Em processos de combustão, uma determinação precisa dos parâmetros envolvendo transferência de calor influencia diretamente os demais fenômenos envolvidos. Dentre os mecanismos de transferência de calor presentes na combustão a radiação térmica é predominante, mas sua correta determinação impõe uma elevada complexidade, principalmente quando se trata da solução de meios participantes. O cálculo envolve propriedades de absorção que variam com a temperatura e o comprimento de onda, sendo então necessária a utilização de modelos espectrais para obter bons resultados com um baixo tempo computacional. Para o cálculo da transferência radiante, existem diversos modelos espectrais, desde modelos de simples implementação, como, por exemplo, o GG (gás cinza) e o WSGG (soma-ponderada-dos-gases-cinza), até modelos com um grau elevado de detalhamento, como o SLW (soma-ponderada-dos-gases-cinza baseado em linhas espectrais) e o CW (número de onda cumulativa). Como os modelos com maior grau de detalhamento são de complexa implementação, alguns autores preferem empregar modelos simplistas, como o GG (gás cinza), apenas por questões de conveniência, mesmo em detrimento da qualidade dos resultados. Uma forma de executar o cálculo da radiação térmica sem simplificações é levar em conta as absorções em cada comprimento de onda, sendo esses cálculos denominados integração linha-por-linha (LBL), por executar o cálculo da transferência radiante em cada linha de absorção, o que gera resultados benchmark, podendo ser utilizados para avaliar os diversos modelos existentes. Este trabalho tem por objetivo verificar e sintetizar a aplicação dos modelos espectrais, em configurações envolvendo concentração e temperatura não uniformes, onde são realizados cálculos em um meio contendo CO2, H2O e fuligem. São avaliados os modelos GG, WSGG, SLW e CW. Dentre os modelos avaliados, o que apresenta os melhores resultados para as condições apresentadas é o modelo WSGG. De forma a aprimorar o modelo WSGG, uma nova implementação para a solução de misturas é apresentada, a qual apresenta correlações para o H2O e para o CO2 geradas individualmente, possibilitando misturas com qualquer razão de concentração, mostrando que o modelo apresenta bons resultados em diversas situações e é uma boa opção para a solução de problemas de combustão. / In combustion processes a good determination of the heat transfer parameters are of great importance because of its direct influence in the computation of the chemical reactions rate in the process and, consequently, in the formation of the combustion products. Among the processes of heat transfer in combustion, thermal radiation is predominant, and their determination can be a very complex task, especially with participating medium. The analysis involves absorption properties that vary with the temperature and wavelength, and therefore it is necessary to use spectral models to ensure good results with low computational time. There are several spectral models developed along the years, since the simplistic models such as the GG (gray gas) and WSGG (weighted-sum-of-gray-gases), to more advanced methods such as the SLW (spectral line weighted-sum-of-gray-gases) and CW (cumulative wavenumber). Due advanced models are in general a hard task to implement, the option is to use simplified models, for example the GG, even working with considerably errors. In order to quantify these solutions, for temperature and concentration conditions of the absorbing species, it is necessary to implement the radiation heat transfer taking into account the absorption at each wavelength through line-by-line (LBL) integration, being this solution the exact one, or, the benchmark solution, which it is used to evaluate the spectral models. In this study, the LBL integration is carried out to evaluate some of the existing models in a non-isothermal and inhomogeneous medium containing CO2, H2O and soot. The work involves the GG, WSGG, SLW and CW spectral models. For the presented cases, the best results occur with WSGG model. In order to improve the WSGG model a new implementation for the mixture solution is presented, which solves the correlations for H2O and CO2 generated individually, enabling mixtures containing any concentration ratio, showing the good agreement of the spectral model at any condition, being the WSGG a good option to solve combustion problems.

