-
A Model Hierarchy for Predicting the Flow in Stirred Tanks with Physics-Informed Neural Networks
Authors:
Veronika Trávníková,
Daniel Wolff,
Nico Dirkes,
Stefanie Elgeti,
Eric von Lieres,
Marek Behr
Abstract:
This paper explores the potential of Physics-Informed Neural Networks (PINNs) to serve as Reduced Order Models (ROMs) for simulating the flow field within stirred tank reactors (STRs). We solve the two-dimensional stationary Navier-Stokes equations within a geometrically intricate domain and explore methodologies that allow us to integrate additional physical insights into the model. These approac…
▽ More
This paper explores the potential of Physics-Informed Neural Networks (PINNs) to serve as Reduced Order Models (ROMs) for simulating the flow field within stirred tank reactors (STRs). We solve the two-dimensional stationary Navier-Stokes equations within a geometrically intricate domain and explore methodologies that allow us to integrate additional physical insights into the model. These approaches include imposing the Dirichlet boundary conditions (BCs) strongly and employing domain decomposition (DD), with both overlap** and non-overlap** subdomains. We adapt the Extended Physics-Informed Neural Network (XPINN) approach to solve different sets of equations in distinct subdomains based on the diverse flow characteristics present in each region. Our exploration results in a hierarchy of models spanning various levels of complexity, where the best models exhibit l1 prediction errors of less than 1% for both pressure and velocity. To illustrate the reproducibility of our approach, we track the errors over repeated independent training runs of the best identified model and show its reliability. Subsequently, by incorporating the stirring rate as a parametric input, we develop a fast-to-evaluate model of the flow capable of interpolating across a wide range of Reynolds numbers. Although we exclusively restrict ourselves to STRs in this work, we conclude that the steps taken to obtain the presented model hierarchy can be transferred to other applications.
△ Less
Submitted 7 March, 2024;
originally announced March 2024.
-
Neural Networks vs. Splines: Advances in Numerical Extruder Design
Authors:
Jaewook Lee,
Sebastian Hube,
Stefanie Elgeti
Abstract:
We present a novel application of neural networks to design improved mixing elements for single-screw extruders. Specifically, we propose to use neural networks in numerical shape optimization to parameterize geometries. Geometry parameterization is crucial in enabling efficient shape optimization as it allows for optimizing complex shapes using only a few design variables. Recent approaches often…
▽ More
We present a novel application of neural networks to design improved mixing elements for single-screw extruders. Specifically, we propose to use neural networks in numerical shape optimization to parameterize geometries. Geometry parameterization is crucial in enabling efficient shape optimization as it allows for optimizing complex shapes using only a few design variables. Recent approaches often utilize CAD data in conjunction with spline-based methods where the spline's control points serve as design variables. Consequently, these approaches rely on the same design variables as specified by the human designer. While this choice is convenient, it either restricts the design to small modifications of given, initial design features - effectively prohibiting topological changes - or yields undesirably many design variables. In this work, we step away from CAD and spline-based approaches and construct an artificial, feature-dense yet low-dimensional optimization space using a generative neural network. Using the neural network for the geometry parameterization extends state-of-the-art methods in that the resulting design space is not restricted to user-prescribed modifications of certain basis shapes. Instead, within the same optimization space, we can interpolate between and explore seemingly unrelated designs. To show the performance of this new approach, we integrate the developed shape parameterization into our numerical design framework for dynamic mixing elements in plastics extrusion. Finally, we challenge the novel method in a competitive setting against current free-form deformation-based approaches and demonstrate the method's performance even at this early stage.
△ Less
Submitted 9 February, 2023;
originally announced February 2023.
