Exploring the synergistic potential of response surface methodology based multi-objective optimization in the performance-emission-stability trade-off envelope of an existing diesel engine

The present study reveals synergistic possibilities of split-injection techniques under different exhaust gas recirculation (EGR) profiles to meet the emission-performance-stability trade-off from existing diesel engines under the Bharat Stage constraints. The coefficient of variance for indicated mean effective pressure (COVIMEP) and Exergy efficiency was selected as engine stability and performance indicators. At the same time, emission responses are marked by nitrogen oxide and hydrocarbon-particulate matter (NHC-PM) profile. The study undertook a comprehensive design-of-experiment (DoE) effort to identify operational limitations and subsequently determine the parametric design space. Trade-off zones have been optimized by meta-model-based objective function formation and desirability maximization approach. It seeks to find a robust optimization by reducing the level of uncertainty and standard errors of estimation during model generation based on the DoE selection and response surface methodology (RSM) approach. The study attempts to create a niche in the archetypical DoE-RSM-based engine response calibration endeavors by invoking several robustness fitness metrics. From the pilot experimental results, EGR has shown its potential as a reducing agent of NHC emission. Whereas in the case of split injecting strategies, the NHC-PM-exergy trade-off can be observed. The response parameters subsequent to the best Pareto solution corresponding to the highest desirability obtained from the optimization study were registered 5.2 g/kW-hr of NOx, 0.9 g/kW-hr of Soot, 23% of Exergy efficiency with the COVIMEP of 4.9. Optimum predicted sets obtained from optimization strategies have been compared with the respective pilot experimentation and concluded that the multifactor at a time (MFAT) strategy through RSM showed a significantly better result than the one factor at a time (OFAT) strategy.