Machine learning for predictive modelling of the performance of automobile engine operating without coolant thermostat

Abstract

In tropical regions with relatively high temperatures, it is common practice for automobile dealers of preowned vehicles, usually imported from Europe and America, and roadside mechanics to remove their engine coolant thermostats to prevent overheating. This study investigates the effect of coolant thermostat removal on the performance of an automobile engine and further develops predictive models to forecast engine performance operating with and without coolant thermostat. A 2ZR engine mounted in a Toyota Corolla was used as the experimental specimen to achieve this objective. Several parameters were measured, and others were computed from the measured data. Results show that the automobile engine's average coolant and transmission oil temperature decreased by 31.69% and 28.90%, respectively when the thermostat was removed. Furthermore, the specific fuel consumption and carbon monoxide emission increased by 62.96% and 737%, respectively, when the engine operated without a coolant thermostat. The indicated thermal and brake thermal efficiencies decreased by 14.60% and 12.02%, respectively. The practice of removing automobile engine coolant thermostats renders the engine inefficient, uneconomical, and counterproductive.