Em processos de combustão, uma determinação precisa dos parâmetros envolvendo transferência de calor influencia diretamente os demais fenômenos envolvidos. Dentre os mecanismos de transferência de calor presentes na combustão a radiação térmica é predominante, mas sua correta determinação impõe uma elevada complexidade, principalmente quando se trata da solução de meios participantes. O cálculo envolve propriedades de absorção que variam com a temperatura e o comprimento de onda, sendo então necessária a utilização de modelos espectrais para obter bons resultados com um baixo tempo computacional. Para o cálculo da transferência radiante, existem diversos modelos espectrais, desde modelos de simples implementação, como, por exemplo, o GG (gás cinza) e o WSGG (soma-ponderada-dos-gases-cinza), até modelos com um grau elevado de detalhamento, como o SLW (soma-ponderada-dos-gases-cinza baseado em linhas espectrais) e o CW (número de onda cumulativa). Como os modelos com maior grau de detalhamento são de complexa implementação, alguns autores preferem empregar modelos simplistas, como o GG (gás cinza), apenas por questões de conveniência, mesmo em detrimento da qualidade dos resultados. Uma forma de executar o cálculo da radiação térmica sem simplificações é levar em conta as absorções em cada comprimento de onda, sendo esses cálculos denominados integração linha-por-linha (LBL), por executar o cálculo da transferência radiante em cada linha de absorção, o que gera resultados benchmark, podendo ser utilizados para avaliar os diversos modelos existentes. Este trabalho tem por objetivo verificar e sintetizar a aplicação dos modelos espectrais, em configurações envolvendo concentração e temperatura não uniformes, onde são realizados cálculos em um meio contendo CO2, H2O e fuligem. São avaliados os modelos GG, WSGG, SLW e CW. Dentre os modelos avaliados, o que apresenta os melhores resultados para as condições apresentadas é o modelo WSGG. De forma a aprimorar o modelo WSGG, uma nova implementação para a solução de misturas é apresentada, a qual apresenta correlações para o H2O e para o CO2 geradas individualmente, possibilitando misturas com qualquer razão de concentração, mostrando que o modelo apresenta bons resultados em diversas situações e é uma boa opção para a solução de problemas de combustão. / In combustion processes a good determination of the heat transfer parameters are of great importance because of its direct influence in the computation of the chemical reactions rate in the process and, consequently, in the formation of the combustion products. Among the processes of heat transfer in combustion, thermal radiation is predominant, and their determination can be a very complex task, especially with participating medium. The analysis involves absorption properties that vary with the temperature and wavelength, and therefore it is necessary to use spectral models to ensure good results with low computational time. There are several spectral models developed along the years, since the simplistic models such as the GG (gray gas) and WSGG (weighted-sum-of-gray-gases), to more advanced methods such as the SLW (spectral line weighted-sum-of-gray-gases) and CW (cumulative wavenumber). Due advanced models are in general a hard task to implement, the option is to use simplified models, for example the GG, even working with considerably errors. In order to quantify these solutions, for temperature and concentration conditions of the absorbing species, it is necessary to implement the radiation heat transfer taking into account the absorption at each wavelength through line-by-line (LBL) integration, being this solution the exact one, or, the benchmark solution, which it is used to evaluate the spectral models. In this study, the LBL integration is carried out to evaluate some of the existing models in a non-isothermal and inhomogeneous medium containing CO2, H2O and soot. The work involves the GG, WSGG, SLW and CW spectral models. For the presented cases, the best results occur with WSGG model. In order to improve the WSGG model a new implementation for the mixture solution is presented, which solves the correlations for H2O and CO2 generated individually, enabling mixtures containing any concentration ratio, showing the good agreement of the spectral model at any condition, being the WSGG a good option to solve combustion problems.

