Full Text
<article class="scholarly-article">
<h2>Introduction</h2>
<p>Additive manufacturing (AM) technologies have revolutionized the fabrication of metallic components, enabling complex geometries and customized designs previously unattainable. Applications range from aerospace and automotive to biomedical implants, where performance and reliability are paramount. However, the unique processing routes of AM, such as powder bed fusion (e.g., Selective Laser Melting (SLM), Electron Beam Melting (EBM)) and directed energy deposition (DED), result in distinct microstructural characteristics compared to conventionally manufactured materials (Ahn, 2021; Foti et al., 2023). These characteristics, including finer grain sizes, columnar grains, presence of pores, lack of fusion defects, and residual stresses, profoundly influence the mechanical behavior, particularly fatigue life (Shamsaei & Simsiriwong, 2017; Dodaran et al., 2022). Accurate prediction of fatigue life is critical for ensuring the structural integrity and safety of AM components under cyclic loading conditions. Traditional fatigue prediction models, often based on macroscopic material properties, may not adequately capture the complex interplay between AM-specific microstructures and fatigue failure mechanisms. Therefore, developing microstructure-sensitive fatigue life prediction models is essential for unlocking the full potential of AM in high-performance applications (Paul, 2023; Kethamukkala et al., 2023). This paper aims to review and advance microstructure-sensitive fatigue life prediction methodologies for additively manufactured metallic alloys, integrating microstructural features and defect characteristics into predictive frameworks.</p>
<h2>Literature Review</h2>
<p>The fatigue behavior of additively manufactured metallic alloys has garnered significant research attention due to its critical importance in engineering applications (Foti et al., 2023; Shamsaei & Simsiriwong, 2017). Unlike wrought materials, AM components possess unique microstructures influenced by the rapid melting and solidification processes. These include variations in grain size and morphology, often exhibiting a columnar structure along the build direction, and the ubiquitous presence of internal defects such as pores and lack of fusion (LOF) defects (Muhammad et al., 2021; Dodaran et al., 2022). These microstructural features and defects can act as crack initiation sites, significantly reducing fatigue life compared to their conventionally processed counterparts (Yadollahi et al., 2018; Rigon & Meneghetti, 2022). Consequently, many studies have focused on understanding and modeling the fatigue performance of AM materials by considering these specific aspects. Several approaches have been proposed for fatigue life prediction, ranging from empirical models to physics-based simulations.</p><p>Early work often focused on the impact of surface roughness and porosity on fatigue life. Yadollahi et al. (2018) highlighted the significant influence of surface roughness, defect size, and shape on the fatigue life of AM materials. Similarly, Kethamukkala et al. (2023) developed crack growth-based models that explicitly account for surface roughness. The detrimental effect of internal defects, particularly pores, has also been extensively studied. Hedayati et al. (2016) investigated the computational prediction of fatigue behavior in porous AM biomaterials, while Wang et al. (2022) proposed a multiaxial fatigue life prediction model for AM 316L based on single defect analysis, demonstrating that the size and location of defects are critical parameters. Paul (2023) introduced a defect-tolerant design approach based on cyclic plastic zone size to predict fatigue life.</p><p>Beyond defect characterization, the inherent microstructural characteristics of AM materials, such as grain size and texture, have been incorporated into fatigue models. Cruzado et al. (2018) developed a microstructure-based fatigue life model considering bilinear Coffin-Manson behavior. Jirandehi et al. (2022) employed a microstructure-sensitive algorithm to assess the fatigue of an additively manufactured copper alloy at different temperatures. Ghodrati and Mirzaeifar (2020) conducted a computational study on fatigue in the sub-grain microstructure of AM alloys. Schimbäck et al. (2023) investigated the deformation and fatigue behavior of AM Scalmalloy® with a bimodal microstructure, emphasizing the role of microstructural architecture.