OverFeat Network Algorithm for Fabric Defect Detection in Textile Industry

Automation of systems emerged since the beginning of 20th century. In the early days, the automation systems were developed with a fixed algorithm to perform some specific task in a repeated manner. Such fixed automation systems are revolutionized in recent days with an artificial intelligence program to take decisions on their own. The motive of the proposed work is to train a textile industry system to automatically detect the defects presence in the generated fabrics. The work utilizes an OverFeat network algorithm for such training process and compares its performances with its earlier version called AlexNet and VGG. The experimental work is conducted with a fabric defect dataset consisting of three class images categorised as horizontal, vertical and hole defects.

Research on fracture behaviour of the adhesive sealant based on energy failure criterion for TFT-LCD

The adhesive sealant is a crucial structure connecting color filters and thin film transistors in liquid crystal panels. Research on the fracture progress of the connection structure is heavily needed in reliability evaluation engineering. In this work, three types of adhesive sealants with different widths were tested by the uniaxial tensile experiment to obtain their fracture process curves, which conformed to the brittle fracture characteristics described by the bilinear cohesion zone model. Then, according to the theory of engineering fracture mechanics, the Dugdale-Barenblatt plastic zone model was employed to analyze the adhesive sealant with hole defects, and it was simplified to mode ? fracture mechanics problem. Calculating with finite element numerical simulation, the numerical relationship between the stress field of the internal defect and the external stress of the material was obtained, and the brittle fracture behavior model was deduced as related to the defect size. Applying the model to the adhesive sealant, the average error of the model value after the correction was reduced from 7.98-12.13% to 6.84-7.53%, and the overall error was only within 15%. The model includes the material’s basic characteristics and the defect’s size that affect the fracture process, provides a theoretical basis for predicting the fracture of the sealant and improving the strength of bonded joints, thus is of great significance for material application and fracture analysis in engineering.

Research and Application of Neutron Detection Technology in Inspection of Void Defects in Turbine Runner Chamber

After a long-term operation of the turbine runner chamber, there may be voids between the steel lining and the concrete. If it is not discovered and treated in time, the safe and stable operation of the unit will be affected. In engineering practice, the hammering method is often used to detect hole defects, and the accuracy is low. The neutron detection technology is proposed to detect the void defects of the steel lining, quantitatively display the void location and size, and improve the accuracy of void defect detection.

Comparative analysis of the wear performance of spindle hook teeth during field work

The wear performance of the spindle hook teeth during field work was compared, and the causes of the wear failure of the spindle hook teeth were analyzed. Samples of three kinds of spindle were obtained from the fixed installation position under continuous field operation conditions. The hardness, phase structure, elemental composition, and micromorphology of the spindle hook teeth were characterized using microhardness, X-ray diffractometer (XRD), energy spectrum analyzer (EDS), and scanning electron microscope (SEM) after cutting of the spindles. Results show that the coating hardness and element penetration zone of No.3 spindle hook tooth are the largest, and the surface coating phase structure of the three kinds of spindle is mainly a body-centered cubic structure of Cr (211). Micro-crack and hole defects exist in the coating of all three kinds of spindle. The thicknesses of the coating of No.1, No.2, and No.3 spindles are 74, 100, and 130 μm, respectively. During the field operation, the wear of the spindle hook tooth coating is caused by abrasive wear and fatigue wear, while the wear of the substrate is the result of the combined effect of abrasive wear and oxidation wear. Extracting the wear area and width of spindle hook teeth shows that the wear area of all three kinds of spindle hook teeth increases exponentially and the wear width changes linearly with the increase in field operation area.

Impact of Electromagnetic Stirring on the Gas Metal Arc Welding of an MAR-M247 Superalloy

In this paper, the impact of electromagnetic stirring (EMS) on the gas metal arc welding (GATW) of an MAR-M247 superalloy was investigated. Results revealed that, without electromagnetic stirring, it was easy for carbides in the heat-affected zone (HAZ) of the weld bead to liquefy during welding, leading to weld bead cracks. Electromagnetic stirring refined the grains in the HAZ and the weld bead, leading to grain strengthening and subsequently resulting in the effective improvement in the hardness of the weld bead. In addition, electromagnetic stirring significantly facilitated the formation of the weld bead by the removal of large inclusions which in turn effectively improved crack resistance at the joint. It also accelerated the floating up of gas holes thereby reducing the generation of gas hole defects.

Influence of Screw-Hole Defect Size on the Biomechanical Properties of Feline Femora in an Ex Vivo Model

Objective The study aims to evaluate the biomechanical properties of feline femora with craniocaudal screw-hole defects of increasing diameter, subjected to three-point bending and torsion to failure at two different loading rates. Study Design Eighty femoral pairs were harvested from adult cat cadavers. For each bending and torsional experiment, there were five groups (n = 8 pairs) of increasing craniocaudal screw-hole defects (intact, 1.5 mm, 2.0 mm, 2.4 mm, 2.7mm). Mid-diaphyseal bicortical defects were created with an appropriate pilot drill-hole and tapped accordingly. Left and right femora of each pair were randomly assigned to a destructive loading protocol at low (10 mm/min; 0.5 degrees/s) or high rates (3,000 mm/min; 90 degrees/s) respectively. Stiffness, load/torque-to-failure, energy-to-failure and fracture morphology were recorded. Results Defect size to bone diameter ratio was significantly different between defect groups within bending and torsional experiments respectively (intact [0%; 0%], 1.5 mm [17.8%; 17.1%], 2.0 mm [22.8%; 23.5%], 2.4 mm [27.8%; 27.6%], 2.7 mm [31.1%; 32.4%]) (p < 0.001). No significant differences in stiffness and load/torque-to-failure were noted with increasing deficit sizes in all loading conditions. Screw-hole (2.7 mm) defects up to 33% bone diameter had a maximum of 20% reduction in bending and torsional strength compared with intact bone at both loading rates. Stiffness and load/torque-to-failure in both bending and torsion were increased in bones subjected to higher loading rates (p < 0.001). Conclusion Screw-hole defects up to 2.7 mm did not significantly reduce feline bone failure properties in this ex vivo femoral study. These findings support current screw-size selection guidelines of up to 33% bone diameter as appropriate for use in feline fracture osteosynthesis.

Common issues in the hetero-epitaxial seeding on SiC substrates in the sublimation growth of AlN crystals

Aluminium nitride (AlN) is a futuristic material for efficient next-generation high-power electronic and optoelectronic applications. Sublimation growth of AlN single crystals with hetero-epitaxial approach using silicon carbide substrates is one of the two prominent approaches emerged, since the pioneering crystal growth work from 1970s. Many groups working on this hetero-epitaxial seeding have abandoned AlN growth altogether due to lot of persistently encountered problems. In this article, we focus on most of the common problems encountered in this process such as macro- and micro-hole defects, cracks, 3D-nucleation, high dislocation density, and incorporation of unintentional impurity elements due to chemical decomposition of the substrate at very high temperatures. Possible ways to successfully solve some of these issues have been discussed. Other few remaining challenges, namely low-angle grain boundaries and deep UV optical absorption, are also presented in the later part of this work. Particular attention has been devoted in this work on the coloration of the crystals with respect to chemical composition. Wet chemical etching gives etch pit density (EPD) values in the order of 105 cm-2 for yellow-coloured samples, while greenish coloration deteriorates the structural properties with EPD values of at least one order more.