Este trabalho analisa numericamente a transferência de calor radiativa em uma chama turbulenta de metano-ar. São resolvidas equações de conservação de massa, quantidade de movimento, energia, espécies químicas gasosas e fuligem, e variância da flutuação de temperatura em coordenadas cilíndricas axissimétricas. O modelo de combustão é o Eddy Break-Up – Arrhenius, com reação de combustão em duas etapas. O modelo de turbulência é o k −e padrão. A modelagem das interações turbulência-radiação (TRI - do inglês: Turbulence-Radiation Interactions) considera a “correlação combinada entre coeficiente de absorção e temperatura” e a “autocorrelação de temperatura”. O termo fonte de calor radiativo é calculado com o método de ordenadas discretas, considerando os modelos de gás cinza (GC) e da soma-ponderada-de-gases-cinza (WSGG – do inglês: weighted-sum-of-gray-gases) com correlações clássicas e recentes. O modelo linha-por-linha, considerado benchmark, também é empregado no cálculo daquele termo fonte, porém em cálculos desacoplados entre radiação e dinâmica de fluidos computacional (CFD - do inglês: Computational Fluid Dynamics), com o objetivo de avaliar os modelos WSGG e GC. Primeiramente, estudou-se o efeito da radiação térmica dos gases H2O e CO2 através dos modelos GC e WSGG, em cálculos acoplados radiação-CFD. Os resultados mostraram que os campos de temperatura e do termo fonte de calor radiativo, a transferência de calor para a parede da câmara e a fração radiativa, foram sensíveis aos diferentes modelos, enquanto o efeito sobre as concentrações das espécies foi de menor relevância para o modelo de combustão considerado. Os resultados obtidos com o modelo WSGG mais recente ficaram mais próximos dos dados experimentais da literatura, enquanto que a consideração das interações TRI melhorou esta concordância. As principais contribuições das interações TRI foram sobre a temperatura máxima e a fração radiativa, concordando com resultados da literatura. Os efeitos radiativos da fuligem juntamente com os gases também foram estudados, sendo importantes sobre o termo fonte de calor radiativo somente na região onde a fuligem estava presente (aumento de 30%). O fluxo de calor radiativo sobre a parede radial da câmara aumentou 25% na região de maior concentração de fuligem. A contribuição dos gases para a transferência radiativa foi de 92% e a da fuligem foi de 8%. Ao comparar os resultados dos modelos WSGG e GC com a solução benchmark, considerando o meio composto por gases, o modelo WSGG mais recente foi o que apresentou os melhores resultados (erro máximo 22,49%, médio 4,72%), enquanto ao considerar o meio composto por gases e fuligem, os erros foram menores (máximo 11,07%, médio 2,95%). / This work analyses numerically the thermal radiation heat transfer on a methane-air turbulent non-premixed flame. Conservation equations for mass, momentum, gaseous chemical species and soot, energy, and temperature variance, are solved in axisymmetric coordinates. The combustion model is Eddy Break-Up – Arrhenius, with two steps for the combustion reaction. Turbulence is modeled by standard k −e model. Consideration of TRI (Turbulence-Radiation Interactions) effects is made through a methodology that considers both cross-correlation between absorption coefficient and temperature and temperature self-correlation. The radiative heat source term is calculated with the discrete ordinates method, considering the gray gas model (GG) and the weighted-sum-of-gray-gases model (WSGG) based on classical and recent correlations. The benchmark solution obtained by the line-by-line model is also employed to calculate that source term, but in decoupled radiation-CFD (Computational Fluid Dynamics) calculations, with the objective of evaluating WSGG and GG models. Firstly, it was studied the effects of thermal radiation from the gases H2O and CO2 employing GG and WSGG models, and then the influence of TRI was also studied, both in coupled radiation- CFD calculations. Results pointed that temperature and radiative heat source fields, as well as wall heat transfer rates and radiative fraction, were significantly affected by thermal radiation, as well as by the different models and by TRI, while the influence on species concentrations was minor, for the combustion model employed. Numerical results obtained considering the recent WSGG model correlations were closer to experimental data from literature, and consideration of TRI into calculations improved that agreement. The main TRI contributions were the decrease on flame peak temperature and the increase on radiative fraction, in agreement with literature data. Radiative effects of the mixture of soot and gases were also studied, showing to be important for the radiative heat source only in the region with presence of soot (increase of 30%). Radiative heat flux on chamber wall increased 25% locally in the region with the highest soot concentration. Contribution of gases and soot for the net radiative transfer was 92% and 8%. Comparing the results obtained with WSGG and GG models with the benchmark solution in decoupled radiation-CFD calculations, considering the media composed by CO2 and H2O, the recent WSGG reached the best results (maximum error of 22.49%, average error of 4.72%), while considering the media composed by gases and soot, errors were reduced (maximum of 11.07% and average of 2.95%).

