A novel method to apply artificial neural network (ANN) for both chemical kinetics reduction and source term evaluation is introduced and tested in direct numerical simulation (DNS) and large eddy simulation (LES) of reactive flows. To gather turbulence affected flame data for ANN training, a new computation-economical method, called 1D pseudo-velocity disturbed flame (PVDF), is developed and used to generate thermo-chemical states independent of the modeled flame. Then a back-propagation ANN is trained using scaled conjugate gradient algorithm to memorize the sample states with reduced orders. The new method is employed in DNS and LES modeling of H 2 /air and C 3 H 8 /air premixed flames experiencing various levels of turbulence. The test result shows that compared to traditional computation with full mechanism and direct integration, this method can obtain quite large speed-ups with adequate prediction accuracy. [7,8] are often used, while ANN method, a extremely new systemic technique, is seldom seen in previous work. The procedure of using QSS assumption has been shown in Figure 1. However, QSS species concentrations should be first resolved which includes a large amount of algebraic iterations. Therefore, the net efficiency could be undermined [9], and also calculation may be failed if iterations do not converge.
ANN, kinetics reduction, LES, DNSAs for the second respect, several approaches to accelerate chemical sources evaluation have already been presented, such as look-up table (LUT) [10] and in situ-adaptive tabulation (ISAT) [11]. However, they are both based on tabulation technique which needs huge memory and a large number of check-up and interpolation operations. Artificial neural network (ANN) method, although it was proposed previously to handle above drawbacks, has got great progress through Sen and Menon's work [12,13]. It has been successfully applied to account for chemical kinetics in LES modeling of turbulent premixed flame with speed-ups even more than 10.In this paper, we tend to extend the application of ANN method, using it to not only calculate chemical sources but also reduce the detailed mechanism. The initial idea is to construct the direct mapping between non-QSS species concentrations and their reaction rates at plenty of thermal