Thermal response analysis and parameter prediction of additively manufactured polymers. (25th July 2022)

• Record Type: Journal Article
• Authors: Moslemi, Navid; Abdi, Behzad.; Gohari, Soheil; Sudin, Izman; Atashpaz-Gargari, E.; Redzuan, Norizah; Ayob, Amran; Burvill, Colin; Su, Meini; Arya, Farid
• Is Part Of: Applied thermal engineering. Volume 212(2022)
• Publication Date: 2022-07-25
• Subjects: Finite element analysis -- Artificial Neural Network -- Polymers -- Additive manufacturing -- 3D printing; Heat engineering -- Periodicals; Heating -- Equipment and supplies -- Periodicals; Periodicals; 621.40205
• DOI: 10.1016/j.applthermaleng.2022.118533
• Languages: English
• ISSNs: 1359-4311

Abstract

Highlights: Thermal response of extrusion process in 3D printing obtained via numerical models. Phase change and steady-state energy balance in the standoff region are modelled. The finite element results are validated experimentally and analytically. The effects of processing parameters are studied using validated numerical models. An artificial neural network trained by numerical models for further prediction. Abstract: Fused Deposition Modelling (FDM), is an additive manufacturing technology where polymers are extruded using appropriate processing parameters to achieve suitable bonding while ensuring that overheating does not occur. Among processing parameters, polymer inlet temperature, nozzle size, extrusion speed, and air cooling speed are significantly effect on the extrusion process at the distance between the build plate and the nozzle tip (standoff region). This study aims to evaluate the influences of the processing parameters on the thermal behavior and phase change zone of Polyamide 12 (PA12) and Acrylonitrile Butadiene Styrene (ABS) polymers at standoff region. A nonlinear three-dimensional (3D) finite element (FE) model was developed by implementing an apparent heat capacity model using the Heat Transfer Module in COMSOL® Multiphysics software. FE results in the standoff region were validated by experimental tests, concerning various nozzle sizes and extrusion speed. The validated numerical results demonstrated that there is a complex correlation between processing parameters and thermal behaviors such as phase change and temperature distribution in the standoff region. The FE results were then employed in training an artificial neural network (ANN). A well-established compromise between the trained ANN and the FE results demonstrates that the trained ANN can be employed in the prediction of further thermal and glass transition behavior using subsequent processing parameters.