The presence of defects like gas bubble in fabricated parts is inherent in the selective laser sintering process and the prediction of bubble shrinkage dynamics is crucial. In this paper, two artificial intelligence (AI) models based on Decision Trees algorithm were constructed in order to predict bubble dissolution time, namely the Ensemble Bagged Trees (EDT Bagged) and Ensemble Boosted Trees (EDT Boosted). A metadata including 68644 data were generated with the help of our previously developed numerical tool. The AI models used the initial bubble size, external domain size, diffusion coefficient, surface tension, viscosity, initial concentration, and chamber pressure as input parameters, whereas bubble dissolution time was considered as output variable. Evaluation of the models’ performance was achieved by criteria such as Mean Absolute Error (MAE), Root Mean Squared Error (RMSE) and coefficient of determination (R2). The results showed that EDT Bagged outperformed EDT Boosted. Sensitivity analysis was then conducted thanks to the Monte Carlo approach and it was found that three most important inputs for the problem were the diffusion coefficient, initial concentration, and bubble initial size. This study might help in quick prediction of bubble dissolution time to improve the production quality from industry.
22 May 2012 in Archives of Computational Methods in Engineering
This paper characterizes and analyzes the microstructures of injection-molded polypropylene parts reinforced with 20 wt% of hemp fibers in order to understand the process induced variations in thermomechanical properties. In-thickness fiber orientation and fiber content were determined by X-ray tomography along the flow. The fiber content along the flow path was also determined by direct fiber content measurements after matrix dissolution, showing an increase of 2%/100 mm for a 2.2 mm-thick plate due to fiber migration during the filling stage. A typical shell/core structure for fiber orientation in injection molding was observed, but with a very clear transition between the layer solidified under high shear rates and the core in which the fiber content was reduced by more than 50%. The orientation of hemp fibers is lower than the one of glass fibers, especially in thickness direction. However, the overall fiber orientation in the injection direction induces significant anisotropic thermomechanical properties, which cannot be explained by simple micromechanical models that consider isotropic mechanical properties for hemp fibers. These phenomena must be taken into account in process simulation codes for injection molding to better predict thermomechanical properties as well as part shrinkage and warpage to design molds.
Commonly isolated carbon nanotubes in suspension have been modelled as a perfectly straight structure. Nevertheless, single-wall carbon nanotubes (SWNTs) contain naturally side-wall defects and, in consequence, natural bent configurations. Hence, a semi-flexile filament model with a natural bent configuration was proposed to represent physically the SWNT structure. This continuous model was discretized as a non-freely jointed multi-bead–rod system with a natural bent configuration. Using a Brownian dynamics algorithm the dynamical mechanical contribution to the linear viscoelastic response of naturally bent SWNTs in dilute suspension was simulated. The dynamics of such system shows the apparition of new relaxation processes at intermediate frequencies characterized mainly by the activation of a mild elasticity. Storage modulus evolution at those intermediate frequencies strongly depends on the flexibility of the system, given by the rigidity constant of the bending potential and the number of constitutive rods.
1 November 2020 in Polymer Degradation and Stability