par Di Matteo, Mariano
Président du jury Parente, Alessandro
Promoteur Hendrick, Patrick
Publication Non publié, 2025-09-15
Thèse de doctorat
Résumé : Aviation industry is making tremendous efforts to reduce aircraft emissions. Engine manufacturers are focusing on improving engine performance, mass and efficiency while reducing emissions and climate change effects. These efforts extend to all engine subsystems. Among these, the lubrication system requires continuing research and development to continue functioning under increasing demanding operational conditions.The lubrication system is a closed-loop oil recirculating system composed of three main subsystems: the oil supply line, the oil recovery line, and the vent line. The vent line ejects the air used to seal bearing chambers and gearboxes. However, the air vented overboard transports also oil droplets. These droplets originate from inside the lubricating system where the sealing air enters in contact with the lubricating oil.A centrifugal separator, also called breather or de-oiler, separates air and oil and recovers the oil before venting cleaner air overboard. This process reduces oil emissions and oil consumption. The centrifugal separator removes oil droplets using centrifugal force. A hollow body, through which the flow passes, is set into rotation. The air, loaded with oil droplets, flows through it, causing the higher density liquid to be separated by centrifugal action. Often, a porous structure is placed inside the separator to enhance the separation process. However, this separation process also introducespressure losses into the vent system. The breather design must therefore balance efficient oil separation with minimal air pressure loss.The performance evaluation of new breather prototypes and the study of strategies to evaluate its performance is the objective of this study. This objective is pursued through a combination of experimental and numerical activities. A dedicated test bench has been developed to replicate the operating conditions of the vent line. The experimental campaign is focused on the characterization of multiple breather prototypes and oil droplets measurements, while the numerical analysis is conducted to support the interpretation of results and to provide performance predictions under varying operating conditions.The test bench is designed to assess the breathers’ performance parameters, including separation efficiency, pressure drop across the breather, and the oil droplets size in the flow after the separation. The test bench allows control over the functional parameters that define the operating conditions such as variation of air flow rate, oil flow rate, rotational speed of the separator, and droplets size distribution in the inletmixture, allowing a systematic investigation of breather behaviour across a range of operating conditions.The prototypes analyzed during the campaign differ mainly in the configuration of their internal porous structures. Experimental testing is employed to quantify how these structural variations influence the overall performance of the devices.The evaluation of the performance required the development of methodologies and strategies. To assess the influence of droplet size on separation efficiency, a systematic investigation is carried out. This is achieved by testing the breathers under controlled conditions using different particle size distributions (PSDs). The PSDs are generated and introduced into the system through a controlled injection process. Droplet size measurements are performed using a laser diffraction system, allowing for accurate characterization of the inlet flow. The measurement of oil consumption is carried out by the use of a gravimetric analysis. A filter bag collect the oil exiting the separator and the difference in weight of the filter before and after a measurement is proportional to the oil consumption. This measurement is performed for the different operating condition allowed by the test bench. The pressure drops are measured by keeping constant rotational speed of the breather and varying the inlet air flow rate. Finally the size of the droplets at the exit of the separator is measured again with the laser diffraction. These measurements are made in the different operating conditions allowed by the test bench.The experimental results indicate how the breather performances are influenced by operating parameters such as rotational speed, airflow rate, oil flow rate, and droplet size distribution. An increase in rotational speed is associated with a rise in pressure losses, while concurrently reducing oil consumption, thanks to a better separation. Increasing the airflow rate leads to both higher pressure losses and increased oil consumption. The oil flow rate does not exhibit a significant impact on pressure losses; however its influence on the consumption depends strongly on the characteristicsof the droplet size distribution. Between the variables considered, droplet size is the most critical factor affecting oil consumption. Specifically, smaller droplets are associated with higher oil consumption, indicating reduced separation efficiency for a finer droplets distribution. The presence of a porous media inside the breather enhances the separation efficiency, introducing consequential pressure losses in certain configurations. The results also show a correlation between the size of droplets escaping separation and the geometry of the porous structure inside the breather.In parallel with the experimental campaign, a numerical study is carried out to extend the analysis. While the experimental results offer direct insight into the operational performance of the separators, the numerical simulations are employed to develop a predictive model and to help the design of next generation aero-engine breather configurations.The numerical study is divided into two phases. The first focused on the study of the pressure losses across the separator. A numerical model is developed to predict pressure losses under different operating conditions. This model applies a momentum source approach to simulate the porous structure inside the separator, based on the Darcy-Forchheimer equation. Consequently, it requires the determination of the permeability coefficient of the porous structure and the turbulence related coefficient. Initially, these coefficients are identified using experimental results, allowing for a direct correlation between measured data and model predictions. However, since this approach remains highly dependent on experimental input, a second method, based entirely on numerical analysis is also introduced. A representative element of volume (REV) of the porous material is analyzed, and the pressure losses across this element are computed. These pressure losses are then used to determine the permeability and turbulence coefficients for the numerical model.In both approaches, the model systematically underestimates pressure losses at higher air velocities, indicating that the flow conditions at these air velocities exceeded the regime described by the Darcy-Forchheimer equation. The second aspect investigated numerically is the trajectory of droplets through the porous medium. For this analysis is employed the Discrete Phase Model (DPM) of the commercial software Fluent. By solving the flow for homogeneous droplet distributions, the behavior of droplets inside the porous medium is analyzed, allowing for the evaluation of transport mechanisms and potential separation efficiency.The study focuses on the residence time of droplets inside the porous structure, as a longer residence time increases the likelihood of separation. The results show that as droplet size increases, residence time also increases. However, the residence time is strongly related to the impact with the metallic grid of the porous medium. The larger particles are more likely to impact against the structure and therefore remain trapped longer inside the mesh. Smaller particles are more easily transported by the continuum phase.

Documents en relation

DI-fusion