Metal AM Research Articles

Multivariate process capability analysis for evaluating metal additive manufacturing via electron beam melting

Abstract Electron beam melting (EBM) as one of the relatively new metal AM techniques showed promising and increasing applications. Therefore, there is a need to evaluate the quality of the EBM process using its critical quality characteristics. However, EBM and different AM process parts have many functionally or statistically correlated quality characteristics. Consequently, the quality characteristics of the EBM process should be evaluated together. Therefore, this research aims to evaluate the quality of the EBM process using a multivariate process capability index (MPCI). In this study, the dimensional accuracy in different directions is considered as a quality characteristics. The proposed methodology involves producing a large sample of small specimens of square shape using EBM technology. Three critical dimensions of the specimen in the X, Y, and Z axis are investigated as quality characteristics. The dimensions of quality characteristics are measured using a precise measurement device. The normality and stability assumptions of the collected data are investigated using skewness measure, and multivariate process control chart respectively. Then a large sample of the multivariate normal data is simulated using computer software to estimate the percent of nonconforming (PNC) from the established specification limits, which is used to estimate MPCI. Finally, the capable tolerance of the process is estimated and the sensitivity analysis of variation is investigated. The results show the capability of the EBM process under different specification limits designations. Estimating MPCI revealed that the EBM process is capable under very coarse limits only. Moreover, the sensitivity analysis showed that variation in quality characteristics data is very sensitive for MPCI estimation, especially variation in width quality characteristic.

AMGPT: A large language model for contextual querying in additive manufacturing

Generalized large language models (LLMs) such as GPT-4 may not provide specific answers to queries formulated by materials science researchers. These models may produce a high-level outline but lack the capacity to return detailed instructions on manufacturing and material properties of novel alloys. We introduce “AMGPT”, a specialized LLM text generator designed for metal AM queries. The goal of AMGPT is to assist researchers and users in navigating a curated corpus of literature. Instead of training from scratch, we employ a pre-trained Llama2-7B model from Hugging Face in a Retrieval-Augmented Generation (RAG) setup, utilizing it to dynamically incorporate information from ∼50 AM papers and textbooks in PDF format. Mathpix is used to convert these PDF documents into TeX format, facilitating their integration into the RAG pipeline managed by LlamaIndex. A query retrieval function has also been added, enabling the system to fetch relevant literature from Elsevier journals based on the context of the query. Expert evaluations of this project highlight that specific embeddings from the RAG setup accelerate response times and maintain coherence in the generated text.

Build Plate Roughness Study on Part Bonding for the Laser Powder Bed Fusion Process

AbstractAdditive manufacturing (AM) processes have emerged as valuable partners in conventional manufacturing, facilitating the production of low-batch components with complex geometries across diverse industries. However, despite ongoing advancements in various AM technologies, consistently achieving reliable and defect-free components remains a challenge. In powder metal AM, the use of substrates or build plates to support the entire build plays a crucial role in ensuring build stability. Build plate preparation typically involves surface grinding followed by finishing sanding, leading to variations in surface roughness between different manufacturing runs. This study aimed to elucidate the bonding characteristics at the build plate-part interface by investigating the porosity and build plate-part strength at different substrate surface roughness. To this end, a multi-roughness build plate was designed and fabricated for tensile testing via laser powder bed fusion (LPBF) processing of upright specimens. The specimens were subjected to computed tomography (CT) scans for porosity assessment, followed by tensile tests to evaluate the mechanical performance at the build plate-part interface (bp-p). CT inspection revealed no porosity at the interface for any roughness level. Furthermore, analysis of the tensile behavior in relation to substrate roughness (Ra values of 0.8 μm, 1.4 μm, 3.5 μm, and 4.4 μm) did not reveal statistically significant differences.

A review of the current state-of-the-art on in situ monitoring in electron beam powder bed fusion

The industrial development of electron beam powder bed fusion (PBF-EB) is relatively younger and much more limited in terms of global widespread and revenues compared to laser powder bed fusion (PBF-L). Nevertheless, PBF-EB has been adopted in some of the most successful industrial case studies of metal AM, as it provides specific benefits and capabilities that make it a key enabling technology in a variety of industrial applications. Moreover, the recent years have seen a rapid evolution with new actors and new systems entering the market, together with a considerable increase of research and innovation programs. A field of major interest is the development and continuous improvement of in situ sensing and monitoring methods to anticipate the detection of defects, to predict the final quality of the part, and to rethink product qualification procedures. The technological features of the PBF-EB process have motivated the development of solutions that differ from the ones in PBF-L. Some of them have reached a good maturity level, being recently integrated into industrial machines, while others still deserve further research. This study explores the current state-of-the-art on in situ and in-line monitoring of the PBF-EB process, aiming to provide an up-to-date overview of the major differences with respect to PBF-L, currently available methods and their performances, as well as open issues, challenges to be tackled, and perspective for future research and industrial developments.

