In the field of advanced near-final manufacturing of metal materials and components, arc additive manufacturing can significantly enhance design freedom and manufacturing diversification, and it is easier to realize the integrated molding manufacturing of medium and large parts. By combining factors such as anisotropic microstructure, widely distributed defects, deep accumulated residual stresses and complex surface roughness, the integrated evaluation method of Process-Structure-Properties-Performance (PSPP) will become an important topic and effective strategy to promote the wide application of metal additive manufacturing technology in the field of large-scale structural metal parts manufacturing.
https://doi.org/10.1080/17452759.2024.2390495
近日，吴圣川课题组在Virtual and Physical Prototyping(IF=10.6)上发表题为"State-of-art review on the process-structure-properties-performance linkage in wire arc additive manufacturing"邀请综述论文。 轨道交通运载系统全国重点实验室博士生张涵为论文第一作者，吴圣川研究员、刘军江助理教授和太行实验室雷力明研究员为共同通讯作者。 该研究工作得到太行实验室重点课题项目（C2024-1-0405）和国基金大科学装置联合基金（U2032121）支持。
Background:
Additive manufacturing (AM) technology breaks the constraints of traditional manufacturing models by enabling the construction of 3D parts layer by layer from digital models, increasing design freedom and flexibility while reducing market expectations for products from concept to physical product. To meet the technical challenges of manufacturing large and complex components, Wire Arc Additive Manufacturing (WAAM) has won the favor of the industrial manufacturing industry due to its high efficiency, low cost, excellent material utilization, and remarkable eco-friendliness. WAAM technology is considered a "green technology" by streamlining intermediate links, reducing material loss, and effectively shortening the manufacturing cycle and subsequent processing time. Up to now, WAAM technology has been successfully applied to a variety of engineering materials, showing great application potential and broad development prospects, and this paper systematically summarizes the current research and development of WAAM technology.
Therefore, it is particularly important to carry out an integrated process-structure-performance-life (PSPP) evaluation, focusing on the comprehensive effects of microstructure, defects, residual stresses and roughness, due to the challenges of rapid cooling and thermal cycling, such as porosity, cracking, and high residual stress, which weaken the mechanical strength and fatigue durability of components, which are directly related to service behavior. We know that traditional research relies on experimental analysis, theoretical modeling, and numerical solutions to construct PSPP models, which have obvious limitations and large deviations in exploring the intricate multi-factor coupling relationship between metal additive process parameters, influencing factors, and mechanical properties, which makes it difficult to meet the requirements of accuracy and ease of use when predicting macroscopic mechanical properties and service life.
As a powerful tool for resolving the complexity of high-dimensional data, machine learning (ML) has shown unprecedented research potential in the field of additive manufacturing, especially in WAAM technology, in recent years, Figure 1 shows a schematic diagram of machine learning in the field of WAAM to establish PSPP relationships. In general, the application boundary of PSPP is constantly expanding with the intervention of ML, which integrates the key elements of the whole life cycle from material preparation to component service in an unprecedented way, and opens up new ideas for the integrated evaluation of process, structure, performance and service life.
This paper attempts to conduct a comprehensive, systematic and in-depth analysis of the latest progress of WAAM technology. By focusing on the process, process-related defects, microstructure, residual stress, and surface roughness, the quantitative evaluation of relevant mechanical properties and service life can be realized. This paper demonstrates the unique advantages of the integrated evaluation framework of process-structure-performance-service life (PSPP) in WAAM technology, and also discusses its current limitations and challenges, so as to jointly promote PSPP evaluation as an important node in the transformation and upgrading of the manufacturing industry.
Fig.1 Schematic diagram of the PSPP relationship of WAAM technology based on machine learning
Introduction
This paper first briefly introduces WAAM as an advanced manufacturing method that combines arc welding and additive manufacturing, and gas tungsten arc welding (GTAW), gas metal arc welding (GMAW) and plasma arc welding (PAW) are the three forms of WAAM process. Subsequently, the process parameters, characteristics and application scenarios of each were introduced. Despite the many advantages of WAAM over traditional manufacturing methods, especially in terms of production efficiency, there are still some problems, such as the difficulty of the WAAM process to produce near-final parts with the required characteristics and precision, which hinders its large-scale industrial application. Hybrid Arc Additive Manufacturing (Hybrid-WAAM) is a process that uses two or more sequential processes to manufacture a finished part, improving the performance of WAAM by applying an auxiliary field. These additive manufacturing technologies offer a wide range of options for the production of metal-based materials, and the right technical pathways can be selected according to specific needs and application scenarios.
Subsequently, the manufacturing process control of the WAAM process is described in detail. With sensor monitoring, process parameters can be adjusted in real time. These dynamic adjustments are essential to maintain optimal conditions throughout the material deposition process, ensuring the structural integrity and dimensional accuracy of the final product. In addition to the monitoring of the WAAM process, multi-dimensional and multi-scale characterization techniques are critical in the monitoring of WAAM components and are key to precise control and optimization of the microstructure during the manufacturing process, as shown in Figure 2. In addition, the ML model establishes a quantitative mapping relationship within the PSPP framework, and makes predictions to further improve the process by utilizing historical data from real-time monitoring and offline monitoring, so as to achieve rapid iteration, low-cost development, and wide application of high-performance materials.
