Crack Sensitivity Control of Nickel-Based Laser Coating Based on Genetic Algorithm and Neural Network

Hastelloy X exhibits outstanding thermal fatigue resistance, making it a promising material for repairing 50CrVA landing gear via directed energy deposition (DED). However, the substantial differences in composition and thermophysical properties between 50CrVA and Hastelloy X pose challenges by affecting interfacial microstructure and surface quality. This study investigates the effect of DED process parameters (laser power p, powder feed rate f, scanning speed v, and overlap rate) on the dilution ratio (η), microscopic morphology, surface flatness (ζ), and porosity of Hastelloy X claddings on a 50CrVA substrate. An optimization methodology integrating thermal–flow coupled simulation models and orthogonal experiments is developed to fabricate high-quality claddings. Furthermore, the corrosion–wear performance of the claddings is evaluated. The results indicate that the η of a single track increases with higher p or lower f, while it first increases and then decreases with the increase in v. Ablation marks tend to occur at excessive p or insufficient f, while low v causes surface ripples. The ζ of a single layer initially improves and subsequently deteriorates with increasing overlap rate. Porosity is significantly influenced by p and f. The optimal p, f, v, and overlap rate are 1600 W, 2.4 g/min, 240 mm/min, and 55%, respectively. The wear resistance of the cladding is nearly identical to that of the substrate, while corrosion resistance is significantly improved. This work provides a theoretical foundation for high-performance repair of 50CrVA landing gear in aircraft.

Human body is highly sensitive and repairing often incurs pain and expenses. Strength of the materials degraded by poor joint (either weld or link). New material technology is proposed many biomaterials for repairing bone and tissue and also many bio-implantation applications. Especially bioactive material like bioactive glass is used for biomedical applications for replacement and repairing organs in human body. This research work focuses on suggesting material of S53P4 bioactive glass Nano-coated Zirconium dioxide for manufacturing artificial knee implant for fixing in human body. The substrate of Zirconium dioxide is Nano-coated with S53P4 bioactive glass by means of laser cladding process. The laser cladding process parameters were optimized by Taguchi method to enhance mechanical properties like compressive strength, wear resistance and microhardness of Zirconium dioxide implant material. The key parameters like Laser Power (1 kW, 2 kW, 3 kW and 4 kW), beam diameter (2 mm, 3 mm, 4 mm and 5 mm), powder feed rate (10 g/min, 15 g/min, 20 g/min and 25 g/min) and scanning speed (3 mm/s, 4 mm/s, 5 mm/s and 6 mm/s) were considered. The optimal parameters result the higher compressive strength and microhardness are obtained as 373 MPa and 898.37 HV0.2 and minimum wear volume is attained as 0.148 mm3 in the Nano-coated implant material.

Optimization of process parameters for laser cladding Stellite6 cobalt-based alloy

A series of laser cladding layers of Ni35 on a medium carbon steel substrate by varying powder feeding rate and laser scanning speed were obtained. An investigation of the effects of the laser scanning speed and the powder feeding rate on microstructure and hardness was presented in this paper. The phase constitution, microstructure and hardness were investigated by X-ray diffractometer, energy dispersion spectroscopy, scanning electron microscope and Vickers hardness tester. The results show that as the scanning speed increases, the ratio of the temperature gradient G to the solidification rate R decreases which results in the finer microstructure of the cladding layer, as the powder feeding rate increases, the microstructure becomes finer due to the decrease of the temperature of the substrate and specific energy per unit mass of the powder. Furthermore, as the scanning speed and the powder feeding rate increase, the hardness of the cladding layer increases. This is because that the finer microstructure of the cladding layer results in the increase of grain boundary area, which results in the increase in the ability of deformation resistance. An intuitive and facilitative method which combining the line that joining the turning points of the hardness profile to show the effect rules of the laser scanning and the powder feeding rate on the range of dilution and the region of heat affect zone (HAZ) is adopted. This method is helpful to the study of the bonding interface, the control of the dilution range, the microstructure and properties of the cladding layer.

Laser cladded NiCrBSi alloy coating was fabricated on the surface of a 42CrMo roll using a 6 kW fiber laser. The effects of the laser power, scanning speed and feeding rate on the cladding layer form, size, dilution rate, microstructure, and hardness of coating were studied by using optical microscopy (OM), scanning electron microscopy (SEM) and a microhardness tester. The results show that the microstructure, size, dilution rate and hardness of the cladding layer had changed with an increase in laser power, powder feeding rate and scanning speed. The appropriate parameters of the laser cladding experiment are as follows: the laser power is 2000 W, powder feed rate 20 g × min-1, the scanning speed 4 mm × s-1. The cladding layer and the substrate exhibit good metallurgical bonding under the above processing parameters. The microstructure of the cladding layer is fine, the dilution rate is 9.8 wt.-%, and the microhardness of the cladding layer is 710.7 HV.

