J9
Meshing

Back to overview

17:35
conference time (CEST, Berlin)
Reliable Meshing of 10 000 Parts
26/10/2021 17:35 conference time (CEST, Berlin)
Room: J
M. Lautsch (Lautsch Finite Elemente GmbH, DEU)
M. Lautsch (Lautsch Finite Elemente GmbH, DEU)
We present a tetra mesher for multipart meshing purposes. The method is fully automated and rejects no parts if some simple rules are obeyed. The result is a ready to compute, conformal mesh for all parts and for the fluid outside. This method does not rely on Delaunay ideas but is based on wrapping-, segmentation- and marching volume techniques. At first we give a precise definition of what a part is. Starting with a special single tetra which covers the entire input geometry an adaptive refinement process creates a mesh which is fine near this geometry. This refinement process iterates the subdivision of the first tetra to eight tetras of the same shape and therefore of the same quality. This is just like in the cubic case. This division pattern is not applicable to arbitrary tetras, but for our starting one it works. The Marching Tetra method gives tetras which are assigned to parts. Since our mesh is conformal, imprints, i.e. triangles between tetras of different part-membership, can be derived easily. These imprints allow straight forward load definitions. We discuss the criteria which are necessary for mesh refinement and show how to achieve geometry correctness and good element quality simultaneously, by using edge- and face- collapse algorithms and by edge swapping methods. A special chapter is devoted to the generation of feature lines, i.e. sharp edges and lines where three parts meet. To treat feature points is an extension of this work. The Marching Tetra Multigrid method works for any number of parts. The input geometry must allow an answer to the question for any point, whether this point is inside or outside. If this point is inside for more than one part, we need a precise answer which part wins. This is the way Boolean subtract is performed automatically for overlapping parts. Some case studies of multipart problems are presented in the final section.
multipart meshing, marching Volume, wrapping, multigid, mesh improvement
17:55
conference time (CEST, Berlin)
Mesh Adaptation Based on Taylor Micro-scale for Aeroacoustics Simulation
26/10/2021 17:55 conference time (CEST, Berlin)
Room: J
L. Delmas (MSC Software France - groupe Hexagon, FRA); A. Poulos, C. Legendre (Free Field Technologies, BEL)
L. Delmas (MSC Software France - groupe Hexagon, FRA); A. Poulos, C. Legendre (Free Field Technologies, BEL)
Due to the large size and physical complexity of modern industrial aeroacoustics problems, the use of uniform meshes fine enough to achieve the desired accuracy is unfeasible, even with the computation power available nowadays. To reduce such a computational cost to affordable levels, engineers must be creative and locally refine the meshes using physical considerations, sometimes unknown a priori. In practical terms, since numerical predictions for aeroacoustics heavily rely on accurate source estimations from the flow information, this translate in back-and-forth iterations of: (i) Computational Fluid Dynamics (CFD) solutions used to estimate noise sources; then (ii) to be propagated with acoustic solvers to predict noise levels in the far field. If the noise level computed is not accurate enough, the mesh is refined, and this process is repeated until convergence is reached. Added to this complexity, for resolving the flow, a Large Eddy Simulation (LES) is frequently used requiring further mesh refinement applied across large parts of the computational domain. This intensifies even more the cost associated with the overall process because in such complex geometry an adequate mesh must be determined to perform a valid and numerically stable LES. This iterative process is not only tedious and computational expensive but also it consumes engineering time. A solution to this problem is adaptive mesh refinement techniques based on physical and relevant indicators, such techniques can save both computational and engineering time whilst it guarantees accurate numerical solutions. For this reason, a novel methodology for adaptive mesh refinement is proposed based on the Taylor micro-scale, where most of the noise-generating eddies are properly resolved. Although such a micro-scale reaches local minima, i.e. especially near the walls, some techniques are proposed to avoid excessive mesh refinement via the use of limiters. The methodology is evaluated in the context of a hybrid aeroacoustic simulation, where the flow solution is obtained using a CFD solution to then compute equivalent sources to be imposed in a Finite Element (FE) acoustic propagation solver. The results obtained are further compared with results from a manually refined mesh, both in terms of cost as well as accuracy.
aeroacoustics, cfd, LES, finite elements
18:15
conference time (CEST, Berlin)
Autonomous Hexahedral Meshing Using Artificial Intelligence
26/10/2021 18:15 conference time (CEST, Berlin)
Room: J
A. Patel (Illinois Rocstar LLC, USA); S. Pemberton (Illinois Rocstar LLC, USA); M. Safdari (Ansys USA); W. Quadros (Sandia National Laboratories, USA)
A. Patel (Illinois Rocstar LLC, USA); S. Pemberton (Illinois Rocstar LLC, USA); M. Safdari (Ansys USA); W. Quadros (Sandia National Laboratories, USA)