-
Investigation of reinforcement learning for shape optimization of profile extrusion dies
Authors:
Clemens Fricke,
Daniel Wolff,
Marco Kemmerling,
Stefanie Elgeti
Abstract:
Profile extrusion is a continuous production process for manufacturing plastic profiles from molten polymer. Especially interesting is the design of the die, through which the melt is pressed to attain the desired shape. However, due to an inhomogeneous velocity distribution at the die exit or residual stresses inside the extrudate, the final shape of the manufactured part often deviates from the…
▽ More
Profile extrusion is a continuous production process for manufacturing plastic profiles from molten polymer. Especially interesting is the design of the die, through which the melt is pressed to attain the desired shape. However, due to an inhomogeneous velocity distribution at the die exit or residual stresses inside the extrudate, the final shape of the manufactured part often deviates from the desired one. To avoid these deviations, the shape of the die can be computationally optimized, which has already been investigated in the literature using classical optimization approaches.
A new approach in the field of shape optimization is the utilization of Reinforcement Learning (RL) as a learning-based optimization algorithm. RL is based on trial-and-error interactions of an agent with an environment. For each action, the agent is rewarded and informed about the subsequent state of the environment. While not necessarily superior to classical, e.g., gradient-based or evolutionary, optimization algorithms for one single problem, RL techniques are expected to perform especially well when similar optimization tasks are repeated since the agent learns a more general strategy for generating optimal shapes instead of concentrating on just one single problem.
In this work, we investigate this approach by applying it to two 2D test cases. The flow-channel geometry can be modified by the RL agent using so-called Free-Form Deformation, a method where the computational mesh is embedded into a transformation spline, which is then manipulated based on the control-point positions. In particular, we investigate the impact of utilizing different agents on the training progress and the potential of wall time saving by utilizing multiple environments during training.
△ Less
Submitted 23 December, 2022;
originally announced December 2022.
-
Seamless Integration of Analysis and Design: Automatic CAD Reconstruction of Post-Analysis Geometries
Authors:
Sebastian Hube,
Roxana Pohlmann,
Stefanie Elgeti
Abstract:
A key step during industrial design is the passing of design information from computer aided design (CAD) to analysis tools (CAE) and vice versa. Here, one is faced with a severe incompatibility in geometry representation: While CAD is usually based on surface representations, analysis mostly relies on volumetric representations. The forward pass, i.e., converting CAD data to computational meshes,…
▽ More
A key step during industrial design is the passing of design information from computer aided design (CAD) to analysis tools (CAE) and vice versa. Here, one is faced with a severe incompatibility in geometry representation: While CAD is usually based on surface representations, analysis mostly relies on volumetric representations. The forward pass, i.e., converting CAD data to computational meshes, is well understood and established. However, the same does not hold for the inverse direction, i.e., CAD reconstruction of deformed geometries resulting, e.g., from shape optimization. The few reconstruction methods reported mainly rely on spline fitting, in particular methods that rely on creating new splines simililar to shape reconstruction from 3D imaging. In contrast, this paper studies a novel approach that reuses the CAD data given in the initial design. We show that this concept leads to precise shape reconstructions while also preserving the initial notion of features defined during design. Furthermore, reusing the initial CAD representation reduces the shape reconstruction problem to a shape modification problem. We study this unique feature and show that it enables the reconstruction of CAD data from computational meshes by composing each spline in the initial CAD data with a single, global deformation spline. While post-processing is needed for use in current CAD software, this novel approach not only allows creating watertight models, but also enables reconstructing complete CAD models even from defeatured computational meshes.
△ Less
Submitted 17 July, 2023; v1 submitted 9 May, 2022;
originally announced May 2022.
