Modelling and simulation of soot generation and transport
1
Citation
0
Reference
20
Related Paper
Citation Trend
Abstract:
Soot released from fires not only causes danger to lives and property damage, but also effects fire spread by altering the radiation characteristics of fire effluents. In many situations, it is the soot concentration that controls the fire development. Therefore, soot modelling is of great importance in fire safety science. This necessitates the development of a global and general soot model within fire field models that can simulate the amount of soot generated and transported in large-scale fires in order to obtain an accurate soot concentration distribution within the building.
A soot transport model, called Multi-Particle-Size (MPS) model, has been developed in this study to improve the prediction of soot particle behaviour during transportation by considering the uneven soot mass size distributions and gravitational settling force on soot particles. The efficiency of the MPS model was investigated by simulating soot movements in three real experiments. The first two validation experiments were cable fires in a large-scale enclosed corridor and the third experiment analysed the soot produced from a soot generator in a warehouse with a high ceiling. The soot layers predicted by the MPS model matched the measurements/observation better than that from the Conventional Model in which the soot generation is modelled with a constant soot yield (CY) value and soot particles are treated as a gaseous combustion product.
A global soot generation model, called Beta soot generation (BSG) model has also been developed for non-premixed laminar flames. By making use of the characteristics of the beta function, the model has been extended to turbulent flames in the pre-scribed probability density function (PDF) approach with low cost in terms of computational resources. The model was validated by two turbulent methane and ethylene pool fires. The simulation results demonstrated that the soot volume fractions produced by the BSG model were in good agreement with the experimental data.
Further, the two new models have been integrated into a single soot model called BSG+MPS model. The performance of the model was examined by predicting the soot generation and transport in a large-scale enclosed corridor. The BSG+MPS model improved the prediction of soot concentration distribution in the corridor compared with the CY +MPS model.
Finally, the entire work is summarised and future work is suggested.Cite
Lagrangian particle tracking
Large-Eddy Simulation
Particle (ecology)
Cite
Citations (29)
Large-Eddy Simulation
Cite
Citations (17)
Cite
Citations (21)
Particle (ecology)
Number density
Cite
Citations (26)
Particle (ecology)
Cite
Citations (1)
Discrete sectional method (DSM)-based soot models are computationally demanding since several transport equations are required to be solved for sectional soot variables using large chemical kinetic mechanisms. In this paper, the sectional soot model is coupled with Flamelet Generated Manifold (FGM) tabulated chemistry to achieve computationally efficient chemistry reduction for combustion simulations. Different approaches are explored for incorporating soot-gas phase coupling with FGM chemistry, and a comprehensive assessment of these FGM-DSM approaches is conducted for their predictive accuracy and computational performance in simulations of laminar ethylene counterflow flames. The accuracy of soot prediction with FGM chemistry is found to be sensitive to the tabulated concentrations of PAH species. An unaccounted consumption of PAH originating from PAH-based processes leads to significant overprediction of soot. In this context, the performance of two strategies (i) solving the transport equation of PAH species (ii) including a contribution of soot nucleation during the manifold generation stage is compared. The latter approach showed relatively better accuracy and computational efficiency than the former. Numerical results further reveal that the strategy of including the complete soot model during the manifold generation stage reproduces the detailed chemistry solutions most accurately. The influence of unsteadiness on predictive capabilities of FGM-DSM approaches is also investigated by imposing time-dependent strain rates in unsteady simulations. The computational performance analysis indicates that by adopting FGM chemistry, up to two orders of magnitude reduction in CPU time can be achieved, based on the choice of section number and simulation approach. Promising results obtained for FGM-DSM strategies in one-dimensional configuration, provide a good outset for extending the application of FGM for soot estimation in turbulent flames.
Cite
Citations (11)
Extinction (optical mineralogy)
Entrainment (biomusicology)
Cite
Citations (18)
Cite
Citations (6)
Soot formation in combustion systems is a growing concern due to its adverse environmental and health effects. It is considered to be a tremendously complicated phenomenon which includes multiphase flow, thermodynamics, heat transfer, chemical kinetics, and particle dynamics. Although various numerical approaches have been developed for the detailed modeling of soot evolution, most industrial device simulations neglect or rudimentarily approximate soot formation due to its high computational cost. Developing accurate, easy to use, and computationally inexpensive numerical techniques to predict or estimate soot concentrations is a major objective of the combustion industry. In the present study, a supervised Artificial Neural Network (ANN) technique is applied to predict the soot concentration fields in ethylene/air laminar diffusion flames accurately with a low computational cost. To gather validated data, eight different flames with various equivalence ratios, inlet velocities, and burner geometries are modeled using the CoFlame code (a computational fluid dynamics (CFD) parallel combustion and soot model) and the Lagrangian histories of soot-containing fluid parcels are computed and stored. Then, an ANN model is developed and optimized using the Levenberg-Marquardt approach. Two different scenarios are introduced to validate the network performance; testing the prediction capabilities of the network for the same eight flames that are used to train the network, and for two new flames that are not within the training data set. It is shown that for both of these cases the ANN is able to predict the overall soot concentration field very well with a relatively low integrated error.
Cite
Citations (22)
The emission of NOx from aero-gas turbine combustors, which in the present generation of
designs consists mainly of thermal NOx ' is of great concern due to its potential damage to the
stratospheric ozone layer. Soot production in gas turbine combustors is also undesirable since
it is both the major source of exhaust smoke and, more importantly, the principal agent in
thermal radiation to the combustor liner. Furthermore thermal radiation from the soot redistributes
energy in the combustor, modifying the temperature field. This consequently
affects the production of other pollutants, notably that of thermal NOx> since the production
rate is especially sensitive to temperature.
Mathematical models for predicting gas turbine combustor emissions can be divided, in
general terms, into two main groups, Methods based on zonal (or modular) approach and on
CFD modelling. CFD modelling allows the use of computation intensive multi-dimensional
Navier-Stokes codes but cannot account for detailed chemistry which is responsible for
emissions. On the other hand, although the modular approaches make significant assumptions
about the mean flowfield and mixing, they employ detailed chemical kinetics.
The work reported in this thesis seeks to develop a model for emission predictions in the gas
turbine combustor which combines the advantage of both the modular approach and CFD
modelling. The strategy was based on a pdf calculation using the Monte-Carlo simulation
technique because the chemical source term is in closed form for the approach and the
solution procedure requires a CFD based calculation. Averaging of the particle properties was
on an extended zonal or planar basis in order to reduce computational effort. The predictions
are evaluated against available experimental results and other predictions employing more
conventional approaches.
Since the pdf method allows the modelling of slow chemistry and simultaneous influence of
multiple scalars, the thermal NO x production rate was implemented considering the effect of
NO concentration itself. Predicted exit NOx concentration was higher than the measured exit
level. It has been thought that this discrepancy is mainly due to neglecting radioactive heat loss
for temperature calculations.
The modelling of soot formation and oxidation has proved more problematic since the
assumption that soot is simply perturbation to the gaseous field, analogous to the NO
concentration, and temperature may be accurately described by single adiabatic flamelet are
no longer valid at elevated pressure and temperature conditions. Soot bum-out is
under-predicted. The computed mean soot oxidation is less than 10% of the maximum
production levels, even when OH is considered to be oxidising species in addition to
O2 ••
Although high soot formation rate was predicted as a result of neglecting radioactive loss and
using single perturbed flamelet calculation, the main uncertainties come from instantaneous
soot oxidation rate and the particle size effect which influence the…
Cite
Citations (1)