Manual inspection of onba earing casting defects is not realistic and unreliable,particularly in the case of some micro-level anomalies which lead to major defects on a large scale.To address these challenges,we propo...Manual inspection of onba earing casting defects is not realistic and unreliable,particularly in the case of some micro-level anomalies which lead to major defects on a large scale.To address these challenges,we propose BearFusionNet,an attention-based deep learning architecture with multi-stream,which merges both DenseNet201 and MobileNetV2 for feature extraction with a classification head inspired by VGG19.This hybrid design,figuratively beaming from one layer to another,extracts the enormity of representations on different scales,backed by a prepreprocessing pipeline that brings defect saliency to the fore through contrast adjustment,denoising,and edge detection.The use of multi-head self-attention enhances feature fusion,enabling the model to capture both large and small spatial features.BearFusionNet achieves an accuracy of 99.66%and Cohen’s kappa score of 0.9929 in Kaggle’s Real-life Industrial Casting Defects dataset.Both McNemar’s and Wilcoxon signed-rank statistical tests,as well as fivefold cross-validation,are employed to assess the robustness of our proposed model.To interpret the model,we adopt Grad-Cam visualizations,which are the state of the art standard.Furthermore,we deploy BearFusionNet as a webbased system for near real-time inference(5-6 s per prediction),which enables the quickest yet accurate detection with visual explanations.Overall,BearFusionNet is an interpretable,accurate,and deployable solution that can automatically detect casting defects,leading to significant advances in the innovative industrial environment.展开更多
The impact of casting defects on the weldability of K4951 superalloy was investigated using tungsten inert gas(TIG)welding.The as-cast K4951 superalloy samples with prefabricated U-shaped grooves of varying depths and...The impact of casting defects on the weldability of K4951 superalloy was investigated using tungsten inert gas(TIG)welding.The as-cast K4951 superalloy samples with prefabricated U-shaped grooves of varying depths and widths were TIG welded,and the microstructures,cracks morphology,and precipitated phases were analyzed using optical microscope,scanning electron microscope,transmission electron microscope,and energy dispersive X-ray spectrometer.The results reveal that the dimensions of casting defects significantly affect the weldability of K4951.Deep defects(greater than 2 mm)lead to rapid crack propagation,while wider defects can moderate the propagation process of cracks.Elemental segregation and the formation of precipitated phases,such as MC carbides,are observed in the fusion zone,contributing to welding cracks.An optimal groove aspect ratio(depth-to-width)between 0.2 and 0.5 minimizes crack formation tendency and enhances tensile strength,resulting in a mixed brittle-ductile fracture mode of joint after high-temperature tensile testing.展开更多
Automatic surface defect detection is a critical technique for ensuring product quality in industrial casting production.While general object detection techniques have made remarkable progress over the past decade,cas...Automatic surface defect detection is a critical technique for ensuring product quality in industrial casting production.While general object detection techniques have made remarkable progress over the past decade,casting surface defect detection still has considerable room for improvement.Lack of sufficient and high-quality data has become one of the most challenging problems for casting surface defect detection.In this paper,we construct a new casting surface defect dataset(CSDD)containing 2100 high-resolution images of casting surface defects and 56356 defects in total.The class and defect region for each defect are manually labeled.We conduct a series of experiments on this dataset using multiple state-of-the-art object detection methods,establishing a comprehensive set of baselines.We also propose a defect detection method based on YOLOv5 with the global attention mechanism and partial convolution.Our proposed method achieves superior performance compared to other object detection methods.Additionally,we also conduct a series of experiments with multiple state-of-the-art semantic segmentation methods,providing extensive baselines for defect segmentation.To the best of our knowledge,the CSDD has the largest number of defects for casting surface defect detection and segmentation.It would benefit both the industrial vision research and manufacturing applications.Dataset and code are available at https://github.com/Kerio99/CSDD.展开更多
Casting technology of thin-wall TiAl alloy turbochargers was studied by investment casting and numerical simulation.Misruns and gas holes were the main defects observed in preliminary work due to the poor fluidity of ...Casting technology of thin-wall TiAl alloy turbochargers was studied by investment casting and numerical simulation.Misruns and gas holes were the main defects observed in preliminary work due to the poor fluidity of alloy,and to gas entrapment.In order to eliminate these defects,cast parameters,such as centrifugal rotation rate and mould preheating temperature,were optimized by numerical simulation,meanwhile,the structure of the shell mould was optimized to improve the filling capacity of TiAl alloy.Pouring experiments were carried out by vacuum induction melting furnace equipped with a water-cooled copper crucible based on the above optimization.The quality of the TiAl alloy casting was analyzed by fluorescent penetrant inspection and X-ray detection.The results show that a centrifugal rotation rate of 200 rpm,mould preheating temperature of 600°C,shell preparation through organic fiber addition can dramatically improve the mould filling capacity,and integrated turbochargers were finally prepared.展开更多
