Predictive Chemical and Statistical Modeling of Particulate Matter Formation in Turbulent Combustion with Application to Aircraft Engines

Dr. Heinz Pitsch | Stanford University

WP-1574

Objective

WP-1574 Project Graphic

Large-eddy simulation of a modern Pratt & Whitney gas turbine combustor. The combustor bulkhead is to the left of the flame. Fuel and air enter the combustor through the injector/swirler assembly, which has three different air passages. Fuel droplets are shown in green. The remaining color representation shows iso-surfaces of the temperature.

Soot formation in aircraft engines is of great concern due to the health and environmental impacts of soot. The design of aircraft propulsion systems would benefit from a computational methodology that can accurately predict soot emissions. This project sought to develop these simulation capabilities in the framework of large eddy simulation (LES). The complexity of the soot formation process and its strong interactions with gas-phase combustion chemistry and turbulence necessitated a comprehensive approach involving three key research areas: chemical modeling of soot surface reactions, especially oxidation, statistical modeling of soot particle distributions, and soot modeling in LES of turbulent combustion.

Back to Top

Technical Approach

Soot growth and oxidation processes were investigated using theoretical approaches applicable to large reaction systems. The work focused on the thermodynamics and kinetics of aromatic oxyradicals, the key intermediates in oxidation of soot.

A novel moment method called the hybrid method of moments (HMOM) was developed for predicting moments of the soot particle number density function. This formulation described soot particle size using both volume and surface area and is able to account for bimodal distributions of soot particles sizes. An integrated LES model for soot evolution in turbulent nonpremixed  flames was developed. The model components include a soot model based on HMOM, an extended flamelet/progress variable gas-phase combustion model, and a model to account for the slow chemistry of soot precursors. The development of these modeling approaches was aided by the results of a first-of-its-kind direct numerical simulation (DNS) study of soot nucleation and growth using a polycyclic aromatic hydrocarbon (PAH) based model for soot inception.

Back to Top

Results

The principal conclusion of the soot oxidation study is that only the outer rings of aromatic oxyradicals are able to undergo oxidation, whereas the inner rings resist oxidation in combustion environments. The researchers also found that the variability of thermodynamic stability and decomposition rate of aromatic oxyradicals can be explained and correlated with substrate aromaticity. These findings will support the development of practical rules for predicting oxidation rate coefficients for aromatics, thereby enhancing models of soot oxidation.

The LES soot model was validated in a laboratory scale natural gas jet diffusion flame. Compared to experimental measurements, the LES provided a reasonable prediction of the maximum soot volume fraction. Factors extrinsic to the soot model formulation were found to be major sources of error and uncertainty in soot prediction. These factors include modeling of small scale gas-phase mixing rates and specification of the kinetic mechanism for PAH formation and growth.

Back to Top

Benefits

The overall outcome of this project is a significant advancement in the sophistication of soot modeling approaches for LES. Using these approaches, the superiority of LES to Reynolds-averaged Navier Stokes (RANS) methods for soot modeling was demonstrated. The LES model permits the effects of turbulence and gas-phase chemistry on soot formation to be assessed qualitatively. Sources of error in soot prediction were clearly identified, allowing future modeling studies to be targeted to these areas.

Back to Top

Points of Contact

Principal Investigator

Dr. Heinz Pitsch

Stanford University

Phone: 650-736-1995

Fax: 650-725-3525

Share