Ndt & E International最新文献_第6页

Translation of MFL and UT data by using generative adversarial networks: A comparative study 使用生成式对抗网络翻译 MFL 和 UT 数据：比较研究

IF 4.1 2区材料科学 Q1 MATERIALS SCIENCE, CHARACTERIZATION & TESTING

Ndt & E International

Pub Date : 2024-10-23 DOI: 10.1016/j.ndteint.2024.103246

Jiatong Ling , Xiang Peng , Matthias Peussner , Kevin Siggers , Zheng Liu

Magnetic flux leakage (MFL) and ultrasonic testing (UT) are widely used in-line inspection technologies to detect corrosion defects along pipelines. The integration of MFL and UT data has the potential to provide complementary insights that facilitate a comprehensive assessment of pipeline integrity. However, due to the inherent dissimilarity with their underlying physical principles, these techniques yield notable disparities in signal characteristics, posing challenges in integrating these multimodal data. This study aims to establish a translation mapping between MFL and UT signals to achieve consistent physical interpretations across the two modalities. Thus, this study explored the feasibility of generative adversarial network (GAN) based models encompassing both supervised and unsupervised translation approaches contingent on the availability of aligned data. Furthermore, two translation modes, MFL-UT and UT-MFL, were analyzed separately to understand the effectiveness of the translation direction. The experimental results demonstrate satisfactory performance for both aligned and unaligned data translation, with the UT-MFL translation direction yielding superior results. Overall, the translation approaches pave the way for future applications, especially in subsequent data analysis tasks such as registration, comparison, and fusion of multimodal data.

磁通量泄漏（MFL）和超声波测试（UT）是广泛使用的在线检测技术，用于检测管道的腐蚀缺陷。磁通量泄漏和超声波检测数据的整合有可能提供互补的见解，从而促进对管道完整性的全面评估。然而，由于其基本物理原理存在固有的差异，这些技术产生的信号特征存在明显的差异，这给整合这些多模态数据带来了挑战。本研究旨在建立 MFL 和 UT 信号之间的转换映射，以实现两种模式之间一致的物理解释。因此，本研究探索了基于生成对抗网络（GAN）的模型的可行性，该模型包含监督和非监督翻译方法，取决于是否有对齐的数据。此外，还分别分析了 MFL-UT 和 UT-MFL 两种翻译模式，以了解翻译方向的有效性。实验结果表明，对齐和非对齐数据翻译的性能都令人满意，UT-MFL 翻译方向的结果更优。总之，这些平移方法为未来的应用，尤其是后续的数据分析任务（如多模态数据的注册、比较和融合）铺平了道路。

{"title":"Translation of MFL and UT data by using generative adversarial networks: A comparative study","authors":"Jiatong Ling , Xiang Peng , Matthias Peussner , Kevin Siggers , Zheng Liu","doi":"10.1016/j.ndteint.2024.103246","DOIUrl":"10.1016/j.ndteint.2024.103246","url":null,"abstract":"<div><div>Magnetic flux leakage (MFL) and ultrasonic testing (UT) are widely used in-line inspection technologies to detect corrosion defects along pipelines. The integration of MFL and UT data has the potential to provide complementary insights that facilitate a comprehensive assessment of pipeline integrity. However, due to the inherent dissimilarity with their underlying physical principles, these techniques yield notable disparities in signal characteristics, posing challenges in integrating these multimodal data. This study aims to establish a translation mapping between MFL and UT signals to achieve consistent physical interpretations across the two modalities. Thus, this study explored the feasibility of generative adversarial network (GAN) based models encompassing both supervised and unsupervised translation approaches contingent on the availability of aligned data. Furthermore, two translation modes, MFL-UT and UT-MFL, were analyzed separately to understand the effectiveness of the translation direction. The experimental results demonstrate satisfactory performance for both aligned and unaligned data translation, with the UT-MFL translation direction yielding superior results. Overall, the translation approaches pave the way for future applications, especially in subsequent data analysis tasks such as registration, comparison, and fusion of multimodal data.</div></div>","PeriodicalId":18868,"journal":{"name":"Ndt & E International","volume":"149 ","pages":"Article 103246"},"PeriodicalIF":4.1,"publicationDate":"2024-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142553896","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Development of a flexible phased array electromagnetic acoustic testing system with array pickups 开发带阵列拾波器的柔性相控阵电磁声学测试系统

IF 4.1 2区材料科学 Q1 MATERIALS SCIENCE, CHARACTERIZATION & TESTING

Ndt & E International

Pub Date : 2024-10-21 DOI: 10.1016/j.ndteint.2024.103263

Jie Deng, Yuange Zhang, Yinqiang Qu, Cuixiang Pei, Tianhao Liu, Zhenmao Chen

The thermal barrier coatings (TBCs) used to protect blades of heavy-duty gas turbines is prone to debonding defects during fabrication and service. It is difficult to detect the debonding defects in the TBC system non-destructively and in situ due to the curved shape of blade and the narrow space between the blades setting in the gas turbine. In this paper, a flexible phased array electromagnetic acoustic (FPA-EMA) testing system with array pickups is developed, which is capable to drive a flexible phased array electromagnetic acoustic transducer (FPA-EMAT) to generate and receive ultrasonic wave directly in the metallic substrate of TBCs, and is potential to be applied to the inspection of debonding defects in the TBC system of gas turbine blades. A phased array excitation unit, a multi-channel signal receiving unit and a new signal post-processing algorithm are developed for the new EMA testing system to enhance the surface acoustic wave in the layered structure and to improve the low conversion efficiency of the conventional EMA testing system. In addition, a bias magnetic field coil fed with long pulse current of large amplitude is adopted to make the new EMAT thin and flexible in order to be applied in a narrow space. The interfering noise and geometric size problems brought by the multiple excitation and pickup channels are well addressed in the new system. The good performances of the developed FPA-EMA testing system were verified experimentally and show the system of good potential to be applied to NDT of practical curved structures.

