07
Nov
DECADE project, multi-geometry automatic mesh
in News
Comments
DECADE- Digital Engineering for AirCraft Aerodynamic Design is an ambitious project reimagining aerospace education for the engineers of tomorrow! At the core of this transformative program is CEASIOMpy, a system enabling students to do complex aircraft design calculations in a fully digital environment.
To deepen their learning, students also engaged in (semi-)automatic mesh generation using Fidelity Pointwise amongst others software tools, allowing them to explore high control meshing on simplified aircraft configurations.
We are proud to share our contribution to the project.
Introduction
Fidelity Pointwise (PW) is a powerful software when complex geometries or particular needs are requested in CFD simulation.
The ability to modify locally every type of mesh gives the possibility to reach high-end quality mesh. High-quality meshing is not always the top priority; instead, speed and reliability are often more critical in a CFD workflow.
In such cases, not all is lost: with Fidelity Pointwise, you can script commands to fully automate both individual tasks and complex processes. This allows for significant time savings while maintaining consistent mesh quality, even when dealing with varied geometries or specific challenges. This capability ensures a level of flexibility and efficiency that other software might struggle to achieve without compromising the mesh quality. In this study, developed in collaboration with Cadence, Airinnova, CFSE, KTH, and Vinnova, we propose a Fidelity Pointwise script that has been developed to automate as much as possible high-quality mesh over 4 different geometries.
Geometries
The proposed geometries are very different between them: each one of them presents at least one element that is not found in the previous one making it challenging to create one and only one script that works for all of them.
Figure 1 shows the comparison between two of the four used geometries.
Figure 1: Side by side comparison of two different geometry that can be meshed with one Fidelity Pointwise script.
By individuating commonality and group quilts under the same name, Fig.2, the script individuates surfaces shared by different geometries easily. Thus a preprocessing choosing coherent names along the several geometries is necessary to have a smooth algorithm. The more similar the geometries are between each other, the more automatic the script will be. Exception elements will be coded individually.
Figure 1: Side by side comparison of two different geometry that can be meshed with one Fidelity Pointwise script.
Mesh
The size of the mesh is parametric: by choosing the geometry an automatic ”medium” mesh will be generated. It is also possible to change the size of the biggest element size we want in our mesh and automatically all the related connectors will scale respectively.
This allows a rapid and easy refinement or coarsening of the mesh so that grid convergence studies can be run flawlessly.
The relation between the largest element size in the mesh and the size of several other connectors has been determined based on best practices. It’s important to highlight that these proportions can be easily fine-tuned as needed.
An example of how the code adapts to different geometries is shown in Fig.3.
