Determining optimal bead central angle by applying machine learning to wire arc additive manufacturing (WAAM)

Objectives: Fabrication of a cylinder-shaped High-Strength Low Alloyed steel wall component through wire (metal cored) arc additive manufacturing and assess the homogeneity and integrity of the built component. Method: Mechanical testing and microstructural characterizations such as optical microscopy, scanning electron microscopy and electron backscattered diffraction techniques were performed at distinct regions (namely, bottom and top). Findings: The conducted tensile test at the bottom area recorded a tensile strength of 864 MPa, which is 5.9 % higher than the top region but with a lower total elongation. The microhardness study revealed that the average hardness at the bottom region was 266 HV0.5, while the top region was 241 HV0.5. The microstructural analysis conducted in both areas indicates that the bottom region exhibits better hardness and strength due to the presence of a high value of geometrically necessary dislocations, High-Angle Grain Boundaries and increased grains with reduced size. However, the built component proves to have better homogeneity with marginal deviations in the properties estimated. Novelty: Solid wire was used as a predominant filler material for the Wire-Arc Additive Manufacturing (WAAM) process, which requires high heat input to melt the material. Hence, to reduce the heat input, metal-cored wire was used instead of solid wire for the Gas Metal Arc -WAAM process. Keywords: Gas metal arc welding, Metal cored wire, Mechanical properties, Direct energy deposition, Wire-arc additive manufacturing

A defect detection system for wire arc additive manufacturing using incremental learning

Additive manufacturing has the potential to revolutionize industrial hardware and unlock efficiency gains through the fabrication of geometries and architectures not possible by conventional processing. Wire Arc Additive manufacturing (WAAM) process is a class of directed energy deposition process enabling higher build rate and allowing custom wire feedstock allowing spatial variation of microstructure. The larger spot size and lower speed creates a larger melt pool, reducing residual stress and often time creating directional/columnar microstructure for Ni-superalloy. The use of flexible platform provides freedom in deposition strategy, which can accommodate complex substrates, including feature addition onto existing structures, non-flat layers, and repair methods. However, the certification of the final component needs to match the strength requirement. Hence, the quality of the component produced is stringently monitored to avoid buildup of residual stress, cracks, porosity and to reduce detrimental segregated phases commonly observed during alloy solidification. Simulation plays a huge role in predicting the melt pool dimension and can be used to optimize the process parameter. Similar development is also required to perform physics-based modeling of microstructural development in WAAM that can predict the microstructural features during solidification and can be used to optimize the process more effectively. To move toward this goal, Raytheon Technologies Research Center together with Siemens worked to create a set of computational tools to control the process parameters, enable on-line measurements and acquisition with feedback to the optimized process parameters, and eventually track material evolution through each step of the additive process. Computational fluid dynamics is used for accurate prediction and calibration of the thermal field during WAAM process. and phase field models for microstructure evolution as a function of processing parameters to establish a connection between additive parameters and the final microstructure. Here we report cellular automata (CA) model development to predict the dendritic microstructure evolution with surface and bulk nuclei for single track and multiple layers. The CA model was developed to account for secondary element addition and predict segregation, local melting, and latent heat release as well as prediction of Euler angles from orientation information and validated against experiments. This framework was utilized to tailor spatially-varying composition in a part by appropriately controlling the microstructure evolution during the additive process. Functionally graded Haynes 282 alloy with high Cr content at the surface was tested for oxidation and mechanical properties. An advanced physics-based reaction-diffusion model predicting the simultaneous creation of chromium oxide and alumina is developed and validated to extend life expectancy of the WAAM manufactured high temperature part. A machine-learning data-driven framework establishing the process-structure relationship from a dataset of real microstructure images and corresponding process history data has been developed and implemented in the NX Siemens design system. The digital twin configuration along with the tool path generation enabled prediction of WAAM component buildup time and the techno-economic analysis provided a favorable option for all 4 cases with 15-40% cost reduction.

Bending fatigue properties of structural steel fabricated through wire arc additive manufacturing (WAAM)

Additive manufacturing is a very popular word in the present day as various researchers are focused on this field all around the world. Every new researcher coming to this field of additive manufacturing is busy in developing new methods and techniques to make it more versatile and efficient in terms of applicability. Wire arc additive manufacturing (WAAM) process, a new addition to the AM processes, is growing rapidly in the field of the metal additive manufacturing process because of its high deposition rate, low equipment cost compared to laser and electron beam, saving lead time, high BTF (buy-to-fly) ratio, no limit on part size, and many other advantages. In this study, a wire arc additive manufacturing system based on cold metal transfer gas metal arc welding technology was used to fabricate a solid block component with ER70S6 MS alloy welding wire. This study presents the investigation on anisotropy behavior of resulted microstructure and mechanical properties including tensile, impact, and hardness of the manufactured block component. The samples were taken along the parallel and perpendicular to the welding direction. The tensile properties of the samples were found almost uniform in both the directions with values of the ultimate tensile strength (UTS), yield strength (YS), and elongation (%E) found to be 510 ± 5 MPa, 373 ± 5 MPa, and 37% ± 2%, respectively. The average Vickers hardness values at 10 kg load in parallel and perpendicular directions were found around 80 and 72 kg/mm2. The impact toughness at −30℃ was found to be 148 and 170 J in the parallel and perpendicular direction. The toughness obtained is much higher than the required minimum impact strength of 80 J at the subzero temperature as per the AWS requirement for the used ER70S6 welding wire. The microstructure observed in welded samples is ferrite and bainite which consists inside and along the prior austenite grain boundaries. Mostly grains were elongated in the welding direction. From the present work, it is evident that the near-net-shape component fabricated with the WAAM process has quite acceptable mechanical properties without much effect of directionality.