-
Computing the jump-term in space-time FEM for arbitrary temporal interpolation
Authors:
Eugen Salzmann,
Florian Zwicke,
Stefanie Elgeti
Abstract:
One approach with rising popularity in analyzing time-dependent problems in science and engineering is the so-called space-time finite-element method that utilized finiteelements in both space and time. A common ansatz in this context is to divide the mesh in temporal direction into so-called space-time slabs, which are subsequently weakly connected in time with a Discontinuous Galerkin approach.…
▽ More
One approach with rising popularity in analyzing time-dependent problems in science and engineering is the so-called space-time finite-element method that utilized finiteelements in both space and time. A common ansatz in this context is to divide the mesh in temporal direction into so-called space-time slabs, which are subsequently weakly connected in time with a Discontinuous Galerkin approach. The corresponding jumpterm, which is responsible for imposing the weak continuity across space-time slabs can be challenging to compute, in particular in the context of deforming domains. Ensuring a conforming discretization of the space-time slab at the top and bottom in time direction simplifies the handling of this term immensely. Otherwise, a computationally expensive and error prone projection of the solution from one time-level to another is necessary. However, when it comes to simulations with deformable domains, e.g. for free-surface flows, ensuring conforming meshes is quite laborious. A possible solution to this challenge is to extrude a spatial mesh in time at each time-step resulting in the so-called time-discontinuous prismatic space-time (D-PST) method. However, this procedure is restricted to finite-elements of 1st order in time. We present a novel algorithmic approach for arbitrarily discretized meshes by flip** the mesh in time-direction for each time-step. This ansatz allows for a simple evaluation of the jump-term as the mesh is always conforming. The cost of flip** the mesh around its symmetry plane in time scales with the number of nodes, which makes it computationally cheaper than an additional update of the mesh to enforce conformity or the evaluation of a projection. We validate the approach on various physical problems with and without deforming domains.
△ Less
Submitted 3 May, 2022;
originally announced May 2022.
-
A level-set based space-time finite element approach to the modelling of solidification and melting processes
Authors:
Leonardo Boledi,
Benjamin Terschanski,
Stefanie Elgeti,
Julia Kowalski
Abstract:
We present a strategy for the numerical solution of convection-coupled phase-transition problems, with focus on solidification and melting. We solve for the temperature and flow fields over time. The position of the phase-change interface is tracked with a level-set method, which requires knowledge of the heat-flux discontinuity at the interface. In order to compute the heat-flux jump, we build up…
▽ More
We present a strategy for the numerical solution of convection-coupled phase-transition problems, with focus on solidification and melting. We solve for the temperature and flow fields over time. The position of the phase-change interface is tracked with a level-set method, which requires knowledge of the heat-flux discontinuity at the interface. In order to compute the heat-flux jump, we build upon the ghost-cell approach and extend it to the space-time finite element method. This technique does not require a local enrichment of the basis functions, such as methods like extended finite elements, and it can be easily implemented in already existing finite element codes. Verification cases for the 1D Stefan problem and the lid-driven cavity melting problem are provided. Furthermore, we show a more elaborate 2D case in view of complex applications.
△ Less
Submitted 7 May, 2021;
originally announced May 2021.
-
A novel approach to fluid-structure interaction simulations involving large translation and contact
Authors:
Daniel Hilger,
Norbert Hosters,
Fabian Key,
Stefanie Elgeti,
Marek Behr
Abstract:
In this work, we present a novel method for the mesh update in flow problems with moving boundaries, the phantom domain deformation mesh update method (PD-DMUM). The PD-DMUM is designed to avoid remeshing; even in the event of large, unidirectional displacements of boundaries. The method combines the concept of two mesh adaptation approaches: (1) The virtual ring shear-slip mesh updatemethod (VR-S…
▽ More
In this work, we present a novel method for the mesh update in flow problems with moving boundaries, the phantom domain deformation mesh update method (PD-DMUM). The PD-DMUM is designed to avoid remeshing; even in the event of large, unidirectional displacements of boundaries. The method combines the concept of two mesh adaptation approaches: (1) The virtual ring shear-slip mesh updatemethod (VR-SSMUM); and (2) the elastic mesh update method (EMUM). As in the VR-SSMUM, the PD-DMUMextends the fluid domain by a phantom domain; the PD-DMUM can thus locally adapt the element density. Combined with the EMUM, the PD-DMUMallows the consideration of arbitrary boundary movements. In this work, we apply the PD-DMUM in two test cases. Within the first test case, we validate the PD-DMUM in a 2D Poiseuille flow on a moving background mesh. Subsequently the fluid-structure interaction (FSI) problem in the second test case serves as a proof of concept. More, we stress the advantages of the novel method with regard to conventional mesh update approaches.
