Optimization and Improvement of Gas Spring Design in An Energy Storing Prosthetic Knee

Abstract

This paper discusses the optimization and improvement of gas spring design. The gas spring is used as the suspension component of an energy storing prosthetic knee. The gas spring replaces the quadricep muscles of a transfemoral amputee. Deterministic and stochastic optimizations are proposed in this research. Both models are used to determine the optimal design variables of the gas spring: cylinder diameter, cylinder length, extension stroke, and compression stroke. The optimal design variables that result from the deterministic optimization model must be further analyzed to determine the effect of their variation from the objective function. A Monte Carlo simulation is used to determine the effect of such variations and make improvements when necessary. The process capability index (Cp) is used as a criteria to make this improvement while considering the contribution in variation of the design variables to the objective function. Stochastic optimization is proposed to find the optimal design variables by taking into consideration the randomness of its parameters. The objective function of stochastic optimization is to maximize the capability process. Both the Monte Carlo simulation and stochastic optimization were solved using Oracle Crystal Ball Software. From the simulation, the reduction of the compression and extension stroke standard deviations resulted in a 30% improvement in the energy storage standard deviation. The Cp was also improved by about 70%, from 0.99 to 1.44. The stochastic optimization resulted in extension and compression strokes shorter than deterministic optimization, with a 1.55 process capability index.