KTU Open Journal Systems (Kaunas University of technology)
Not a member yet
13395 research outputs found
Sort by
A Comparative Study of Metaheuristic Optimization Approaches to Optimize Laser Welding Process Parameter with Pre-Set Weld Size Magnitude for AISI 416 and AISI 440 FSe Stainless Steels
No full textOptimization methods are used to accurately predict laser welding process parameters, helping to save material effort and time in determining the desired output variables. Based on a mathematical model, parameter selection is considered a binding optimization problem. The work involved is closely related to evolutionary optimization algorithms. This article proposes highly effective meta-heuristic methods: the GA (Genetic Algorithm), the JAYA optimization algorithm, and the MDE (Modified Differential Evolution) algorithm, which optimize the parameters of the laser welding to achieve the desired size for the weld. The performance of these three methods is evaluated on laser welds for AISI 416 and AISI 440 FSe stainless steels. With the same initial conditions, the MDE algorithm outperforms the other algorithms, GA and JAYA algorithms, regarding the best fitness value after ten runs. Thus, the MDE algorithm is used to optimize three parameters: Laser Power (LP), Welding Speed (WS), and Fiber Diameter (FD) to achieve two desired welding dimensions: the Width of the Weld Zone (WWZ) and the Penetration Depth of the Weld (PDW) for laser welds
Tribological Characterization of AlCrN, TiAlN, TiSiN and AlTiN Coatings Against Mold Steel
No full textThis study investigates the tribological performance of TiAlN, AlTiN, AlCrN, and TiSiN coatings under boundary lubrication conditions using a tribometer. The findings indicate significant reductions in average friction coefficients compared to uncoated tungsten carbide (0.2436), with TiSiN demonstrating the lowest friction coefficient of 0.2111, thus showcasing its superior performance. Optical microscopy and profilometry analyses further reveal that TiSiN coating effectively minimizes surface roughness and wear tracks, suggesting enhanced wear resistance. While the coatings successfully reduce friction, they tend to increase wear on the counter materials. This study highlights the critical role of these coatings in industrial applications, emphasizing the need to balance friction reduction with wear enhancement for optimal performance
Effect of Nanoscale Amorphization on Edge Dislocation Emission from a Bifurcated Crack Tip in Deformed Nanocrystalline Solids
No full text The effect of nanoscale amorphization at the triple junction of grain boundaries on edge dislocation emission from a bifurcation crack tip in nanocrystalline materials has been suggested and theoretically described. A corresponding mechanical model has been established, and the exact analytical solution of the modified model was obtained using the complex potential method of elastic mechanics. The resultant force acting on the dislocation was calculated, and the analytical expression for the critical stress intensity factor corresponding to dislocation emission was obtained based on the dislocation emission criterion. The influence of the size, position, strength of nanoscale amorphization, and bifurcation crack shape on the critical stress intensity factor was discussed using numerical analysis. The study found that an increase in the angle between the main crack and the branched crack makes it more difficult for dislocations to emit from the bifurcation crack tip. The critical dislocation emission angle is independent of the angle between the main crack and the branched crack. The presence of nanoscale amorphization can reduce the high stress field near the bifurcation crack tip, making it difficult for dislocations to emit from the bifurcation crack tip, thereby reducing the toughness of the material caused by dislocation emission
Effect of Surface Thermoelastic Deformation on the Performance of the Hydrodynamic Big-Size Step Bearing
No full textThe multiscale lubrication analysis is presented for estimating the performance of the hydrodynamic big-size step bearing by incorporating the effects of the surface thermoelastic deformation and the lubricant molecule layers physically adsorbed to the bearing surface. The numerical calculation results show that in the condition of heavy loads and high sliding speeds, the effect of the surface thermoelastic deformation can reduce the minimum surface separation by 1 to 2 orders, while the effect of the physically adsorbed layer on the bearing surface significantly increases the minimum surface separation especially for the strong fluid-bearing surface interaction; The effect of the surface thermoelastic deformation largely modifies both the film pressure profile and the surface separation profile in the bearing; It also obviously changes the friction coefficient of the bearing. The effect of the physically adsorbed layer significantly influences the friction coefficient of the bearing only in the condition of heavy loads and high sliding speeds, which yields very low surface separations.