用于保护重型燃气轮机叶片的隔热涂层（TBC）在制造和使用过程中容易出现脱粘缺陷。由于燃气轮机叶片的弯曲形状和叶片之间的狭窄空间，很难对 TBC 系统的脱粘缺陷进行非破坏性的现场检测。本文开发了一种带阵列拾波器的柔性相控阵电磁声学（FPA-EMA）测试系统，该系统能够驱动柔性相控阵电磁声学换能器（FPA-EMAT）直接在 TBC 的金属基体上产生和接收超声波，有望应用于燃气轮机叶片 TBC 系统脱粘缺陷的检测。新型 EMA 检测系统采用相控阵激励单元、多通道信号接收单元和新型信号后处理算法，以增强分层结构中的表面声波，改善传统 EMA 检测系统转换效率低的问题。此外，还采用了馈入大振幅长脉冲电流的偏置磁场线圈，使新型电磁超声测试系统变得轻薄灵活，以便在狭窄空间中应用。新系统很好地解决了多激励和拾波通道带来的干扰噪声和几何尺寸问题。实验验证了所开发的 FPA-EMA 测试系统的良好性能，并表明该系统具有应用于实际曲面结构无损检测的良好潜力。

{"title":"Development of a flexible phased array electromagnetic acoustic testing system with array pickups","authors":"Jie Deng, Yuange Zhang, Yinqiang Qu, Cuixiang Pei, Tianhao Liu, Zhenmao Chen","doi":"10.1016/j.ndteint.2024.103263","DOIUrl":"10.1016/j.ndteint.2024.103263","url":null,"abstract":"<div><div>The thermal barrier coatings (TBCs) used to protect blades of heavy-duty gas turbines is prone to debonding defects during fabrication and service. It is difficult to detect the debonding defects in the TBC system non-destructively and in situ due to the curved shape of blade and the narrow space between the blades setting in the gas turbine. In this paper, a flexible phased array electromagnetic acoustic (FPA-EMA) testing system with array pickups is developed, which is capable to drive a flexible phased array electromagnetic acoustic transducer (FPA-EMAT) to generate and receive ultrasonic wave directly in the metallic substrate of TBCs, and is potential to be applied to the inspection of debonding defects in the TBC system of gas turbine blades. A phased array excitation unit, a multi-channel signal receiving unit and a new signal post-processing algorithm are developed for the new EMA testing system to enhance the surface acoustic wave in the layered structure and to improve the low conversion efficiency of the conventional EMA testing system. In addition, a bias magnetic field coil fed with long pulse current of large amplitude is adopted to make the new EMAT thin and flexible in order to be applied in a narrow space. The interfering noise and geometric size problems brought by the multiple excitation and pickup channels are well addressed in the new system. The good performances of the developed FPA-EMA testing system were verified experimentally and show the system of good potential to be applied to NDT of practical curved structures.</div></div>","PeriodicalId":18868,"journal":{"name":"Ndt & E International","volume":"149 ","pages":"Article 103263"},"PeriodicalIF":4.1,"publicationDate":"2024-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142531807","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A Mobius transformation-based liftoff effect reduction method for crack classification and prediction in eddy current testing 基于莫比乌斯变换的升力效应减弱法，用于涡流测试中的裂纹分类和预测

IF 4.1 2区材料科学 Q1 MATERIALS SCIENCE, CHARACTERIZATION & TESTING

Ndt & E International

Pub Date : 2024-10-21 DOI: 10.1016/j.ndteint.2024.103261

Yu Li , Zihan Xia , Saibo She , Yuchun Shao , Yinchao Yang , Wuliang Yin

In eddy current (EC) defect evaluation, accurate determination of the crack depth is crucial. The lift-off signal on the complex impedance plane is not necessarily horizontal and mostly arc-shaped, which makes it difficult to separate from crack signals, which can have various orientations depending on probe characteristics and frequencies. Phase rotation is normally used to align the lift-off signal to the real axis and defect signal is read from the imaginary axis. However, this approach proves challenging to separate lift-off and defect signals within the complex plane in many cases.

To address this issue, this study proposes and implements a Mobius transformation that linearizes the arc-shaped lift-off signals on the complex plane, enabling more accurate extraction of crack signals unaffected by lift-off signal height. Furthermore, it is demonstrated through k-nearest neighbors (kNN) that the transformed crack signals can still be effectively quantified.