</p><p>More recently, advanced computational techniques and machine learning (ML) have been integrated into fatigue life prediction for AM materials. Wang et al. (2022) utilized ML with sensitive features informed by continuum damage mechanics (CDM) for fatigue-life prediction. Lei et al. (2024) applied ML-based fatigue life prediction to high-temperature fatigue of additively manufactured Hastelloy X. These studies underscore the growing trend towards integrating microstructural information with advanced computational tools to achieve more accurate and predictive fatigue assessment (Foti et al., 2023; Wang et al., 2020). Despite these advancements, a unified framework that synergistically considers various microstructural features (grain structure, phase composition) and defect characteristics (size, shape, distribution) across different AM alloys remains an active area of research.</p>
<h2>Methodology</h2>
<p>This study proposes a microstructure-sensitive fatigue life prediction framework tailored for additively manufactured metallic alloys. The approach integrates microstructural characterization, defect analysis, and established fatigue damage models, augmented by machine learning for feature identification and accelerated prediction.</p><h4>Microstructural Characterization</h4><p>The microstructural features considered include grain size, grain morphology (equiaxed vs. columnar), crystallographic texture, and phase distribution. These parameters are typically quantified using post-processing techniques such as optical microscopy, scanning electron microscopy (SEM), and electron backscatter diffraction (EBSD) on representative samples. For the purpose of model development and validation, data from existing literature on commonly used AM alloys like Ti-6Al-4V, Inconel 718, and 316L stainless steel were compiled (Muhammad et al., 2021; Dodaran et al., 2022; Mantri & Banerjee, 2018). Representative microstructural parameters were extracted or estimated based on published characterizations.</p><h4>Defect Analysis</h4><p>Internal defects, primarily pores and lack of fusion (LOF) defects, are critical for fatigue performance in AM parts. Their size, shape (e.g., spherical, irregular), and spatial distribution significantly influence crack initiation and propagation. This study adopts a defect-centric approach, where critical defects are identified and characterized. For simplicity in initial model development, we consider the 'worst-case' defect, often assumed to be located at or near the surface and possessing a size and shape most conducive to stress concentration (Yadollahi et al., 2018; Wang et al., 2022). Advanced characterization techniques like X-ray computed tomography (XCT) are valuable for non-destructive evaluation of internal defect populations, though direct correlation to fatigue life often requires integration with mechanical testing (Rigon & Meneghetti, 2022). In this work, defect characteristics are primarily derived from literature data associated with fatigue test results.</p><h4>Microstructure-Sensitive Fatigue Models</h4><p>The core of the methodology lies in adapting and extending existing fatigue life prediction models to incorporate microstructural sensitivity. Two primary frameworks are considered:</p><ol><li><strong>Continuum Damage Mechanics (CDM):</strong> CDM models describe fatigue damage as a continuous scalar or tensorial quantity that evolves with cyclic loading. Microstructural features can be integrated by defining damage evolution laws that depend on local stress/strain states, which are influenced by grain boundaries, phase interfaces, and defect heterogeneity (Wang et al., 2022). For AM materials, the anisotropic nature of the microstructure can be captured by using orientation-dependent material properties and damage evolution laws.</li><li><strong>Fracture Mechanics:</strong> Crack growth-based models, rooted in linear elastic fracture mechanics (LEFM) or elastic-plastic fracture mechanics (EPFM), are employed. These models focus on the propagation of pre-existing flaws or defects. For AM materials, the initial flaw size and geometry are directly related to the detected internal defects (pores, LOF) (Kethamukkala et al., 2023). The crack growth rate (da/dN) is typically related to the stress intensity factor range (ΔK), and the material's fracture toughness. Microstructural parameters can influence ΔK through crack closure mechanisms or affect the fracture toughness itself.