Em processos de combustão, uma determinação precisa dos parâmetros envolvendo transferência de calor influencia diretamente os demais fenômenos envolvidos. Dentre os mecanismos de transferência de calor presentes na combustão a radiação térmica é predominante, mas sua correta determinação impõe uma elevada complexidade, principalmente quando se trata da solução de meios participantes. O cálculo envolve propriedades de absorção que variam com a temperatura e o comprimento de onda, sendo então necessária a utilização de modelos espectrais para obter bons resultados com um baixo tempo computacional. Para o cálculo da transferência radiante, existem diversos modelos espectrais, desde modelos de simples implementação, como, por exemplo, o GG (gás cinza) e o WSGG (soma-ponderada-dos-gases-cinza), até modelos com um grau elevado de detalhamento, como o SLW (soma-ponderada-dos-gases-cinza baseado em linhas espectrais) e o CW (número de onda cumulativa). Como os modelos com maior grau de detalhamento são de complexa implementação, alguns autores preferem empregar modelos simplistas, como o GG (gás cinza), apenas por questões de conveniência, mesmo em detrimento da qualidade dos resultados. Uma forma de executar o cálculo da radiação térmica sem simplificações é levar em conta as absorções em cada comprimento de onda, sendo esses cálculos denominados integração linha-por-linha (LBL), por executar o cálculo da transferência radiante em cada linha de absorção, o que gera resultados benchmark, podendo ser utilizados para avaliar os diversos modelos existentes. Este trabalho tem por objetivo verificar e sintetizar a aplicação dos modelos espectrais, em configurações envolvendo concentração e temperatura não uniformes, onde são realizados cálculos em um meio contendo CO2, H2O e fuligem. São avaliados os modelos GG, WSGG, SLW e CW. Dentre os modelos avaliados, o que apresenta os melhores resultados para as condições apresentadas é o modelo WSGG. De forma a aprimorar o modelo WSGG, uma nova implementação para a solução de misturas é apresentada, a qual apresenta correlações para o H2O e para o CO2 geradas individualmente, possibilitando misturas com qualquer razão de concentração, mostrando que o modelo apresenta bons resultados em diversas situações e é uma boa opção para a solução de problemas de combustão. / In combustion processes a good determination of the heat transfer parameters are of great importance because of its direct influence in the computation of the chemical reactions rate in the process and, consequently, in the formation of the combustion products. Among the processes of heat transfer in combustion, thermal radiation is predominant, and their determination can be a very complex task, especially with participating medium. The analysis involves absorption properties that vary with the temperature and wavelength, and therefore it is necessary to use spectral models to ensure good results with low computational time. There are several spectral models developed along the years, since the simplistic models such as the GG (gray gas) and WSGG (weighted-sum-of-gray-gases), to more advanced methods such as the SLW (spectral line weighted-sum-of-gray-gases) and CW (cumulative wavenumber). Due advanced models are in general a hard task to implement, the option is to use simplified models, for example the GG, even working with considerably errors. In order to quantify these solutions, for temperature and concentration conditions of the absorbing species, it is necessary to implement the radiation heat transfer taking into account the absorption at each wavelength through line-by-line (LBL) integration, being this solution the exact one, or, the benchmark solution, which it is used to evaluate the spectral models. In this study, the LBL integration is carried out to evaluate some of the existing models in a non-isothermal and inhomogeneous medium containing CO2, H2O and soot. The work involves the GG, WSGG, SLW and CW spectral models. For the presented cases, the best results occur with WSGG model. In order to improve the WSGG model a new implementation for the mixture solution is presented, which solves the correlations for H2O and CO2 generated individually, enabling mixtures containing any concentration ratio, showing the good agreement of the spectral model at any condition, being the WSGG a good option to solve combustion problems.

