Pei-feng Hsu

Experimental and Numerical Study of Premixed Combustion Within Nonhomogeneous Porous Ceramics

Abstract An experimental and numerical investigation of premixed methane combustion within a nonhomogeneous porous ceramic was performed. The burner consisted of two porous ceramic cylinders of equal length and diameter that were stacked together and insulated around the circumference. Four series of experiments were carried out to determine the lean limit using three different pore sizes in the downstream ceramic cylinder (SBR). The pore size in the upstream ceramic cylinder was constant in all four cases. A new definition of the lean limit was introduced to account for the effects of the porous ceramic. The burners were tested over the range of lean limit 0.41 < φ≤ 0.68 and the numerical simulations were performed over 0.43 < φ ≤ 1.0. The results demonstrated that porous ceramic burners provide a range of stable burning rates at a constant φ, The maximum flame speed inside the burners was much higher than the premixed, freely burning adiabatic laminar flame (free flame) speed. The lean limits in the por...

An Integral Formulation of Transient Radiative Transfer

A time-dependent integral formulation is developed for modeling transient radiative transfer. The development is based on a rigorous analysis of the wave propagation process inside the participating media. The physical significance of the present integral formulation is the consideration of the time-dependent domain of computation, which is different from the domain disturbed by radiation (i.e., the wave front envelope). Numerical computations are performed for the medium that is an absorbing and isotropically scattering one-dimensional plane slab geometry. The spatial and temporal incident radiation and radiative flux distributions are presented for different boundary conditions and for uniform and nonuniform property distribution

Analysis of light pulse transport through two-dimensional scattering and absorbing media

In this paper, a two-dimensional transient radiation transport algorithm is developed to analyze the ultra-short light pulse propagation in an anisotropically scattering rectangular medium. The time-dependent discrete ordinates method is used with the high order upwind piecewise parabolic interpolation scheme. The upwind difference scheme is needed to resolve the propagating wave front. This method extends from a prior one-dimensional model to treat the two-dimensional geometry through a Strang-type dimensional splitting. The accuracy and efficiency of this algorithm are studied. Comparisons with one-dimensional case and a parametric study show the flexibility of this method for homogeneous and non-homogeneous media.

MEASUREMENTS OF THERMAL CONDUCTIVITY AND OPTICAL PROPERTIES OF POROUS PARTIALLY STABILIZED ZIRCONIA

Experimental measurements of the high-temperature (300 < T < 800 K) thermal conductivity of highly porous (porosity greater than 80%), partially stabilized zirconia (PSZ) were performed. A method of simultaneously inverting conductivity and extinction coefficient from the experimental data is presented. The effect of natural convection within the porous plates with heating from below was found to be negligible. The thermal conductivity integral (TCI) method was incorporated into the inversion of conductivity and radiative properties from the diffusion approximation of the combined radiation and conduction heat flux measurement. The measured conductivity decreased slightly as the pore size of the PSZ increased. The extinction coefficient decreased with increased pore size, and for pore size greater than 0.6 mm the trend had good agreement with the geometric optics limit prediction.

The necessity of using detailed kinetics in models for premixed combustion within porous media

Abstract Models for premixed combustion within porous inert media (PIM) are complicated by the highly nonlinear radiative exchange terms in the energy equation for the solid matrix in addition to the stiffness of the set of gas phase equations. Therefore, prior researchers have simulated the gas-phase reactions using single-step chemistry. In the present work, predictions are made using both single-step and multistep kinetics mechanisms. It is concluded that it is essential to use multistep kinetics if accurate predictions of the temperature distributions, energy release rates, and total energy release are sought. Obviously, this is also true if predictions of the composition profiles and emissions are sought. Single-step kinetics is shown to be adequate for predicting all the flame characteristics except the emissions for the very lean conditions under which equilibrium favors the more complete combustion process dictated by global chemistry. The first predictions of NO and CO emissions from PIM burners are presented and compared with experimental data. The model predicts the CO emissions very accurately and predicts the NO trend correctly but overpredicts the NO emissions for φ > 0.8. The present multistep PIM burner model does not accurately reproduce the data for the burning speed and NO emissions for nondilute mixtures. These discrepancies can be only partially attributed to experimental uncertainties and/or imprecise knowledge of the properties of the solid matrix. Thus, it is concluded that important aspects of the physical processes within PIM combustors are not well simulated at present.

