Knowledge-based Modelling of Additive Manufacturing for Sustainability Performance Analysis and Decision Making

Research output: Book/ReportDoctoral thesisCollection of Articles

Abstract

Additive manufacturing (AM) has been considered viable for complex geometries, topology optimized parts, and parts that are otherwise difficult to produce using conventional manufacturing processes. Despite the advantages, one of the prevalent challenges in AM has been the poor capability of producing functional parts at production volumes that are competitive with traditional manufacturing. Modelling and simulation are powerful tools that can help shorten the design-build-test cycle by enabling rapid analysis of various product designs and process scenarios. Nevertheless, the capabilities and limitations of traditional and advanced manufacturing technologies do define the bounds for new product development. Thus, it is important that the designers have access to methods and tools that enable them to model and simulate product performance and associated manufacturing process performance to realize functional high value products. The motivation for this dissertation research stems from ongoing development of a novel high temperature superconducting (HTS) magnet assembly, which operates in cryogenic environment. Its complexity requires the convergence of multidisciplinary expertise during design and prototyping. The research applies knowledge-based modelling to aid manufacturing process analysis and decision making in the design of mechanical components of the HTS magnet. Further, it explores the feasibility of using AM in the production of the HTS magnet assembly. The developed approach uses product-process integrated modelling based on physical experiments to generate quantitative and qualitative information that define process-structure-property-performance interactions for given material-process combinations. The resulting interactions are then integrated into a graph-based model that can aid in design space exploration to assist early design and manufacturing decision-making. To do so, test components are fabricated using two metal AM processes: wire and arc additive manufacturing and selective laser melting. Metal alloys (stainless steel, mild steel, high-strength low-alloyed steel, aluminium, and copper alloys) commonly used in structural applications are tested for their mechanical-, thermal-, and electrical properties. In addition, microstructural characterization of the alloys is performed to further understand the impact of manufacturing process parameters on material properties. The integrated modelling approach combines the collected experimental data, existing analytical and empirical relationships, and other data-driven models (e.g., finite element models, machine learning models) in the form of a decision support system that enables optimal selection of material, manufacturing technology, process parameters, and other control variables for attaining desired structure, property, and performance characteristics of the final printed component. The manufacturing decision making is performed through implementation of a probabilistic model i.e., a Bayesian network model, which is robust, modular, and can be adapted for other manufacturing systems and product designs. The ability of the model to improve throughput and quality of additive manufacturing processes will boost sustainable manufacturing goals.
Original languageEnglish
Place of PublicationTampere
PublisherTampere University
ISBN (Electronic)978-952-03-2757-6
ISBN (Print) 978-952-03-2756-9
Publication statusPublished - 2023
Publication typeG5 Doctoral dissertation (articles)

Publication series

NameTampere University Dissertations - Tampereen yliopiston väitöskirjat
Volume745
ISSN (Print)2489-9860
ISSN (Electronic)2490-0028

Fingerprint

Dive into the research topics of 'Knowledge-based Modelling of Additive Manufacturing for Sustainability Performance Analysis and Decision Making'. Together they form a unique fingerprint.

Cite this