Effect of improved accelerating method on efficient chemistry calculations in diesel engine

With detailed chemical kinetics being employed in combustion simulations, its major computational challenge is the time-intensive nature of chemical kinetics integration due to the large number of chemical species and wide range of chemical timescales involved. In this work, an extended tabulated dynamic chemistry approach with dynamic pruning method is carried out to simulate complex spray combustion for non-premixed combustion process. The thought of extended tabulated dynamic chemistry approach with dynamic pruning is achieved by selecting the optimum acceleration method as well as its error tolerances at different combustion stages depending on combustion characteristics involving the low-temperature combustion. The present method is applied to realistically complex combustion systems involving spray flame of n-heptane fuel and non-premixed combustion engine. Computation efficiency of the proposed method is compared with the results using different accelerating methods, including dynamical adaptive chemistry, in situ adaptive tabulation, and coupled method of tabulated dynamical adaptive chemistry. The results show that transient computational cost will decrease for low-temperature combustion by reducing ambient oxygen concentration clearly in spray flame. Meanwhile, very low computational efficiency is presented once the autoignition occurs, especially at the initial oxygen concentration of 21%. Based on the feature, extended tabulated dynamic chemistry approach with dynamic pruning with different dynamic adaptive chemistry error tolerances is proposed to improve computational efficiency. Extended tabulated dynamic chemistry approach with dynamic pruning with larger error tolerance ε D A C = 0 . 2 improves around two times for decreased amplitude of transient computational cost at high-temperature combustion stage, and at the same time, the computational accuracy is also improved by comparing the important intermediate species obtained by direct integration. For applications in diesel engine, the results show that extended tabulated dynamic chemistry approach with dynamic pruning can accurately capture the first-stage ignition feature that determines the high-temperature combustion stage. In addition, extended tabulated dynamic chemistry approach with dynamic pruning with the smaller in situ adaptive tabulation error tolerance of 0.001 only used at the high-temperature combustion stage significantly improves the performance on diesel engine simulation with a larger chemistry mechanism. The present method further significantly improves computational efficiency with an overall speedup factor of 10 with high-accuracy compared with result using direct integration.