Analysis of the fractal characteristics for combustion instability in a premixed natural gas engine

To investigate the influence of gas injection timing (GIT) on the combustion instability of a premixed natural gas engine, experiments were conducted under low load conditions using various GITs. Multifractal and multiscale entropy analyses were employed to examine the fractal characteristics and complexity of the experimental time series for indicated mean effective pressure (IMEP) and heat release (Q) at different scales. Statistical analysis and return maps of the IMEP and Q time series were utilized to verify the results. The findings revealed that the combustion process of the natural gas engine demonstrates clear fractal characteristics at different scales. A strong correlation is found between the combustion instability and the fractal characteristics. Furthermore, the probability densities of the IMEP and Q time series exhibit super-Gaussian distributions. Retarding the GIT results in an initial increase, followed by a decrease in the difference value of the Hurst index and singular spectrum width. The mapping point distributions of the IMEP and Q time series initially disperse and subsequently concentrate. The fractal complexity and chaotic characteristics of combustion instability initially strengthen and then gradually diminish. Moreover, under lower load conditions, the anti-persistent correlation becomes more pronounced, and the intermittence and complexity of the fractal characteristics also intensify, signifying a more significant impact of GIT on the combustion instability of the natural gas engine. Notably, when the GIT is approximately 60 degrees CA after top dead center, the combustion process exhibits stronger fractal characteristics, accompanied by a greater dispersion degree of the mapping points. This study provides a theoretical basis for enhancing the lean-burn stability of natural gas engines.