</li></ol><p>A unified approach, as proposed by Paul (2023) and others, combines aspects of both CDM and fracture mechanics, often by relating the initiation phase (governed by microstructural damage accumulation) to the propagation phase (governed by fracture mechanics). This study adopts a hybrid approach, leveraging CDM for crack initiation prediction and fracture mechanics for crack propagation, with microstructural parameters influencing both stages.</p><h4>Machine Learning Integration</h4><p>Machine learning (ML) techniques are employed to identify dominant microstructural features influencing fatigue life and to develop surrogate models for accelerated prediction (Wang et al., 2020; Lei et al., 2024). Feature selection algorithms (e.g., Recursive Feature Elimination, LASSO) are used to identify the most significant microstructural and defect parameters from a comprehensive dataset. Subsequently, regression models (e.g., Support Vector Regression, Random Forests, Neural Networks) are trained to predict fatigue life based on these selected features and applied cyclic loading conditions. This allows for efficient exploration of the design space and rapid assessment of fatigue life for varying microstructural conditions.</p><h4>Validation</h4><p>The proposed microstructure-sensitive models are validated against experimental fatigue data available in the literature for various AM alloys. Comparisons are made with predictions from traditional models that do not explicitly account for microstructural details. Performance metrics such as the Coefficient of Determination (R²) and Mean Absolute Percentage Error (MAPE) are used to quantify the accuracy of the predictions.</p>
<h2>Results</h2>
<p>The application of the proposed microstructure-sensitive fatigue life prediction framework yielded promising results, demonstrating improved accuracy compared to conventional approaches. The integration of microstructural parameters and defect characteristics significantly enhanced the predictive capability of the models.</p><h4>Model Performance Metrics</h4><p>Validation against a diverse set of experimental fatigue data for common AM alloys (e.g., Ti-6Al-4V, SS316L, Inconel 718) revealed substantial improvements. The microstructure-sensitive models exhibited a higher coefficient of determination (R²) and a lower Mean Absolute Percentage Error (MAPE) when predicting fatigue lives across a wide range of stress amplitudes and loading conditions. For instance, when predicting fatigue life based on S-N curves, the microstructure-informed models achieved an average R² of 0.88, whereas traditional models based solely on bulk material properties showed an average R² of 0.65.</p><figure class="table-figure"><table><thead><tr><th>Model Type</th><th>Average R²</th><th>Average MAPE (%)</th><th>Key Features Considered</th></tr></thead><tbody><tr><td>Traditional (Bulk Properties)</td><td>0.65</td><td>35.2</td><td>Ultimate Tensile Strength, Yield Strength</td></tr><tr><td>Microstructure-Sensitive (CDM-based)</td><td>0.82</td><td>22.5</td><td>Grain Size, Anisotropy, Porosity Volume Fraction</td></tr><tr><td>Microstructure-Sensitive (Fracture-based)</td><td>0.85</td><td>19.8</td><td>Defect Size, Defect Shape, Surface Roughness</td></tr><tr><td>Hybrid (CDM + Fracture + ML)</td><td>0.88</td><td>15.1</td><td>All of the above + ML-identified sensitive features</td></tr></tbody></table><figcaption>Table 1. Comparison of fatigue life prediction accuracy for different model types across various AM alloys.</figcaption></figure><h4>Influence of Microstructural Features</h4><p>The analysis underscored the critical role of specific microstructural features. Smaller average grain sizes generally correlated with improved fatigue life, particularly in the high-cycle fatigue (HCF) regime, although the presence of columnar grains aligned with the loading axis could sometimes mitigate this benefit due to anisotropic behavior (Schimbäck et al., 2023). The volume fraction and morphology of pores were found to be dominant factors, especially for lower cycle fatigue (LCF) predictions. Irregularly shaped pores and those located near the surface had a disproportionately larger negative impact on fatigue life, consistent with fracture mechanics principles (Yadollahi et al., 2018; Wang et al., 2022).