Design of Experiments for Effects and Interactions during Brake Emissions Testing Using High-Fidelity Computational Fluid Dynamics

摘要

The investigation and measurement of particle emissions from foundation brakes require the use of a special adaptation of inertia dynamometer test systems. To have proper measurements for particle mass and particle number, the sampling system needs to minimize transport losses and reduce residence times inside the brake enclosure. Existing models and spreadsheets estimate key transport losses (diffusion, turbophoretic, contractions, gravitational, bends, and sampling isokinetics). A significant limitation of such models is that they cannot assess the turbulent flow and associated particle dynamics inside the brake enclosure; which are anticipated to be important. This paper presents a Design of Experiments (DOE) approach using Computational Fluid Dynamics (CFD) to predict the flow within a dynamometer enclosure under relevant operating conditions. The systematic approach allows the quantification of turbulence intensity, mean velocity profiles, and residence times. The factors of the DOE include: a) airflow level, b) brake size, c) rotor style, d) caliper position, e) brake rotation, f) brake rotational speed, and g) fixture style. Numerical simulations are performed using NGA, a high-order, multi-physics large-eddy simulation code. Particles are tracked individually in a Lagrangian manner. The CFD code is coupled with a conservative immersed boundary method to handle complex geometries. The second part of the study investigates the flow behaviour and the associated isokinetics near the sampling plane in the different nozzles that feed the air samples to the various instruments. In order to better understand the transport and fate of solid particles, the model uses a log-normal particle size distribution between 0.55 μm and 20 μm.