Data assimilation optimization for the evaluation of inverse mixing and convection flows

Abstract

A data assimilation approach is developed for evaluation of fluid-thermal systems wherein a complete specification of the boundary conditions is not known a priori and experimental diagnostics are restricted to a limited region of the flowfield. The methodology is applied to the configuration of a heated jet injected into a laminar boundary layer where the jet temperature and velocity are unknowns. The closed loop method selects the initial locations for the experimental measurements of mean temperature within the flowfield using diode laser absorbance. Diode laser absorption measurements near 761 nm of molecular oxygen are recorded to characterize the time varying temperature of the two dimensional jet. The Response Surface Models built using two-dimensional unsteady Navier-Stokes simulations are used to predict the jet exit temperature and velocity, and select the second set of locations for the experimental measurements. The two sets of experimental data are used to generate the final prediction for the jet exit temperature and velocity. The jet exit velocity is correctly predicted to within the experimental uncertainty; however, the jet exit temperature is over-estimated by 9% to 23%.

Combined experimental and numerical approach to model, design and optimize thermal processes

“…There are circumstances where experiment data can be obtained over only a limited region because of access, time or other limitations. In such cases, numerical modeling and experimentation may be used together to solve an inverse problem in order to define and quantify the boundary conditions and then proceed to the numerical simulation of the complete problem [11,12]. An example of this problem is given in Fig.…”