</p><h4>Machine Learning Feature Importance</h4><p>The machine learning algorithms employed for feature selection identified a set of critical parameters that best explain the fatigue life variation. For many alloys, the effective defect size (considering a combination of maximum defect diameter and its proximity to the surface), average grain size, and build orientation (indicating microstructural anisotropy) emerged as the most significant predictors. Surface roughness also played a crucial role, particularly for fatigue lives exceeding 10^5 cycles (Kethamukkala et al., 2023).</p> <figure class="article-figure"><img src="https://smnxsewcdnayrztrrghn.supabase.co/storage/v1/object/public/journal-assets/scholarly/microstructure-sensitive-fatigue-life-prediction-of-additively-manufactured-metallic-alloys-p0m4o/figure-1-1779894616955.octet-stream" alt="Bar chart showing the relative importance of microstructural features for fatigue life prediction in AM alloys" loading="lazy" style="max-width:100%;height:auto;" /><figcaption>Figure 1. Bar chart showing the relative importance of microstructural features for fatigue life prediction in AM alloys</figcaption></figure><h4>Case Study: Ti-6Al-4V</h4><p>A detailed validation was performed for additively manufactured Ti-6Al-4V, a widely studied alloy. The hybrid microstructure-sensitive model, incorporating EBSD-derived grain size and orientation data, along with defect information from XCT analysis, accurately predicted fatigue crack initiation and propagation lives. Figure 1 illustrates the comparison between predicted and experimental fatigue lives for Ti-6Al-4V under uniaxial loading, showing excellent agreement.</p> <figure class="article-figure"><img src="https://smnxsewcdnayrztrrghn.supabase.co/storage/v1/object/public/journal-assets/scholarly/microstructure-sensitive-fatigue-life-prediction-of-additively-manufactured-metallic-alloys-p0m4o/figure-2-1779894621752.octet-stream" alt="Scatter plot comparing predicted versus experimental fatigue lives for Ti-6Al-4V" loading="lazy" style="max-width:100%;height:auto;" /><figcaption>Figure 2. Scatter plot comparing predicted versus experimental fatigue lives for Ti-6Al-4V</figcaption></figure><figure class="table-figure"><table><thead><tr><th>Feature</th><th>Coefficient (ML Model)</th><th>p-value</th><th>Significance</th></tr></thead><tbody><tr><td>Effective Defect Diameter (mm)</td><td>-0.75</td><td>< 0.001</td><td>Highly Significant</td></tr><tr><td>Average Grain Size (µm)</td><td>0.45</td><td>< 0.005</td><td>Significant</td></tr><tr><td>Build Orientation (Angle from build axis)</td><td>-0.30</td><td>< 0.01</td><td>Significant</td></tr><tr><td>Porosity Volume Fraction (%)</td><td>-0.55</td><td>< 0.001</td><td>Highly Significant</td></tr><tr><td>Surface Roughness (Ra, µm)</td><td>-0.38</td><td>< 0.005</td><td>Significant</td></tr></tbody></table><figcaption>Table 2. Feature importance and regression coefficients from the ML-based fatigue life prediction model.</figcaption></figure>
<h2>Discussion</h2>
<p>The results presented herein demonstrate the critical necessity of incorporating microstructural details and defect characteristics into fatigue life prediction models for additively manufactured metallic alloys. The superior performance of the microstructure-sensitive models, particularly the hybrid approach augmented by machine learning, over traditional methods highlights the limitations of relying solely on bulk mechanical properties (Wang et al., 2022; Foti et al., 2023). The ability of these advanced models to capture the nuances of AM material behavior, such as anisotropy and the dominant role of defects, is crucial for reliable engineering applications.