrary liftoff distance to size long surface-breaking cracks in ferrous metals

We propose a model-based inversion method to size long surface-breaking cracks in ferrous metals using alternative current field measurement (ACFM) data at an arbitrary liftoff distance. This method employs conjugate gradients optimization to invert measured surface ACFM signal to crack depth. To alleviate the adverse effect of sensor liftoff uncertainty on crack sizing, we propose a blind de-convolution algorithm for reconstructing respective surface ACFM crack signal. In this algorithm, the partially known filter function associated with the sensor liftoff is estimated from which the surface crack signal can be restored. The validity of the proposed inversion method is demonstrated by comparing the actual and predicted depths of several simulated and machine-made long cracks in mild steel test blocks.     

Optimizing the mechanical and physical properties of thermoplastic starch via tuning the molecular microstructure through co‐plasticization by sorbitol and glycerol

Nowadays, environmental hazards caused by plastic wastes are a major concern in academia and industry. Utilization of biodegradable polymers derived from renewable sources for replacing common petroleum‐based plastics is a potential solution for reducing the problem. In this regard, starch has become one of the most promising alternatives to non‐biodegradable polymers for depleting plastic waste thanks to its low expense, abundance, renewability and biodegradability. However, the main drawbacks of starch are its poor processability, weak mechanical properties and severe hydrophilicity. In this work, thermoplastic starch (TPS) samples have been prepared using glycerol and sorbitol as co‐plasticizers in a laboratory co‐rotating twin screw extruder. Based on the mechanical test results, glycerol caused higher elongation to break but had lower tensile strength and elastic modulus compared to sorbitol plasticized starch. Fourier transform infrared spectroscopy and DSC results indicated that the hydrogen bond interaction between starch chains and plasticizers could be improved by replacing glycerol by sorbitol, which resulted in higher resistance against retrogradation proved by XRD results. TGA illustrated that the higher the sorbitol to glycerol ratio was, the more stable was the TPS. Using a proper amount of plasticizers (42 wt% total plasticizer, sorbitol to glycerol ratio 2:1) led to the preparation of a TPS sample with optimized properties including enhanced mechanical properties, high thermal stability, strong hydrogen bond formation and high resistance against retrogradation. © 2017 Society of Chemical Industry     

Thermo-oxidative degradation during sintering of polyethylene particles

Polymer sintering is not only a well-established procedure for producing functional polymeric parts, but it is also the basis for the relatively new additive manufacturing technique, selective laser sintering. Although studying the impact of thermo-oxidative degradation during sintering has significant practical importance, few studies have focused on this aspect of the sintering process. In the present work, we have investigated the active thermo-oxidative degradation mechanisms during sintering of high-density polyethylene (HDPE) particles, the conditions that promote them, and their respective impact on the morphological evolution of the polyethylene particles. To perform a comprehensive study, we have complemented the rheological, thermal, chemical, and morphological analysis of the sintered HDPE particles with the study of their ensemble pore structure. We observed two distinct degradation regimes. At the beginning, the relatively low concentration of oxygen in the particles led to the dominance of branching and resulted in a pore structure evolution limited to surface relaxation. In the second regime and with the diffusion of more oxygen, chain scission became the dominant route. In this regime, the emergence of highly mobile short chains markedly accelerated the pore evolution.     

Experimental and numerical evaluation of rigid column to baseplate connection under cyclic loading

One of the most important connections in steel structures is column to the baseplate connection. This kind of connection has a complex nature, due to different behaviors of the constituents, including baseplate, grout, foundation, and anchor bolts. Studying the impact of this connection in general behavior of structure seems to be an essential issue, especially during earthquake, which is more likely to be associated with plastic hinge forming. The main goal of this work is to provide a hybrid modeling for describing hysteresis behavior of column to the baseplate connection. Within hybrid modeling, a mechanical model is accomplished by using neural network model components. Our findings reveal that hybrid models are able to show pinched hysteresis complex behavior of baseplate connections. Also, the proposed hybrid developed model is robust and reliable approaches for predicting behaviors of a recent designed connection.     

Data-dependent dissimilarity measure: an effective alternative to geometric distance measures

Nearest neighbor search is a core process in many data mining algorithms. Finding reliable closest matches of a test instance is still a challenging task as the effectiveness of many general-purpose distance measures such as \(\ell _p\)-norm decreases as the number of dimensions increases. Their performances vary significantly in different data distributions. This is mainly because they compute the distance between two instances solely based on their geometric positions in the feature space, and data distribution has no influence on the distance measure. This paper presents a simple data-dependent general-purpose dissimilarity measure called ‘\(m_p\)-dissimilarity’. Rather than relying on geometric distance, it measures the dissimilarity between two instances as a probability mass in a region that encloses the two instances in every dimension. It deems two instances in a sparse region to be more similar than two instances of equal inter-point geometric distance in a dense region. Our empirical results in k-NN classification and content-based multimedia information retrieval tasks show that the proposed \(m_p\)-dissimilarity measure produces better task-specific performance than existing widely used general-purpose distance measures such as \(\ell _p\)-norm and cosine distance across a wide range of moderate- to high-dimensional data sets with continuous only, discrete only, and mixed attributes.