Physics-driven digital twin for laser powder bed fusion on GPUs

Stephanie Ferreira; Benjamin Klein; André Stork; Dieter W. Fellner

doi:10.23967/eccomas.2022.221

Physics-driven digital twin for laser powder bed fusion on GPUs

Stephanie Ferreira^*, Benjamin Klein, André Stork, Dieter W. Fellner

^*Korrespondierende/r Autor/-in für diese Arbeit

Institut für Computer Graphik und Wissensvisualisierung (7110)

Publikation: Beitrag in Buch/Bericht/Konferenzband › Beitrag in einem Konferenzband › Begutachtung

Abstract

Metal Additive Manufacturing (AM) processes such as Laser Powder Bed Fusion (LPBF) suffer from part distortion due to the localized melting and resolidification of the metal powder, which introduces stresses and strains. Despite becoming more and more important as a manufacturing process, options for simulating the printing process to predict the distortions are limited, especially because existing solutions often require very long computation times. In this work, we present the results of an implementation of the inherent strain method on graphics processing units (GPUs) that exploits the massive parallelism of the many GPU cores to speed up the simulations considerably compared to CPU-based implementations.

Originalsprache	englisch
Titel	ECCOMAS Congress 2022 - 8th European Congress on Computational Methods in Applied Sciences and Engineering
Herausgeber (Verlag)	Scipedia S.L.
Seitenumfang	9
DOIs	https://doi.org/10.23967/eccomas.2022.221
Publikationsstatus	Veröffentlicht - 1 Nov. 2022
Veranstaltung	8th European Congress on Computational Methods in Applied Sciences and Engineering: ECCOMAS 2022 - Oslo, Oslo, Norwegen Dauer: 5 Juni 2022 → 9 Juni 2022 https://www.eccomas2022.org/frontal/default.asp https://www.eccomas.org/2021/01/22/3542/

Konferenz

Konferenz	8th European Congress on Computational Methods in Applied Sciences and Engineering
Kurztitel	ECCOMAS CONGRESS 2022
Land/Gebiet	Norwegen
Ort	Oslo
Zeitraum	5/06/22 → 9/06/22
Internetadresse	https://www.eccomas2022.org/frontal/default.asp https://www.eccomas.org/2021/01/22/3542/

Zugriff auf Dokument

10.23967/eccomas.2022.221Lizenz: CC BY-NC-SA 4.0

Dieses zitieren

Ferreira, S, Klein, B, Stork, A & Fellner, DW 2022, Physics-driven digital twin for laser powder bed fusion on GPUs. in ECCOMAS Congress 2022 - 8th European Congress on Computational Methods in Applied Sciences and Engineering. Scipedia S.L., 8th European Congress on Computational Methods in Applied Sciences and Engineering, Oslo, Norwegen, 5/06/22. https://doi.org/10.23967/eccomas.2022.221

@inproceedings{9ad5e866e96a4f24926da0d9815d7d8c,

title = "Physics-driven digital twin for laser powder bed fusion on GPUs",

abstract = "The inherent strain method [1] is known and used for simulating - on a physical basis - deformation effects that occur in laser powder bed fusion (LPBF). The method is based on the idea of the inherent strain which is a black-box representation of strains caused by the process of metal printing inherentto the printer, printer settings and material used. This inherent strain can be generated from experimental data. Thus, a Digital Twin of the process combines data-driven and physics-driven aspects.In this work, we present an implementation of the inherent strain method on graphics processing units (GPUs) that exploits the massive parallelism of the many GPU cores to speed up the simulations considerably – some results show up to two orders of magnitude speed-up compared to CPU-based implementations. Compared to GPU-based simulation with unstructured finite elements on tetra-hedral meshes [2], our implementation leverages the characteristics of structured meshes in terms of indexing and compactness of the resulting stiffness matrix.To simulate the deformations, the part is discretized in a set of regular finite elements (hexahedral elements of the same shape and size). The computation of the resulting deformation is parallelized in two ways: Firstly, each layer addition is simulated with a highly parallel finite element simulation.Secondly, each of these layer additions results in an increment of the total result. The calculations of the increments are independent from each other, thus, all increments can be computed in parallel.",

keywords = "Physically based simulation, General Purpose Computation on Graphics Processing Unit (GPGPU), 3D Printing",