Este trabalho analisa numericamente a transferência de calor radiativa em uma chama turbulenta de metano-ar. São resolvidas equações de conservação de massa, quantidade de movimento, energia, espécies químicas gasosas e fuligem, e variância da flutuação de temperatura em coordenadas cilíndricas axissimétricas. O modelo de combustão é o Eddy Break-Up – Arrhenius, com reação de combustão em duas etapas. O modelo de turbulência é o k −e padrão. A modelagem das interações turbulência-radiação (TRI - do inglês: Turbulence-Radiation Interactions) considera a “correlação combinada entre coeficiente de absorção e temperatura” e a “autocorrelação de temperatura”. O termo fonte de calor radiativo é calculado com o método de ordenadas discretas, considerando os modelos de gás cinza (GC) e da soma-ponderada-de-gases-cinza (WSGG – do inglês: weighted-sum-of-gray-gases) com correlações clássicas e recentes. O modelo linha-por-linha, considerado benchmark, também é empregado no cálculo daquele termo fonte, porém em cálculos desacoplados entre radiação e dinâmica de fluidos computacional (CFD - do inglês: Computational Fluid Dynamics), com o objetivo de avaliar os modelos WSGG e GC. Primeiramente, estudou-se o efeito da radiação térmica dos gases H2O e CO2 através dos modelos GC e WSGG, em cálculos acoplados radiação-CFD. Os resultados mostraram que os campos de temperatura e do termo fonte de calor radiativo, a transferência de calor para a parede da câmara e a fração radiativa, foram sensíveis aos diferentes modelos, enquanto o efeito sobre as concentrações das espécies foi de menor relevância para o modelo de combustão considerado. Os resultados obtidos com o modelo WSGG mais recente ficaram mais próximos dos dados experimentais da literatura, enquanto que a consideração das interações TRI melhorou esta concordância. As principais contribuições das interações TRI foram sobre a temperatura máxima e a fração radiativa, concordando com resultados da literatura. Os efeitos radiativos da fuligem juntamente com os gases também foram estudados, sendo importantes sobre o termo fonte de calor radiativo somente na região onde a fuligem estava presente (aumento de 30%). O fluxo de calor radiativo sobre a parede radial da câmara aumentou 25% na região de maior concentração de fuligem. A contribuição dos gases para a transferência radiativa foi de 92% e a da fuligem foi de 8%. Ao comparar os resultados dos modelos WSGG e GC com a solução benchmark, considerando o meio composto por gases, o modelo WSGG mais recente foi o que apresentou os melhores resultados (erro máximo 22,49%, médio 4,72%), enquanto ao considerar o meio composto por gases e fuligem, os erros foram menores (máximo 11,07%, médio 2,95%). / This work analyses numerically the thermal radiation heat transfer on a methane-air turbulent non-premixed flame. Conservation equations for mass, momentum, gaseous chemical species and soot, energy, and temperature variance, are solved in axisymmetric coordinates. The combustion model is Eddy Break-Up – Arrhenius, with two steps for the combustion reaction. Turbulence is modeled by standard k −e model. Consideration of TRI (Turbulence-Radiation Interactions) effects is made through a methodology that considers both cross-correlation between absorption coefficient and temperature and temperature self-correlation. The radiative heat source term is calculated with the discrete ordinates method, considering the gray gas model (GG) and the weighted-sum-of-gray-gases model (WSGG) based on classical and recent correlations. The benchmark solution obtained by the line-by-line model is also employed to calculate that source term, but in decoupled radiation-CFD (Computational Fluid Dynamics) calculations, with the objective of evaluating WSGG and GG models. Firstly, it was studied the effects of thermal radiation from the gases H2O and CO2 employing GG and WSGG models, and then the influence of TRI was also studied, both in coupled radiation- CFD calculations. Results pointed that temperature and radiative heat source fields, as well as wall heat transfer rates and radiative fraction, were significantly affected by thermal radiation, as well as by the different models and by TRI, while the influence on species concentrations was minor, for the combustion model employed. Numerical results obtained considering the recent WSGG model correlations were closer to experimental data from literature, and consideration of TRI into calculations improved that agreement. The main TRI contributions were the decrease on flame peak temperature and the increase on radiative fraction, in agreement with literature data. Radiative effects of the mixture of soot and gases were also studied, showing to be important for the radiative heat source only in the region with presence of soot (increase of 30%). Radiative heat flux on chamber wall increased 25% locally in the region with the highest soot concentration. Contribution of gases and soot for the net radiative transfer was 92% and 8%. Comparing the results obtained with WSGG and GG models with the benchmark solution in decoupled radiation-CFD calculations, considering the media composed by CO2 and H2O, the recent WSGG reached the best results (maximum error of 22.49%, average error of 4.72%), while considering the media composed by gases and soot, errors were reduced (maximum of 11.07% and average of 2.95%).