Temperature-based measurement interpretation of the MX3D Bridge

The MX3D Bridge is the world’s first structure produced using metal additive manufacturing (AM, i.e. 3D printing), deposited through wire and arc AM (WAAM). The long-term behaviour of structures built using metal AM is unknown and therefore requires the development of techniques that enable this investigation, to facilitate future widespread adoption of these novel manufacturing processes. The MX3D Bridge is instrumented with a comprehensive sensor network (accelerometers, displacement gauges, inclinometers, load cells, strain gauges and thermistors) to enable condition monitoring through the implementation of data-driven anomaly detection techniques. The structure was exposed to environmental and operational variability (EOV) over an 8-week commissioning period, which demonstrated that its response is predominantly temperature driven. Towards the end of the data collection period the decorative end swirls on the bridge were removed, which is taken as a proxy for damage on a structure built using WAAM. EOV is known to mask the damage-sensitive features used in anomaly detection techniques, such as those produced through the end swirl detachment. This paper explores the use of the temperature-based measurement interpretation (TB-MI) approach to: (i) predict the thermal response of the bridge using the iterative regression-based thermal response prediction (IRBTRP) methodology; (ii) remove the influence of EOV from damage sensitive features; and (iii) detect an anomaly event (e.g. the end swirl removal). Accurate predictions are obtained for the thermal response of the MX3D Bridge across 70 sensor measurement signals through the IRBTRP methodology, with 1.2% and 6.4% average prediction errors for the training/validation and unseen testing subperiods, respectively. Moving principal component analysis and cointegration are used with both the measured and thermal response corrected sensor signals for anomaly detection. The detection of the simulated damage is shown to be significantly improved through the removal of EOV, with damage being detected earlier, with greater certainty and across a wider range of sensor clusters (i.e. groups of sensors). The TB-MI approach and IRBTRP methodology can therefore be used to accurately predict the thermal response and detect damage on components and structures produced using metal AM, enabling the advantages of novel AM techniques to be realised within the built environment.

LPBF metal AM platform for serial production

EOS adds to its midframe industrial 3D printing portfolio.

An experimental and numerical study on the influence of interlayer time interval in wire-arc additive manufacturing process

Wire arc additive manufacturing is the derivative of the metal AM process based on arc welding and a preferred technology for manufacturing large metallic components with medium to low complexity. The major problem with the WAAM process arises due to severe heat accumulation within the built structure and repeated heating-cooling cycles during layer-by-layer material deposition. Therefore, the interlayer time interval influences the microstructure and mechanical properties of the components. This paper has characterized the impact of interlayer cooling time on the microstructure and mechanical properties of SS-316L build. The local thermal cycle causes a non-uniform cooling rate, which alters the structure of the grains. The grain size decreases with increased interlayer time interval due to a faster cooling rate brought on by the low inter-pass temperature. A wider and smaller wall was deposited with less time interval. X-ray diffraction analysis confirms the formation of austenite and ferrite phases in all WAAM components with a slight difference in intensity. The wall constructed with a maximum time interval has a maximum value of average Vickers micro-hardness 244.00 HV0.5 and possesses the least coefficient of friction due to its improved property by the formation of fine grains. The maximum volume of materials was worn out from the wall deposited with lesser cooling time. The designed numerical model predicts the peak temperature and heat accumulation within the built structure which further relates to the morphological and mechanical properties of printed parts.

Influence of the physicochemical properties of reusable powders on the mechanical properties of Ti6Al4V manufactured via laser powder bed fusion additive manufacturing

Laser powder bed fusion additive manufacturing (L-PBF AM) technology holds notable advantages regarding raw material reusability. Nonetheless, a notable gap exists in extensive research assessing how the reuse of powders influences their intrinsic physicochemical traits as well as the mechanical attributes of the fabricated components. This study focuses on introducing an innovative approach to Ti6Al4V powder recycling in L-PBF AM, involving the incorporation of virgin spherical powder to maintain powder characteristics prior to printing. This results in sustained mechanical properties throughout seven cycles of printing (P7), with the fifth print (P5) showcasing the optimal mechanical performance (strength of 1185 MPa and elongation of 6.3%). Transmission electron microscopy (TEM) analysis was performed before and after the P5 tensile experiments，revealing a high amount of primary twinning {101¯1} in LPBF. This study offers both theoretical insights and practical validation, thereby lending substantial support to the advancement of metal AM industrialization.

