Open Access Open Access  Restricted Access Subscription or Fee Access

Optimization and Polynomial Chaos-Based Uncertainty Analysis of Additively Manufactured Polymer Composites



Additive manufacturing (AM) of composite structures includes various sources of uncertainties at multiple levels from design to manufacturing stage. This affects adversely by the formation of void contents, weak bead bonding in AM parts, and reducing structural properties, which often hinders the widespread application of AM components. Fused deposition modeling (FDM) is the most widely used AM process because of its simplicity and affordability. In this paper, a multiscale experimental and uncertainty quantification framework have been developed considering various FDM process parameters to achieve optimum structural properties. Filament level extrusion process was first studied by experimental investigation of different fiber distribution (0-10 wt. %) and then by genetic algorithm based optimizer. At the manufacturing level, an experimental scheme was designed for bead, lamina and laminate level properties. The adjusted process parameters for 5 wt. % carbon fiber reinforced polylactic acid (PLA) composite shows minimum variability and maximum structural property at a print temperature of 220℃, platform temperature of 80℃, and print speed of 20 /. The dimensional accuracy and void contents support this results. Physics based surrogate models were also developed at each manufacturing stages to improve the computational efficiency and accurately model the uncertainties. A nonintrusive polynomial chaos based uncertainty analysis was carried out on the surrogate models by considering voids, material variations and imperfect bonding between the beads. The uncertainty model further revealed the distribution of different categories of voids (intra-bead, inter-bead and interfacial-bead), which paved the way for reduced design phase prior to manufacturing. The different types of voids relating to independent uncertainty sources can be encountered by addressing improved fibermatrix bonding for intra-bead voids, modified nozzle geometry for inter-bead voids and optimized printing speed for the interfacial-bead void.


Full Text:



  • There are currently no refbacks.