Este trabalho analisa numericamente a transferência de calor radiativa em uma chama turbulenta de metano-ar. São resolvidas equações de conservação de massa, quantidade de movimento, energia, espécies químicas gasosas e fuligem, e variância da flutuação de temperatura em coordenadas cilíndricas axissimétricas. O modelo de combustão é o Eddy Break-Up – Arrhenius, com reação de combustão em duas etapas. O modelo de turbulência é o k −e padrão. A modelagem das interações turbulência-radiação (TRI - do inglês: Turbulence-Radiation Interactions) considera a “correlação combinada entre coeficiente de absorção e temperatura” e a “autocorrelação de temperatura”. O termo fonte de calor radiativo é calculado com o método de ordenadas discretas, considerando os modelos de gás cinza (GC) e da soma-ponderada-de-gases-cinza (WSGG – do inglês: weighted-sum-of-gray-gases) com correlações clássicas e recentes. O modelo linha-por-linha, considerado benchmark, também é empregado no cálculo daquele termo fonte, porém em cálculos desacoplados entre radiação e dinâmica de fluidos computacional (CFD - do inglês: Computational Fluid Dynamics), com o objetivo de avaliar os modelos WSGG e GC. Primeiramente, estudou-se o efeito da radiação térmica dos gases H2O e CO2 através dos modelos GC e WSGG, em cálculos acoplados radiação-CFD. Os resultados mostraram que os campos de temperatura e do termo fonte de calor radiativo, a transferência de calor para a parede da câmara e a fração radiativa, foram sensíveis aos diferentes modelos, enquanto o efeito sobre as concentrações das espécies foi de menor relevância para o modelo de combustão considerado. Os resultados obtidos com o modelo WSGG mais recente ficaram mais próximos dos dados experimentais da literatura, enquanto que a consideração das interações TRI melhorou esta concordância. As principais contribuições das interações TRI foram sobre a temperatura máxima e a fração radiativa, concordando com resultados da literatura. Os efeitos radiativos da fuligem juntamente com os gases também foram estudados, sendo importantes sobre o termo fonte de calor radiativo somente na região onde a fuligem estava presente (aumento de 30%). O fluxo de calor radiativo sobre a parede radial da câmara aumentou 25% na região de maior concentração de fuligem. A contribuição dos gases para a transferência radiativa foi de 92% e a da fuligem foi de 8%. Ao comparar os resultados dos modelos WSGG e GC com a solução benchmark, considerando o meio composto por gases, o modelo WSGG mais recente foi o que apresentou os melhores resultados (erro máximo 22,49%, médio 4,72%), enquanto ao considerar o meio composto por gases e fuligem, os erros foram menores (máximo 11,07%, médio 2,95%). / This work analyses numerically the thermal radiation heat transfer on a methane-air turbulent non-premixed flame. Conservation equations for mass, momentum, gaseous chemical species and soot, energy, and temperature variance, are solved in axisymmetric coordinates. The combustion model is Eddy Break-Up – Arrhenius, with two steps for the combustion reaction. Turbulence is modeled by standard k −e model. Consideration of TRI (Turbulence-Radiation Interactions) effects is made through a methodology that considers both cross-correlation between absorption coefficient and temperature and temperature self-correlation. The radiative heat source term is calculated with the discrete ordinates method, considering the gray gas model (GG) and the weighted-sum-of-gray-gases model (WSGG) based on classical and recent correlations. The benchmark solution obtained by the line-by-line model is also employed to calculate that source term, but in decoupled radiation-CFD (Computational Fluid Dynamics) calculations, with the objective of evaluating WSGG and GG models. Firstly, it was studied the effects of thermal radiation from the gases H2O and CO2 employing GG and WSGG models, and then the influence of TRI was also studied, both in coupled radiation- CFD calculations. Results pointed that temperature and radiative heat source fields, as well as wall heat transfer rates and radiative fraction, were significantly affected by thermal radiation, as well as by the different models and by TRI, while the influence on species concentrations was minor, for the combustion model employed. Numerical results obtained considering the recent WSGG model correlations were closer to experimental data from literature, and consideration of TRI into calculations improved that agreement. The main TRI contributions were the decrease on flame peak temperature and the increase on radiative fraction, in agreement with literature data. Radiative effects of the mixture of soot and gases were also studied, showing to be important for the radiative heat source only in the region with presence of soot (increase of 30%). Radiative heat flux on chamber wall increased 25% locally in the region with the highest soot concentration. Contribution of gases and soot for the net radiative transfer was 92% and 8%. Comparing the results obtained with WSGG and GG models with the benchmark solution in decoupled radiation-CFD calculations, considering the media composed by CO2 and H2O, the recent WSGG reached the best results (maximum error of 22.49%, average error of 4.72%), while considering the media composed by gases and soot, errors were reduced (maximum of 11.07% and average of 2.95%).