Autonomous hexahedral mesh generation is considered a holy grail in the meshing community. Hexahedral meshes are known to reduce the number of elements needed for comparable accuracy to tetrahedra, leading to savings in time. Hexahedral meshes are especially desired for finite element analyses of highly elastic and plastic structural domains. Unlike tetrahedral meshing, full-hex or hex-dominant meshing is mathematically global in scope. Existing automated hex meshing methodologies can handle limited classes of geometries and often yield sub-optimal mesh quality. To date, successful hex meshing is limited to semi-autonomous methods and requires significant human intervention due to the complexity of the process. They lack the required reliability and robustness. Currently, the reliable hex(-dominant) meshing of geometry requires manual decomposition of geometry into hex meshable regions. This process is slow and requires significant meshing expertise. To address these issues, Illinois Rocstar LLC (IR) in collaboration with Sandia National Laboratories (SNL) is developing the Auto-Hex plugin which will be a part of SNL CUBIT meshing software. Auto-Hex uses state-of-the-art geometric reasoning and artificial intelligence for automatic geometric decomposition and robust hexahedral meshing of complex geometries. Using geometric reasoning capabilities, boundary representation (B-Rep) data along with extracted skeletal object from given geometry is analyzed to identify the most prominent feature locations in the geometry. Skeletal objects reduce the volumetric representation of geometry to surface equivalent, making it easier to analyze complex 3D geometries. Skeletal meshes help identify T-junctions, sweep directions, and through holes in the geometry. A robust reinforcement learning network utilizes the extracted geometric and skeletal features to identify the best web-cuts in the correct order that result in hex or hex-dominant mesh in the given geometry. The reinforcement learning network learns using real-time feedback from the meshing environment (CUBIT) and strengthens the geometric decomposition ability over time. The need for labeled training data is eliminated since the state-of-the-art reinforcement learning network directly interacts with the CUBIT meshing environment and learns the effective geometric decomposition policy using feedback received from CUBIT. We believe that the proposed technology will be a pioneer in the field of truly automated hex meshing and invaluable to the broader modeling and simulation community. Auto-Hex plugin aims to utilize artificial intelligence to teach it behavior of meshing experts when decomposing a geometry into hex meshable components. Such a trained model will be used to predict decompositions on complex geometries while ensuring the quality of the mesh which is otherwise, out of the reach for non-meshing experts in the simulation community.
Hexahedral Meshing, Finite Element Analysis, Computational Fluid Dynamics, Machine Learning, Autonomous Meshing, Geometry Decomposition, Skeletal
18:35
conference time (CEST, Berlin)
Creating a 20-Node Hexahedral Element Model : An Innovative Solution to an Old Problem
26/10/2021 18:35 conference time (CEST, Berlin)
Room: J
J. Pura, J.Leedom (MSC Software, USA); J. Kofoid (Northrop Grumman, USA)
J. Pura, J.Leedom (MSC Software, USA); J. Kofoid (Northrop Grumman, USA)
One of the many critical analysis tasks in aircraft design is the evaluation of structural stability, and since the invention of finite element analysis (FEA), engineers around the world have been attempting to perform analysis with as few elements (and thus the lowest amount of computational cost) as possible. One of those “sweet spots” between mathematically accurate results and lowest number of elements is the 20-node Hexahedral element (also known as Hex mesh). Unfortunately, from a historical perspective, the process required to develop geometry suitable for a hexahedral mesh has been complicated and time-intensive – almost to the point of negating the time saved by faster solutions due to the hexahedral mesh. With the availability of MSC Apex’s hex meshing toolsets, engineers now have a quick and innovative solution to the creation of high-quality mesh for structural analysis. Ever since hex meshing was invented, comparisons have been made in many industries as to the accuracy of results between hexahedral elements and tetrahedral elements [1][2][3]. Most analysts prefer hex elements because the results are generally more accurate, but sometimes avoid using them because of the extra cost (measured in hours of engineering work) needed to create them. Therefore, engineers would like to use tetrahedral mesh for analysis due to project schedule. In certain circumstances, though, a high-quality hexahedral mesh is necessary and extra time must be taken to develop this type of mesh in favor of result accuracy. Users from Northrop Grumman Mission Systems evaluated MSC Apex’s hex meshing functionality by conducting a benchmark workflow to compare time spent in using Patran and MSC Apex, finding approximate 9 hours productivity gains with the new workflow. In this white paper, the details of the benchmark workflow, MSC Apex’s hex mesh toolset, and time saving of each step are introduced. Prior to MSC Apex’s advancements in the hex meshing workflow, users used Patran as a pre-processor and MSC Nastran as a finite element solver in order to create the hexahedral mesh for a variety of its electronics products. This legacy workflow, measured on an evaluation-worthy geometry part, previously took 16 hours to complete the pre-processing in Patran and solve using MSC Nastran. Using MSC Apex on this same geometry file, this workflow was reduced to 7 hours (55% time saved), yielded more uniform results, and allowed the engineer additional time spent on inspecting and refining the mesh quality, which was not a luxury allowed using the traditional workflow with Patran. Additionally, the Nastran BDF file size was smaller f om MSC Apex, due to the advancements in hex meshing, allowing for fewer elements to achieve the desired results. MSC Apex’s advancements in hexahedral mesh creation & validation were applied during this workflow evaluation, including Geometry Cleanup, Solid Split, Hex Mesh Creation, Generative Meshing Updates, and Hex Mesh Diagnostics Tool. Included below is a detailed description of this geometry file and workflow evaluation, including a description of how these time savings were achieved.
meshing,FEA,simulation,structures
×

[TITLE]

[LISTING

[ABSTRACT]

[DATE]

[ROOM]

[KEYWORDS]