ISBN 9783843950909

978-3-8439-5090-9, Reihe Thermodynamik

Alexander P. Schindler
Partitioned Modeling of Transient Conjugate Heat Transfer

305 Seiten, Dissertation Universität Stuttgart (2022), Softcover, A5

Zusammenfassung / Abstract

The structural components in modern gas turbines of aero engines are exposed to high temperature gas flow. In order to ensure safe and efficient operation, a proper cooling of the relevant components is essential. The thermal interaction between the solid structure and the adjacent fluids (combustion gas and coolant) in conjunction with a highly dynamic operation mode may lead to thermal stresses and variable clearances affecting the product’s durability.

The demand for fast and cost-effective development yields integral virtual design procedures to become increasingly important. Finite element solvers are most commonly used to compute the temperature distribution and the resulting thermal stresses in the solid structure, while computational fluid dynamics simulations of the adjacent flow are mostly based on finite volume solvers.

The research presented in this thesis aims to evaluate and enhance partitioned approaches that enable the coupling of domain-specialized solvers to model the behavior of the complete conjugate system and thus improve the prediction quality. Special focus is put on adequately addressing the time scale disparity of the convection-dominated thermal processes in the fluid and the heat conduction in the solid.

For this purpose, a Python coupling program has been developed that allows to analyze, compare, and improve various partitioned approaches in one flexible framework. The approaches are classified with regard to their performance, stability, and suitability for modeling time-disparate phenomena. Steady and transient heat transfer is analyzed in geometries comprising composite solids, a plate in a channel, and a rotor-stator cavity. Parameters concerning transmission conditions, relaxation, the influence of geometry and its discretization, interpolation and data mapping, as well as material data and flow characteristics are evaluated with regard to their influence on convergence, robustness, and accuracy of the coupled solution.