This study optimized laser cladding process parameters by combining BP neural network (BPNN) and genetic algorithm (GA) based on the multi-layer cladding nano Al₂O₃-13 wt.%TiO₂ coating. The model structure was trained according to the orthogonal test results. The BPNN model of closed loop control temperature of the molten pool, ultrasonic vibration frequency and preheating temperature of the incubator with bonding strength and microhardness of the coating was constructed. Moreover, single-objective and multi-objective parameter optimizations were carried out to bonding strength and microhardness of the coating based on GA. Results demonstrated that the predicted values of models are very close to test values, indicating that the constructed model was accurate and reliable. The maximum bonding strength and the maximum microhardness of the coating after GA optimization are 70.7 MPa and 2025.5 HV, respectively. When weights of bonding strength and microhardness of the coating are the same, the coating achieved the optimal comprehensive performances under the 2472.0 ^oC of closed-loop control temperature of the molten pool, 31.9 kHZ of ultrasonic vibration frequency and 400 ^oC of preheating temperature of incubator. The corresponding bonding strength and microhardness are 69.1 MPa and 1835.5 HV, respectively.

The aim of the present work is to establish the relationship between laser cladding (LC) process parameters (power, process speed, and powder feed rate) and AISI D2 tool steel metallurgical transformations, with the objective of optimizing the processing conditions during real reparation. It has been deposited H13 tool steel powder on some steel substrates with different initial metallurgical status (annealed or tempered) using a coaxial LC system. The microstructure of the laser clad layer and substrate heat-affected zone (HAZ) was characterized by optical microscopy, scanning electron microscopy, and electron backscattered diffraction. Results show that the process parameters (power, process speed, feed rate, etc.) determine the dimensions of the clad layer and are related to the microstructure formation. Although it is simple to obtain geometrically acceptable clads (with the right shape and dimensions) in many cases some harmful effects occur, such as carbide dilution and non-equilibrium phase formation, which modify the mechanical properties of the coating. Specifically, the presence of retained austenite in the substrate–coating interface is directly related to the cooling rate and implies a hardness diminution that must be avoided. It has been verified that initial metallurgical state of the substrate has a big influence in the final result of the deposition. Tempered substrates imply higher laser absorption and heat accumulation than the ones in annealed condition. This produces a bigger HAZ. For this reason, it is necessary to optimize the process conditions for each repair in order to improve the working behaviour of the component.

A genetic neural network (GNN) model was developed to predict the phase behavior of microemulsion (ME), lamellar liquid crystal (LC), and coarse emulsion forming systems (W/O EM and O/W EM) depending on the content of separate components in the system and cosurfactant nature. Eight pseudoternary phase triangles, containing ethyl oleate as the oil component and a mixture of two nonionic surfactants and n-alcohol or 1,2-alkanediol as a cosurfactant, were constructed and used for training, testing, and validation purposes. A total of 21 molecular descriptors were calculated for each cosurfactant. A genetic algorithm was used to select important molecular descriptors, and a supervised artificial neural network with two hidden layers was used to correlate selected descriptors and the weight ratio of components in the system with the observed phase behavior. The results proved the dominant role of the chemical composition, hydrophile-lipophile balance, length of hydrocarbon chain, molecular volume, and hydrocarbon volume of cosurfactant. The best GNN model, with 14 inputs and two hidden layers with 14 and 9 neurons, predicted the phase behavior for a new set of cosurfactants with 82.2% accuracy for ME, 87.5% for LC, 83.3% for the O/W EM, and 91.5% for the W/O EM region. This type of methodology can be applied in the evaluation of the cosurfactants for pharmaceutical formulations to minimize experimental effort.

A theoretical and experimental study on laser cladding of Ni60 alloy on 56NiCrMoV7 steel is presented for remanufacturing of machine parts. Due to multiparameter and nonlinearity of the laser cladding process, it is a challenge for developing an accurate method to predict the geometry dimension and the forming quality of multitrack coatings. In this paper, the nonlinear relationships between the process parameters (laser power, powder feeding rate, scanning speed, and overlap rate) and the geometry dimensions (melt width and melt height) as well as the forming qualities (dilution rate and microhardness) of the Ni60 alloy coating are established using the response surface methodology. Statistical analysis gives the influences of each process parameter and their interaction effects on the geometry dimensions and forming qualities of coatings. Process optimizations are also carried out to acquire the desired coating. The verification experiments under optimized conditions are conducted, which indicates that the results obtained by the prediction models agree well with the actual values. Metallographic observation shows that the grain sizes of the microstructure in the coating are different at different zones. X-ray diffraction and scanning electron microscopy results display that the microstructure of the Ni60 coating is mainly composed of austenite phase and (Cr, Fe)7C3 phase. The microhardness of the Ni60 coating is about twice that of 56NiCrMoV7 steel, and the compressive strength of the material with Ni60 coating is higher than that of the 56NiCrMoV7 material. Finally, the practical application of the presented method for repairing the damaged surfaces of a hammer head is proceeded successfully.

