Evaluating and optimizing of steam ejector performance considering heterogeneous condensation using machine learning framework

In the context of global warming and pollution concerns, refrigeration systems have become pivotal in energy conversion system. Within this realm, ejector types that harness renewable energy resources emerge as promising alternatives, offering a pathway towards environmentally conscious and resilient energy practices. Under specific conditions, condensation within the heat exchanger results in diverse droplet sizes at the ejector inlet, inducing homogeneous-heterogeneous condensation (HMTC) and heterogeneous condensation (HTC) phenomena. This study aims to evaluate and improve the performance of steam ejectors by investigating and optimizing the effects of homogeneous condensation (HMC), HTC, HMTC, and evaporation processes using a machine learning (ML) framework. The drone squadron optimization (DSO), Particle Swarm Optimization (PSO), and Genetic Algorithm (GA) are chosen and used in the ML framework to find the optimal droplet radius and number. Findings predict that the presence of 1018 1/kg droplets with a radius of 0.02 μm (Optimal mode) at the inlet results in a 2.6 % increase in the entrainment ratio (Er) and a 6.9 % reduction in the entropy generation compared to the baseline mode. Generally, the research reveals that HTC exhibits superior performance compared to prevailing theories, leading to enhanced ejector performance.