Bibliographic record
Abstract
RESUME Un des defis a relever pour les aerodynamiciens numericiens est de developper des methodes representant le plus fidelement possible la dynamique des fluides. L’augmentation des ressources de calcul disponibles permet maintenant a la dynamique des fluides numerique de representer et resoudre adequatement ces problemes. Les travaux presentes dans ce memoire se concentrent sur le developpement d’une methode pour resoudre les equations de Navier-Stokes sur des geometries complexes. Le logiciel utilise pour faire ces simulations est celui developpe a Polytechnique Montreal, NSCODE. Deux objectifs sont definis pour le projet: developper une methode permettant la resolution de geometries complexes utilisant des maillages partageant une surface et demontrer la robustesse de la methode en lien a des applications de type industriel. Dans le but d’augmenter les capacites de la methode, une revue de litterature du developpement de la methode dans differents groupes de recherche, tels la NASA ou l’ONERA, a ete faite. La methode chimere, aussi connue sous son appellation anglaise «Overset», est choisie pour sa grande flexibilite a supporter des geometries complexes. Elle permet de mailler les differentes composantes d’une geometrie de facon independante entre celles-ci. Cela permet donc de simplifier la generation des maillages, etape complexe dans le processus de la dynamique des fluides numerique. La methode chimere fait l’assemblage entre les differents maillages, utilisant des fonctions d’interpolation pour creer la communication entre eux. Une premiere version de la methode avait precedemment ete implementee au sein du solveur NSCODE, mais n’avait ete validee que sur des geometries dont les differentes composantes etaient entierement entourees de fluide. Pour des geometries complexes, il n’est toutefois pas possible de proceder ainsi, et les maillages doivent pouvoir se superposer sur la surface de la geometrie. Trois axes de developpement permettant d’elargir les capacites de la methode actuelle sont identifies. Premierement, la methode telle qu’implementee presentait un algorithme de decoupe de geometries (traduction libre du terme anglais «hole cutting») sommaire, echouant sur des geometries concaves. Un algorithme utilisant une triangulation Delaunay contrainte pour modeliser la geometrie est venu renforcir cette etape de la methode chimere. Deuxiemement, pour supporter des maillages qui se superposent sur la meme geometrie, l’interpolation dans les regions visqueuses a ete etudiee. Principalement, les particularites liees au solveur, soit une discretisation centree aux cellules et un schema de dissipation artificielle requerant 2 voisins, sont venues influencer les choix pour la methode.----------ABSTRACT Aerodynamics engineers aspire to develop methods that represent with as much fidelity as possible fluid dynamics. With the fast growth of computational resources, Computational Fluid Dynamics (CFD) tools can now rely on high fidelity methods to solve these problems. This thesis focuses on the development of a method to solve the Navier-Stokes equations over complex geometries. The flow solver developed at Polytechnique Montreal, NSCODE, is the software used to perform the simulations. Two objectives are defined: develop a method to simulate complex geometries using surface conforming meshes and demonstrate its robustness with respect to industrial type applications. A literature review is conducted to evaluate the maturation of the overset method inside different research groups, notably the NASA and the ONERA. Also known as the Chimera method, it is selected based on its capacity to handle such difficult geometries. It allows to mesh different components individually, which ensures maximum grid quality. The mesh generation process is then simplified, which is regarded as a tedious and time consuming aspect of CFD. The overset method proceeds to perform the assembly of the different components together. Communication between these meshes is assured by using interpolation functions. An initial version of the overset method had previously been implemented inside NSCODE. Its validation was partially done, as it was only used for fully separated geometries. For complex geometries, this condition can not always be met, and the method must be able to treat meshes that overlap on the surface. Three development axis are identified to increase the capabilities of the current implementation. First, the hole cutting algorithm in place, while being a fast method, lacks of versatility towards more complex cases. Concave geometries lead to non valid grid assembly. An algorithm is developed to replace it, which uses a constrained Delaunay triangulation to represent accurately the internal geometry. Second, in order to support meshes with overlapping surfaces, a study of the interpolation in the viscous region is performed. Focus is given to the particularities of the flow solver, mainly the cell centred scheme as well as an artificial dissipation scheme, to influence the chosen methods. Two aspects are analyzed: the mesh generation for these meshes and the proper treatment of the boundary condition. A limitation is proposed to the mesh generation, to help ensure adequate grid assemblies and valid interpolation donors. Third, the manner to compute the aerodynamic forces and moments is addressed. A weighted panel method is introduced to avoid the double integration in overlapping regions.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".