OPTIMIZING THE DESIGN OF A FLY WHEEL USING MACHINE LEARNING

Flywheels are an inertial storage device for energy. It is a mechanical energy absorber and acts as a storage device which stores energy whenever the energy supply is more than the demand, and then releases it when demand for energy exceeds the supply. The flywheel in machines functions as an accumulator, which stores energy when energy input is higher than the demand and releases it when there is a demand for energy higher than the energy input. The internal combustion engine is based on flywheels. The load placed on the flywheel grows and the stresses increase, so too do the loads and stress. The model of the steering wheel is designed using the CATIA tool, and then imported into ANSYS to be analyzed. The Finite Element Analysis is utilized to calculate the stress in the flywheel. The analysis of the flywheel was conducted on a single component. On the massive flywheel with cast iron (Ultimatestress-214Mpa Density-7510 kg/m3 Poisons Ratio-0.23) the stresses in the flywheel are analyzed and estimated. The web type also analyzes the same material. The third type studies the steering wheel wire analyzes the stress within the steering wheel and then compares the results of 3 steering wheels. The radio steering wheel was modelled with modeling software like CATIA and ANSYS and the results taken and subsequently an analysis of the exact direction of the steering wheel and the proper speed could be identified. Based on the results, machine learning technique i.e., a neural network program to study strain and stress that is known as Generalized Regression Neural Network (GRNN) was designed. This process involves defining certain input parameters (geo, speed and thickness) and output parameters that are pre-defined are immediately available. (weight, strain and stress)