△ Less
Submitted 13 January, 2021;
originally announced January 2021.
-
SplineLib: A Modern Multi-Purpose C++ Spline Library
Authors:
Markus Frings,
Norbert Hosters,
Corinna Müller,
Max Spahn,
Christoph Susen,
Konstantin Key,
Stefanie Elgeti
Abstract:
This paper provides the description of a novel, multi-purpose spline library. In accordance with the increasingly diverse modes of usage of splines, it is multi-purpose in the sense that it supports geometry representation, finite element analysis, and optimization. The library features reading and writing for various file formats and a wide range of spline manipulation algorithms. Further, a new…
▽ More
This paper provides the description of a novel, multi-purpose spline library. In accordance with the increasingly diverse modes of usage of splines, it is multi-purpose in the sense that it supports geometry representation, finite element analysis, and optimization. The library features reading and writing for various file formats and a wide range of spline manipulation algorithms. Further, a new efficient and objective-oriented algorithm for B-spline basis function evaluation is included. All features are available by a spline-type independent interface. The library is written in modern C++ with CMake as build system. This enables it for usage in typical scientific applications. It is provided as open-source library.
△ Less
Submitted 27 February, 2020;
originally announced February 2020.
-
Optimizing micro-tiles in micro-structures as a design paradigm
Authors:
Pablo Antolin,
Annalisa Buffa,
Elaine Cohen,
John F. Dannenhoffer,
Gershon Elber,
Stefanie Elgeti,
Robert Haimes,
Richard Riesenfeld
Abstract:
In recent years, new methods have been developed to synthesize complex porous and micro-structured geometry in a variety of ways. In this work, we take these approaches one step further and present these methods as an efficacious design paradigm. Specifically, complex micro-structure geometry can be synthesized while optimizing certain properties such as maximal heat exchange in heat exchangers, o…
▽ More
In recent years, new methods have been developed to synthesize complex porous and micro-structured geometry in a variety of ways. In this work, we take these approaches one step further and present these methods as an efficacious design paradigm. Specifically, complex micro-structure geometry can be synthesized while optimizing certain properties such as maximal heat exchange in heat exchangers, or minimal weight under stress specifications. By being able to adjust the geometry, the topology and/or the material properties of individual tiles in the micro-structure, possibly in a gradual way, a porous object can be synthesized that is optimal with respect to the design specifications. As part of this work, we exemplify this paradigm on a variety of diverse applications.
△ Less
Submitted 4 July, 2019;
originally announced July 2019.
-
Simplex space-time meshes in two-phase flow simulations
Authors:
Violeta Karyofylli,
Markus Frings,
Stefanie Elgeti,
Marek Behr
Abstract:
In this paper, we present the numerical solution of two-phase flow problems of engineering significance with a space-time finite element method that allows for local temporal refinement. Our basis is the method presented in [3], which allows for arbitrary temporal refinement in preselected regions of the mesh. It has been extended to adaptive temporal refinement that is governed by a quantity that…
▽ More
In this paper, we present the numerical solution of two-phase flow problems of engineering significance with a space-time finite element method that allows for local temporal refinement. Our basis is the method presented in [3], which allows for arbitrary temporal refinement in preselected regions of the mesh. It has been extended to adaptive temporal refinement that is governed by a quantity that is part of the solution process, namely, the interface position in two-phase flow. Due to local effects such as surface tension, jumps in material properties, etc., the interface can, in general, be considered a region that requires high flexibility and high resolution, both in space and in time. The new method, which leads to tetrahedral (for 2D problems) and pentatope (for 3D problems) meshes, offers an efficient yet accurate approach to the underlying two-phase flow problems.