Fig.2 Multi-dimensional and multi-scale correlation imaging characterization of fatigue damage behavior of Ti6Al4V alloy formed by WAAM
The dynamic processes involved in WAAM are complex, and determining the process-structure-performance-service (PSPP) relationship is critical to understanding the WAAM process. For example, researchers have developed multiscale models to define PSPP relationships. The mechanical properties and service life of components produced by WAAM technology are significantly influenced by four main factors: microstructure, internal defects, residual stresses, and surface roughness, as shown in Figure 4. These factors are closely related to process parameters, and the relationship between these influencing factors and process parameters is comprehensively reviewed, such as the classical Murakami model and the Manson-Coffin formula. Research generally explores the mechanisms of interaction between them through conventional experiments and physical modeling, with the aim of establishing a relationship between process, structure and performance, and service life. This approach is designed to provide insights into the optimization of WAAM processes and component designs. The integrated evaluation based on physical information can better understand the potential interrelationship between process, structure, performance, and service behavior, so as to realize multi-scale dynamic process modeling.
The efficiency and accuracy of predicting the mechanical properties of metals in additive manufacturing based on traditional empirical models and limited data are severely challenged by a variety of factors, such as anisotropic structure, wide defect distribution, deep residual stress, and complex surface roughness. In recent years, with the development of big data and artificial intelligence, machine learning (ML) methods have provided opportunities for effectively dealing with complex nonlinear relationships between high-dimensional physical quantities. As shown in Figure 3, this study uses machine learning techniques to input traditional empirical models and datasets with the aim of establishing a correlation mechanism between process-structure-performance-service behavior to control process parameters and optimize component quality.
Fig.3. Schematic diagram of the PSPP relationship of metal components fabricated by WAAM process
Fig.4 Schematic diagram of the PSPP relationship of WAAM metal components driven by machine learning
Although WAAM has broad application prospects and great development potential, it also faces many challenges. (1) Modeling of complex PSPP relationships of WAAM components. In order to accurately correlate the process with the structure, performance, and service behavior of components, it is necessary to combine new technologies and methods, such as in-situ observation and high-resolution imaging, as well as high-fidelity simulation, to obtain rich training data. A multi-scale ML model and efficient algorithm for WAAM process were further developed, and the modeling technology based on physical information was enhanced to control the performance of components by determining the PSPP relationship. (2) Real-time monitoring and control of WAAM production. High-speed data acquisition, sensor fusion, and in-situ monitoring techniques such as high-resolution X-ray imaging and ultrasonic inspection can be used to enhance defect detection and quality control. And the development of adaptive control systems and predictive modeling, combined with cloud computing and edge computing, can manage the complexity and computing requirements of control systems. (3) WAAM product standardization and the introduction of physical mechanisms. To improve the quality and quantity of training samples, data augmentation techniques can be combined. These techniques allow for the integration and standardization of diverse data at different scales and physical quantities from experimental data, simulation results, or prior knowledge. In addition, ML models driven by physical mechanisms can be built to improve the accuracy, interpretability, generalization ability, and reduced data dependency of the models.
When it comes to the cutting edge of additive manufacturing, we can't ignore Selective Laser Melting (SLM). The core principle shared by both is the precise construction of complex 3D structures based on a layer-by-layer cumulative construction strategy. Specifically, SLM technology uses a high-energy laser beam as a heat source to precisely control the melting and solidification process of metal powders, and "draws" a design blueprint point by point and layer by layer in three-dimensional space, achieving extremely high manufacturing accuracy and direct molding of complex structures. The WAAM technology uses the principle of arc welding, through continuous wire feeding and precisely controlled arc heat input, to realize the layer-by-layer deposition and fusion of metal wires, and construct a solid structure with high strength and good compactness.
Looking forward to the future, it is expected that WAAM technology can play a bridge role, not only to promote the in-depth application and optimization of the comprehensive evaluation system of PSPP in its application field, but also to expand to other additive manufacturing technologies including SLM, and to build a comprehensive evaluation system across technology platforms. The realization of this goal will rely on in-depth scientific research and extensive practical exploration, aiming to promote the development of the entire additive manufacturing industry to a higher level of intelligence and precision, and bring far-reaching changes and significant leaps to the field of industrial manufacturing.