Reducer failure was analyzed and by use of BP neural network in the paper. Model of failure diagnosis was established. By using genetic algorithms, the value of neural networks, the threshold, and the network structure were optimized. Genetic neural network model was applied to the system design of remote reducer fault diagnosis. To compare training error curve of BP neural network with genetic neural network, it was shown that genetic neural network in the training of speed and accuracy higher than the neural network training model.

This paper gets a further study to cladding of Ni-based Titanium Carbide in the surface of die steel of 718, This paper introduces the effect of laser process parameters on the titanium carbide and nickel solid solution and cemented carbide in cladding layer; on that basis, this paper studied the effect of the different of powder paving thickness, laser power and scanning velocity and other parameters on microstructure and properties of cladding layer. By controlling the powder-bed depth, using suitable laser cladding power and scanning speed, we can get the cladding layer of compact structure, no holes or no cracks. Experiments were carried out to get suitable process parameters, and analyzed its mechanism. It has certain significance guiding to improve the laser cladding of 718 die steel quality,

Analysis and prediction of Cu-Sn-Ti alloy deposited on 316 L steel by coaxial laser cladding

To study the surface morphology of laser cladding, Workbench simulated the influence of laser power and scanning speed on the width and height of the cladding layer numerically, as well as the temperature field change and residual stress distribution of the cladding layer. The simulation results reveal that the melting height and width of the cladding layer are inversely proportional to the scanning speed. When the scanning speed is from V = 3 mm/s to V = 5 mm/s, the Al cladding layer’s melting width and melting height are reduced by 15.59% and 20.8%, respectively. A positive correlation exists between the melting height and width of the cladding layer and the laser power. When the laser power changes from P = 23 w to P = 27 w, the welding width and height of the A1 cladding layer increase by 6.55% and 55.56%, respectively. The melting height and width of the second cladding layer are generally higher than those of the bottom cladding layer. The pre-experiment screening process parameters ranges are laser power P (23 w–27 w) and scanning speed (3 mm/s–8 mm/s). Based on the Minitab response surface central composite method, the most notable influence on the melting height and width is revealed to be the powder-feeding rate and laser power, respectively. The response surface analysis method establishes the regression prediction models of melting width and height. The predicted value of melting width was 95.68%, and the predicted value of melting height was 82.26%. The results show that the values of cladding width and height are within the 95% prediction interval, proving that the regression model is correct.

A western North Pacific tropical cyclone (TC) intensity prediction scheme has been developed based on climatology and persistence (CLIPER) factors as potential predictors and using genetic neural network (GNN) model. TC samples during June–October spanning 2001–2010 are used for model development. The GNN model input is constructed from potential predictors by employing both a stepwise regression method and an Isometric Mapping (Isomap) algorithm. The Isomap algorithm is capable of finding meaningful low-dimensional architectures hidden in their nonlinear high-dimensional data space and separating the underlying factors. In this scheme, the new developed model, which is termed the GNN-Isomap model, is used for monthly TC intensity prediction at 24- and 48-h lead times. Using identical modeling samples and independent samples, predictions of the GNN-Isomap model are compared with the widely used CLIPER method. By adopting different numbers of nearest neighbors, results of sensitivity experiments show that the mean absolute prediction errors of the independent samples using GNN-Isomap model at 24- and 48-h forecasts are smaller than those using CLIPER method. Positive skills are obtained as compared to the CLIPER method with being above 12 % at 24 h and above 14 % at 48 h. Analyses of the new scheme suggest that the useful linear and nonlinear prediction information of the full pool of potential predictors is excavated in terms of the stepwise regression method and the Isomap algorithm. Moreover, the GNN is built by integrating multiple individual neural networks with the same expected output and network architecture is optimized by an evolutionary genetic algorithm, so the generalization capacity of the GNN-Isomap model is significantly enhanced, indicating a potentially better operational weather prediction.

NOVEL STAGING TOOL FOR LOCALIZED PROSTATE CANCER: A PILOT STUDY USING GENETIC ADAPTIVE NEURAL NETWORKS