Three-dimensional thermo-hydrodynamic analysis of gas turbine combustion chambers is of great importance in the power generation industry to achieve higher efficiency and reduced emissions. However, it is prohibitive to use a comprehensive full-detailed mechanism in their simulation algorithms because of the huge CPU time and memory space requirements. Many reduction approaches are available in the literature to remedy this problem. Here a new approach is presented to reduce large detailed or skeletal mechanisms of oxidation of hydrocarbon fuels to a low-cost skeletal mechanism. The method involves an integrated procedure including a Sensitivity Analysis (SA) and a procedure of Gradual Evaluation of Ignition Error (GEIE). The sensitivity analysis identifies reactions which have less effect on the flame temperature (T_f) and also those with less effect on the NO concentration (X_NO). Using the GEIE procedure also identifies reactions that have less effect on the ignition delay time (τ_ign). In this process, three cutoff limits are selected for T_f, X_NO, and τ_ign. The procedure is validated and examined for two different hydrocarbon fuels, i.e., methane and kerosene. The detailed mechanism of GRI-3.0 is used for methane, to produce a low-cost skeletal mechanism containing 118 reactions and 39 species. Similarly, a validated skeletal mechanism for kerosene including 382 reactions and 106 species is used to generate a low-cost skeletal mechanism including only 180 reactions and 79 species. The accuracy of the obtained skeletal mechanisms was investigated to predict the ignition delay and the flame temperature for ranges of inlet temperatures (T₀) of 1000–1800 K, combustion pressures (p_c) of 1.0–30.0 atm, and equivalence ratios (ϕ) of 0.5–2.0 using a homogeneous IGNITION model. In addition, the applicability of the produced mechanisms to predict oxidation parameters such as flame temperature, velocity of burnt gas, concentration of the main fuel species, some minor radicals, and other selected species was investigated and validated for both skeletal mechanisms using homogeneous models PSR and PREMIXED over a range of different T₀ (300–1800 K), p_c (1.0–30.0 atm), and ϕ values (0.5–2.0). Comparisons show that the two new skeletal mechanisms have a good agreement with similar known base mechanisms but offer a significant gain in terms of computational cost.

中文翻译：

---

使用敏感性分析和逐步评估点火延迟误差来产生精确的低成本骨架机制，用于高温条件下烃类燃料的氧化

燃气轮机燃烧室的三维热流体动力学分析对于发电行业实现更高的效率和减少排放至关重要。但是，由于其巨大的CPU时间和内存空间要求，因此在其仿真算法中禁止使用全面的详细机制。文献中提供了许多减少方法来解决此问题。在这里，提出了一种新的方法，可以将碳氢化合物燃料的大型详细或骨架氧化机理降低为低成本的骨架机理。该方法涉及包括敏感性分析（SA）和点火误差逐步评估（GEIE）的过程的综合过程。敏感性分析确定了对火焰温度影响较小的反应（T_f）以及对NO浓度影响较小的那些（X _NO）。使用GEIE程序还可以识别对点火延迟时间（_τign）影响较小的反应。在此过程中，为T _f，X _NO和_τign选择了三个截止极限。。对于两种不同的碳氢燃料，即甲烷和煤油，对该程序进行了验证和检查。GRI-3.0的详细机理用于甲烷，以产生包含118个反应和39个物种的低成本骨架机理。类似地，已验证的包括382个反应和106种物质的煤油骨架机制被用于生成仅包含180个反应和79种物质的低成本骨架机制。研究了获得的骨架机制的准确性，以预测在入口温度（T ₀）为1000–1800 K，燃烧压力（p _c）的情况下的点火延迟和火焰温度。）在1.0–30.0 atm范围内，当量比（ϕ）在0.5-2.0之间（使用均质IGNITION模型）。此外，使用均相模型PSR，研究并验证了所产生机理对预测氧化参数（如火焰温度，燃烧气体的速度，主要燃料物质的浓度，一些次要自由基和其他选定物质）的适用性，并针对两种骨骼机理进行了验证。和PREMIXED在不同的T ₀（300–1800 K），p _c（1.0–30.0 atm）和ϕ值（0.5–2.0）的范围内。比较表明，这两种新的骨骼机制与类似的已知基本机制具有很好的一致性，但是在计算成本方面却有显着提高。