Search results

An effective machine vision-based method for micro-crack detection of solar cell can economically improve the qualified rate of solar cells. However, how to extract features which have strong generalization and data representation ability at the same time is still an open problem for machine vision-based methods.

A micro-crack detection method based on adaptive deep features and visual saliency is proposed in this paper. The proposed method can adaptively extract deep features from the input image without any supervised training. Furthermore, considering the fact that micro-cracks can obviously attract visual attention when people look at the solar cell’s surface, the visual saliency is also introduced for the micro-crack detection.

Comprehensive evaluations are implemented on two existing data sets, where subjective experimental results show that most of the micro-cracks can be detected, and the objective experimental results show that the method proposed in this study has better performance in detecting precision.

First, an adaptive deep features extraction scheme without any supervised training is proposed for micro-crack detection. Second, the visual saliency is introduced for micro-crack detection.

This paper aims to improve the generalization capability of feature extraction scheme by introducing a micro-cracks detection method based on self-learning features. Micro-cracks detection of multicrystalline solar cell surface based on machine vision is fast, economical, intelligent and easier for on-line detection. However, the generalization capability of feature extraction scheme adopted by existed methods is limited, which has become an obstacle for further improving the detection accuracy.

A novel micro-cracks detection method based on self-learning features and low-rank matrix recovery is proposed in this paper. First, the input image is preprocessed to suppress the noises and remove the busbars and fingers. Second, a self-learning feature extraction scheme in which the feature extraction templates are changed along with the input image is introduced. Third, the low-rank matrix recovery is applied to the decomposition of self-learning feature matrix for obtaining the preliminary detection result. Fourth, the preliminary detection result is optimized by incorporating the superpixel segmentation. Finally, the optimized result is further fine-tuned by morphological postprocessing.

Comprehensive evaluations are implemented on a data set which includes 120 testing images and corresponding human-annotated ground truth. Specifically, subjective evaluations show that the shape of detected micro-cracks is similar to the ground truth, and objective evaluations demonstrate that the proposed method has a high detection accuracy.

First, a self-learning feature extraction method which has good generalization capability is proposed. Second, the low-rank matrix recovery is combined with superpixel segmentation for locating the defective regions.

The purpose of this paper is to evaluate theoretically and numerically the stress and stress intensity factor (SIF) at the time of propagation of the crack in bi-material. The problem is formulated using two thin materials which are bound by a cracked adhesive at the tip and having a micro-crack in one of these two materials.

The plane stresses and the SIF will be determined as a function of two parameters (Poisson’s ratio and Shear modulus). The numerical analysis is carried out on a flat element, having a main crack in one of these ends, and a micro-crack varies in the vicinity of this main crack. The problem is analyzed by the finite element method and processed by computational software (ABAQUS).

The numerical and theoretical analysis allowed the author to determine and compare the values of plane stresses and SIF in each area of the material.

The theoretical analysis of SIF is based mainly on a mathematical calculation of equations of plane stresses; these equations are determined by development of complex analytical functions of bi-materials given by other researchers. Using the numerical method, several models are modeled by changing the micro-crack position relative to the main crack to determine the plane stresses and SIF for each position.

The interaction between rock mass structural planes and dynamic stress levels is important to determine the stability of rock mass structures in underground geotechnical engineering. In this work, the authors aim to focus on the degradation effects of fracture geometric parameters and unloading stress paths on rock mechanical properties.

A three-dimensional Particle Flow Code (PFC3D) was used for a systematic numerical simulation of the strength failure and cracking behavior of granite specimens containing prefabricated cracks under conventional triaxial compression and triaxial unilateral unloading. The authors demonstrated the unique mechanical response of prefabricated fractured rock under two conditions. The crack initiation, propagation, and coalescence process of pre-fissured specimens were analyzed in detail.

The authors show that the prefabricated cracks and unilateral unloading conditions not only deteriorate the mechanical strength but also have significant differences in failure modes. The degrading effect of cracks on model strength increases linearly with the decrease of the dip angle. Under the condition of true triaxial unilateral unloading, the deterioration effect of peak strength of rock is very significant, and unloading plays a role in promoting the instability failure of rock after peak, making the rock earlier instability failure. Associating with the particle vector diagram and crack coalescence process, the authors find that model failure mode under unilateral loading conditions is obviously distinct from that in triaxial loading. The peak strain in the unloading direction increases sharply, resulting in a new shear slip.

This study is expected to improve the understanding of the strength failure and cracking behavior of fractured rock under unilateral unloading.