A model for the rapid prediction of small feature width producible by Metal AM

In this paper, an analytical parametric model is presented for the swift prediction of small features producible in the X-Y plane through the Metal AM Laser Powder Bed Fusion (LPBF) process. This model can be employed to design and create surface textures on Additive Manufactured parts without the need for costly post-processing steps. The Rosenthal equation is the basis for the model, which considers both the build parameters of the LPBF process and the thermo-physical properties of the materials. The initial model was constructed and assessed using one LPBF machine followed by the implementation of a tuning method utilizing the Limited Memory Algorithm for Bound Constrained Optimization to enhance the model’s accuracy. Overall, the findings suggest that with a simple optimization step based on a single printed tuning sample, precise analytical models can be established for specific LPBF machines and materials combinations.

EXplainable artificial intelligence for automatic defect detection in additively manufactured parts using CT scan analysis

AbstractAdditive Manufacturing (AM) and in particular has gained significant attention due to its capability to produce complex geometries using various materials, resulting in cost and mass reduction per part. However, metal AM parts often contain internal defects inherent to the manufacturing process. Non-Destructive Testing (NDT), particularly Computed Tomography (CT), is commonly employed for defect analysis. Today adopted standard inspection techniques are costly and time-consuming, therefore an automatic approach is needed. This paper presents a novel eXplainable Artificial Intelligence (XAI) methodology for defect detection and characterization. To classify pixel data from CT images as pores or inclusions, the proposed method utilizes Support Vector Machine (SVM), a supervised machine learning algorithm, trained with an Area Under the Curve (AUC) of 0.94. Density-Based Spatial Clustering with the Application of Noise (DBSCAN) is subsequently applied to cluster the identified pixels into separate defects, and finally, a convex hull is employed to characterize the identified clusters based on their size and shape. The effectiveness of the methodology is evaluated on Ti6Al4V specimens, comparing the results obtained from manual inspection and the ML-based approach with the guidance of a domain expert. This work establishes a foundation for automated defect detection, highlighting the crucial role of XAI in ensuring trust in NDT, thereby offering new possibilities for the evaluation of AM components.

Detailed evaluation of topographical effects of Hirtisation post-processing on electron beam powder bed fusion (PBF-EB) manufactured Ti-6Al-4V component

Metal additive manufacturing surface topographies are complex and challenging to characterise due to e.g. steep local slopes, re-entrant features, varying reflectivity and features of interest in vastly different scale ranges. Nevertheless, average height parameters such as Ra or Sa are commonly used as sole parameters for characterisation. In this paper, a novel method for selecting relevant parameters for evaluation is proposed and demonstrated using a case study where the smoothing effects after three processing steps of the electro chemical post-process Hirtisation of a metal AM surface are quantified. The method uses a combination of conventional areal texture parameters, multiscale analysis and statistics and can be used to efficiently achieve a detailed and more relevant surface topography characterisation. It was found that the three process steps have different effects on the surface topography regarding the types and sizes of features that were affected. In total, Sdq was reduced by 97 %, S5v was reduced by 81 % and Sa was reduced by 78 %. A surface texture with much lower average roughness, less deep pits and less steep slopes was produced, which is expected to be beneficial for improved fatigue properties.

Machine-to-machine variability of roughness and corrosion in additively manufactured 316L stainless steel

Numerous studies on the corrosion response of metal AM have been conducted. Nonetheless, the specimens being tested are commonly ground or polished to remove the outer as-built surface. If metal AM is truly going to be employed as a transformative technology that can produce complex shapes that do not require traditional machining, then the material needs to be evaluated in the as-built state. The reality is that AM alloys have shown significant inconsistencies regarding as-built surface texture, topology, and residual stress. One metric that has shown significant unpredictability in the literature is the susceptibility to localized corrosion of typically passive alloys, such as stainless steel (SS). There are a large number of studies that have attempted to understand the corrosion response of metal AM materials, but these studies are typically performed on materials that have been printed on a single machine and often mechanically polished to a smooth finish. This study compares the corrosion response of as-built AM, laser-beam powder bed fusion (LB-PBF), 316L SS parts that have been fabricated on five different machines. The majority of this work focused on understanding the susceptibility to localized corrosion of AM metals with respect to machine-dependent variables, namely surface roughness, and build angle. Surface roughness data was collected using scanning white light triangulation, laser scanning confocal microscopy, and coherence scanning interferometry. The results show that there is significant variability (ptwo-tail < 0.05) in the susceptibility to local corrosion initiation of LB-PBF 316L SS samples built on different machines. Surface oxides were probed with electron dispersive spectroscopy and revealed that variations in local corrosion susceptibility likely arise from differences in the stability of the passive film caused by chemical segregation, unique microstructure, and tortuous roughness features at the as-built surfaces. The variability of roughness and corrosion properties from test samples printed on different machines was corroborated by property measurements performed at five different testing sites, proving reproducibility of the data. Most importantly, this study shows that if the as-built surface layer is removed through grinding or electropolishing the machine-to-machine variation observed in the corrosion susceptibility is reduced.

