The quenching process is an essential part of the manufacturing process of aircraft landing gear main fittings. It provides the required material properties to achieve the desired strength to withstand high loads and corrosive environments. Unfortunately, quenching comes with undesired effects such as quenched induced distortion, which might impact the productivity of the manufacturers. Control and prediction of distortion remain an important field of engineering application. Notably, the numerical tools capable of dealing with the complex interactions between the thermal, metallurgical, and mechanical fields during quenching still rely on the definition of the thermal boundary conditions. Therefore, it is essential to accurately characterize and model the thermal boundary conditions of the process of interest, i.e., the facilities where large landing gear main fittings are processed.
The conditions encountered under industrial environments, often distant from the highly controlled laboratory setups, increase the challenge of accurately characterizing and modeling the thermal boundary conditions. For instance, pit furnace layouts require the heat-treated load to be extracted and then transferred by air to the quench bath which alters the starting condition before immersion even occurs. Moreover, uneven cooling might be caused by the heat treatment rig itself, poor circulation of the fluid, or even local conditions in the working space of the bath itself. Such differences encountered in the daily practice have led multiple researchers to question the applicability of standard testing on real workpieces.
This project implements industry-oriented quench probes, plate-like and hollow cylinder like, to characterize the quenching process of the industrial partner. The experimental means in this study successfully assessed the bath heterogeneity, the influence of probe orientation, and the position on the tooling. Likewise, the influence of the quench probe, in terms of geometry and size, was analyzed by comparing data from different test specimens.
The thermal boundary conditions were estimated by solving the inverse heat transfer problem for both the air transfer step and immersion step. The surface temperature is reconstructed first before retrieving the surface heat fluxes from a direct simulation using the identified temperature as a boundary condition. The chosen methodology produces accurate results in the form of a relative error below 5% with regards to the experimentally acquired cooling curves. The resulting database, which accounts for the same factors as the experimental work outlined above, represents a unique source of representative coefficients of an industrial scale process, currently unavailable in the literature.
Furthermore, mathematical models of the thermal boundary conditions are proposed for both steps (air transfer and immersion) accounting for the influence of the quench probe size. The models are based on characteristic points delimiting the evolution of the different cooling stages. During immersion, the characteristic points are linked to the physical phenomena associated with the transition between the boiling regimes (film, transition, and nucleate boiling). Therefore, each cooling regime of a typical boiling curve is modeled separately. The validation of the proposed models against the cooling curves of an independent test produced a relative error below 8%, a slight increase when compared with the inverse results, yet comparable with data in the literature and therefore deemed acceptable for the conditions of study.
Salazar Jimenez, G. (Author),
Champliaud (Supervisor),
Jahazi (Co-supervisor) &
Tahan (Co-supervisor),
24 Apr 2026Student thesis: Doctoral thesis › Doctorate in Engineering: Engineering