For high crack-susceptibility materials such as Inconel 713LC (IN713LC) nickel alloy, fabricating crack-free components using the laser powder bed fusion (LPBF) technique represents a significant challenge because of the complex interactions between the effects of the main processing parameters, namely, the laser power and scanning speed. Accordingly, this study aims to build up a methodology which combines simulation model and experimental approach to fabricate high-density (>99.9%) IN713LC components using LPBF process.

The present study commences by performing three-dimensional (3D) heat transfer finite element simulations to predict the LPBF outcome (e.g. melt pool depth, temperature and mushy zone extent) for 33 representative sample points chosen within the laser power and scanning speed design space. The simulation results are used to train a surrogate model to predict the LPBF result for any combination of the processing conditions within the design space. Then, experimental trials were performed to choose the proper hatching space and also to define the high crack susceptibility criterion. The process map is then filtered in accordance with five quality criteria, namely, avoiding the keyhole phenomenon, improving the adhesion between the melt pool and the substrate, ensuring single-scan-track stability, avoiding excessive melt pool evaporation and suppressing the formation of micro-cracks, to determine the region of the process map which improves the relative density of the IN713LC component and minimizes the micro-cracks. The optimal processing conditions are used to fabricate IN713LC specimens for tensile testing purposes.

The optimal processing conditions predicted by simulation model are used to fabricate IN713LC specimens for tensile testing purposes. Experimental results show that the tensile strength and elongation of 3D-printed IN713LC tensile bar is higher than those of tensile bar made by casting. The yield strength of 791 MPa, ultimate strength of 995 MPa, elongation of 12%, and relative density of 99.94% are achieved.

The present study proposed a systematic methodology to find the processing conditions that are able to minimize the formation of micro-crack and improve the density of the high crack susceptivity metal material in LPBF process.

This study aimed to overcome the challenging issues involved in providing high-precision eigensolutions. The accurate prediction of the buckling load bearing capacity under different crack damage locations, sizes and numbers, and analysing the influence mechanism of crack damage on buckling instability have become the needs of theoretical research and engineering practice. Accordingly, a finite element method was developed and applied to solve the elastic buckling load and buckling mode of curved beams with crack damage. However, the accuracy of the solution depends on the quality of mesh, and the solution inevitably introduces errors due to mesh. Therefore, the adaptive mesh refinement method can effectively optimise the mesh distribution and obtain high-precision solutions.

For the elastic buckling of circular curved beams with cracks, the section damage defect analogy scheme of a circular arc curved beam crack was established to simulate the crack size (depth), position and number. The h-version finite element mesh adaptive analysis method of the variable section Euler–Bernoulli beam was introduced to solve the elastic buckling problem of circular arc curved beams with crack damage. The optimised mesh and high-precision buckling load and buckling mode solutions satisfying the preset error tolerance were obtained.

The results of testing typical examples show that (1) the established section damage defect analogy scheme of circular arc curved beam crack can effectively realise the simulation of crack size (depth), position and number. The solution strictly satisfies the preset error tolerance; (2) the non-uniform mesh refinement in the algorithm can be adapted to solve the arbitrary order frequencies and modes of cracked cylindrical shells under the conditions of different ring wave numbers, crack positions and crack depths; and (3) the change in the buckling mode caused by crack damage is applicable to the study of elastic buckling under various curved beam angles and crack damage distribution conditions.

This study can provide a novel strategy for the adaptive mesh refinement for finite element analysis of elastic buckling of circular arc curved beams with crack damage. The adaptive mesh refinement method established in this study is fundamentally different from the conventional finite element method which employs the user experience to densify the meshes near the crack. It can automatically and flexibly generate a set of optimised local meshes by iteratively dividing the fine mesh near the crack, which can ensure the high accuracy of the buckling loads and modes. The micro-crack in curved beams is also characterised by weakening the cross-sectional stiffness to realise the characterisation of locations, depths and distributions of multiple crack damage, which can effectively analyse the disturbance behaviour of different forms of micro-cracks on the dynamic behaviour of beams.

Much progress has been made in the past few years in improving the corrosion‐resistance of decorative nickel‐chromium‐plated articles, especially those subjected to outdoor exposure. There now exists a wide range of alternative finishes, all of which are alleged to improve the corrosion‐resistance to some extent. To evaluate these finishes completely, manufacturers are required to spend much time, money and energy. Manufacturers generally wish to improve their standard of corrosion‐resistance of decorative deposits with no increase in cost, and with some of the more complex systems for corrosion protection this is now possible by an overall reduction in nickel thickness.

The purpose of this paper is to provide a systematic method to perform long‐term reliability assessment of gold (Au) and copper (Cu) ball bonds in fineline ball grid array package. Also with the aim to study the apparent activation energies (E_aa) and its associated wearout mechanisms of both Au and Cu wire in semiconductor device packaging. This paper discusses the influence of wire type on the long‐term reliability and mechanical performance after several component reliability stress tests.

