Torrey R. Hayden

Parameter Estimation for a Turbulent Buoyant Jet with Rotating Cylinder Using Approximate Bayesian Computation

Assigning accurate boundary, initial, and geometric conditions, as well as fluid properties, in numerical simulations of real-world applications continues to be a major challenge in computational fluid dynamics research. However, recent advances have provided new and increasingly sophisticated methods for accurately determining these conditions and properties. This paper focuses on a data-driven parameter estimation technique called Approximate Bayesian Computation (ABC), which allows numerical simulation parameters to be determined from experimental or other “truth” data. In this proof-of-concept study, the ABC approach is demonstrated for a rotating cylinder above a high-temperature turbulent buoyant jet, which is an engineering problem of relevance to the commercial aerospace industry. Here the “truth” observations come from a two-dimensional Reynolds-averaged Navier Stokes (RANS) simulation that serves as a known test case against which other simulations are compared. In particular, the “truth” case has known jet inflow and cylinder rotational velocities, and we show that the ABC approach is able to correctly identify the true values of these parameters. This success indicates that ABC can be extended to real-world engineering systems and that, in the future, ABC will allow experimental observations to accurately drive the selection of boundary, initial, and geometric conditions, as well as fluid properties, in numerical simulations.

Characterization of the Output from a Catalytic Combustor Using Wavelength Modulation Spectroscopy

We characterize temperature and water vapor mole fraction in the output of a catalytic combustor. Catalytic combustion is used extensively in power production, propulsion, and industrial treatments. Wavelength modulation spectroscopy enables sensitive, non-intrusive, and fast measurements of temperature and water vapor mole fraction in the flow above the combustor. The sensor is first validated under known conditions in a controlled furnace. We then measure the temporal and spatial profiles of temperature and water vapor mole fraction at fuel lean conditions in a model combustor to quantify stability and uniformity, and develop boundary conditions for computational fluid dynamic studies of the flow.

Large amplitude wavelength modulation spectroscopy for sensitive measurements of broad absorbers

We demonstrate wavelength modulation spectroscopy of CO 2 at >30 atm using a fast-scanning MEMS laser. The technique shows promise for sensitive measurements of species with broadband absorption features, such as high-pressure gases and large molecules.