Latin hypercube sampling and the propagation of uncertainty in analyses of complex systems

The following techniques for uncertainty and sensitivity analysis are briefly summarized: Monte Carlo analysis, differential analysis, response surface methodology, Fourier amplitude sensitivity test, Sobol’ variance decomposition, and fast probability integration. Desirable features of Monte Carlo analysis in conjunction with Latin hypercube sampling are described in discussions of the following topics: (i) properties of random, stratified and Latin hypercube sampling, (ii) comparisons of random and Latin hypercube sampling, (iii) operations involving Latin hypercube sampling (i.e. correlation control, reweighting of samples to incorporate changed distributions, replicated sampling to test reproducibility of results), (iv) uncertainty analysis (i.e. cumulative distribution functions, complementary cumulative distribution functions, box plots), (v) sensitivity analysis (i.e. scatterplots, regression analysis, correlation analysis, rank transformations, searches for nonrandom patterns), and (vi) analyses involving stochastic (i.e. aleatory) and subjective (i.e. epistemic) uncertainty. Published by Elsevier Science Ltd.

/pdf/latin-hypercube-sampling-and-the-propagation-of-uncertainty-k6f019s9h6.pdf

Latin Hypercube Sampling and the Propagation of Uncertainty in Analyses of Complex Systems

▪ Abstract The issue of the physical mechanism(s) that control the efficiency with which the density field in stably stratified fluid is mixed by turbulent processes has remained enigmatic. Similarly enigmatic has been an explanation of the numerical value of ∼0.2, which is observed to characterize this efficiency experimentally. We review recent work on the turbulence transition in stratified parallel flows that demonstrates that this value is not only numerically predictable but also that it is expected to be a nonmonotonic function of the Richardson number that characterizes preturbulent stratification strength. This value of the mixing efficiency appears to be characteristic of the late-time behavior of the turbulent flow that develops after an initially laminar shear flow has undergone the transition to turbulence through an intermediate instability of Kelvin-Helmholtz type.

Mixing efficiency in stratified shear flows

The subject of ocean turbulence is in a state of discovery and development with many intellectual challenges. This book describes the principal dynamic processes that control the distribution of turbulence, its dissipation of kinetic energy and its effects on the dispersion of properties such as heat, salinity, and dissolved or suspended matter in the deep ocean, the shallow coastal and the continental shelf seas. It focuses on the measurement of turbulence, and the consequences of turbulent motion in the oceanic boundary layers at the sea surface and near the seabed. Processes are illustrated by examples of laboratory experiments and field observations. The Turbulent Ocean provides an excellent resource for senior undergraduate and graduate courses, as well as an introduction and general overview for researchers. It will be of interest to all those involved in the study of fluid motion, in particular geophysical fluid mechanics, meteorology and the dynamics of lakes.

https://assets.cambridge.org/97805218/35435/frontmatter/9780521835435_frontmatter.pdf

The Turbulent Ocean

Uncertainty and sensitivity analyses for systems that involve both stochastic (i.e., aleatory) and subjective (i.e., epistemic) uncertainty are discussed. In such analyses, the dependent variable is usually a complementary cumulative distribution function (CCDF) that arises from stochastic uncertainty; uncertainty analysis involves the determination of a distribution of CCDFs that results from subjective uncertainty, and sensitivity analysis involves the determination of the effects of subjective uncertainty in individual variables on this distribution of CCDFs. Uncertainty analysis is presented as an integration problem involving probability spaces for stochastic and subjective uncertainty. Approximation procedures for the underlying integrals are described that provide an assessment of the effects of stochastic uncertainty, an assessment of the effects of subjective uncertainty, and a basis for performing sensitivity studies. Extensive use is made of Latin hypercube sampling, importance sampling and reg...