A Numerical Investigation of Premixed Combustion Within Porous Inert Media

A numerical investigation of premixed combustion within a highly porous inert medium is reported. Specifically, results of a numerical model using detailed chemical kinetics and energy exchange between the flowing gas and the porous solid are presented. The current formulation differs from prior models of this type of combustion in that multistep kinetics is used and a better description of the thermophysical properties of the solid is applied in the present model. It was found that the preheating effect increases strongly with increasing convectiue heat transfer and with increasing effective thermal conductivity of the solid

Transient radiative transfer in three-dimensional homogeneous and non-homogeneous participating media

The transient radiative transfer is studied by an integral equation model. The salient feature of the integral formulation is the revelation of the time-dependent domain of integration (or domain of influence). This transient model can be reduced to simulate the steady state radiative transfer and, thus, it is a general formulation. To demonstrate the effectiveness of the model, the transient radiative transfer in a rectangular volume with absorbing and isotropic scattering medium is considered. Diffuse irradiation enters at one boundary surface. The other five boundaries are cold and black surfaces. The spatial and temporal distributions of the integrated intensity and radiative flux are presented for different radiative property distributions. Numerical quadratures used are the discrete rectangular volume method and YIX method. The results show good agreement between the two methods. The discrete rectangular volume method is a variation of the general quadrature method.

Effects of multiple scattering and reflective boundary on the transient radiative transfer process

Abstract A time-dependent Monte Carlo method is developed for modeling the transient radiative transfer within absorbing and scattering media. It is well known that the transmitted and reflected signal from a short-pulse incident radiation to the scattering medium has very long, extended temporal spread. However, two important but yet to be clarified issues are: (1) the relative and combined influences of the multiple scattering and reflective boundary on the transient radiative transfer process; and (2) the effect of the incident temporal pulse width and pulse shape on the radiative transport. It is found that the pulse width and shape has little effect on the transient radiative transport process when the pulse temporal width is much smaller than the pulse transmission time through the one-dimensional slab. The stochastic simulation of time-of-flight of photon compares very well with the results obtained by a recently developed Volterra integral equation formulation of the transient radiative transfer.

Radiative transfer by the YIX method in nonhomogeneous, scattering, and nongray media

A numerical solution of the integral equation of radiative heat transfer using the YIX method involving a mixture of highly anisotropic scattering particles and a nongray absorbing gas is presented. To validate the three-dimensional calculation, bench mark solutions are established on a model problem using a high-order accuracy method, the product integration method (PIM). Various effects, e.g., the discrete ordinates sets, first integration point of the YIX quadrature, optical thickness of the medium, grid sizes, and spectral resolution on the accuracy of the three-dimensional calculation are discussed. Results for three-dimensional calculations are presented. For all cases, the pressure variation has less significant effect on the results than those by particle density or temperature variations. The three-dimensional nonhomogeneous cases have different trends of variation in radiative flux and divergence due to their nonuniform particle density distribution and nonisothermal participating medium. The use of the YIX method with discrete ordinates for the multidimensional calculations of highly anisotropic scattering and spectrally-dependent medium is shown to be accurate and flexible.

Parallel Computing of Three-Dimensional Monte Carlo Simulation of Transient Radiative Transfer in Participating Media

A three-dimensional transient radiative transport model using Monte Carlo (MC) method was implemented in a Beowulf-class parallel computer system. The main purpose of this study is to simulate the light pulse transport inside the absorbing and scattering media such as biological tissues. The threedimensional results for the case of infinite height and width but finite thickness configuration have good agreement with our prior one-dimensional MC results. The major advantage of MC method is its flexibility and simplicity to simulate the photon movement in arbitrary geometry and complex boundary condition. Since the error bound of MC method is inversely proportional to the square root of the number of statistical samplings, it requires a large number of samples to reach the satisfactory accuracy. Therefore, the primary drawback of MC methods is that it is a computationally intensive method. However, the method is very adaptable to parallel computing, i.e., coarse grained algorithm. Parallel computing is introduced to improve the performance of this method. The parallel system is based on a cluster of commodity-class processors with a standard MassagePassing-Interface (MPI). MPI is a library of subprograms. It is chosen as a tool to implement parallel programming for its portability to other similar systems and the very low cost compared to the conventional supercomputer systems. NOMENCLATURE a absorption coefficient, 1/m c propagation speed of radiation transport in the medium, m/s G incident radiation or integrated intensity, W/m2