在电涡流（EC）缺陷评估中，准确确定裂纹深度至关重要。复阻抗平面上的离析信号不一定是水平的，大多呈弧形，因此很难将其与裂纹信号区分开来，而裂纹信号可能因探头特性和频率的不同而有不同的方向。通常采用相位旋转的方法将离析信号对准实轴，然后从虚轴读取缺陷信号。为解决这一问题，本研究提出并实施了一种莫比乌斯变换（Mobius transformation），该变换可在复平面上对弧形离析信号进行线性化处理，从而更准确地提取不受离析信号高度影响的裂纹信号。此外，通过 k-nearest neighbors (kNN) 方法证明，变换后的裂纹信号仍可有效量化。

引用次数: 0

Low-frequency magnetic incremental permeability for the non-destructive evaluation of hardness profile after carburization treatment with large case hardening depth 用于无损评估大表面淬火深度渗碳处理后硬度曲线的低频磁增量渗透技术

IF 4.1 2区材料科学 Q1 MATERIALS SCIENCE, CHARACTERIZATION & TESTING

Ndt & E International

Pub Date : 2024-10-19 DOI: 10.1016/j.ndteint.2024.103248

Hicham Lberni , Benjamin Ducharne , Hélène Petitpré , Jean-François Mogniotte , Yves Armand Tene Deffo , Fan Zhang , Christophe Gallais

Carburization treatment with large case hardening depth is a technical process to enhance steel parts' surface hardness and wear resistance. Accurate evaluation of this metallurgical treatment is crucial to prevent critical mechanical failures. Low-frequency magnetic incremental permeability (LF-MIP) emerges as a non-destructive surface technique well-suited for this purpose in the case of ferromagnetic parts. Although correlations between magnetic indicators obtained through LF-MIP characterization and deep carburization treatment have been demonstrated, they remain qualitative. In this study, we propose an innovative method to assess the entire hardness profile based on LF-MIP characterization. Experimental results and simulation data are integrated into a reference chart used for post-processing, enabling the prediction of hardness profiles for specimens in a blind test. With a relative Euclidean distance of less than 6 % between the method's predictions and destructive tests conducted on specimens treated with medium, deep, and intense intensities, we consider the method validated.

大表面硬化深度渗碳处理是一种提高钢零件表面硬度和耐磨性的技术工艺。准确评估这种冶金处理对防止关键机械故障至关重要。低频磁增量渗透率（LF-MIP）是一种非破坏性表面技术，非常适合铁磁性零件。虽然通过 LF-MIP 表征获得的磁性指标与深渗碳处理之间的相关性已经得到证实，但它们仍然是定性的。在本研究中，我们提出了一种基于 LF-MIP 表征评估整个硬度曲线的创新方法。实验结果和模拟数据被整合到一个用于后处理的参考图表中，从而能够预测盲测试样的硬度曲线。由于该方法的预测结果与对试样进行中度、深度和强度处理后进行的破坏性试验之间的相对欧氏距离小于 6%，因此我们认为该方法是有效的。

{"title":"Low-frequency magnetic incremental permeability for the non-destructive evaluation of hardness profile after carburization treatment with large case hardening depth","authors":"Hicham Lberni , Benjamin Ducharne , Hélène Petitpré , Jean-François Mogniotte , Yves Armand Tene Deffo , Fan Zhang , Christophe Gallais","doi":"10.1016/j.ndteint.2024.103248","DOIUrl":"10.1016/j.ndteint.2024.103248","url":null,"abstract":"<div><div>Carburization treatment with large case hardening depth is a technical process to enhance steel parts' surface hardness and wear resistance. Accurate evaluation of this metallurgical treatment is crucial to prevent critical mechanical failures. Low-frequency magnetic incremental permeability (LF-MIP) emerges as a non-destructive surface technique well-suited for this purpose in the case of ferromagnetic parts. Although correlations between magnetic indicators obtained through LF-MIP characterization and deep carburization treatment have been demonstrated, they remain qualitative. In this study, we propose an innovative method to assess the entire hardness profile based on LF-MIP characterization. Experimental results and simulation data are integrated into a reference chart used for post-processing, enabling the prediction of hardness profiles for specimens in a blind test. With a relative Euclidean distance of less than 6 % between the method's predictions and destructive tests conducted on specimens treated with medium, deep, and intense intensities, we consider the method validated.</div></div>","PeriodicalId":18868,"journal":{"name":"Ndt & E International","volume":"149 ","pages":"Article 103248"},"PeriodicalIF":4.1,"publicationDate":"2024-10-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142531804","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Improving the industrial defect recognition in radiographic testing by pre-training on medical radiographs 通过对医学射线照片的预先培训，提高射线照片检测中的工业缺陷识别能力

IF 4.1 2区材料科学 Q1 MATERIALS SCIENCE, CHARACTERIZATION & TESTING

Ndt & E International

Pub Date : 2024-10-18 DOI: 10.1016/j.ndteint.2024.103260

Han Yu, Xingjie Li, Huasheng Xie, Xinyue Li, Chunyu Hou

Deep learning methodologies have gained substantial traction for defect recognition in industrial radiographic testing including welds, castings and other fields. Regardless of the deep learning utilized, it has emerged as a standard configuration to use a model pre-trained from ImageNet to accelerate convergence and enhance recognition accuracy. However, there is a significant gap between the domain of natural images and industrial radiographs, raising the question of whether there might be a superior pre-training method than relying on ImageNet pre-training. Fortunately, medical radiographs are more similar to industrial radiographs than natural images because of the same imaging method. In this paper, we initially utilize numerous medical radiographic images from CheXpert dataset to train a pre-trained CNN model. Then, we apply this model to four distinct tasks within two radiographic testing scenarios to validate its advantages and generalization capabilities. Finally, experiments on multiple datasets indicate that our method brings more benefits than ImageNet pre-training or training from scratch, with a F1 score improvement of 3.41 %–13.72 % for defect classification and a mIoU improvement of 1.05 %–6.58 % for defect segmentation. It demonstrates that pre-training from medical radiographs provides a cost-free improvement for all kinds of tasks in industrial defect recognition.