Polishing of metal 3D printed parts with complex geometry: Visualizing the influence on geometrical features using centrifugal disk finishing.

Parts produced with metal additive manufacturing often suffer from a poor surface finish. Surface finishing techniques are effective to improve the quality of 3D printed surfaces, however they have as downsides that they also slightly change the geometry of the part, in an unpredictable way. This effect on the geometrical features of complex parts has received little attention. In this research, we illustrate a method to visualize the impact of surface finishing techniques on geometrical features, as well as their effectiveness on parts with high shape-complexity, by using centrifugal disk finishing as a case study. We designed and 3D printed test parts with different features using selective laser melting, which were coated with a blue metal lacquer prior to polishing. After polishing, the blue lacquer was eroded away from the spots that were easily reached by the polishing process, yet had remained on the surfaces that could not be reached by the process. We used measurements of material removal and image processing of the remaining blue lacquer on the surfaces to analyze these effects. Using this method, we were able to derive a number of specific design guidelines that can be incorporated while designing metal AM parts for centrifugal disk finishing. We suggest that this visualization method can be applied to different polishing methods to gain insight into their influence, as well as being used as an aid in the design process.

High surface quality additive manufacturing process of titanium alloy with composite heat source

In order to solve the problems of high surface roughness and low dimensional accuracy of titanium alloy metal-wire fuse additive manufacturing, this paper proposes a high surface quality and dimensional accuracy of titanium alloy additive manufacturing with composite heat source process. On the basis of orthogonal experiments, titanium alloy wires with diameter of 0.3 mm Ti-6Al-4V were deposited in a small molten pool by laser and Joule heat source. Thin-walled titanium alloy parts with high surface quality and dimensional accuracy were obtained, and the quality, microstructure, and mechanical properties were tested. The measurement areas are shown that the average roughness Ra is 1.65 μm in the direction deposition, and the dimensional accuracy error is less than 0.28 mm. On this basis, the microstructure and mechanical properties of titanium alloy thin-walled were analyzed. The microstructures are net-like, with microhardness of about 330HV. The ultimate tensile strength of titanium alloy thin-walled parts is 917.56 MPa, and the elongation can reach 20.68%, which is comparable to the strength of casting and forging fracture. The results show that the proposed additive manufacturing process of metal-wire with composite heat source has the potential of further metal AM in high surface quality parts.

Digital Twin Science of Metal AM and Creation of Materials by Super-Thermal Field

Digital Twin Science of Metal AM and Creation of Materials by Super-Thermal Field

Experimental Setup for Fatigue Testing of Additively Manufactured Specimens

Poor fatigue life is a huge issue of additively manufactured parts, despite the unique qualities characterizing this manufacturing process (such as low waste of material and geometry freedom). Fatigue life is strongly affected by both surface defects and internal defects, metal AM is characterized by extremely poor surface quality, internal porosities and lack of fusions. For this reason, many researchers investigated methods to improve manufacts quality. The most promising methods are surface finishing treatments and thermal treatments which provide an enhancement of fatigue behavior. A focal point of the research should be evaluating the respective contribution of surface treatments and thermal treatments. In order to evaluate the effectiveness of surface treatment, it is necessary to highlight the surface quality contribution in terms of fatigue life thus a specific testing method is necessary. Rotating beam fatigue test fits this requirement because each point of the specimen’s surface is subjected to the maximum stress. The aim of this work is to present the experimental setup for rotating beam fatigue testing that has been used to evaluate the fatigue behavior of AM SLM IN718 specimens.

