Complexity of selective laser melting : a new optimisation approach

The current research deals with the complexity of Selective Laser Melting (SLM) as a Rapid Manufacturing technology. The first part of the thesis contains a formal, generic optimisation methodology. The second part describes performed experiments in order to improve some important aspects of SLM. The described formal methodology consists of five tools: factor screening, dimensional analysis, transformations, experimental design and a genetic algorithm. The benefits and possibilities of these tools are described. Next the proposed methodology is presented and a mass-spring-damper system is used to demonstrate the synergetic possibilities. The complexity of SLM prohibits us to simply demonstrate the complete methodology as a whole. Collaboration between different research institutes seems the only possibility to do so. Still, many ideas used in the methodology can be used in the optimisation of simple experiments, as demonstrated in the second part. The first series of experiments is set up to facilitate the production of overhanging structures. Without precautionary measures, such structures are impossible to build with SLM and thus limit the freedom of design. By scanning single lines and single planes on a loose powder bed, different processing regions have been distinguished. The use of a genetic algorithm and of transformations has been demonstrated for a regression analysis. The obtained information is used to build overhanging structures with an improved quality. Finally, a fractal crystal has been built as a demonstration. The second series of experiments demonstrates the use of experimental design to optimise the density of parts on an own made machine with different laser systems. As such, it has been proved that the old Nd:YAG laser has insufficient properties to melt titanium-alloys completely, while a new demo fibre laser proved to perform very well.status: publishe