深度学习方法在工业射线检测（包括焊缝、铸件和其他领域）的缺陷识别中获得了广泛的应用。无论采用哪种深度学习方法，使用从 ImageNet 中预先训练好的模型来加速收敛并提高识别准确率已成为一种标准配置。然而，自然图像领域与工业射线照片领域之间存在巨大差距，这就提出了一个问题：是否有比依赖 ImageNet 预训练更优越的预训练方法？幸运的是，由于采用了相同的成像方法，医学射线照片与工业射线照片比自然图像更加相似。在本文中，我们首先利用 CheXpert 数据集中的大量医学放射图像来训练一个预训练 CNN 模型。然后，我们将该模型应用于两个放射测试场景中的四个不同任务，以验证其优势和泛化能力。最后，在多个数据集上的实验表明，我们的方法比 ImageNet 预训练或从头开始训练更有优势，在缺陷分类方面，F1 分数提高了 3.41 %-13.72 %，在缺陷分割方面，mIoU 提高了 1.05 %-6.58 %。这表明，在工业缺陷识别的各种任务中，通过医学射线照片进行预训练可以实现无成本的改进。

{"title":"Improving the industrial defect recognition in radiographic testing by pre-training on medical radiographs","authors":"Han Yu, Xingjie Li, Huasheng Xie, Xinyue Li, Chunyu Hou","doi":"10.1016/j.ndteint.2024.103260","DOIUrl":"10.1016/j.ndteint.2024.103260","url":null,"abstract":"<div><div>Deep learning methodologies have gained substantial traction for defect recognition in industrial radiographic testing including welds, castings and other fields. Regardless of the deep learning utilized, it has emerged as a standard configuration to use a model pre-trained from ImageNet to accelerate convergence and enhance recognition accuracy. However, there is a significant gap between the domain of natural images and industrial radiographs, raising the question of whether there might be a superior pre-training method than relying on ImageNet pre-training. Fortunately, medical radiographs are more similar to industrial radiographs than natural images because of the same imaging method. In this paper, we initially utilize numerous medical radiographic images from CheXpert dataset to train a pre-trained CNN model. Then, we apply this model to four distinct tasks within two radiographic testing scenarios to validate its advantages and generalization capabilities. Finally, experiments on multiple datasets indicate that our method brings more benefits than ImageNet pre-training or training from scratch, with a F1 score improvement of 3.41 %–13.72 % for defect classification and a mIoU improvement of 1.05 %–6.58 % for defect segmentation. It demonstrates that pre-training from medical radiographs provides a cost-free improvement for all kinds of tasks in industrial defect recognition.</div></div>","PeriodicalId":18868,"journal":{"name":"Ndt & E International","volume":"149 ","pages":"Article 103260"},"PeriodicalIF":4.1,"publicationDate":"2024-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142531803","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Metallic material microstructure grain size measurements from backscattering signals in ultrasonic array data sets 从超声阵列数据集中的反向散射信号测量金属材料的微观结构晶粒尺寸

IF 4.1 2区材料科学 Q1 MATERIALS SCIENCE, CHARACTERIZATION & TESTING

Ndt & E International

Pub Date : 2024-10-16 DOI: 10.1016/j.ndteint.2024.103251

Weixin Wang, Jie Zhang, Paul D. Wilcox

Ultrasound backscattering signals from material microstructures can be used to evaluate the material microstructure grain size. This typically involves making pulse-echo immersion measurements at multiple locations using a focused ultrasonic transducer in order to obtain an accurate estimate of the root-mean-square amplitude of the back-scattered signal at a specified focal position. However, this restricts some practical applications of using such techniques in, for example, on-line measurements in high-value manufacturing and in-service inspections where multiple immersion measurements are not feasible to use. The main benefit of using ultrasonic phased arrays is that one array probe at one position can focus ultrasound beams at multiple points using different focal laws either physically or in data postprocessing. Potentially this means that accurate grain size measurements can be obtained from a single array measurement. In this paper, the classic backscattering method for conventional transducers is adapted to be used for full matrix capture datasets from an ultrasonic array. Three-dimensional ultrasonic models are developed in the proposed inverse process to measure material microstructure grain size. Experimental validations were performed on two metallic materials: copper (EN1652) and bright mild steel (BS970). A good agreement is shown between the experimentally measured grain sizes from array data and metallography measurements. Compared to the classic pulse-echo immersion back-scattering measurements, the proposed method enables accurate measurement of grain size in a direct contact configuration at fewer locations. This has potential to make on-line grain size measurements possible.