Introduction to teamwork
In recent years, Wu Shengchuan's research group at Southwest Jiaotong University has developed a series of original devices for the internal damage characterization of advanced material structures based on in-situ tension and compression, fretting fatigue, contact wear, axial/rotational fatigue and ultra-high temperature/ultra-low temperature/strong corrosion based on the three-dimensional imaging technology of synchrotron radiation light source, proposed an image mechanics method (3D + time + damage) for internal damage of additive manufacturing and welded components, and established a three-parameter K-T evaluation diagram (or W-W parameter model) of stress level-defect criticality-fatigue life. The in-situ ultra-high cycle fatigue test technology based on high spatiotemporal resolution was identified as a third-party verification method by the national standard "Ultra-high Cycle Fatigue Ultrasonic Fatigue Test Method", and the relevant results were published in journals such as Nature, Acta Mater, Int J Mech Sci, Int J Fatigue, Int J Fatigue, etc., and many ESI highly cited papers were selected as 16 authorized inventions. He has published the papers "Advanced Light Source Characterization Technology for Fatigue Damage Behavior of Materials" and "Advanced Materials and Structural Integrity for Additive Manufacturing", and the experimental data and models have been adopted by Thermo Fisher Scientific, an international leading scientific research software, and he is also an editorial board member of Int J Fatigue and the founding editorial board of Tomo Mater Struct, and some of the results have supported the first prize of China Quality Technology Award in 2022 (ranked 2).
Related References:
[1] Wu ZK, Wu SC, Qian WJ, Zhang HO, Zhu H, Chen QY, et al. Structural integrity issues of additively manufactured railway components: Progress and challenges. Eng Fail Anal 2023;149:107265. https://doi.org/10.1016/j.engfailanal.2023.107265. （综述，ESI高被引）
[2] Peng X, Wu SC, Qian WJ, Bao JG, Hu YN, Huang ZX, et al. The potency of defects on fatigue of additively manufactured metals. Int J Mech Sci 2022;221:107185. https://doi.org/10.1016/j.ijmecsci.2022.107185. （ESI高被引）
[3] Qian WJ, Wu SC, Wu ZK, Ahmed S, Zhang W, Qian GA, et al. In situ X-ray imaging of fatigue crack growth from multiple defects in additively manufactured AlSi10Mg alloy. Int J Fatigue 2022;155:106616. https://doi.org/10.1016/j.ijfatigue.2021.106616. （ESI高被引）
[4] Wu ZK, Wu SC, Bao JG, Qian WJ, Karabal S, Sun W, et al. The effect of defect population on the anisotropic fatigue resistance of AlSi10Mg alloy fabricated by laser powder bed fusion. Int J Fatigue 2021;151:106317. https://doi.org/10.1016/j.ijfatigue.2021.106317. （ESI高被引）
[5] Bao HY, Wu SC, Wu ZK, Kang GZ, Peng X, Withers PJ. A machine-learning fatigue life prediction approach of additively manufactured metals. Eng Fract Mech 2021;242:107508. https://doi.org/10.1016/j.engfracmech.2020.107508. （ESI高被引）
[6] Hu YN, Wu SC, Withers PJ, Zhang J, Bao HYX, Fu YN, et al. The effect of manufacturing defects on the fatigue life of selective laser melted Ti-6Al-4V structures. Mater Des 2020;192:108708. https://doi.org/10.1016/j.matdes.2020.108708. （ESI高被引）
[7] Hu YN, Wu SC, Wu ZK, Zhong XL, Ahmed S, Karabal S, et al. A new approach to correlate the defect population with the fatigue life of selective laser melted Ti-6Al-4V alloy. Int J Fatigue 2020;136:105584. https://doi.org/10.1016/j.ijfatigue.2020.105584. （ESI高被引）
[8] Zhang H, Li RS, Liu JJ, Wang KY, Qian WJ, Shi L, et al. State-of-art review on the process-structure-properties-performance linkage in wire arc additive manufacturing. Virtual Phys Prototyp 2024; 19:e2390495. https://doi.org/10.1080/17452759.2024.2390495. （邀请综述）
[9] Hu YN, She YF, Wu SC, Kan QH, Yu H, Kang GZ. Critical physics informed fatigue life prediction of laser 3D printed AlSi10Mg alloys with mass internal defects. Int J Mech Sci 2024:109730. https://doi.org/10.1016/j.ijmecsci.2024.109730.
[10] Wu ZK, Wu SC, Duan YS, Huang K, He WT, Du DF, et al. In situ X-ray tomography of fracture behaviour in low-porosity L-PBF AlSi10Mg alloy with laser shock peening. Virtual Phys Prototyp 2023; 18:e2273955. https://doi.org/10.1080/17452759.2023.2273955.
[11] Wu ZK, He ZA, Wu SC, Gao X, Lei L, Liu CK, et al. Rotating bending fatigue mechanisms of L-PBF manufactured Ti-6Al-4V alloys using in situ X-ray tomography. Int J Fatigue 2023;176:107876. https://doi.org/10.1016/j.ijfatigue.2023.107876.
[12] Wu ZK, Wu SC, Gao X, Lin Y, Xue YL, Withers PJ. The role of internal defects on anisotropic tensile failure of L-PBF AlSi10Mg alloys. Sci Rep 2023;13:14681. https://doi.org/10.1038/s41598-023-39948-z.
Exclusive proof of authorization
Note: The content of this article is exclusively licensed by the research team.
Reprinted by Chen Changjun of the Yangtze River Delta G60 Laser Alliance