Experimental Study on the Ignition/Extinction Performances of a Central Staged Direct Mixing Combustor and the Construction of an Engineering Prediction Model

The ignition and extinction performances of an engine’s combustion chamber are crucial for its overall performance, reliability, fuel efficiency, and environmental impact. This study focuses on a central staged dual-swirl direct mixing combustor’s ignition and extinction performances. The goal is to construct a network surrogate prediction model estimating the ignition and extinction boundaries in the central staged dual-swirl direct mixing combustor. The experimental results indicate that the minimum ignition fuel–air ratio of the combustor initially increases and then decreases with increasing inlet temperature. In contrast, the extinction fuel–air ratio decreases with increasing inlet temperature. The ignition fuel–air ratio decreases as the inlet pressure increases, while the extinction fuel–air ratio also decreases with increasing inlet pressure. The minimum ignition fuel–air ratio initially decreases and then increases as the inlet flow rate increases. Similarly, the extinction fuel–air ratio decreases with increasing inlet flow rate. Additionally, both the ignition and extinction fuel–air ratios decrease with increasing pilot and main stage swirl intensity. The constructed neural network model comprises an output layer representing the ignition and extinction fuel–air ratios and an input layer including variables such as inlet temperature, pressure, flow rate, and pilot and main swirl intensity. During training, the model achieved a final relative error of 0.4%, with a relative error of 0.3% for the ignition fuel–air ratio and 0.5% for the extinction fuel–air ratio. During validation, the relative error was 1.7%, with a relative error of 1.1% for the ignition fuel–air ratio and 2.1% for the extinction fuel–air ratio. The neural network model demonstrates its effectiveness in accurately predicting the numerical values of the ignition and extinction fuel–air ratios, indicating its potential for engineering predictions in the context of central staged dual-swirl direct mixing combustors.