The problem with development of the support vector machine (SVM) classifiers using modified particle swarm optimization (PSO) algorithm and their ensembles has been considered. Solving this problem would allow fulfilling the high-precision data classification, especially Big Data classification, with the acceptable time expenditures. The modified PSO algorithm conducts a simultaneous search of the type of kernel functions, the parameters of the kernel function and the value of the regularization parameter for the SVM classifier. The idea of particles' «regeneration» served as the basis for the modified PSO algorithm. In the implementation of this algorithm, some particles change the type of their kernel function to the one which corresponds to the particle with the best value of the classification accuracy. The offered PSO algorithm allows reducing the time expenditures for the developed SVM classifiers, which is very important for Big Data classification problem. In most cases such SVM classifier provides the high quality of data classification. In exceptional cases the SVM ensembles based on the decorrelation maximization algorithm for the different strategies of the decision-making on the data classification and the majority vote rule can be used. Also, the two-level SVM classifier has been offered. This classifier works as the group of the SVM classifiers at the first level and as the SVM classifier on the base of the modified PSO algorithm at the second level. The results of experimental studies confirm the efficiency of the offered approaches for Big Data classification.

Owing to global trends, automation with the fourth industrial revolution, and environmentally friendly policies owing to certified emission reduction, wire arc additive manufacturing (WAAM) is attracting attention in various industrial fields. In addition, the demands of non-ferrous metals such as aluminum and magnesium alloys are attracting attention from industries that require lightweight, large, and complex metallic parts that can be fabricated through WAAM. In this study, additive manufacturing (AM), WAAM, and current issues on the WAAM of non-ferrous metals particularly aluminum and magnesium alloys are discussed.

Additive manufacturing (AM) techniques, such as wire arc additive manufacturing (WAAM), offer unique advantages in producing large, complex structures with reduced lead time and material waste. However, their application in fatigue-critical applications requires a thorough understanding of the material properties and behavior. Due to the layered nature of the manufacturing process, WAAM structures have different microstructures and mechanical properties compared to their substrate counterparts. This study investigated the mechanical behavior and fatigue performance of Ti-6Al-4V fabricated using WAAM compared to the substrate material. Tensile and low-cycle fatigue (LCF) tests were conducted on both materials, and the microstructure was analyzed using optical microscopy and scanning electron microscopy (SEM). The results showed that the WAAM material has a coarser and more heterogeneous grain structure, an increased amount of defects, and lower ultimate tensile strength and smaller elongation at fracture. Furthermore, strain-controlled LCF tests revealed a lower fatigue strength of the WAAM material compared to the substrate, with crack initiation occurring at pores in the specimen rather than microstructural features. Experimental data were used to fit the Ramberg–Osgood model for cyclic deformation behavior and the Manson–Coffin–Basquin model for strain-life curves. The fitted models were subsequently used to compare the two material conditions with other AM processes. In general, the quasi-static properties of WAAM material were found to be lower than those of powder-based processes like selective laser melting or electron beam melting due to smaller cooling rates within the WAAM process. Finally, two simplified estimation models for the strain-life relationship were compared to the experimentally fitted Manson–Coffin–Basquin parameters. The results showed that the simple “universal material law” is applicable and can be used for a quick and simple estimation of the material behavior in cyclic loading conditions. Overall, this study highlights the importance of understanding the mechanical behavior and fatigue performance of WAAM structures compared to their substrate counterparts, as well as the need for further research to improve the understanding of the effects of WAAM process parameters on the mechanical properties and fatigue performance of the fabricated structures.

Wire arc additive manufacturing of a novel ATZM31 Mg alloy: Microstructure evolution and mechanical properties

A boosted SVM classifier trained by incremental learning and decremental unlearning approach

Wire-arc additive manufacturing (WAAM) complex components can be built-up layer by layer from metallic construction materials. In this investigation two different WAAM processes with high built-up rates (CMT and pulsed GMAW) were compared in terms of geometry formation and component properties. The reference is a rectangular thin-walled geometry made of the austenitic stainless steel 316LSi (1.4430). During the welding process, the temperature development in the weld layer was measured. The experimental comparison of CMT (\({\text{R}}_{\text{m}}=630 {\text{MPa}}\)) and pulsed GMAW (\({\text{R}}_{\text{m}}=605 {\text{MPa}}\)) is completed by the determination of the mechanical properties using micro-tensile tests. Furthermore, the additive-manufactured walls were cold rolled with a subsequent heat treatment or hot rolled to provide proof of formability and forming induced property improvement.KeywordsWAAMCMTPulsed GMAWAdditive manufacturingForming behavior