A fineline ball grid array (FBGA) package with Cu and Au wire bonds was assembled with green molding compound and substrate. Samples are subjected for long‐term high temperature storage bake test at elevated temperatures of 150°C, 175°C and 200°C. Long‐term reliability plots (lognormal plots) are established and E_aa of both ball bonds are determined from Arrhenius plots. Detailed failure analysis has been conducted on failed sample and HTSL failure mechanisms have been proposed.

Reliability results show Au ball bond in FBGA package is observed with higher hour‐to‐failure compared to Cu ball bonds. The E_aa value of high temperature storage life (HTSL) reliability for Au ball bond is lower than Cu ball bond. Typical HTSL failure mechanism of Au ball bond is induced by micro‐voiding and AuAl intermetallic compound (IMC) micro‐cracks while CuAl IMC micro‐cracking (induced by Cl⁻ corrosion attack and micro‐cracking) caused wearout opens in Cu ball bond. These test results affirm the test‐to‐failure data collected is a useful method for lifetime prediction and E_aa calculation.

The paper reveals higher reliability performance of Cu ball bond in FBGA flash memory package which can be deployed in flash memory FBGA packaging with optimised package bill of materials.

The test‐to‐failure methodology is a useful technique for wearout reliability prediction and E_aa calculation.

Moderately thick circular cylindrical shells are widely used as supporting structures or storage cavities in structural engineering, rock engineering, and aerospace engineering. In practical engineering, shells often work with micro-cracks or defects. The existence of micro-crack damage may result in the disturbance of dynamic behaviours and even induce accidental dynamic disasters. The free vibration frequency and mode are important parameters for the dynamic performance and damage identification analysis. In particular, stiffness weakening of the local damage region leads to significant changes in the vibration mode, which makes it difficult for the mesh generated in the conventional finite element method to capture a high-precision solution of the local oscillation.

In response to the above problems, this study developed an adaptive finite element method and a crack damage characterisation method for moderately thick circular cylindrical shells. By introducing the inverse power iteration method, error estimation, and mesh subdivision refinement technique for the analysis of finite element eigenvalue problems, an adaptive computation scheme was constructed for the free vibration problem of moderately thick circular cylindrical shells with circumferential crack damage.

Based on typical numerical examples, the established adaptive finite element solution for the free vibration of moderately thick circular cylindrical shells demonstrated its suitability for solving the high-precision free vibration frequency and mode of cylindrical shell structures. The any order frequency and mode shape of cracked cylindrical shells under the conditions of different ring wave numbers, crack locations, crack depths, and multiple cracks were successfully solved. The influences of the location, depth, and number of cracks on the disturbance of dynamic behaviours were analysed.

This study can be used as a reference for the adaptive finite element solution of free vibration of moderately thick circular cylindrical shells with cracks and lays the foundation for further development of a high-performance computation method suitable for the dynamic disturbance and damage identification analysis of general cracked structures.

The Ni-based superalloy IN-738 LC is known to be susceptible to porosity and different types of cracking during the build-up process and, thus, challenging to manufacture using selective laser melting (SLM). Determining a feasible set of operating parameters for SLM of nickel-based superalloys involves new approach to experimental design based on the Doehlert method that assists in determining an optimal (feasible) set of operating parameters for SLM of IN-738 LC powder alloy.

The SLM parameters are evaluated in terms of their effectiveness in obtaining the microstructure with a porosity content of <0.5 per cent and without micro-cracking. The experimental approach is exemplified with the Doehlert matrix response variable, relative density, by comparing Archimedes method with microstructural assessments of pores and cracks from image analysis. The effect of heat treatment (HT) and hot isostatic pressing (HIP) on the microstructure of the SLMed IN-738 LC powder alloy has been examined and the consequential tensile response characterised.

By using optimised process parameters (low heat input, medium scanning speed and small hatching distance) which provides medium energy density, samples of IN-738 LC with a macroscopic porosity <0.5 per cent and free of micro-cracks can be manufactured by SLM. The results indicate that HIP of SLMed material did not lead to a noticeable effect on mechanical properties compared to HT of SLMed material suggesting that the level of both porosity and crack density might be already below the detection limit for the mere heat-treated material.

SLM processing parameters (power, scan speed, hatching distance) for IN-738 LC were successfully optimised after only 14 experiments using Doehlert design. Two independent methods, Archimedes method and image analysis, were used in this study to assess relative density of SLM-produced samples with sets of processing parameters showing coherency in prediction with predicted response by Doehlert design.