 
Failure Analysis of Deep-Drawn Single Piece Pressure Vessels
No full textThis study presents a comprehensive failure analysis of fully deep-drawn, thin-walled pressure vessels used in fire extinguisher applications, focusing on the effects of deep drawing and welding processes. Hydrostatic burst pressure testing, tensile and microhardness measurements, microstructural evaluations, and coupled thermomechanical finite element analysis (FEA) were performed to identify the root causes of failure and assess the influence of weld-induced residual stresses.
The experimental results revealed that the weld heat-affected zone (HAZ), particularly at the curved head section of the vessel, exhibited reduced microhardness and structural integrity due to recrystallization and grain refinement. Hydrostatic testing confirmed that failure typically initiated in this thinned, weld-affected region, where residual tensile stresses were also found to be concentrated.
Numerical simulations further substantiated the experimental observations. Thermal-mechanical FEA demonstrated the presence of tensile residual stresses in the HAZ and identified stress concentrations aligned with experimentally observed crack zones. A two-stage FEA approach, incorporating both thermal and structural analyses, was used to simulate weld heat input and internal pressure loading. The resulting stress distributions and crack propagation patterns, evaluated using a semi-elliptical crack model, revealed that Mode I crack opening was dominant, especially at the head section.
Comparative analysis of critical stress intensity factors between the shell and head sections showed that the head region had a significantly lower threshold, explaining its susceptibility to catastrophic fracture. The numerical predictions showed strong correlation with experimental hydrostatic burst test results, validating the use of FEA in predicting failure mechanisms in deep-drawn welded vessels.
Overall, the study highlights the critical role of weld-induced residual stresses and geometric thinning in determining failure zones in pressure vessels. The integration of experimental and numerical techniques offers a robust framework for evaluating structural integrity and improving the safety of welded deep-drawn pressure vessels in industrial applications
FACENet: A Fusion Atrous and Channel Enhancement Network for Remote Sensing Image Instance Segmentation
No full textThe instance segmentation task has been widely used in remote sensing. However, existing remote sensing instance segmentation models may lead to incomplete mask segmentation in complex and diverse background environments. In addition, commonly used feature fusion methods struggle to handle instances of different sizes well and predominantly suffer from loss of semantic information, failing to segment the mask accurately. To solve these problems, we propose a fusion atrous and channel enhancement network (FACENet) for the remote sensing image (RSI) instance segmentation. Specifically, we first replace the FPN with the FACE-FPN, which produces a more detailed pyramid by increasing the receptive field at the feature level. Second, we propose a semantic enhancement module for mining the rich semantic information of the underlying features. Then, we enhance the model\u27s adaptability to complex object deformations by introducing deformable convolution. Experiments on the iSAID, NWPU VHR-10, and HRSID datasets demonstrate that our proposed FACENet outperforms SOLOv2 in terms of average accuracy by 5.1%, 12.9%, and 7.6%, respectively, and beats other instance segmentation models
Improved Agricultural Machinery Navigation Algorithm Based on Machine Learning and Machine Vision Technology
No full textThe automatic navigation of agricultural machinery is one of the important directions in intelligent agriculture research. To realize the automatic production of agricultural machinery, the automatic planning of the navigation route for agricultural machinery is the key. Considering the complexity of the agricultural production environment, the agricultural machinery navigation model is constructed based on binocular vision technology, and the optimized BP network is used to calibrate the binocular vision model. Considering the difficulty in crop identification by traditional machine vision technology, RGB space technology is used to complete image segmentation and noise processing. The optimized S-RANSAC algorithm is used to extract image features. The experimental results showed that in the multi-algorithm agricultural rice field image feature matching test, the S-RANSAC algorithm accurately identified the color difference, shape difference, and hydrological environment difference of seedlings. In contrast, other algorithms were unable to identify complex environmental features. At the same time, in the complex agricultural environment positioning test, the maximum error of the S-RANSAC algorithm was 4.16m, which was better than 5.17m of SURF and had the best positioning performance. It can be seen that the proposed technology has excellent application effects in practical scenarios, providing important technical references for the intelligent development of agriculture and the innovation of visual navigation technology
