Mechanical components and tools in modern industry are facing increasing performance requirements leading to the growing need for advanced materials and thus, for modern frictional systems. In the last decades, the Pulsed Laser Deposition (PLD) has emerged as an unique tool to grow high quality mono- as well as multilayers surfaces in metallic/ceramic systems. Building up a knowledge base of tribological properties of industrially-scaled, room temperature deposited PLD hard coatings are the most important step for the application of these coatings in engineering design. Although single-layer coatings find a range of applications, there are an increasing number of applications where the properties of a single material are not sufficient. One way to surmount this problem is to use a multilayer coating. Application of metallic interlayers improves adhesion of nitride hard layer in multilayer systems, which has been used in PVD processes for many years, however, the PLD technique gives new possibilities to produce system comprising many bilayers at room temperature. Tribological coatings consisted of 2, 4 and 16 bilayers of Cr/CrN and Ti/TiN type were fabricated with the Pulsed Laser Deposition (PLD) technique in the presented work. It is found in transmission electron examinations on thin foils prepared from cross-section that both nitride-based multilayer structures studied are characterized by small columnar crystallite sizes and high defect density, what might rise their hardness but compromise coating adhesion. The intermediate metallic layers contained larger sized and less defective columnar structure compared to the nitride layers, which should improve the coatings toughness. Switching from single layer to multi-layer metal/nitride composition improved resistance to delamination.
In this paper, a new Multi-Layer Perceptron Neural Network (MLP NN) classifier is proposed for classifying sonar targets and non-targets from the acoustic backscattered signals. Besides the capabilities of MLP NNs, it uses Back Propagation (BP) and Gradient Descent (GD) for training; therefore, MLP NNs face with not only impertinent classification accuracy but also getting stuck in local minima as well as lowconvergence speed. To lift defections, this study uses Adaptive Best Mass Gravitational Search Algorithm (ABGSA) to train MLP NN. This algorithm develops marginal disadvantage of the GSA using the bestcollected masses within iterations and expediting exploitation phase. To test the proposed classifier, this algorithm along with the GSA, GD, GA, PSO and compound method (PSOGSA) via three datasets in various dimensions will be assessed. Assessed metrics include convergence speed, fail probability in local minimum and classification accuracy. Finally, as a practical application assumed network classifies sonar dataset. This dataset consists of the backscattered echoes from six different objects: four targets and two non-targets. Results indicate that the new classifier proposes better output in terms of aforementioned criteria than whole proposed benchmarks.
In the extra-thick coal seams and multi-layered hard roofs, the longwall hydraulic support yielding, coal face spalling, strong deformations of goaf-side entry, and severe ground pressure dynamic events typically occur at the longwall top coal caving longwall faces. Based on the Key strata theory an overburden caving model is proposed here to predict the multilayered hard strata behaviour. The proposed model together with the measured stress changes in coal seam and underground observations in Tongxin coal mine provides a new idea to analyse stress changes in coal and help to minimise rock bursts in the multi-layered hard rock ground. Using the proposed primary Key and the sub-Key strata units the model predicts the formation and instability of the overlying strata that leads to abrupt dynamic changes to the surrounding rock stress. The data obtained from the vertical stress monitoring in the 38 m wide coal pillar located adjacent to the longwall face indicates that the Key strata layers have a significant influence on ground behaviour. Sudden dynamically driven unloading of strata was caused by the first caving of the sub-Key strata while reloading of the vertical stress occurred when the goaf overhang of the sub-Key strata failed. Based on this findings several measures were recommended to minimise the undesirable dynamic occurrences including pre-split of the hard Key strata by blasting and using the energy consumption yielding reinforcement to support the damage prone gate road areas. Use of the numerical modelling simulations was suggested to improve the key theory accuracy.