材料微结构的超声波反向散射信号可用于评估材料微结构的晶粒尺寸。这通常需要使用聚焦超声换能器在多个位置进行脉冲回波浸入式测量，以获得特定焦点位置的反向散射信号均方根振幅的精确估算值。然而，这限制了此类技术的一些实际应用，例如在高价值制造和在役检查中的在线测量中，使用多次浸入测量是不可行的。使用超声相控阵的主要好处是，在一个位置上的一个阵列探头可以使用不同的焦距法将超声波束聚焦在多个点上，无论是在物理上还是在数据后处理中。这可能意味着通过单个阵列测量即可获得精确的粒度测量结果。在本文中，传统传感器的经典反向散射方法被调整用于超声阵列的全矩阵捕获数据集。在拟议的逆过程中开发了三维超声波模型，用于测量材料的微观结构晶粒尺寸。对两种金属材料：铜（EN1652）和光亮低碳钢（BS970）进行了实验验证。实验结果表明，阵列数据测量的晶粒尺寸与金相测量结果之间具有良好的一致性。与传统的脉冲回波浸入式反向散射测量相比，所提出的方法能够在更少的位置以直接接触的方式精确测量晶粒尺寸。这使在线晶粒尺寸测量成为可能。

{"title":"Metallic material microstructure grain size measurements from backscattering signals in ultrasonic array data sets","authors":"Weixin Wang, Jie Zhang, Paul D. Wilcox","doi":"10.1016/j.ndteint.2024.103251","DOIUrl":"10.1016/j.ndteint.2024.103251","url":null,"abstract":"<div><div>Ultrasound backscattering signals from material microstructures can be used to evaluate the material microstructure grain size. This typically involves making pulse-echo immersion measurements at multiple locations using a focused ultrasonic transducer in order to obtain an accurate estimate of the root-mean-square amplitude of the back-scattered signal at a specified focal position. However, this restricts some practical applications of using such techniques in, for example, on-line measurements in high-value manufacturing and in-service inspections where multiple immersion measurements are not feasible to use. The main benefit of using ultrasonic phased arrays is that one array probe at one position can focus ultrasound beams at multiple points using different focal laws either physically or in data postprocessing. Potentially this means that accurate grain size measurements can be obtained from a single array measurement. In this paper, the classic backscattering method for conventional transducers is adapted to be used for full matrix capture datasets from an ultrasonic array. Three-dimensional ultrasonic models are developed in the proposed inverse process to measure material microstructure grain size. Experimental validations were performed on two metallic materials: copper (EN1652) and bright mild steel (BS970). A good agreement is shown between the experimentally measured grain sizes from array data and metallography measurements. Compared to the classic pulse-echo immersion back-scattering measurements, the proposed method enables accurate measurement of grain size in a direct contact configuration at fewer locations. This has potential to make on-line grain size measurements possible.</div></div>","PeriodicalId":18868,"journal":{"name":"Ndt & E International","volume":"149 ","pages":"Article 103251"},"PeriodicalIF":4.1,"publicationDate":"2024-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142441735","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Analyzing the permeability distribution of multilayered specimens using pulsed eddy-current testing with multi-scale 1D-ResNet 利用脉冲涡流测试和多尺度 1D-ResNet 分析多层试样的渗透性分布

IF 4.1 2区材料科学 Q1 MATERIALS SCIENCE, CHARACTERIZATION & TESTING

Ndt & E International

Pub Date : 2024-10-16 DOI: 10.1016/j.ndteint.2024.103247

Xinnan Zheng, Saibo She, Zihan Xia, Lei Xiong, Xun Zou, Kuohai Yu, Rui Guo, Ruoxuan Zhu, Zili Zhang, Wuliang Yin

The mechanical properties of steel are determined by its microstructure, which is closely related to its permeability profile. In thermal processing, layered structures are formed in steel and different layers have different mechanical and magnetic properties. Therefore, it is crucial to propose a practical method to monitor the change of permeability profile along the depth, which can indicate the evolution of the microstructure of steel during thermal processing, such as hot rolling. This paper presents a method for determining the layered structure and permeability profile of the steel by using pulsed eddy current testing (PECT), which offers better penetration ability. An analytical model has been deduced for calculating the time-domain pulsed eddy current (PEC) response from a Hall sensor of a triple-layer conductor system based on the inverse Laplace transform. It is found the Tau (

τ

) curve is closely related to the permeability profile of the conductor. For the inverse solution, the Simultaneous Iterative Reconstruction Technique (SIRT) is utilized to determine the permeability profile of the multilayered specimens from the measured response. The approximate Jacobian matrix (sensitivity matrix) is obtained by the perturbation method based on the Tau curve. However, the permeability profile suffers from smoothing effect and sharp features are lost. Deep learning (DL) algorithm based on the Multi-Scale 1D-ResNet model is therefore introduced to address this issue. Numerical simulations and experiments have been performed to evaluate the proposed method for permeability profile estimation with various materials and thicknesses. The DL method can achieve an accurate estimation of the plate permeability profile with a relative error under 5%.