Uncertainty and sensitivity analysis in the presence of stochastic and subjective uncertainty

Trends and needs in experimentation and numerical simulation for LWR safety

The direct numerical simulation (DNS) of a temporally-growing mixing layer has been carried out, for a variety of initial conditions at various Richardson and Prandtl numbers, by means of a pseudo-spectral technique; the main objective being to elucidate how the entrainment and mixing processes in mixing-layer turbulence are altered under the combined influence of stable stratification and thermal conductivity. Stratification is seen to significantly modify the way by which entrainment and mixing occur by introducing highly-localized, convective instabilities, which in turn cause a substantially different three-dimensionalization of the flow compared to the unstratified situation. Fluid which was able to cross the braid region mainly undisturbed (unmixed) in the unstratified case, pumped by the action of rib pairs and giving rise to well-formed mushroom structures, is not available with stratified flow. This is because of the large number of ribs which efficiently mix the fluid crossing the braid region. More efficient entrainment and mixing has been noticed for high Prandtl number computations, where vorticity is significantly reinforced by the baroclinic torque. In liquid sodium, however, for which the Prandtl number is very low, the generation of vorticity is very effectively suppressed by the large thermal conduction, since only small temperature gradients, and thus negligible baroclinic vorticity reinforcement, are then available to counterbalance the effects of buoyancy. This is then reflected in less efficient entrainment and mixing. The influence of the stratification and the thermal conductivity can also be clearly identified from the calculated entrainment coefficients and turbulent Prandtl numbers, which were seen to accurately match experimental data. The turbulent Prandtl number increases rapidly with increasing stratification in liquid sodium, whereas for air and water the stratification effect is less significant. A general law for the entrainment coefficient as a function of the Richardson and Prandtl numbers is proposed, and critically assessed against experimental data.

Numerical investigation of the entrainment and mixing processes in neutral and stably-stratified mixing layers

Computational Fluid Dynamics for nuclear applications: from CFD to multi-scale CMFD

A temporally-growing mixing layer has been directly simulated with a pseudospectral technique, for initial bulk Richardson numbers from 0.0 to 0.2 and for Prandtl numbers from 0.00535 to 2.2. Several different initial conditions for the velocity fluctuations were imposed. For the two-dimensional (2-D) case only purely-deterministic conditions were used, whereas purely-deterministic, combined deterministic-random, or purely-random conditions were imposed in the three-dimensional (3-D) cases. The numerical procedure allowed fields with very different characteristic lengths to be resolved, with spectral accuracy maintained. The evolution of the velocity, active (temperature), and passive scalar fields were followed independently by adaptively redistributing collocation points in the regions of high shear and rapid scalar variations. The vertical boundary conditions were imposed at infinity to eliminate any boundary-layer effects and an exponential mapping was used to translate infinite physical space into fi...

Numerical investigation of the formation of three-dimensional structures in stably-stratified mixing layers

The paper discusses the results of a detailed direct numerical simulation study of condensing stratified flow, involving a sheared steam-water interface under various thermal and turbulent conditions. The flow system comprises a superheated steam and subcooled water flowing in opposite directions. The transport equations for the two fluids are alternately solved in separate domains and then coupled at the interface by imposing mass, momentum, and energy jump conditions with phase change. The effects induced by changes in the interfacial shear were analyzed by comparing the relevant statistical flow properties. New scaling laws for the normalized heat transfer coefficient (HTC), K + , have been derived for both the steam and liquid phases. The steam-side law is found to compare with the passive-scalar law obtained hitherto by (Lakehal et al.(2003, "Direct Numerical Simulation of Turbulent Heat Transfer Across a Mobile, Sheared Gas-Liquid Interfaces, " ASME J. Heat Transfer, 125, pp. 1129-1139) in that HTC scales with Pr -3/5 . A close inspection of the transfer rates on the liquid side reveals a consistent relationship between K + , the local wave deformation or curvature and the interfacial shear stress. The surface divergence model of Banerjee et al. (2004, "Surface Divergence Models for Scalar Exchange Between Turbulent Streams," Int. J. Multiphase Flow, 30(8), pp. 965-977) is found to apply in the liquid phase, too.

George Yadigaroglu

Papers

Trends and needs in experimentation and numerical simulation for LWR safety

Numerical investigation of the entrainment and mixing processes in neutral and stably-stratified mixing layers

Computational Fluid Dynamics for nuclear applications: from CFD to multi-scale CMFD

Numerical investigation of the formation of three-dimensional structures in stably-stratified mixing layers

Direct Numerical Simulation of Condensing Stratified Flow