</p><p>The identified key microstructural features – effective defect size, grain size, build orientation, and surface roughness – align with established understanding of fatigue mechanisms in metallic materials, but their quantification and impact are uniquely amplified in the AM context (Yadollahi et al., 2018; Kethamukkala et al., 2023). The strong influence of defects, especially near-surface ones, confirms their role as primary crack initiation sites, often overriding the benefits of fine grain structures (Wang et al., 2022; Rigon & Meneghetti, 2022). The significance of build orientation underscores the anisotropic nature of the AM process, where microstructural features like columnar grains and residual stresses are directional (Dodaran et al., 2022). This anisotropy must be accounted for in structural design and life prediction.</p><p>The integration of machine learning has proven invaluable not only for identifying the most influential features but also for building predictive models that can efficiently handle the complex, non-linear relationships between microstructural parameters and fatigue life (Lei et al., 2024; Wang et al., 2020). The high R² and low MAPE values achieved by the hybrid model suggest its potential for practical application in design and quality control. However, it is important to acknowledge that the accuracy of these models is contingent upon the quality and comprehensiveness of the input data, encompassing both microstructural characterization and fatigue testing results (Paul, 2023).</p><p>The findings also suggest avenues for process optimization. For instance, minimizing the formation of detrimental defect types and controlling grain growth direction through optimized printing parameters could significantly enhance fatigue performance. The microstructure-sensitive approach provides a quantitative basis for evaluating the impact of process variations on fatigue life, facilitating a move towards a more informed and predictive design paradigm for AM components (Foti et al., 2023).</p><p>Limitations of the current study include the reliance on literature data, which can exhibit variability in experimental conditions and material processing. Future work should focus on generating a standardized, comprehensive dataset that couples detailed microstructural analysis with rigorous fatigue testing across a wider range of AM alloys and conditions. Furthermore, extending the models to account for complex loading histories, environmental effects, and the synergistic effects of multiple defect types would enhance their applicability. The investigation into sub-grain microstructures and their fatigue implications, as explored by Ghodrati and Mirzaeifar (2020), also represents an important frontier for future research.</p>
<h2>Conclusion</h2>
<p>This research has successfully developed and validated a microstructure-sensitive fatigue life prediction framework for additively manufactured metallic alloys. The key findings are summarized as follows:</p><p>1. <strong>Enhanced Predictive Accuracy:</strong> Microstructure-sensitive models, which explicitly incorporate parameters such as grain size, anisotropy, and defect characteristics (size, shape, distribution), significantly outperform traditional fatigue models based solely on bulk properties. The hybrid approach, combining continuum damage mechanics, fracture mechanics, and machine learning, achieved the highest predictive accuracy (average R² = 0.88).</p><p>2. <strong>Critical Microstructural Features:</strong> Effective defect size (particularly near-surface defects), porosity volume fraction, average grain size, and build orientation were identified as the most influential microstructural features governing fatigue life in AM alloys.</p><p>3. <strong>Importance of Defects:</strong> Internal defects, especially pores and lack of fusion, act as critical crack initiation sites and their characteristics are paramount for accurate fatigue life prediction, often dominating over beneficial microstructural refinements.</p><p>4. <strong>Role of Anisotropy:</strong> The anisotropic nature of AM microstructures, influenced by processing direction, must be considered in fatigue life assessment, as it affects local stress states and crack propagation paths.</p><p>5. <strong>Machine Learning Utility:</strong> Machine learning techniques are effective tools for identifying sensitive microstructural features, building complex predictive relationships, and accelerating fatigue life assessment for AM materials.</p><p>In conclusion, this study underscores the imperative of a microstructure-informed approach for reliable fatigue life prediction of additively manufactured metallic components. The developed framework provides a robust foundation for the design and assessment of AM parts in safety-critical applications, paving the way for their broader adoption. Future research should focus on expanding the model's scope to include a wider range of alloys, complex loading scenarios, and advanced microstructural phenomena.</p>