The present study comprised an intensive investigation of the two newly proposed parameterisation forms for the wind input source term S[subscript]in (Donelan et a1., 2006) and the wave dissipation source term S[subscript]ds (Young and Babanin, 2006) proposed on the basis of the recent experimental findings at Lake George, New South Wales, Australia in 1997-2000. The main objective of this study was to obtain advanced spectral forms for the wind input source function S[subscript]in and wave spectral dissipation source function S[subscript]ds, which satisfy important physical constraints. A new approach was developed to achieve the objectives of this study, within the strong physical framework. This approach resulted in a new balance scheme between the energy source terms in the wave model, mentioned before as the split balance scheme (Badulin, 2006). The wave-induced stress was defined as the main physical constraint for a new wave model including recently suggested source functions for the wind input and wave dissipation source terms. Within this approach, a new methodology was developed for correction of the wind input source function S[subscript]in. Another important physical constraint was the consistency between the wave dissipation and the wind energy input to the waves. The new parameter, the dissipation rate, R, was introduced in this study, as the ratio of the wave dissipation energy to the wind input energy. The parameterisation form of the dissipation rate is presented as a function of the inverse wave age U ₁₀ / c[subscript]p Some aspects of wave spectral modelling regarding the shape of the wave spectrum and spectral saturation were revised. The two-phase behaviour of the spectral dissipation function was investigated in terms of the functional dependency of the coefficients a for the inherent wave breaking term and b for the forced dissipation term. The present study found that the both coefficients have functional dependence on the inverse wave age U ₁₀ / c[subscript]p and the spectral frequency. Based on the experimental data by Young and Babanin (2006), a new directional spreading function of bimodal shape was developed for the wave dissipation source term. The performance of the new spectral functions of the wind input S[subscript]in(f) and the wave dissipation S[subscript]ds(f) source terms was assessed using a new third-generation two-dimensional research wave model WAVETIME-I. The model incorporating the corrected source functions was able to reproduce the existing experimental data. / Thesis (Ph.D.) -- University of Adelaide, School of Civil, Environmental and Mining Engineering, 2009

The present study comprised an intensive investigation of the two newly proposed parameterisation forms for the wind input source term S[subscript]in (Donelan et a1., 2006) and the wave dissipation source term S[subscript]ds (Young and Babanin, 2006) proposed on the basis of the recent experimental findings at Lake George, New South Wales, Australia in 1997-2000. The main objective of this study was to obtain advanced spectral forms for the wind input source function S[subscript]in and wave spectral dissipation source function S[subscript]ds, which satisfy important physical constraints. A new approach was developed to achieve the objectives of this study, within the strong physical framework. This approach resulted in a new balance scheme between the energy source terms in the wave model, mentioned before as the split balance scheme (Badulin, 2006). The wave-induced stress was defined as the main physical constraint for a new wave model including recently suggested source functions for the wind input and wave dissipation source terms. Within this approach, a new methodology was developed for correction of the wind input source function S[subscript]in. Another important physical constraint was the consistency between the wave dissipation and the wind energy input to the waves. The new parameter, the dissipation rate, R, was introduced in this study, as the ratio of the wave dissipation energy to the wind input energy. The parameterisation form of the dissipation rate is presented as a function of the inverse wave age U ₁₀ / c[subscript]p Some aspects of wave spectral modelling regarding the shape of the wave spectrum and spectral saturation were revised. The two-phase behaviour of the spectral dissipation function was investigated in terms of the functional dependency of the coefficients a for the inherent wave breaking term and b for the forced dissipation term. The present study found that the both coefficients have functional dependence on the inverse wave age U ₁₀ / c[subscript]p and the spectral frequency. Based on the experimental data by Young and Babanin (2006), a new directional spreading function of bimodal shape was developed for the wave dissipation source term. The performance of the new spectral functions of the wind input S[subscript]in(f) and the wave dissipation S[subscript]ds(f) source terms was assessed using a new third-generation two-dimensional research wave model WAVETIME-I. The model incorporating the corrected source functions was able to reproduce the existing experimental data. / Thesis (Ph.D.) -- University of Adelaide, School of Civil, Environmental and Mining Engineering, 2009

