Non-destructive testing （NDT） has been widely used in many fields, we can easily see it is being used in shipbuilding, aerospace, weapons manufacturing and so on. ICT is one of the best non destructive testing metho...Non-destructive testing （NDT） has been widely used in many fields, we can easily see it is being used in shipbuilding, aerospace, weapons manufacturing and so on. ICT is one of the best non destructive testing methods currently, but it has not been used widely, because it requires much compute and costs much time. Defect location is one of the most important processing steps in the digital image analysis. Defect location would correspondingly reduce the time spent in the testing. We may only require locating the defects in some case. So, we divide the CT images into several little blocks which the square is equal, and then calculate the fractal on each block. By determining the value and connected region number of the fractal dimension, we can locate defects of the image. The results show that the block ffactal dimension is a useful and time-saving defect location method.展开更多
In industrial manufacturing,accurate and efficient identification of product surface defects is essential for ensuring product quality,optimizing the production process and reducing cost.However,complex and diverse de...In industrial manufacturing,accurate and efficient identification of product surface defects is essential for ensuring product quality,optimizing the production process and reducing cost.However,complex and diverse defect morphologies and the need for fine-grained description present significant challenges.General image description methods based on large visual language models often struggle to provide accurate defect type and location information for specific areas such as steel surface defect recognition.To address this,a defect identification method for the Qwen2.5-VL-3B large model based on LoRA finetuning is proposed.We built a specialized dataset covering six key steel surface defects-cracks,impurities,plaques,pitting,scale penetration,and scratches-and refined the model through efficient low-rank adaptation.Experimental results demonstrate that the fine-tuned Qwen2.5-VL-3B model significantly improves industrial defect recognition,accurately identifying defect types and locations,thus overcoming limitations of general large models and providing an efficient solution for industrial inspection.展开更多
An underground roadway usually contains defects of various types,and when the roadway is subjected to external loading,the locations of those defects influence the roadway by differing degrees.In this study,to study h...An underground roadway usually contains defects of various types,and when the roadway is subjected to external loading,the locations of those defects influence the roadway by differing degrees.In this study,to study how the locations of defects affect crack propagation in a roadway,specimens with tunnel-type voids were made using polymethyl methacrylate,and the stress wave produced by a bullet impacting an incident rod was used as the impact load.Meanwhile,the variations in crack speed,displacement,and dynamic stress intensity factor during crack propagation were obtained using an experimental system of digital laser dynamic caustics,and the commercial software ABAQUS was used for numerical simulations.From the experiments and numerical simulations,the crack propagation path was verified and the impact fracture behavior of a semicircular-arch roadway with different defect positions was presented.The results show that when the pre-fabricated crack is on the central axis of the sample,the crack propagation is purely mode I;when the pre-fabricated crack is 5 mm from the central axis,the crack propagation alternates between mode I and a mixture of modes I and II;when the pre-fabricated crack is at the edge of the semicircular-arch roadway,the crack propagation follows the I-II mixed mode.展开更多
基金Supported by National Nature Science Foundation of China(No.61171177)
文摘Non-destructive testing （NDT） has been widely used in many fields, we can easily see it is being used in shipbuilding, aerospace, weapons manufacturing and so on. ICT is one of the best non destructive testing methods currently, but it has not been used widely, because it requires much compute and costs much time. Defect location is one of the most important processing steps in the digital image analysis. Defect location would correspondingly reduce the time spent in the testing. We may only require locating the defects in some case. So, we divide the CT images into several little blocks which the square is equal, and then calculate the fractal on each block. By determining the value and connected region number of the fractal dimension, we can locate defects of the image. The results show that the block ffactal dimension is a useful and time-saving defect location method.
文摘In industrial manufacturing,accurate and efficient identification of product surface defects is essential for ensuring product quality,optimizing the production process and reducing cost.However,complex and diverse defect morphologies and the need for fine-grained description present significant challenges.General image description methods based on large visual language models often struggle to provide accurate defect type and location information for specific areas such as steel surface defect recognition.To address this,a defect identification method for the Qwen2.5-VL-3B large model based on LoRA finetuning is proposed.We built a specialized dataset covering six key steel surface defects-cracks,impurities,plaques,pitting,scale penetration,and scratches-and refined the model through efficient low-rank adaptation.Experimental results demonstrate that the fine-tuned Qwen2.5-VL-3B model significantly improves industrial defect recognition,accurately identifying defect types and locations,thus overcoming limitations of general large models and providing an efficient solution for industrial inspection.
基金This work was financed by the State Key Development Program for Basic Research of China(2016YFC0600903)the National Natural Science Foundation of China(51774287).
文摘An underground roadway usually contains defects of various types,and when the roadway is subjected to external loading,the locations of those defects influence the roadway by differing degrees.In this study,to study how the locations of defects affect crack propagation in a roadway,specimens with tunnel-type voids were made using polymethyl methacrylate,and the stress wave produced by a bullet impacting an incident rod was used as the impact load.Meanwhile,the variations in crack speed,displacement,and dynamic stress intensity factor during crack propagation were obtained using an experimental system of digital laser dynamic caustics,and the commercial software ABAQUS was used for numerical simulations.From the experiments and numerical simulations,the crack propagation path was verified and the impact fracture behavior of a semicircular-arch roadway with different defect positions was presented.The results show that when the pre-fabricated crack is on the central axis of the sample,the crack propagation is purely mode I;when the pre-fabricated crack is 5 mm from the central axis,the crack propagation alternates between mode I and a mixture of modes I and II;when the pre-fabricated crack is at the edge of the semicircular-arch roadway,the crack propagation follows the I-II mixed mode.
