Abstract A dynamic adaptive combustion modeling framework based on chemical explosive mode analysis (CEMA) is proposed to account for different flame features such as local auto-ignition, premixed and non-premixed flamelets in diesel spray flames. The proposed modeling strategy is achieved by assigning zone-dependent combustion models on-the-fly to different flame zones segmented using a CEMA-based approach. An approximate CEMA formulation is developed to approximate the eigenvalue of the chemical explosive mode with high computational efficiency in three-dimensional (3-D) turbulent flame simulations. The utility of the CEMA-based criterion for dynamic flame segmentation is first demonstrated using CEMA-based adaptive chemistry by applying different reduced chemistry to different flame zones. The capability of the dynamic adaptive combustion modeling strategy is then demonstrated in large eddy simulations (LES) of turbulent lifted n-dodecane spray flames. Specifically, inert mixing is used for chemically inactive zones, and the well-mixed combustion model with finite rate chemistry is applied in the pre-ignition zone to capture the two-stage ignition as well as premixed reaction fronts. Adaptive mesh refinement (AMR) is further adopted near the premixed reaction fronts to capture the local flame structure and flame propagation speed. For the post-ignition zone, a recently developed tabulated flamelet model (TFM) is applied and compared with the flamelet progress variable (FPV) method. It is shown that CEMA-based adaptive chemistry induces small errors to the statistically-averaged flame structures, as CEMA is an effective and robust approach for on-the-fly flame segmentation. It is further seen that the CEMA-based adaptive modeling strategy more accurately predicts the ignition delay time and flame lift-off length compared with the low-cost flamelet models such as TFM and FPV, while the computational cost is substantially lower compared with the well-mixed combustion model using finite rate chemistry.

中文翻译：

---

基于化学爆炸模态分析的喷射火焰动态自适应燃烧建模

摘要 提出了一种基于化学爆炸模式分析（CEMA）的动态自适应燃烧建模框架，以解决柴油喷射火焰中的局部自燃、预混和非预混火焰等不同火焰特征。所提出的建模策略是通过将依赖区域的燃烧模型动态分配到使用基于 CEMA 的方法分割的不同火焰区域来实现的。开发了近似 CEMA 公式，以在三维 (3-D) 湍流火焰模拟中以高计算效率近似化学爆炸模式的特征值。基于 CEMA 的动态火焰分割标准的效用首先使用基于 CEMA 的自适应化学进行演示，方法是将不同的还原化学应用于不同的火焰区域。然后在湍流提升的正十二烷喷雾火焰的大涡模拟 (LES) 中证明了动态自适应燃烧建模策略的能力。具体而言，惰性混合用于化学惰性区，具有有限速率化学的良好混合燃烧模型应用于预点火区以捕获两阶段点火以及预混反应前沿。在预混反应前沿附近进一步采用自适应网格细化 (AMR) 来捕获局部火焰结构和火焰传播速度。对于后点火区，应用了最近开发的表格火焰模型 (TFM) 并与火焰进程变量 (FPV) 方法进行了比较。结果表明，基于 CEMA 的自适应化学会对统计平均的火焰结构产生小误差，因为 CEMA 是一种有效且稳健的动态火焰分割方法。进一步看出，与TFM和FPV等低成本小火焰模型相比，基于CEMA的自适应建模策略更准确地预测点火延迟时间和火焰抬升长度，而计算成本与井相比大幅降低- 使用有限速率化学的混合燃烧模型。