△ Less
Submitted 21 March, 2019;
originally announced March 2019.
-
Boundary-Conforming Finite Element Methods for Twin-Screw Extruders: Unsteady - Temperature-Dependent - Non-Newtonian Simulations
Authors:
Jan Helmig,
Marek Behr,
Stefanie Elgeti
Abstract:
We present a boundary-conforming space-time finite element method to compute the flow inside co-rotating, self-wi** twin-screw extruders. The mesh update is carried out using the newly developed Snap** Reference Mesh Update Method (SRMUM). It allows to compute time-dependent flow solutions inside twin-screw extruders equipped with conveying screw elements without any need for re-meshing and pr…
▽ More
We present a boundary-conforming space-time finite element method to compute the flow inside co-rotating, self-wi** twin-screw extruders. The mesh update is carried out using the newly developed Snap** Reference Mesh Update Method (SRMUM). It allows to compute time-dependent flow solutions inside twin-screw extruders equipped with conveying screw elements without any need for re-meshing and projections of solutions - making it a very efficient method. We provide cases for Newtonian and non-Newtonian fluids in 2D and 3D, that show mesh convergence of the solution as well as agreement to experimental results. Furthermore, a complex, unsteady and temperature-dependent 3D test case with multiple screw elements illustrates the potential of the method also for industrial applications.
△ Less
Submitted 4 December, 2018;
originally announced January 2019.
-
The Virtual Ring Shear-Slip Mesh Update Method
Authors:
Fabian Key,
Lutz Pauli,
Stefanie Elgeti
Abstract:
A novel method - the Virtual Ring Shear-Slip Mesh Update Method (VR-SSMUM) - for the efficient and accurate modeling of moving boundary or interface problems in the context of the numerical analysis of fluid flow is presented. We focus on cases with periodic straight-line translation including object entry and exit. The periodic character of the motion is reflected in the method via a map** of t…
▽ More
A novel method - the Virtual Ring Shear-Slip Mesh Update Method (VR-SSMUM) - for the efficient and accurate modeling of moving boundary or interface problems in the context of the numerical analysis of fluid flow is presented. We focus on cases with periodic straight-line translation including object entry and exit. The periodic character of the motion is reflected in the method via a map** of the physical domain onto a closed virtual ring. Therefore, we use an extended mesh, where unneeded portions are deactivated to control the computational overhead. We provide a validation case as well as examples for the applicability of the method to 2D and 3D models of packaging machines.
△ Less
Submitted 10 April, 2018;
originally announced April 2018.
-
Automatic implementation of material laws: Jacobian calculation in a finite element code with TAPENADE
Authors:
Florian Zwicke,
Philipp Knechtges,
Marek Behr,
Stefanie Elgeti
Abstract:
In an effort to increase the versatility of finite element codes, we explore the possibility of automatically creating the Jacobian matrix necessary for the gradient-based solution of nonlinear systems of equations. Particularly, we aim to assess the feasibility of employing the automatic differentiation tool TAPENADE for this purpose on a large Fortran codebase that is the result of many years of…
▽ More
In an effort to increase the versatility of finite element codes, we explore the possibility of automatically creating the Jacobian matrix necessary for the gradient-based solution of nonlinear systems of equations. Particularly, we aim to assess the feasibility of employing the automatic differentiation tool TAPENADE for this purpose on a large Fortran codebase that is the result of many years of continuous development. As a starting point we will describe the special structure of finite element codes and the implications that this code design carries for an efficient calculation of the Jacobian matrix. We will also propose a first approach towards improving the efficiency of such a method. Finally, we will present a functioning method for the automatic implementation of the Jacobian calculation in a finite element software, but will also point out important shortcomings that will have to be addressed in the future.
△ Less
Submitted 21 February, 2017;
originally announced February 2017.
