Directed energy deposition (DED) is a versatile process for coating and repairing high-value components in aerospace, energy and tool manufacturing. However, achieving consistent coating quality remains challenging due to the complex interactions between process parameters, microstructural evolution and resulting properties. Manual measurement of metallurgical features is time-consuming and prone to variability, while the high cost of DED experiments often results in small datasets, limiting systematic, data-driven analysis. To address these challenges, a study was conducted to analyze and model the relationships between DED process parameters and coating outcomes using data-based methods. DED coating experiments were performed with varying process parameters, microhardness was measured using instrumented indentation tests, and microscopic imaging was used to acquire additional metallurgical features. In addition to correlation analysis and feature engineering, resource-efficient modeling was achieved by developing a late fusion framework combining tabular process features with microscopic images. As DED experiments typically yield small datasets, the modeling architecture was designed to avoid overfitting while enabling robust learning from heterogeneous data, leveraging staged training and dimensionality reduction. This approach enabled the prediction of coating properties such as hardness, demonstrating the feasibility of data-driven methods for DED coating quality prediction.<\/p>\n","protected":false},"excerpt":{"rendered":"
Directed energy deposition (DED) is a versatile process for coating and repairing high-value components in aerospace, energy and tool manufacturing. However, achieving consistent coating quality remains challenging due to the complex interactions between process parameters, microstructural evolution and resulting properties. Manual measurement of metallurgical features is time-consuming and prone to variability, while the high cost of DED experiments often results in small datasets, limiting systematic, data-driven analysis. To address these […]<\/p>\n","protected":false},"author":12,"featured_media":0,"template":"","meta":{"_acf_changed":false,"footnotes":""},"publication-type":[30],"class_list":["post-35165","publication","type-publication","status-publish","hentry","publication-type-article"],"acf":[],"publishpress_future_workflow_manual_trigger":{"enabledWorkflows":[]},"_links":{"self":[{"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/publication\/35165","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/publication"}],"about":[{"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/types\/publication"}],"author":[{"embeddable":true,"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/users\/12"}],"version-history":[{"count":0,"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/publication\/35165\/revisions"}],"wp:attachment":[{"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/media?parent=35165"}],"wp:term":[{"taxonomy":"publication-type","embeddable":true,"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/publication-type?post=35165"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}