钢的机械性能由其微观结构决定，而微观结构与钢的磁导率曲线密切相关。在热加工过程中，钢中会形成层状结构，不同的层具有不同的机械和磁性能。因此，提出一种实用的方法来监测沿深度方向的渗透率曲线变化至关重要，这种方法可以显示钢在热加工（如热轧）过程中微观结构的演变。本文提出了一种利用脉冲涡流测试（PECT）确定钢材分层结构和渗透率分布的方法，这种方法具有更好的穿透能力。基于反拉普拉斯变换，本文推导出一个分析模型，用于计算三层导体系统霍尔传感器的时域脉冲涡流（PEC）响应。结果发现 Tau (τ) 曲线与导体的磁导率曲线密切相关。在反求解时，利用同步迭代重构技术（SIRT）从测量响应确定多层试样的渗透率剖面。通过基于 Tau 曲线的扰动法获得近似雅各矩阵（灵敏度矩阵）。然而，渗透率剖面受到平滑效应的影响，失去了鲜明的特征。因此，我们引入了基于多尺度 1D-ResNet 模型的深度学习（DL）算法来解决这一问题。我们进行了数值模拟和实验，以评估针对不同材料和厚度的渗透率剖面估算所提出的方法。DL 方法可以实现板渗透率剖面的精确估算，相对误差低于 5%。

{"title":"Analyzing the permeability distribution of multilayered specimens using pulsed eddy-current testing with multi-scale 1D-ResNet","authors":"Xinnan Zheng, Saibo She, Zihan Xia, Lei Xiong, Xun Zou, Kuohai Yu, Rui Guo, Ruoxuan Zhu, Zili Zhang, Wuliang Yin","doi":"10.1016/j.ndteint.2024.103247","DOIUrl":"10.1016/j.ndteint.2024.103247","url":null,"abstract":"<div><div>The mechanical properties of steel are determined by its microstructure, which is closely related to its permeability profile. In thermal processing, layered structures are formed in steel and different layers have different mechanical and magnetic properties. Therefore, it is crucial to propose a practical method to monitor the change of permeability profile along the depth, which can indicate the evolution of the microstructure of steel during thermal processing, such as hot rolling. This paper presents a method for determining the layered structure and permeability profile of the steel by using pulsed eddy current testing (PECT), which offers better penetration ability. An analytical model has been deduced for calculating the time-domain pulsed eddy current (PEC) response from a Hall sensor of a triple-layer conductor system based on the inverse Laplace transform. It is found the Tau (<span><math><mi>τ</mi></math></span>) curve is closely related to the permeability profile of the conductor. For the inverse solution, the Simultaneous Iterative Reconstruction Technique (SIRT) is utilized to determine the permeability profile of the multilayered specimens from the measured response. The approximate Jacobian matrix (sensitivity matrix) is obtained by the perturbation method based on the Tau curve. However, the permeability profile suffers from smoothing effect and sharp features are lost. Deep learning (DL) algorithm based on the Multi-Scale 1D-ResNet model is therefore introduced to address this issue. Numerical simulations and experiments have been performed to evaluate the proposed method for permeability profile estimation with various materials and thicknesses. The DL method can achieve an accurate estimation of the plate permeability profile with a relative error under 5%.</div></div>","PeriodicalId":18868,"journal":{"name":"Ndt & E International","volume":"149 ","pages":"Article 103247"},"PeriodicalIF":4.1,"publicationDate":"2024-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142539654","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Defect classification and quantification method based on AC magnetic flux leakage time domain signal characteristics 基于交流漏磁通时域信号特征的缺陷分类和量化方法

IF 4.1 2区材料科学 Q1 MATERIALS SCIENCE, CHARACTERIZATION & TESTING

Ndt & E International

Pub Date : 2024-10-13 DOI: 10.1016/j.ndteint.2024.103250

Rongbiao Wang , Yongzhi Chen , Haozhi Yu , Zhiyuan Xu , Jian Tang , Bo Feng , Yihua Kang , Kai Song

Pipelines play a crucial role in industries such as petroleum and nuclear power, where non-destructive testing is essential. Magnetic flux leakage testing methods are widely used for pipeline inspection due to their ability to detect both internal and external defects. However, accurately classifying and evaluating the size of such defects poses challenges due to the complex coupling relationship between ferromagnetic materials and defect magnetic fields. This paper proposes a method for defect classification and quantification based on the time domain characteristics of AC magnetic flux leakage signals. Firstly, the paper explores the shielding effects caused by transient high magnetic permeability within the material on signals from internal defects. It analyzes the differences in signals between internal and external defects. Then, based on the nonlinear attributes of defect signals, the paper proposes a defect classification method based on derivatives analysis of the windowed time-domain signal. Moreover, the study finds that rising time and zero-crossing time can be used to assess the depth of internal and external defects separately, which can decouple the width and depth. Finally, experimental validation confirms the effectiveness of defects classification and quantification. This paper provides a feasible method for evaluating the defect of ferromagnetic materials.

管道在石油和核电等行业中发挥着至关重要的作用，在这些行业中，无损检测至关重要。磁通量泄漏测试方法能够检测内部和外部缺陷，因此被广泛用于管道检测。然而，由于铁磁性材料与缺陷磁场之间复杂的耦合关系，准确分类和评估此类缺陷的大小是一项挑战。本文提出了一种基于交流漏磁通信号时域特征的缺陷分类和量化方法。首先，本文探讨了材料内部瞬态高磁导率对内部缺陷信号的屏蔽效应。它分析了内部和外部缺陷信号的差异。然后，基于缺陷信号的非线性属性，论文提出了一种基于窗口时域信号导数分析的缺陷分类方法。此外，研究还发现上升时间和过零时间可用于分别评估内部和外部缺陷的深度，从而将宽度和深度解耦。最后，实验验证证实了缺陷分类和量化的有效性。本文为铁磁材料的缺陷评估提供了一种可行的方法。