The present study comprised an intensive investigation of the two newly proposed parameterisation forms for the wind input source term S[subscript]in (Donelan et a1., 2006) and the wave dissipation source term S[subscript]ds (Young and Babanin, 2006) proposed on the basis of the recent experimental findings at Lake George, New South Wales, Australia in 1997-2000. The main objective of this study was to obtain advanced spectral forms for the wind input source function S[subscript]in and wave spectral dissipation source function S[subscript]ds, which satisfy important physical constraints. A new approach was developed to achieve the objectives of this study, within the strong physical framework. This approach resulted in a new balance scheme between the energy source terms in the wave model, mentioned before as the split balance scheme (Badulin, 2006). The wave-induced stress was defined as the main physical constraint for a new wave model including recently suggested source functions for the wind input and wave dissipation source terms. Within this approach, a new methodology was developed for correction of the wind input source function S[subscript]in. Another important physical constraint was the consistency between the wave dissipation and the wind energy input to the waves. The new parameter, the dissipation rate, R, was introduced in this study, as the ratio of the wave dissipation energy to the wind input energy. The parameterisation form of the dissipation rate is presented as a function of the inverse wave age U ₁₀ / c[subscript]p Some aspects of wave spectral modelling regarding the shape of the wave spectrum and spectral saturation were revised. The two-phase behaviour of the spectral dissipation function was investigated in terms of the functional dependency of the coefficients a for the inherent wave breaking term and b for the forced dissipation term. The present study found that the both coefficients have functional dependence on the inverse wave age U ₁₀ / c[subscript]p and the spectral frequency. Based on the experimental data by Young and Babanin (2006), a new directional spreading function of bimodal shape was developed for the wave dissipation source term. The performance of the new spectral functions of the wind input S[subscript]in(f) and the wave dissipation S[subscript]ds(f) source terms was assessed using a new third-generation two-dimensional research wave model WAVETIME-I. The model incorporating the corrected source functions was able to reproduce the existing experimental data. / Thesis (Ph.D.) -- University of Adelaide, School of Civil, Environmental and Mining Engineering, 2009

La simulation de la combustion turbulente connait un nouvel essor avec l'introduction de la Simulation aux Grandes Échelles (SGE) qui permet de prédire l'évolution in stationnaire de l'écoulement réactif turbulent. Dans ce contexte la prise en compte du rayonnement soulève des questions d'ordre a la fois fondamental et pratique. En effet les processus physiques du rayonnement et de la combustion sont de nature radicalement différente : la combustion est contrôlée par des échanges locaux sur une durée finie, alors que le rayonnement est instantané et fait intervenir des échanges a distance. En premier lieu il convient de s'interroger sur l'impact de la modélisation SGE de la combustion turbulente sur le rayonnement. Cette question est traitée dans le cadre plus général de l'interaction rayonnement-turbulence. A partir d'études théoriques et numériques, il est montre que cette interaction est faible et qu'une solution SGE peut être directement utilisée pour un calcul radiatif, sans modélisation supplémentaire. Il s'agit ensuite de mettre en place de façon pratique le couplage in stationnaire rayonnement-combustion turbulente. Un point clé est la réduction du temps de calcul pour le rayonnement, et diverses stratégies sont proposées. En particulier un nouveau modèle spectral est introduit, utilisant une technique de tabulation et garantissant un niveau de précision suffisant. Le temps de calcul radiatif a ainsi été réduit de deux ordres de grandeur, permettant la réalisation d'un calcul couple sur une configuration de flamme pré-melangée turbulente. / Simulation of turbulent combustion has gained high potential with the Large Eddy Simulation (LES) approach, allowing to predict unsteady turbulent reactive flows. In this context, taking into account radiation rises new fundamental and practical questions. Indeed the physics involved in radiation and in combustion are completely different : combustion is controlled by local exchanges and finite times whereas radiation is instantaneous and is based on non-local exchanges. In a first step, the impact of LES modelling of turbulent combustion on radiation is regarded. This question is treated in the more general frame of the turbulence-radiation interaction. From theoretical and numerical studies, it is shown that this interaction is weak in the LES context so that LES solutions can be directly coupled to radiative calculations, without further modelling. Then the unsteady coupling of radiation and turbulent combustion is realised. A key point is the reduction of calculation time of radiation, and several strategies are proposed. In particular a new global spectral model is introduced, using a tabulation technique and ensuring a sufficient level of accuracy. The radiative time calculation is finally decreased by two orders of magnitude, enabling the realization of a coupled calculation of a turbulent premixed flame