Application of Automatic Target Species Selection and Global Sensitivity Analysis Methods in Chemical Mechanism Reduction

摘要

The reduced mechanisms driven from detailed mechanisms for fuel areessential to combustion simulation. The directed relation graph with errorpropagation (DRGEP) method for mechanism reduction determines theactive species based on the static target species. Inspired by the betterperformance of automatic target species selection method (ATSS) in dynamicadaptive chemistry simulation and the effectiveness of global sensitivityanalysis (GSA) in assessing the importance of elementary reaction, thepresent work proposes a new reduction procedure. This procedure utilizes theATSS method, which has not been applied to obtaining the overall reducedmechanism. Total 5610 status points are sampled densely during the ignitionprocess to obtain resulting local reduced mechanisms in present work. Theoverall active species are selected according to the total occurrenceprobability. The primary active reactions determined by the active species arefurther eliminated using the GSA method. By comparing the auto-ignitiondelay times (AIDT) of reduced methane mechanisms obtained by DRGEPand new reduction procedures under typical initial conditions. The simulationresults prove the superiority of the combination of ATSS and GSA methods.The ATSS method can avoid improper selection of active species comparedwith traditional DRGEP method.