The effects of Sr addition and pressure increase on the microstructure and casting defects of a low-pressure die cast （LPDC） AISi7Mg0.3 alloy have been studied. Metallographic and image analysis techniques have been...The effects of Sr addition and pressure increase on the microstructure and casting defects of a low-pressure die cast （LPDC） AISi7Mg0.3 alloy have been studied. Metallographic and image analysis techniques have been used to quantitatively examine the microstructural changes and the amount of porosity occurring at different Sr levels and pressure parameters. The results indicate that an increase in the filling pressure induces lower heat dissipation of the liquid close to the die/core surfaces, with the formation of slightly greater dendrite arms and coarser eutectic Si particles. On the other hand, the increase in the Sr level leads to finer microstructural scale and eutectic Si. The analysed variables, within the experimental conditions, do not affect the morphology of eutectic Si particles. Higher applied pressure and Sr content generate castings with lower amount of porosiW. However, as the filling pressure increases the flow of metal inside the die cavity is more turbulent, leading to the formation of oxide films and cold shots. In the analysed range of experimental conditions, the design of experiment methodology and the analysis of variance have been used to develop statistical models that accurately predict the average size of secondary dendrite arm spacing and the amount of porosity in the low-pressure die cast AISiTMg0.3 alloy.展开更多
The current casting surface defect detection algorithms suffer from poor small target defect recognition and imbalance between detection performance and detection time.An improved algorithmic framework for casting def...The current casting surface defect detection algorithms suffer from poor small target defect recognition and imbalance between detection performance and detection time.An improved algorithmic framework for casting defect detection was proposed based on the DEtection TRansformer(DETR)algorithm.The algorithm takes ResNet with an efficient channel attention(ECA)-Net module as the backbone network.In addition,based on the original algorithm architecture,dynamic anchor boxes,improved multi-scale deformable attention module,and SIoU loss function are introduced to improve the sensitivity of transformer structure to input location information and scale size,and the small target defect detection performance is effectively improved.The recognition performance of the algorithm in a self-built casting defect dataset was studied.The improved DETR algorithm has 97.561% accuracy in recognizing two defects,namely sandinclusion and notch,with the detection rate being improved by 65.854% and 17.073% compared with the original DETR and you only look once(Yolo)-V5,respectively.This algorithm verifies the applicability of the transformer architecture target detection algorithm for casting defect detection tasks and provides new ideas for detecting other similar application scenarios.展开更多
In order to study the effect of Zr modification and riser size on microporosity defect distributions in WE54 alloy sand castings, the microporosity volume percentage in Zr-free and Zr-containing WE54 alloy plate casti...In order to study the effect of Zr modification and riser size on microporosity defect distributions in WE54 alloy sand castings, the microporosity volume percentage in Zr-free and Zr-containing WE54 alloy plate castings was determined by density measurement based on Archimedes＇ principle, and the microstructure of the microporosity defects was observed by optical microscopy and scanning electron microscopy. Then by using Procast software, the Niyama criterion was calculated in order to investigate the validity of Niyama criterion on prediction of microporosity defects in WE54 alloy sand castings. It is found from the density measurement results that Zr addition does not affect the microporosity distributions in WE54 alloy castings. While the distribution area of microporosity defect in the plate castings decreases significantly as the riser size increases. Based on the experimental results, a riser selection principle for production of compact WE54 alloy castings is proposed that the solidification modulus of the riser should be greater than that of the casting by 30%, simply mr ≥ 1.3mc. By comparing the experimental and simulating results, it is found that the predicted microporosity regions by Niyama criterion agrees well with experimental results, and a critical Niyama value of 0.4 ℃0.5 s0.5 mm-1 is suggested for prediction of microporosity formation in WE54 alloy sand castings.展开更多
The shrinkage defect of a ductile iron casting is attributed to the volume variations occurring in solidification, which consist of liquid contraction, solidification shrinkage, graphitization expansion, and mold cavi...The shrinkage defect of a ductile iron casting is attributed to the volume variations occurring in solidification, which consist of liquid contraction, solidification shrinkage, graphitization expansion, and mold cavity enlargement. Based on this understanding, a mathematical model for predicting the shrinkage defect of the casting is developed, in which the volume variations of the casting in soli- dification are numerically simulated, especially, the mold cavity enlargement is quantitatively calculated. Moreover, the reliability of the model is verified in production and experiment.展开更多
基金funded by Multimedia University,Cyberjaya,Selangor,Malaysia(Grant Number:PostDoc(MMUI/240029)).
