Development of an ammonia-biodiesel dual fuel combustion engine's injection strategy map using response surface optimization and artificial neural network prediction

Abstract

Abstract The study intends to calibrate the compression ignition (CI) engine split injection parameters as efficiently. The goal of the study is to find the best split injection parameters for a dual-fuel engine that runs on 40% ammonia and 60% biodiesel at 80% load and a constant speed of 1500 rpm with the CRDi system. To optimize and forecast split injection settings, the RSM and an ANN model are created. Based on the experimental findings, the RSM optimization research recommends a per-injection timing of 54 °CA bTDC, a main injection angle of 19 °CA bTDC, and a pilot mass of 42%. As a result, in comparison to the unoptimized map, the split injection optimized calibration map increases BTE by 12.33% and decreases BSEC by 6.60%, and the optimized map reduces HC, CO, smoke, and EGT emissions by 15.68%, 21.40%, 18.82, and 17.24%, while increasing NOx emissions by 15.62%. RSM optimization with the most desirable level was selected for map development, and three trials were carried out to predict the calibrated map using ANN. According to the findings, the ANN predicted all responses with R > 0.99, demonstrating the real-time reproducibility of engine variables in contrast to the RSM responses. The experimental validation of the predicted data has an error range of 1.03–2.86%, which is acceptable.