High performance computing for modelling of stereolithography process

verfasst von: Sandeep Kumar
betreut von: Peter Wriggers
Abstract: In this dissertation, a state-of-the-art 3D computational model has been developed for Stereolithography process to investigate the evolution of properties in a multi-physics framework using Stabilized Optimal Transportation Meshfree (OTM) method based on a continuum approach. In order to accelerate the computational performance, HPC framework of the OTM method has been developed. Stereolithography process is a complex process in the sense that several physical processes are involved therein. In this work, some of the key phenomena incorporated in the modeling framework are highly coupled thermo-chemo-mechanical evolution of resin properties and propagation of the UV laser through the resin. The photopolymerization is driven by the interaction of fluid resin with the UV light and consequently generates heat due to its exothermic nature and resulting in building up of mechanical stresses. The numerical and geometrical complexities arising from these phenomena pose serious challenges and complications in grid-based techniques such as Finite element (FE). Generally, such issues are referred to as mesh distortion. OTM based computational modeling is one solution to these issues. The method is quite new in the field of Stereolithography simulation and it is efficient in capturing the deformations generated during printing process. Moreover, parallelization using MPI with an objective for scalability on large scale CPU clusters reduces the computational efforts. And, the obtained results leads to highly scalable results. The developed tool can be employed to optimize the material and process parameters during the printing process to achieve improved accuracy in the printed parts.
Organisationseinheit(en): Institut für Kontinuumsmechanik
Typ: Dissertation
Anzahl der Seiten: 98
Publikationsdatum: 2022
Publikationsstatus: Veröffentlicht
Elektronische Version(en): https://doi.org/10.15488/12538 (Zugang: Offen)
: Details im Forschungsportal „Research@Leibniz University“

BibTeX

@phdthesis{709246eb8bec4cbc8d5b1f49edc2ba8e,
title = "High performance computing for modelling of stereolithography process",
abstract = "In this dissertation, a state-of-the-art 3D computational model has been developed for Stereolithography process to investigate the evolution of properties in a multi-physics framework using Stabilized Optimal Transportation Meshfree (OTM) method based on a continuum approach. In order to accelerate the computational performance, HPC framework of the OTM method has been developed. Stereolithography process is a complex process in the sense that several physical processes are involved therein. In this work, some of the key phenomena incorporated in the modeling framework are highly coupled thermo-chemo-mechanical evolution of resin properties and propagation of the UV laser through the resin. The photopolymerization is driven by the interaction of fluid resin with the UV light and consequently generates heat due to its exothermic nature and resulting in building up of mechanical stresses. The numerical and geometrical complexities arising from these phenomena pose serious challenges and complications in grid-based techniques such as Finite element (FE). Generally, such issues are referred to as mesh distortion. OTM based computational modeling is one solution to these issues. The method is quite new in the field of Stereolithography simulation and it is efficient in capturing the deformations generated during printing process. Moreover, parallelization using MPI with an objective for scalability on large scale CPU clusters reduces the computational efforts. And, the obtained results leads to highly scalable results. The developed tool can be employed to optimize the material and process parameters during the printing process to achieve improved accuracy in the printed parts.",
author = "Sandeep Kumar",
note = "Doctoral thesis",
year = "2022",
doi = "10.15488/12538",
language = "English",
publisher = "Leibniz Universit{\"a}t Hannover",
school = "Leibniz University Hannover",
}