James C. Sutherland, Babak Goshayeshi


Overview


High-fidelity simulation of turbulent combustion is an expensive endeavor because of the range of length and time scales present and the complexity of the physical phenomena involved (nonlinear coupling between mixing and reaction). LES resolves large length and time scales while modeling unresolved scales. LES also makes significant reductions in the complexity of the thermochemistry to achieve tractability.

Direct Numerical Simulation (DNS) resolves all length and time scales, but becomes computationally prohibitive for anything but the simplest problems at conditions that are not typically relevant to practical applications.

One-Dimensional Turbulence (ODT) is a modeling strategy that resolves the full range of length and time scales (as in DNS) but only in one dimension. The effect of turbulent mixing is modeled stochastically. Because of this, ODT can employ high-fidelity models and compare with simpler models (for use in LES).

 

Goal


The goal of the ODT-based research in coal combustion is to evaluate adequacy of various models for devolatilization, char-oxidation/gasification, soot formation, and gas-phase chemistry against available experimental data.

The ODT model will, thus, serve as a proving ground for models to determine which ones are most appropriate for use in the LES code.

 

Recent Results


Laminar, Single-Particle Combustion

Recently published results1 show the ability for the model to predict ignition delay times for single-particle coal combustion. The model is predictive for several coal types and captures observed trends with particle size and furnace temperature fairly well.

This study also considered the effect of the level of modeling on the ability to predict ignition delay. Some of these results are illustrated in figures 1 and 2, where the ignition delay for Pittsburgh coal (Figure 1) and Black Thunder coal (Figure 2) are shown as a function of the furnace temperature. The experimental data, together with simulation results for two different devolatilization models are shown. The chemical percolation devolatilization (CPD) model is consistently more accurate in predicting ignition delay over coal types and temperatures than the simpler two-step model (Kob).

 

fig1 fig2
Figure 1 - prediction of ignition delay for Pittsburg coal. Figure 2 - prediction of ignition delay for Black Thunder coal.

 

 

Turbulent, Multi-Particle Combustion


We have also recently investigated flame standoff predictions using ODT in a turbulent oxy-coal combustor. Figures 3 and 4 show the PDF of flame standoff observed experimentally and for three different devolatilization models for detailed chemistry (left) and flame-sheet chemistry (right). There are notable differences between the various model pairings. For example, the detailed chemistry - CPD pairing (highest fidelity model pairings) captures the flame standoff PDF quite well. On the other hand, pairing the two-step (TS) and flame-sheet models (very inexpensive models) also captures the experimentally observed flame standoff PDF well - likely due to compensating errors. These results are currently in preparation for submission to a journal.

 

fig3 fig4
Figure 3 - Flame standoff PDF predictions using detailed gas-phase chemistry and three different devolatilization models. Figure 4 - Flame standoff PDF predictions using flame-sheet gas-phase chemistry and three different devolatilization models.

 

1Goshayeshi, Babak, and James C Sutherland. "A comparison of various models in predicting ignition delay in single-particle coal combustion." Combustion and Flame (2014).