文摘Manual inspection of onba earing casting defects is not realistic and unreliable,particularly in the case of some micro-level anomalies which lead to major defects on a large scale.To address these challenges,we propose BearFusionNet,an attention-based deep learning architecture with multi-stream,which merges both DenseNet201 and MobileNetV2 for feature extraction with a classification head inspired by VGG19.This hybrid design,figuratively beaming from one layer to another,extracts the enormity of representations on different scales,backed by a prepreprocessing pipeline that brings defect saliency to the fore through contrast adjustment,denoising,and edge detection.The use of multi-head self-attention enhances feature fusion,enabling the model to capture both large and small spatial features.BearFusionNet achieves an accuracy of 99.66%and Cohen’s kappa score of 0.9929 in Kaggle’s Real-life Industrial Casting Defects dataset.Both McNemar’s and Wilcoxon signed-rank statistical tests,as well as fivefold cross-validation,are employed to assess the robustness of our proposed model.To interpret the model,we adopt Grad-Cam visualizations,which are the state of the art standard.Furthermore,we deploy BearFusionNet as a webbased system for near real-time inference(5-6 s per prediction),which enables the quickest yet accurate detection with visual explanations.Overall,BearFusionNet is an interpretable,accurate,and deployable solution that can automatically detect casting defects,leading to significant advances in the innovative industrial environment.
基金National Natural Science Foundation of China(52201054,52175368)National Science and Technology Major Projects(J2019-VI-0018-0133)+2 种基金Liaoning Provincial Science and Technology Program(2023-BS-019,2023-MS-020)National Key R&D Program of China(2021YFB3700401)Key Specialized Research and Development Break-Through-Unveiling and Commanding the Special Project Program in Liaoning Province(2021JH15)。
文摘The impact of casting defects on the weldability of K4951 superalloy was investigated using tungsten inert gas(TIG)welding.The as-cast K4951 superalloy samples with prefabricated U-shaped grooves of varying depths and widths were TIG welded,and the microstructures,cracks morphology,and precipitated phases were analyzed using optical microscope,scanning electron microscope,transmission electron microscope,and energy dispersive X-ray spectrometer.The results reveal that the dimensions of casting defects significantly affect the weldability of K4951.Deep defects(greater than 2 mm)lead to rapid crack propagation,while wider defects can moderate the propagation process of cracks.Elemental segregation and the formation of precipitated phases,such as MC carbides,are observed in the fusion zone,contributing to welding cracks.An optimal groove aspect ratio(depth-to-width)between 0.2 and 0.5 minimizes crack formation tendency and enhances tensile strength,resulting in a mixed brittle-ductile fracture mode of joint after high-temperature tensile testing.
基金supported by the National Natural Science Foundation of China(U23B2060,62088102)the Key Research and Development Program of China(2020AAA0108305).
文摘Automatic surface defect detection is a critical technique for ensuring product quality in industrial casting production.While general object detection techniques have made remarkable progress over the past decade,casting surface defect detection still has considerable room for improvement.Lack of sufficient and high-quality data has become one of the most challenging problems for casting surface defect detection.In this paper,we construct a new casting surface defect dataset(CSDD)containing 2100 high-resolution images of casting surface defects and 56356 defects in total.The class and defect region for each defect are manually labeled.We conduct a series of experiments on this dataset using multiple state-of-the-art object detection methods,establishing a comprehensive set of baselines.We also propose a defect detection method based on YOLOv5 with the global attention mechanism and partial convolution.Our proposed method achieves superior performance compared to other object detection methods.Additionally,we also conduct a series of experiments with multiple state-of-the-art semantic segmentation methods,providing extensive baselines for defect segmentation.To the best of our knowledge,the CSDD has the largest number of defects for casting surface defect detection and segmentation.It would benefit both the industrial vision research and manufacturing applications.Dataset and code are available at https://github.com/Kerio99/CSDD.
基金financially supported by the Liaoning Natural Science Foundation ( Grant No.20170540888)the Liaoning Science and Technology Project (Grant No.2017221006)
文摘Casting technology of thin-wall TiAl alloy turbochargers was studied by investment casting and numerical simulation.Misruns and gas holes were the main defects observed in preliminary work due to the poor fluidity of alloy,and to gas entrapment.In order to eliminate these defects,cast parameters,such as centrifugal rotation rate and mould preheating temperature,were optimized by numerical simulation,meanwhile,the structure of the shell mould was optimized to improve the filling capacity of TiAl alloy.Pouring experiments were carried out by vacuum induction melting furnace equipped with a water-cooled copper crucible based on the above optimization.The quality of the TiAl alloy casting was analyzed by fluorescent penetrant inspection and X-ray detection.The results show that a centrifugal rotation rate of 200 rpm,mould preheating temperature of 600°C,shell preparation through organic fiber addition can dramatically improve the mould filling capacity,and integrated turbochargers were finally prepared.