Plasma‐Electrolytic Polishing as a Post‐Processing Technology for Additively Manufactured Parts

AbstractAdditive manufacturing (AM) as a key enabling technology provides new options for engineering design of complex parts. Freeform shapes, grid‐like structures and internal cavities enable functional integration, reducing assembly and space requirements and increasing reliability. Metal AM technologies, however, require post‐processing in order to achieve acceptable surface roughness and ‐functionality. Plasma‐electrolytic polishing can, due to its toolless nature and excellent achievable surface properties, be a beneficial option.

Augmenting wire arc additive manufacturing with laser forming for generative realization of complex geometries

Due to its advantages like higher deposition rates and lower costs, Wire Arc Additive Manufacturing (WAAM) is increasingly used to fabricate metal AM parts, particularly big-sized components. However, realization of complex geometrical features, like overhangs, remains one of the major challenges. The present work describes a hybrid approach of intervention of fiber laser intermittently during part fabrication to enhance the geometrical complexity possible. In this two-step process, a component deposited in first stage is subjected to multiple laser passes in second stage to obtain a desired angle/shape. This cycle of deposition and laser forming continues till the final desired geometry is realized. The study includes preliminary trials on deposited specimens to understand dominant laser bending mechanism and decide optimum laser process parameters. Subsequently, a few illustrative components are manufactured using current hybrid process, establishing the effectiveness of this approach in realizing complex shapes, thus laying a foundation for a potential metal 4D printing technology.

A Qualitative and Quantitative Occupational Exposure Risk Assessment to Hazardous Substances during Powder-Bed Fusion Processes in Metal-Additive Manufacturing

Metal-additive manufacturing (AM), particularly the powder-bed fusion (PBF) technique, is undergoing a transition from the short-run production of components to higher-volume manufacturing. The industry’s increased production efficiency is paired with a growing awareness of the risks related to the inhalation of very fine metal powders during PBF and AM processes, and there is a pressing need for a ready-to-use approach to assess the risks and the occupational exposure to these very final metal powders. This article presents a study conducted in an AM facility, which was conducted with the aim to propose a solution to monitor incidental airborne particle emissions during metal AM by setting up an analytical network for a tailored approach to risk assessment. Quantitative data about the respirable and inhalable particle and metal content were obtained by gravimetric and ICP-MS analyses. In addition, the concentrations of airborne particles (10–300 nm) were investigated using a direct reading instrument. A qualitative approach for risk assessment was fulfilled using control banding Nanotool v2.0. The results show that the operations in the AM facility are in line with exposure limit levels for both micron-sized and nano-sized particles. The particulate observed in the working area contains metals, such as chromium, cobalt, and nickel; thus, biological monitoring is recommended. To manage the risk level observed for all of the tasks during the AM process, containment and the supervision of an occupational safety expert are recommended to manage the risk. This study represents a useful tool that can be used to carry out a static evaluation of the risk and exposure to potentially harmful very fine metal powders in AM; however, due to the continuous innovations in this field, a dynamic approach could represent an interesting future perspective for occupational safety.

Robust prediction and validation of as-built density of Ti-6Al-4V parts manufactured via selective laser melting using a machine learning approach

It is well known that slight changes in selective laser melting (SLM) process parameters may alter the outcome of mechanical and physical properties of the as-built material in a drastic and haphazard fashion. To overcome this, reliable property prediction models are most pertinent. In this study, a machine learning approach based on Gaussian Process Regression (GPR) is proposed to predict the relative density of as-built Ti-6Al-4V alloy manufactured via SLM, based on the most common input process parameters such as laser power, scanning speed, hatch spacing, and layer thickness, as well as an integrated input of volumetric energy density. A most comprehensive test dataset to train and verify GPR models was retrieved from literature papers that extensively investigated mechanical and physical properties of additively manufactured Ti-6Al-4V alloy. GPR models with four different kernel functions were analyzed and exponential GPR model with optimized hyperparameters was chosen as the most viable model for predicting as-built density of Ti-6Al-4V alloy. A parametric multiple linear regression (MLR) model was also presented and serves as a benchmark. When inferences were made on newer publication data, the GPR model and the MLR model predicted the densities with mean absolute errors (MAE) of 1.12% and 5.22% respectively. The inferior performance of the MLR model compared emphasizes the need of non-parametric supervised learning technique for SLM. To truly demonstrate the effectiveness of the proposed GPR model in real-world metal AM jobs, 22 experimental samples were printed. Predictions made on all the samples, when compared to their actual density values, resulted in MAE of 0.27%. Clearly, creation of most comprehensive mined data, kernel selection, and rigorous validation and verification of GPR model make this study one of its kind and prove the GPR model's predictive dexterity and the potential impact in the world of additive manufacturing.