Method of Ship Target Oblique Frame Detection in Lightweight SAR Image Based on Recurrent Neural Network
No full textWhen ship targets appear in SAR images at different angles, their shapes and contours may change significantly. At present, target box detection algorithms often match and recognize based on templates with fixed shapes and directions. When the angle of ship targets changes, these templates may no longer be applicable, leading to the decline of detection algorithm performance, and it is difficult to accurately identify and locate targets. Therefore, for the purpose of solving the problem of angle sensitivity, the method of ship target oblique frame detection in lightweight SAR image based on recurrent neural network is studied to improve the effect of ship target oblique frame detection. Using recurrent neural network, the framework of ship target oblique frame detection in lightweight SAR images is established to ensure the detection accuracy, significantly reduce the demand for computing resources, and achieve more efficient detection. In this framework, SAR images are input in the input layer and transmitted to the hidden layer. The lightweight convolutional neural network is used as the hidden layer, and channel attention mechanism is introduced to improve the extraction effect of useful ship target features. The output layer processes the ship target characteristics, predicts the ship target center point heat map, and calculates the oblique frame vertex coordinates of the center point heat map, so as to have better adaptability to the ship targets that tilt or rotate in the SAR image, solve the angle sensitivity problem, and complete the ship target oblique frame detection. The volume Kalman filter algorithm is used to train the recurrent neural network, optimize the network weight, and improve the detection accuracy of ship target oblique frame. Experiments show that this method can effectively extract ship target features. Under different background, this method can accurately detect the slant frame of ship target. Under different occlusion rates, the robustness of the method is better.
Research on Real Time Prediction Method of Kiln Flame Temperature Based on 5G Communication and CA-ResNet50 Fusion Network
No full textAs intermittent kilns, shuttle kilns are often used in the production of daily-use ceramics. The temperature has a significant impact on the products inside the kiln, and currently, most shuttle kilns still rely on human observation of the flame to adjust the temperature, which has uncertainties and limitations. This paper proposes a real-time prediction method for kiln flame temperature based on 5G communication and CA-ResNet50 fusion network, which utilizes the low latency and high bandwidth characteristics of 5G networks to collect real-time data and ensure the correspondence between flame images and temperature. And combine the CA (Coordinate attention) mechanism with the ResNet50 network to improve the network\u27s attention to flame image features, thereby enhancing prediction accuracy. The experimental results show that the proposed method can improve the accuracy of temperature prediction based on flame images, providing new ideas for temperature control in shuttle kilns.
ORPTQ: An Improved Large Model Quantization Method Based on Optimal Quantization Range
No full textQuantization reduces model storage by representing model in low bits. It can help to improve the application capability of transformer-based large models and make them possible to be deployed on resource-limited systems such as PCs and mobile devices. The best weight-only quantization method currently is to use second-order information to fine-tune the weight step by step during the quantization process, compensating for the quantization errors that have occurred. The method can minimize the functional loss of weight due to quantization by adjusting the remaining elements through algebraic transformations in each step. However, the performance of this quantization method will deteriorate rapidly when the adjustment for weight deviates too far from the starting point, especially in low-bit quantization (e.g. 4 bits or fewer). To meet the mathematical prerequisite of this method in the quantization, this paper introduces two parameters α, β to adjust the quantization range based on the second-order method, and presents three approaches to seek their optimal values. The experimental results show that the performance of the proposed method significantly outperforms the original second-order method in low-bit quantization. The code of this paper is available on github.com/t-scen/ORPTQ.