The Ph.D. project addresses turbulent mixing in the presence of stable and unstable density stratifications with a particular focus on high density ratios. This class of mixing phenomena can be found in numerous applied problems, such as chemical processes, the dispersion of accidentally released liquefied gases, the simulation of fires and severe nuclear accidents. One example from the latter field is the accumulation, transport and dispersion of hydrogen in the containment of a LWR. Hydrogen is formed by chemical reactions between the fuel rod cladding material and vapour in case of cooling failure. A reliable prediction of hydrogen distributions in the containment is of crucial importance for the design of countermeasures to avoid detonations endangering containment structures. The latter provides the main motivation to the Laboratory of Thermal Hydraulics of PSI to performing research in this area. The project will provide experimental data for improving CFD models including coarse mesh containment codes for turbulent mixing of fluids with high density ratios.
Despite the technical relevance of the phenomenon, related literature is surprisingly scarce. There are only a few references outlining past shear layer experiments at high density differences (He-N2), and no examples of isokinetic experiments at high density differences were located. Past numerical simulations of the above mentioned mixing scenarios have often deviated from experimental results. This failure is due in part to the violation of the Boussinesq approximation of the first kind, i.e. the assumption that the density variation generates only a small correction to the inertia term of the momentum equation and can be thus neglected, only considering the density variation in conjunction with the acceleration of gravity.
For the measurement of the velocity and concentration fields in the HOMER test section, Particle Image Velocimetry (PIV) and Laser Induced Fluorescence (LIF) will be used. This will be complemented by Laser Doppler Anemometry (LDA) at selected positions in the test section. From the obtained series of instantaneous velocity and concentration fields, data needed for the model development and validation can be derived: (1) distributions of mean velocities, of the velocity fluctuations as well as of the Reynolds stresses, and (2) mean concentrations and concentration fluctuations.
On the basis of the mean and rms values of the velocity and concentration fields, appropriate experimental growth and entrainment laws will be derived and compared with values from the literature as much as possible. Some efforts will be undertaken to derive a universal growth law covering the different density stratifications. This approach will likely be based on a local gradient Richardson number.
The Optical Flow technique will be applied to some of the PIV recordings and will be further developed. It is expected that this new method will allow for an enhanced spatial resolution of the velocity field (compared with the classical cross-correlation approach) in order to resolve small scale coherent structures responsible for the details of the mixing process. This will also support the identification of the – most probably – Richardson number-dependent, and therefore stratification strength-dependent, dominant mixing mechanisms on which there is currently no consensus.