Abstract
Additive Manufacturing, also known as Rapid Prototyping and 3D Printing is a three-dimensional fabrication process, executed by adding materials in layers. Among many different classes of AM processes, Direct Metal Laser Sintering is a widely used metal part manufacturing method. The design, planning and implementation of overall DMLS process and its process parameters are yet to be optimized. To be able to render minimum defects as well as higher quantity of production, it is essential to apply ever developing computer technologies, data storage capabilities and optimization techniques. Typically, the defects on any 3D printed part can alter mechanical properties and shorten its durability. To minimize the defects and produce good quality parts at a mass level, has been a challenge in additive manufacturing industry. In this paper, a framework is presented to utilize game theoretic modelling approach to optimize DMLS process parameters. Online monitoring of DMLS process can identify defects of printed layers and correlate them with temperature signatures. An Artificial Neural Network is trained to predict printing defects and process parameters. predicted model can be further used in a game theoretic playoff matrix to identify the most optimal combination or configuration of DMLS process parameters to minimize defects and maximize the production quantity. The proposed method can also be applied in different domains of additive and advanced manufacturing.
Artificial neural network based modelling approach for municipal solid waste gasification in a fluidized bed reactor
