Keh Chin Chang

A Modified Low-Reynolds-Number Turbulence Model Applicable to Recirculating Flow in Pipe Expansion

A modified low-Reynolds-number k-e turbulence model is developed in this work. The performance of the proposed model is assessed through testing with fully developed pipe flows and recirculating flow in pipe expansion. Attention is specifically focused on the flow region around the reattachment point. It is shown that the proposed model is capable of correctly predicting the near-wall limiting flow behavior while avoiding occurrence of the singular difficulty near the reattachment point as applying to the recirculating flow in sudden-expansion pipe.

Combustion of pulverized coal using waste carbon dioxide and oxygen

Combustion of pulverized coal in CO2O2 as well as in air atmospheres is studied. Predictions using a one-dimensional computer code were compared with actual experimental data from tests conducted by Battelle Columbus Laboratories. The comparison of predicted and measured data for all test cases show that the observed trends of distributions of temperature and of species concentrations are generally predictable. The study confirms that the combustion of pulverized coal can be completed in a CO2O2 atmosphere over a range of CO2-to-O2 mole ratios between 2.23 and 3.65.

Theoretical investigation of transient droplet combustion by considering flame radiation

For a droplet burning in quiescent air and in a high temperature environment, conduction and radiation are two major heat transfer modes. Soot formation has been observed in many experiments on single-droplet combustion of different fuels. As a result, radiation heat transfer should play a dominant role in the burning process. In addition, most liquid fuels are semi-transparent materials. The collocation method is applied to solve the radiative heat transfer equation in regard to the droplet interior. The results, including different cases of droplet diameter square, ignition delay time, with and without considering radiation heat transfer, are compared with the experimental data. The prediction of the diameter square when radiation heat transfer is considered agrees more with the experimental data than the case without considering radiation heat transfer.

Application of a robust β-pdf treatment to analysis of thermal NO formation in nonpremixed hydrogen-air flame

Abstract The assumed β-pdf is widely used in turbulent flame simulation with the moment closure method. However, implementation of the β-pdf in turbulent flame simulation may sometimes encounter singularity difficulties in numerical calculation. The study proposes a robust β-pdf treatment to overcome these numerical difficulties. The proposed β-pdf treatment associated with the partial equilibrium chemistry model and the k -ϵ turbulence model is firstly applied to the case of H 2 /N 2 -air nonpremixed flame to demonstrate its robustness. Then, the present method is applied to the study of thermal NO formation in H 2 -air nonpremixed flames. Comparison of results shows the present model to be capable of yielding fair agreement with the measurements. It is competitive with other more sophisticated methods, such as the pdf evolution method and the conditional moment closure method.

Experimental and Theoretical Study on Hollow-Cone Spray

A theoretical and experimental investigation has been conducted to study the two-phase turbulent structure in an isothermal hollow-cone spray. Mean and fluctuating velocity components, drop number density, as well as drop-size distribution were measured with a nonintrusive diagnostic tool, a two-component phase Doppler particle analyzer. Complete initial conditions required for theoretical calculations were also provided with measurements. Theoretical calculations were made with an Eulerian-Lagrangian formulism. Turbulent dispersion effects were numerically simulated using a Monte Carlo method. Turbulence modulation effects were also taken into account in the modeling. The well-defined experimental data were used to assess the accuracy of the resultant Eulerian-Lagrangian model. Comparisons showed that the theoretical predictions, based upon the Eulerian-Lagrangian model, yielded reasonable agreement with the experimental data. The improvements made by inclusion of the selected turbulence modulation model were insignificant in this work.

Calculation of wall heat transfer in pipeexpansion turbulent flows

Abstract A new modified low-Reynolds-number k-e turbulence [Chang, Hsieh and Chen (CHC)] model, which possesses the proper near-wall limiting behaviors and is free of the singular defect occurring near the reattachment point when applied to separated flows, is examined for use in wall heat transfer problems in flow with pipe expansion geometry. Another eight low-Reynolds-number k-e models, found in open literature, are also examined in this study. Attention is specifically focused on the flow region surrounding the reattachment point. Comparative results show that only the CHC model and the model developed by Abe et al. [Abe, Kondoh and Nagano (AKN model)] can yield satisfactory distributions of the Nusselt number along the wall. However, the CHC model adopted the same model constants as conventionally used for the standard k-e model. Thus, the CHC model is more universal than the AKN model.

Sensitivity study on Monte Carlo solution procedure of two-phase turbulent flow

A parametrical sensitivity study on the stochastic separated flow model which adopts the Lagrangian framework with the Monte Carlo method to track the drops in turbulent flow field is performed. It is found that an approximate 10-μm uniform width of each discrete size interval can adequately represent the spectral effects of drop size distribution in the investigated hollow-cone spray. The number of computational drops required for the statistically stationary solution is greatly dependent on the interval range of PDF domain employed. For the case using the interval range of PDF domain bounded within it is shown that the use of no less than 1000 computational drops for each representative size can yield nearly invariant solution.

Theoretical and Experimental Study on Two-Phase Structure of Planar Mixing Layer

A theoretical and experimental investigation is conducted to study the two-phase turbulent structure in a planar mixing layer with polydispersed drops. Mean and fluctuating velocity components, drop number density, as well as drop size distribution were measured with nonintrusive diagnostics of the two-component phase Doppler particle analyzer. Complete initial conditions required for theoretical calculation were also provided with measurements. Theoretical investigation is made with the Eulerian-Lagrangian formulation. Turbulent dispersion effects are numerically simulated using the Monte Carlo method. The well-defined experimental data have been used to assess the accuracy of the Eulerian-Lagrangian model. The comparisons show that the theoretical predictions, based on the Eulerian-Lagrangian model, yield reasonable agreement with the experimental data.

MULTIGRID COMPUTATION FOR TURBULENT RECIRCULATING FLOWS IN COMPLEX GEOMETRIES

Computation of recirculating turbulent flow in complex geometries is important in engineering but fundamentally difficult. A procedure using a full multigrid and full approximation storage is developed in conjunction with a pressure-based algorithm using curvilinear coordinates and the k-e two-equation turbulence model. This method is applied to several flow problems with different geometries, grid sizes, and convection schemes. Although the multigrid procedure does not yield a convergence rate independent of these factors. Us performance is noticeably less affected than that of the single-grid method. Degradation of the boundary representation during grid restriction is one effect on the multigrid performance. Some features regarding the treatment of the turbulence quantities that help the performance of the multigrid method are identified. A special grid restriction procedure is also introduced, which accommodates the velocity characteristics in the wall region and kelps improve the convergence rate.

Effect of composition change on temperature measurements in a premixed flame by holographic interferometry

Application of holographic interferometry to nonintrusive temperature measurements in flames has been demonstrated as a viable means in engineering practices. However, the effect of composition changes on the previous investigations was often either neglected or assumed by a simpler relationship owing to lack of the distribution information of gas composition. This study investigates the effect of composition changes on temperature measurements using holographic interferometry by analyzing the composition of combustion gas using a gas chromatograph and measuring temperatures with a thermocouple to provide a comparison basis. Results are given for axisymmetric, laminar, premixed propane-air flames. Fuel-rich flame is found to have complex flame structure, and the effect of composition changes on the reconstruction of interferometric temperatures has to be taken into account. On the other hand, the flame structure in the fuel-lean case is much simpler, and using the linear distribution of molar refractivity in reconstruction of interferometric temperatures yields satisfactory results.