金属零件增材制造中细观缺陷及损伤演化的工业CT精确无损评价关键技术研究

In this project, an accurate and non-destructive evaluation method of inherent meso-defects and damage evolution of industrial CT is proposed, aiming at the lack of effective non-destructive testing technology for the detection and control of meso-defects in the existing reinforcement of complex metal components. This method is based on industrial CT technology to detect weak signal of meso-defects and damage evolution characteristics in the complex components. In order to solve the problems caused by the noise and artifacts of industrial CT technology and the errors caused by qualitative and quantitative characterization of meso-defects, an industrial CT image processing method based on multi-frequency narrowband noise statistics and artifact classification suppression is proposed. Through optimizing design the CT image edge degradation model to achieve the meso-defects segmentation and accurate measurement, the body image correlation method is proposed to recognize the meso-defects damage evolution characteristics , the industrial CT measurement results reliability evaluation is built based on CDD curve and noise statistics. Finally, the validity and feasibility of the research results are verified by experiment. It provides effective quality supervision and technical support for the transformation of the metal component material production technology from the research stage to the engineering application promotion, and finally serves the high-end major equipment manufacturing development strategy of our country, to achieve the overall goal of the key components of the production with high reliability and long life.

本项目针对现有复杂金属零件增材制造中细观缺陷检测及控制技术中存在缺乏有效无损检测技术手段，提出一种基于工业CT的增材制造特有细观缺陷及损伤演化精确无损评价方法。这种方法基于工业CT技术实现对复杂零件中细观缺陷及损伤演化特征等弱小信号检测。为了克服基于工业CT技术的噪声及伪影干扰对细观缺陷定性定量及表征造成的误差等问题，提出基于多频窄带噪声统计及伪影分类抑制的工业CT图像处理方法，优化设计CT图像弱边缘退化模型，实现细观尺度缺陷分割及精确测量，提出基于体图像相关法实现细观缺陷损伤演化特征的识别，建立基于CDD曲线与噪声统计结合的工业CT测量结果可靠性评价。最终对所得研究成果的有效性与可行性进行实验验证，为我国金属零件增材制造技术从研究阶段到工程应用推广的转变提供有效质量监督技术支持，最终服务于我国的高端重大装备制造发展战略，实现我国增材制造关键零部件高可靠和长寿命的总体目标。

本项目针对增材制造复杂结构中细观缺陷无损检测的需求，力求解决基于工业CT技术的内部缺陷无损检测特性及方法、CT技术检测能力评价等问题，其难点在于XCT成像原理较为复杂，基于该技术的测量方法的影响因素多，缺乏可靠的测量结果不确定度评价方法及缺陷检出可靠性分析。本项目完成了复杂结构工业CT特有噪声分析及抑制、CT图像弱边缘退化模型、内部尺寸测量及不确定度评价和缺陷定量方法及可靠性评价等研究内容；解决了CT图像弱边缘退化数学建模、CT尺寸测量、细观缺陷定量等关键问题；取得了CT边缘退化模型、等值面分割法尺寸测量结果不确定度评价方法、缺陷自动检出定量可靠性评价方法等成果。在本项目的资助下发表学术论文5篇，其中SCI论文2篇，中文EI论文1篇，中文核心论文1篇；申报发明专利12项，其中已授权3项；研究指标达到了直径大于10mm的小尺寸构件可检出≥0.05mm孔洞类缺陷、测量精度优于±0.05mm、直径大于100mm的小尺寸构件可检出≥0.2mm孔洞类缺陷、测量精度优于±0.2mm等技术指标。其中，小缺陷检出、尺寸测量方法成果已在航天卫星推进器电磁阀焊缝工业CT检测、军用大马力发动机活塞工业CT尺寸测量方面得到部分应用，并且相关研究成果为后续工业计量CT的工程应用奠定了基础。