Homogenization of the Anisotropic Thermal Conductivity of Mesostructures in Material Extrusion
Topic(s) : Manufacturing
Co-authors :
Lukas HOF (GERMANY), Felix FRÖLICH (GERMANY), Florian WITTEMANN (GERMANY), Luise KÄRGER (GERMANY)Abstract :
Material extrusion (MEX) offers high flexibility and design freedom while requiring no part specific tooling. The success of the process and resulting part properties are however highly dependent on the process parameters, the material properties, limitations of the machine and environmental factors.
Print failures are often caused by high warpage, as excessively deformed parts will usually detach from the build plate or collide with the nozzle. Part strength is governed by the strength of the interface between individual bonds. Both phenomena are determined by the thermal history during the manufacturing process. Accurately predicting the thermal history is therefore highly beneficial to prevent print failures and ensure sufficient part properties.
The part consists of a porous mesostructure made up of distinct beads. The properties of a part created by the MEX process are therefore always reduced compared to a monolithic equivalent. This includes the apparent anisotropic thermal conductivity, which is reduced not only by the low conductivity of the air-filled cavities and possibly also by the interface between two beads. In macroscopic numerical analyses, it is not feasible to model every bead across a part, making homogenization of the mesostructure's properties necessary.
The apparent thermal conductivity was previously studied in experimental setups and modelled using a simple one-dimensional approach [1,2]. Due to the simplifications necessary for one-dimensional models, this approach cannot fully consider the influence of the geometry. Elkholy et al. [3] studied the thermal conductivity using the finite volume method (FVM) for simplified mesostructures without considering an additional resistance in the interface between beads. In the present work, the influence of the geometry of beads and cavities, and the bead interfaces are quantified separately. The aim is to provide an accurate and inexpensive method to homogenize the thermal conductivity of the mesostructure. The anisotropic apparent thermal conductivity of printed specimens is measured using a thermal interface materials (TIM) tester according to the ASTM standard [4]. An idealized mesostructure is used to inform a one-dimensional analytic approach for each of the three spatial directions, describing the thermal conductivity of the mesostructure that includes the additional thermal resistivity at the interface. Additionally, a two-dimensional numerical model based on the same idealized structure is developed. A two-dimensional model allows considering the influence of the geometry of the mesostructure as well as the thermal resistance of bead interfaces while maintaining relatively low computational cost. The apparent thermal conductivity is calculated from both models and compared to the experimental data. The influence of both the geometry and the bead interfaces are quantified. Finally, a method to homogenize the thermal conductivity of printed mesostructures is chosen.
Print failures are often caused by high warpage, as excessively deformed parts will usually detach from the build plate or collide with the nozzle. Part strength is governed by the strength of the interface between individual bonds. Both phenomena are determined by the thermal history during the manufacturing process. Accurately predicting the thermal history is therefore highly beneficial to prevent print failures and ensure sufficient part properties.
The part consists of a porous mesostructure made up of distinct beads. The properties of a part created by the MEX process are therefore always reduced compared to a monolithic equivalent. This includes the apparent anisotropic thermal conductivity, which is reduced not only by the low conductivity of the air-filled cavities and possibly also by the interface between two beads. In macroscopic numerical analyses, it is not feasible to model every bead across a part, making homogenization of the mesostructure's properties necessary.
The apparent thermal conductivity was previously studied in experimental setups and modelled using a simple one-dimensional approach [1,2]. Due to the simplifications necessary for one-dimensional models, this approach cannot fully consider the influence of the geometry. Elkholy et al. [3] studied the thermal conductivity using the finite volume method (FVM) for simplified mesostructures without considering an additional resistance in the interface between beads. In the present work, the influence of the geometry of beads and cavities, and the bead interfaces are quantified separately. The aim is to provide an accurate and inexpensive method to homogenize the thermal conductivity of the mesostructure. The anisotropic apparent thermal conductivity of printed specimens is measured using a thermal interface materials (TIM) tester according to the ASTM standard [4]. An idealized mesostructure is used to inform a one-dimensional analytic approach for each of the three spatial directions, describing the thermal conductivity of the mesostructure that includes the additional thermal resistivity at the interface. Additionally, a two-dimensional numerical model based on the same idealized structure is developed. A two-dimensional model allows considering the influence of the geometry of the mesostructure as well as the thermal resistance of bead interfaces while maintaining relatively low computational cost. The apparent thermal conductivity is calculated from both models and compared to the experimental data. The influence of both the geometry and the bead interfaces are quantified. Finally, a method to homogenize the thermal conductivity of printed mesostructures is chosen.