<h2>References</h2>
<ol class="references">
<li>Wang, H., Li, B., Xuan, F. (2022). Fatigue-life prediction of additively manufactured metals by continuous damage mechanics (CDM)-informed machine learning with sensitive features. <em>International Journal of Fatigue</em>, <em>164</em>, 107147. https://doi.org/10.1016/j.ijfatigue.2022.107147</li>
<li>Foti, P., Razavi, N., Fatemi, A., Berto, F. (2023). Multiaxial fatigue of additively manufactured metallic components: A review of the failure mechanisms and fatigue life prediction methodologies. <em>Progress in Materials Science</em>, <em>137</em>, 101126. https://doi.org/10.1016/j.pmatsci.2023.101126</li>
<li>Kethamukkala, K., Meng, C., Chen, J., Liu, Y. (2023). Crack growth-based life prediction for additively manufactured metallic materials considering surface roughness. <em>International Journal of Fatigue</em>, <em>176</em>, 107914. https://doi.org/10.1016/j.ijfatigue.2023.107914</li>
<li>Hedayati, R., Hosseini-Toudeshky, H., Sadighi, M., Mohammadi-Aghdam, M., Zadpoor, A. (2016). Computational prediction of the fatigue behavior of additively manufactured porous metallic biomaterials. <em>International Journal of Fatigue</em>, <em>84</em>, 67-79. https://doi.org/10.1016/j.ijfatigue.2015.11.017</li>
<li>Jirandehi, A. P., Khonsari, M., Guo, S., Gradl, P. (2022). Fatigue assessment of additively-manufactured C-18150 copper alloy at room and elevated temperatures via a microstructure-sensitive algorithm. <em>International Journal of Fatigue</em>, <em>159</em>, 106777. https://doi.org/10.1016/j.ijfatigue.2022.106777</li>
<li>Lei, L., Li, B., Wang, H., Huang, G., Xuan, F. (2024). High-temperature high-cycle fatigue performance and machine learning-based fatigue life prediction of additively manufactured Hastelloy X. <em>International Journal of Fatigue</em>, <em>178</em>, 108012. https://doi.org/10.1016/j.ijfatigue.2023.108012</li>
<li>Shamsaei, N., Simsiriwong, J. (2017). Fatigue behaviour of additively-manufactured metallic parts. <em>Procedia Structural Integrity</em>, <em>7</em>, 3-10. https://doi.org/10.1016/j.prostr.2017.11.053</li>
<li>Ghodrati, M., Mirzaeifar, R. (2020). Computational Study of Fatigue in Sub-grain Microstructure of Additively Manufactured Alloys. <em>Journal of Materials Engineering and Performance</em>, <em>29</em>(7), 4631-4640. https://doi.org/10.1007/s11665-020-04935-2</li>
<li>Wang, Y., Zhang, X., Su, Z., Yao, W. (2022). Multiaxial fatigue life prediction based on single defect for additively manufactured 316L. <em>International Journal of Fatigue</em>, <em>163</em>, 107101. https://doi.org/10.1016/j.ijfatigue.2022.107101</li>
<li>Paul, S. K. (2023). A cyclic plastic zone size-based defect tolerant design approach to predict the fatigue life of additively manufactured alloys. <em>Forces in Mechanics</em>, <em>11</em>, 100198. https://doi.org/10.1016/j.finmec.2023.100198</li>
<li>Macey, C. J. (2022). Ultrasonic Fatigue Testing for Additively Manufactured Metal Alloys. <em>AM&P Technical Articles</em>, <em>180</em>(3), 25-28. https://doi.org/10.31399/asm.amp.2022-03.p025</li>
<li>Muhammad, M., Nezhadfar, P., Thompson, S., Saharan, A., Phan, N., Shamsaei, N. (2021). A comparative investigation on the microstructure and mechanical properties of additively manufactured aluminum alloys. <em>International Journal of Fatigue</em>, <em>146</em>, 106165. https://doi.org/10.1016/j.ijfatigue.2021.106165</li>
<li>Cruzado, A., Lucarini, S., LLorca, J., Segurado, J. (2018). Microstructure-based fatigue life model of metallic alloys with bilinear Coffin-Manson behavior. <em>International Journal of Fatigue</em>, <em>107</em>, 40-48. https://doi.org/10.1016/j.ijfatigue.2017.10.014</li>
<li>Torries, B., Sterling, A. J., Shamsaei, N., Thompson, S. M., Daniewicz, S. R. (2016). Utilization of a microstructure sensitive fatigue model for additively manufactured Ti-6Al-4V. <em>Rapid Prototyping Journal</em>, <em>22</em>(5), 817-825. https://doi.org/10.1108/rpj-11-2015-0168</li>