{"title":"Defect classification and quantification method based on AC magnetic flux leakage time domain signal characteristics","authors":"Rongbiao Wang , Yongzhi Chen , Haozhi Yu , Zhiyuan Xu , Jian Tang , Bo Feng , Yihua Kang , Kai Song","doi":"10.1016/j.ndteint.2024.103250","DOIUrl":"10.1016/j.ndteint.2024.103250","url":null,"abstract":"<div><div>Pipelines play a crucial role in industries such as petroleum and nuclear power, where non-destructive testing is essential. Magnetic flux leakage testing methods are widely used for pipeline inspection due to their ability to detect both internal and external defects. However, accurately classifying and evaluating the size of such defects poses challenges due to the complex coupling relationship between ferromagnetic materials and defect magnetic fields. This paper proposes a method for defect classification and quantification based on the time domain characteristics of AC magnetic flux leakage signals. Firstly, the paper explores the shielding effects caused by transient high magnetic permeability within the material on signals from internal defects. It analyzes the differences in signals between internal and external defects. Then, based on the nonlinear attributes of defect signals, the paper proposes a defect classification method based on derivatives analysis of the windowed time-domain signal. Moreover, the study finds that rising time and zero-crossing time can be used to assess the depth of internal and external defects separately, which can decouple the width and depth. Finally, experimental validation confirms the effectiveness of defects classification and quantification. This paper provides a feasible method for evaluating the defect of ferromagnetic materials.</div></div>","PeriodicalId":18868,"journal":{"name":"Ndt & E International","volume":"149 ","pages":"Article 103250"},"PeriodicalIF":4.1,"publicationDate":"2024-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142531805","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Predicting actual crack size through crack signal obtained by advanced Flexible Eddy Current Sensor using ResNet integrated with CBAM and Huber loss function 利用集成了 CBAM 和 Huber 损失函数的 ResNet，通过先进的柔性涡流传感器获得的裂纹信号预测实际裂纹尺寸

IF 4.1 2区材料科学 Q1 MATERIALS SCIENCE, CHARACTERIZATION & TESTING

Ndt & E International

Pub Date : 2024-10-12 DOI: 10.1016/j.ndteint.2024.103249

Le Quang Trung , Naoya Kasai , Minhhuy Le , Kouichi Sekino

This study presents an advanced FEC sensor, engineered by arranging coils in a co-directional current configuration. Moreover, boasting a compact design, the FEC sensor showcases significantly enhanced spatial resolution, enabling robust detection of small cracks even at low excitation frequencies and mitigating issues of overlapping in adjacent crack detection. Results indicate successful crack detection through voltage and phase measurements, albeit with phase signals demonstrating variation at specific excitation frequencies, complicating the determination of actual crack sizes. Consequently, a novel model is proposed to forecast actual crack sizes, leveraging experimental data from the FEC sensor system. This model integrates a Residual Neural Network (ResNet) architecture with a Convolutional Block Attention Module (CBAM) and utilizes the Huber loss function to minimize errors during model training. Comparative analysis underscores the superior performance of the proposed model in predicting crack length and depth compared to the standalone ResNet, particularly when utilizing the Huber loss function with a δ value of 1.0. Evaluation metrics, encompassing Mean Squared Error (MSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE), illustrate an average accuracy surpassing 95 % for crack size predictions. Consequently, the proposed model demonstrates remarkable performance, significantly reducing the time required to ascertain actual crack sizes by leveraging voltage and phase measurements.

本研究介绍了一种先进的 FEC 传感器，它是通过在同向电流配置中布置线圈而设计的。此外，该 FEC 传感器设计紧凑，空间分辨率显著提高，即使在较低的激励频率下也能稳健地检测出细小裂纹，并缓解了相邻裂纹检测中的重叠问题。结果表明，尽管在特定激励频率下相位信号会发生变化，从而使实际裂纹尺寸的确定变得复杂，但通过电压和相位测量，裂纹检测还是取得了成功。因此，我们提出了一个新模型，利用 FEC 传感器系统的实验数据来预测实际裂纹尺寸。该模型集成了残差神经网络（ResNet）架构和卷积块注意力模块（CBAM），并利用休伯损失函数将模型训练过程中的误差降至最低。对比分析表明，与独立的残差神经网络相比，所提出的模型在预测裂纹长度和深度方面具有更优越的性能，尤其是在使用 δ 值为 1.0 的 Huber 损失函数时。评估指标包括平均平方误差 (MSE)、平均绝对误差 (MAE) 和平均绝对百分比误差 (MAPE)，表明裂纹尺寸预测的平均准确率超过 95%。因此，所提出的模型性能卓越，通过利用电压和相位测量，大大缩短了确定实际裂纹尺寸所需的时间。

