This study aims to confirm the effect of changing the various factors with the directed relation graph error propagation–based methods and to adopt a new approach of the mechanism evaluation in the reduction of biodiesel mechanism. The factors considered in this study were a threshold value, target species, ambient conditions, and the evaluation formula consists of the reduction rates and the maximum error rate of ignition delay to objectively compare the skeletal mechanisms generated under different conditions. For a threshold value, the automatic mechanism reduction process was used to select the appropriate threshold value by applying the relative tolerance and absolute tolerance; so relative tolerance and absolute tolerance represent the factor of the threshold value. Also, the seven steps of mechanism reduction process consist of directed relation graph error propagation, directed relation graph error propagation with sensitivity analysis, peak concentration analysis, full species sensitivity analysis, and A-factor modification. As a result of the mechanism reduction, different relative tolerance and absolute tolerance values should be applied to each step to select the appropriate threshold value. For target species, considering polycyclic aromatic hydrocarbon species as target species shows higher efficiency of mechanism reduction. Also, considering the negative temperature coefficient region as ambient conditions helps the mechanism be reduced efficiently than a wide range of ambient conditions. Finally, the reduced mechanism which had 247 species and 1129 reactions was generated, and the maximum error rate of ignition delay was about 30%. For the applicability of three-dimensional computational fluid dynamics and verification of the reduced mechanism, the compression ignition engine simulation was performed. As a result of three-dimensional computational fluid dynamics, the predicted cylinder pressure, rate of heat release, indicated mean effective pressure, and power were similar to the experimental results. However, the results of carbon monoxide and nitrogen oxide emissions did not match the experimental results.

中文翻译：

---

机理还原因子变化下生物柴油的机理还原与评价研究

本研究旨在通过基于有向关系图误差传播的方法确认改变各种因素的效果，并采用一种新的机制评估方法来减少生物柴油的机理。本研究考虑的因素是阈值、目标物种、环境条件，评估公式由点火延迟的减少率和最大错误率组成，以客观比较不同条件下产生的骨架机制。对于阈值，通过应用相对容差和绝对容差，使用自动机制减少过程来选择合适的阈值；所以相对容差和绝对容差代表阈值的因素。还，机理简化过程的七个步骤包括有向关系图误差传播、有向关系图误差传播与敏感性分析、峰值浓度分析、全物种敏感性分析和 A 因子修正。作为机制减少的结果，应该对每个步骤应用不同的相对容差和绝对容差值以选择合适的阈值。对于目标物种，将多环芳烃物种作为目标物种显示出更高的机理还原效率。此外，将负温度系数区域视为环境条件有助于比广泛的环境条件有效地减少机制。最后，产生了 247 个物种和 1129 个反应的还原机制，点火延迟的最大误差率为30%左右。为了三维计算流体动力学的适用性和简化机构的验证，进行了压燃式发动机模拟。作为三维计算流体动力学的结果，预测的气缸压力、放热率、指示平均有效压力和功率与实验结果相似。然而，一氧化碳和氮氧化物排放的结果与实验结果不符。放热速率、指示平均有效压力和功率与实验结果相似。然而，一氧化碳和氮氧化物排放的结果与实验结果不符。放热速率、指示平均有效压力和功率与实验结果相似。然而，一氧化碳和氮氧化物排放的结果与实验结果不符。