author = "Stephanie Ferreira and Benjamin Klein and Andr{\'e} Stork and Fellner, {Dieter W.}",

note = "Veranstaltungstitel: 8th European Congress on Computational Methods in Applied Sciences and Engineering; 8th European Congress on Computational Methods in Applied Sciences and Engineering : ECCOMAS CONGRESS 2022, ECCOMAS CONGRESS 2022 ; Conference date: 05-06-2022 Through 09-06-2022",

year = "2022",

month = nov,

day = "1",

doi = "10.23967/eccomas.2022.221",

language = "English",

booktitle = "ECCOMAS Congress 2022 - 8th European Congress on Computational Methods in Applied Sciences and Engineering",

publisher = "Scipedia S.L.",

url = "https://www.eccomas2022.org/frontal/default.asp, https://www.eccomas.org/2021/01/22/3542/",

}

TY - GEN

T1 - Physics-driven digital twin for laser powder bed fusion on GPUs

AU - Ferreira, Stephanie

AU - Klein, Benjamin

AU - Stork, André

AU - Fellner, Dieter W.

N1 - Veranstaltungstitel: 8th European Congress on Computational Methods in Applied Sciences and Engineering

PY - 2022/11/1

Y1 - 2022/11/1

N2 - The inherent strain method [1] is known and used for simulating - on a physical basis - deformation effects that occur in laser powder bed fusion (LPBF). The method is based on the idea of the inherent strain which is a black-box representation of strains caused by the process of metal printing inherentto the printer, printer settings and material used. This inherent strain can be generated from experimental data. Thus, a Digital Twin of the process combines data-driven and physics-driven aspects.In this work, we present an implementation of the inherent strain method on graphics processing units (GPUs) that exploits the massive parallelism of the many GPU cores to speed up the simulations considerably – some results show up to two orders of magnitude speed-up compared to CPU-based implementations. Compared to GPU-based simulation with unstructured finite elements on tetra-hedral meshes [2], our implementation leverages the characteristics of structured meshes in terms of indexing and compactness of the resulting stiffness matrix.To simulate the deformations, the part is discretized in a set of regular finite elements (hexahedral elements of the same shape and size). The computation of the resulting deformation is parallelized in two ways: Firstly, each layer addition is simulated with a highly parallel finite element simulation.Secondly, each of these layer additions results in an increment of the total result. The calculations of the increments are independent from each other, thus, all increments can be computed in parallel.

AB - The inherent strain method [1] is known and used for simulating - on a physical basis - deformation effects that occur in laser powder bed fusion (LPBF). The method is based on the idea of the inherent strain which is a black-box representation of strains caused by the process of metal printing inherentto the printer, printer settings and material used. This inherent strain can be generated from experimental data. Thus, a Digital Twin of the process combines data-driven and physics-driven aspects.In this work, we present an implementation of the inherent strain method on graphics processing units (GPUs) that exploits the massive parallelism of the many GPU cores to speed up the simulations considerably – some results show up to two orders of magnitude speed-up compared to CPU-based implementations. Compared to GPU-based simulation with unstructured finite elements on tetra-hedral meshes [2], our implementation leverages the characteristics of structured meshes in terms of indexing and compactness of the resulting stiffness matrix.To simulate the deformations, the part is discretized in a set of regular finite elements (hexahedral elements of the same shape and size). The computation of the resulting deformation is parallelized in two ways: Firstly, each layer addition is simulated with a highly parallel finite element simulation.Secondly, each of these layer additions results in an increment of the total result. The calculations of the increments are independent from each other, thus, all increments can be computed in parallel.

KW - Physically based simulation

KW - General Purpose Computation on Graphics Processing Unit (GPGPU)

KW - 3D Printing

U2 - 10.23967/eccomas.2022.221

DO - 10.23967/eccomas.2022.221

M3 - Conference paper

BT - ECCOMAS Congress 2022 - 8th European Congress on Computational Methods in Applied Sciences and Engineering

PB - Scipedia S.L.

T2 - 8th European Congress on Computational Methods in Applied Sciences and Engineering

Y2 - 5 June 2022 through 9 June 2022

ER -

Physics-driven digital twin for laser powder bed fusion on GPUs

Abstract

Konferenz

Zugriff auf Dokument

Fingerprint

Dieses zitieren