{"title":"Predicting actual crack size through crack signal obtained by advanced Flexible Eddy Current Sensor using ResNet integrated with CBAM and Huber loss function","authors":"Le Quang Trung , Naoya Kasai , Minhhuy Le , Kouichi Sekino","doi":"10.1016/j.ndteint.2024.103249","DOIUrl":"10.1016/j.ndteint.2024.103249","url":null,"abstract":"<div><div>This study presents an advanced FEC sensor, engineered by arranging coils in a co-directional current configuration. Moreover, boasting a compact design, the FEC sensor showcases significantly enhanced spatial resolution, enabling robust detection of small cracks even at low excitation frequencies and mitigating issues of overlapping in adjacent crack detection. Results indicate successful crack detection through voltage and phase measurements, albeit with phase signals demonstrating variation at specific excitation frequencies, complicating the determination of actual crack sizes. Consequently, a novel model is proposed to forecast actual crack sizes, leveraging experimental data from the FEC sensor system. This model integrates a Residual Neural Network (ResNet) architecture with a Convolutional Block Attention Module (CBAM) and utilizes the Huber loss function to minimize errors during model training. Comparative analysis underscores the superior performance of the proposed model in predicting crack length and depth compared to the standalone ResNet, particularly when utilizing the Huber loss function with a δ value of 1.0. Evaluation metrics, encompassing Mean Squared Error (MSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE), illustrate an average accuracy surpassing 95 % for crack size predictions. Consequently, the proposed model demonstrates remarkable performance, significantly reducing the time required to ascertain actual crack sizes by leveraging voltage and phase measurements.</div></div>","PeriodicalId":18868,"journal":{"name":"Ndt & E International","volume":"149 ","pages":"Article 103249"},"PeriodicalIF":4.1,"publicationDate":"2024-10-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142441734","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Intelligent identification of defective regions of voids in tunnels based on GPR data 根据 GPR 数据智能识别隧道空洞缺陷区域

IF 4.1 2区材料科学 Q1 MATERIALS SCIENCE, CHARACTERIZATION & TESTING

Ndt & E International

Pub Date : 2024-10-11 DOI: 10.1016/j.ndteint.2024.103244

Guangyan Cui , Yanhui Wang , Yujie Li , Feifei Hou , Jie Xu

Quantitatively detecting voids behind tunnel linings presents significant challenges in identifying the range of width and depth. This paper develops an innovative method for identifying defective regions of voids based on Ground Penetrating Radar (GPR) data. This method involves three steps: Firstly, the void-identifying-feature-set (VIFS) is constructed by extracting the Amplitude peak (A_T), Signal energy (E_T), and Amplitude peak of the first intrinsic mode function (IMF1) component (A_H) of every A-scan signal. Secondly, the Support Vector Machine (SVM) is utilized to identify defect signals and normal signals, contributing to the width identification of void in the horizontal direction. Thirdly, an innovative Three-Stage-Boundary-Extraction (TSBE) algorithm is proposed to identify the depth range of voids in the vertical direction. Experimental results conducted on both field data and simulated data demonstrated that the Intersection over Union (IOU) value and consumption time of three groups of GPR data (Data I, Data II, and Data V) are 0.739 and 0.888 s, respectively. The average IOU and consumption time of the TSBE algorithm are 0.739 and 0.058 s, respectively.

隧道衬砌背后空洞的定量检测在确定宽度和深度范围方面存在巨大挑战。本文基于地面穿透雷达 (GPR) 数据，开发了一种识别空洞缺陷区域的创新方法。该方法包括三个步骤：首先，通过提取每个 A 扫描信号的振幅峰值 (AT)、信号能量 (ET) 和第一本征模态函数 (IMF1) 分量 (AH) 的振幅峰值，构建空隙识别特征集 (VIFS)。其次，利用支持向量机（SVM）来识别缺陷信号和正常信号，有助于水平方向的空隙宽度识别。第三，提出了一种创新的三阶段边界提取（TSBE）算法，用于识别垂直方向的空洞深度范围。对现场数据和模拟数据进行的实验结果表明，三组 GPR 数据（数据 I、数据 II 和数据 V）的交集大于联合（IOU）值和消耗时间分别为 0.739 秒和 0.888 秒。TSBE 算法的平均 IOU 值和消耗时间分别为 0.739 和 0.058 秒。

{"title":"Intelligent identification of defective regions of voids in tunnels based on GPR data","authors":"Guangyan Cui , Yanhui Wang , Yujie Li , Feifei Hou , Jie Xu","doi":"10.1016/j.ndteint.2024.103244","DOIUrl":"10.1016/j.ndteint.2024.103244","url":null,"abstract":"<div><div>Quantitatively detecting voids behind tunnel linings presents significant challenges in identifying the range of width and depth. This paper develops an innovative method for identifying defective regions of voids based on Ground Penetrating Radar (GPR) data. This method involves three steps: Firstly, the void-identifying-feature-set (<em>VIFS</em>) is constructed by extracting the Amplitude peak (<em>A</em><sub><em>T</em></sub>), Signal energy (<em>E</em><sub><em>T</em></sub>), and Amplitude peak of the first intrinsic mode function (IMF1) component (<em>A</em><sub><em>H</em></sub>) of every A-scan signal. Secondly, the Support Vector Machine (SVM) is utilized to identify defect signals and normal signals, contributing to the width identification of void in the horizontal direction. Thirdly, an innovative Three-Stage-Boundary-Extraction (TSBE) algorithm is proposed to identify the depth range of voids in the vertical direction. Experimental results conducted on both field data and simulated data demonstrated that the Intersection over Union (IOU) value and consumption time of three groups of GPR data (Data I, Data II, and Data V) are 0.739 and 0.888 s, respectively. The average IOU and consumption time of the TSBE algorithm are 0.739 and 0.058 s, respectively.</div></div>","PeriodicalId":18868,"journal":{"name":"Ndt & E International","volume":"149 ","pages":"Article 103244"},"PeriodicalIF":4.1,"publicationDate":"2024-10-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142437935","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0