文摘The effects of Sr addition and pressure increase on the microstructure and casting defects of a low-pressure die cast （LPDC） AISi7Mg0.3 alloy have been studied. Metallographic and image analysis techniques have been used to quantitatively examine the microstructural changes and the amount of porosity occurring at different Sr levels and pressure parameters. The results indicate that an increase in the filling pressure induces lower heat dissipation of the liquid close to the die/core surfaces, with the formation of slightly greater dendrite arms and coarser eutectic Si particles. On the other hand, the increase in the Sr level leads to finer microstructural scale and eutectic Si. The analysed variables, within the experimental conditions, do not affect the morphology of eutectic Si particles. Higher applied pressure and Sr content generate castings with lower amount of porosiW. However, as the filling pressure increases the flow of metal inside the die cavity is more turbulent, leading to the formation of oxide films and cold shots. In the analysed range of experimental conditions, the design of experiment methodology and the analysis of variance have been used to develop statistical models that accurately predict the average size of secondary dendrite arm spacing and the amount of porosity in the low-pressure die cast AISiTMg0.3 alloy.
基金the support of National Natural Science Foundation of China(No.51405002)Anhui Provincial Natural Science Foundation(No.2108085ME173)+2 种基金open funds from Anhui Province Key Laboratory of Metallurgical Engineering&Resources Recycling(No.SKF20-05)Opening Project of Engineering Technology Research Center of Anhui Education Department for Energy Saving and Pollutant Control in metallurgical processOpening Project of Anhui Engineering Laboratory for Intelligent Applications and Security of Industrial Internet(Grant No.IASII21-03)for financial support.
文摘The current casting surface defect detection algorithms suffer from poor small target defect recognition and imbalance between detection performance and detection time.An improved algorithmic framework for casting defect detection was proposed based on the DEtection TRansformer(DETR)algorithm.The algorithm takes ResNet with an efficient channel attention(ECA)-Net module as the backbone network.In addition,based on the original algorithm architecture,dynamic anchor boxes,improved multi-scale deformable attention module,and SIoU loss function are introduced to improve the sensitivity of transformer structure to input location information and scale size,and the small target defect detection performance is effectively improved.The recognition performance of the algorithm in a self-built casting defect dataset was studied.The improved DETR algorithm has 97.561% accuracy in recognizing two defects,namely sandinclusion and notch,with the detection rate being improved by 65.854% and 17.073% compared with the original DETR and you only look once(Yolo)-V5,respectively.This algorithm verifies the applicability of the transformer architecture target detection algorithm for casting defect detection tasks and provides new ideas for detecting other similar application scenarios.
基金supported by the National Basic Research Program of China(973 Program,No.2013CB632202)
文摘In order to study the effect of Zr modification and riser size on microporosity defect distributions in WE54 alloy sand castings, the microporosity volume percentage in Zr-free and Zr-containing WE54 alloy plate castings was determined by density measurement based on Archimedes＇ principle, and the microstructure of the microporosity defects was observed by optical microscopy and scanning electron microscopy. Then by using Procast software, the Niyama criterion was calculated in order to investigate the validity of Niyama criterion on prediction of microporosity defects in WE54 alloy sand castings. It is found from the density measurement results that Zr addition does not affect the microporosity distributions in WE54 alloy castings. While the distribution area of microporosity defect in the plate castings decreases significantly as the riser size increases. Based on the experimental results, a riser selection principle for production of compact WE54 alloy castings is proposed that the solidification modulus of the riser should be greater than that of the casting by 30%, simply mr ≥ 1.3mc. By comparing the experimental and simulating results, it is found that the predicted microporosity regions by Niyama criterion agrees well with experimental results, and a critical Niyama value of 0.4 ℃0.5 s0.5 mm-1 is suggested for prediction of microporosity formation in WE54 alloy sand castings.
文摘The shrinkage defect of a ductile iron casting is attributed to the volume variations occurring in solidification, which consist of liquid contraction, solidification shrinkage, graphitization expansion, and mold cavity enlargement. Based on this understanding, a mathematical model for predicting the shrinkage defect of the casting is developed, in which the volume variations of the casting in soli- dification are numerically simulated, especially, the mold cavity enlargement is quantitatively calculated. Moreover, the reliability of the model is verified in production and experiment.