Wire Arc Additive Manufacturing (WAAM) is one of the directed energy deposition (DED) techniques used for material build-up for cost-effective repair and rebuild of engineering parts used in aerospace, marine and oil and gas industries. It can be considered as an extension of arc welding technique and uses layer-by-layer building strategy — so can be considered to be representative of additive manufacturing technique. The build-up layers have heterogeneous properties in view of the rapid cooling of previous layer and reheating when a new layer is deposited. Assessment of mechanical properties such as tensile strength and in some specific cases the fatigue strength is important to assure compatibility between parent material and WAAM section. In this paper, the tensile properties of typical low carbon steel, ER70S-6 produced through WAAM process is evaluated through automated ball indentation test technique. A Tungsten Carbide (WC) spherical ball of 1.57 mm diameter is used to indent the WAAM built-up sample using a programmed load, unload step sequence on a MTS 810 servo-hydraulic test system. The test is conducted under actuator displacement control and the load-displacement response is recorded for further processing. As the test system compliance can affect the indentation depth response, a local crack opening displacement gage is used to measure the depth of penetration close to the indenter. About five steps of loading and unloading are carried out to obtain the plastic depth of penetration during each unloading segment. The data is processed as per the analysis procedures suggested by Haggag [1]. The true stress-true strain response of the material is thus obtained and compared with the parent material data from the literature. Cyclic indentation studies have been carried out on the same sample to assess the fatigue life of WAAM material. Fatigue cycling under load control is carried out at a low frequency of 0.5 Hz and the depth of penetration is measured as a function of the number of cycles of loading. In addition, the load-crack opening displacement is logged continuously to derive the global stiffness of the material as a function of the number of cycles of loading. This information is used to estimate the failure life for a given cyclic loading condition. Based on such failure data collected from multiple load cycling experiments, a failure life curve is derived for this material.

In order to improve the performance of Support Vector Machine (SVM) classifier for imbalanced data, this paper proposes dynamic cost-sensitive SVM classifier based on chaos particle swarm optimization (CPDC_SVM). Firstly, this paper introduces dynamic cost-sensitive thought to SVM classifier, and gives the method for structuring dynamic cost and cost-sensitive SVM model. Secondly, we propose the evaluation methodology performance for classifier, and adopts decimal base to code the particles. At last, chaos thought is introduced in particle swarm optimization algorithm, and the Algorithm of the dynamic cost-sensitive SVM classifier is given, which improves convergent speed and accuracy of particle swarm optimization, and can optimize dynamic cost-sensitive SVM well, so CPDC_SVM adds effectively the convergence speed and accuracy for the particle swarm optimization algorithm. Experimental results show CPDC_SVM has higher precision than traditional SVM classifier, and dynamic cost and chaos particle swarm optimization can improve the performance for classifier.

The wire arc additive manufacturing (WAAM) process often begins printing on a substrate. In some cases, the substrate could act as a functional part of the WAAM-fabricated parts. In this work, ER70S-6 is directly deposited onto a 316L substrate using the WAAM process. Microstructural studies reveal good fusion and bonding at the interface. Due to the dilution of ER70S-6 and 316L, a gradient change in composition and microstructure was observed in the first six layers. The first layer (Fe-10Cr-5.5Ni-1Mo-1Mn) and the second layer (Fe-4Cr-2Ni-1Mn-0.5Mo) exhibit a bainite microstructure. Ferritic grains were observed from the third layer onwards. The microhardness is high at 336.9 ± 6.3 HV for the first layer and 302.7 ± 5.7 HV for the second layer. A decrease in microhardness was observed with the increase of the building layer until layer six. The yield strength is 356.6 MPa and the UTS is 502.3 MPa for the 316L/ER70S-6 bimetal joint. The digital image correlation (DIC) method is used to analyse the deformation behaviour. During the tensile test, the 316L substrate side yields first, and then the deformation shifts to the ER70S-6 side. This study also holds the potential for the development of novel alloys.

In the wire arc additive manufacturing (WAAM) process, the geometry of single welding beads has significant effects on the stability process and the final quality and shape of manufactured parts. In this paper, the geometry of single welding beads of 308L stainless steel was predicted as functions of process parameters (i.e. welding current I, voltage U, and travel speed v) by using the response surface methodology (RSM). A set of experimental runs was carried out by using the Box-Behnken design method. The adequacy of the developed models was assessed by using an analysis of variance (ANOVA). The results indicate that the RSM allows the predictive models of bead width (BW) and bead height (BH) to be developed with a high accuracy: R2-values of BW and BH are 99.01% and 99.61%, respectively. The errors between the predicted and experimental values for the confirmatory experiments are also lower than 5% that again confirms the adequacy of the developed models. These developed models can efficiently be used to predict the desirable geometry of welding beads for the adaptive slicing principle in WAAM.