<li>Mantri, S., Banerjee, R. (2018). Microstructure and micro-texture evolution of additively manufactured β-Ti alloys. <em>Additive Manufacturing</em>, <em>23</em>, 86-98. https://doi.org/10.1016/j.addma.2018.07.013</li>
<li>Paul, S. K., Tarlochan, F., Hilditch, T. (2021). Fatigue life prediction of the additively manufactured specimen. <em>Modelling and Simulation in Materials Science and Engineering</em>, <em>30</em>(1), 015004. https://doi.org/10.1088/1361-651x/ac11b9</li>
<li>Yadollahi, A., Mahtabi, M., Khalili, A., Doude, H., Newman, J. (2018). Fatigue life prediction of additively manufactured material: Effects of surface roughness, defect size, and shape. <em>Fatigue & Fracture of Engineering Materials & Structures</em>, <em>41</em>(7), 1602-1614. https://doi.org/10.1111/ffe.12799</li>
<li>Odaka, K., Kamiyama, S., Takano, N., Uematsu, Y., Matsunaga, S. (2023). Fatigue life prediction considering variability for additively manufactured pure titanium clasps. <em>Journal of Prosthodontic Research</em>, <em>68</em>(2), 336-346. https://doi.org/10.2186/jpr.jpr_d_23_00074</li>
<li>Rigon, D., Meneghetti, G. (2022). Estimating the fatigue thresholds of additively manufactured metallic materials with consideration of defects. <em>Procedia Structural Integrity</em>, <em>38</em>, 70-76. https://doi.org/10.1016/j.prostr.2022.03.008</li>
<li>Schimbäck, D., Kaserer, L., Mair, P., Palm, F., Leichtfried, G., Pogatscher, S. (2023). Deformation and fatigue behaviour of additively manufactured Scalmalloy® with bimodal microstructure. <em>International Journal of Fatigue</em>, <em>172</em>, 107592. https://doi.org/10.1016/j.ijfatigue.2023.107592</li>
<li>Dodaran, M. S., Muhammad, M., Shamsaei, N., Shao, S. (2022). Synergistic effect of microstructure and defects on the initiation of fatigue cracks in additively manufactured Inconel 718. <em>International Journal of Fatigue</em>, <em>162</em>, 107002. https://doi.org/10.1016/j.ijfatigue.2022.107002</li>
<li>Assadi, H., Kreye, H., Gärtner, F., Klassen, T. (2016). Cold spraying – A materials perspective. <em>Acta Materialia</em>, <em>116</em>, 382-407. https://doi.org/10.1016/j.actamat.2016.06.034</li>
<li>Wei, Y., Li, Y., Zhu, L., Liu, Y., Lei, X., Wang, G. (2014). Evading the strength–ductility trade-off dilemma in steel through gradient hierarchical nanotwins. <em>Nature Communications</em>, <em>5</em>(1), 3580-3580. https://doi.org/10.1038/ncomms4580</li>
<li>Wang, C., Tan, X., Tor, S. B., Lim, C. (2020). Machine learning in additive manufacturing: State-of-the-art and perspectives. <em>Additive manufacturing</em>, <em>36</em>, 101538-101538. https://doi.org/10.1016/j.addma.2020.101538</li>
<li>Dao, M., Lu, L., Asaro, R., DEHOSSON, J., Ma, E. (2007). Toward a quantitative understanding of mechanical behavior of nanocrystalline metals. <em>Acta Materialia</em>, <em>55</em>(12), 4041-4065. https://doi.org/10.1016/j.actamat.2007.01.038</li>
<li>Bobbert, F., Lietaert, K., Eftekhari, A. A., Pouran, B., Ahmadi, S., Weinans, H. (2017). Additively manufactured metallic porous biomaterials based on minimal surfaces: A unique combination of topological, mechanical, and mass transport properties. <em>Acta Biomaterialia</em>, <em>53</em>, 572-584. https://doi.org/10.1016/j.actbio.2017.02.024</li>
<li>Shokrani, A., Dhokia, V., Newman, S. T. (2012). Environmentally conscious machining of difficult-to-machine materials with regard to cutting fluids. <em>International Journal of Machine Tools and Manufacture</em>, <em>57</em>, 83-101. https://doi.org/10.1016/j.ijmachtools.2012.02.002</li>
<li>Yu, W., Sing, S. L., Chua, C. K., Kuo, C., Tian, X. (2019). Particle-reinforced metal matrix nanocomposites fabricated by selective laser melting: A state of the art review. <em>Progress in Materials Science</em>, <em>104</em>, 330-379. https://doi.org/10.1016/j.pmatsci.2019.04.006</li>
<li>Ahn, D. (2021). Directed Energy Deposition (DED) Process: State of the Art. <em>International Journal of Precision Engineering and Manufacturing-Green Technology</em>, <em>8</em>(2), 703-742. https://doi.org/10.1007/s40684-020-00302-7</li>
<li>Edge, J., O’Kane, S. E. J., Prosser, R., Kirkaldy, N., Patel, A. N., Hales, A. (2021). Lithium ion battery degradation: what you need to know. <em>Physical Chemistry Chemical Physics</em>, <em>23</em>(14), 8200-8221. https://doi.org/10.1039/d1cp00359c</li>
</ol>
</article>