Design optimization of cam–follower mechanisms using Rao algorithms and their variants

This paper presents the design optimization of cam–follower mechanisms (CFM) with eccentric roller type follower using recently developed advanced optimization algorithms, namely Rao, SAMP-Rao, and QO-Rao algorithms. Four types of follower motion law i.e., cycloidal, modified harmonic, 3–4–5 degree polynomial, and 4–5–6–7 degree polynomial motion, are considered. The CFM is optimized to minimize three objectives such as the input torque needed to rotate the cam, the radius of the pitch circle of the cam, and the maximum contact stress. The problem has five continuous design variables, namely the roller radius (Rg), the radius of cam base circle (Rb), the distance between the follower bearing and the center of the cam (q), the eccentricity of the follower (e), and the length of follower bearing (b). Eight design constraints related to the geometry of the cam mechanism, the pressure angle, the efficiency of the mechanism, the curvature radius of the pitch curve, and the maximum contact stress, are considered. The computational results obtained using Rao algorithms and their variants are compared with other advanced optimization algorithms such as the salp swarm algorithm (SSA), ant lion optimizer (ALO), moth–flame optimization (MFO), multi verse optimizer (MVO), evaporation rate water cycle algorithm (ER-WCA), grey wolf optimizer (GWO), and mine blast algorithm (MBA). The comparison of optimization results reveals that the optimum value of a fitness function obtained using Rao algorithms and their variants is superior to GWO, ALO, MFO, SSA, ER-WCA, MBA, and MVO for all four cases considered. The optimum fitness function value of the CFM with case III is reduced by 3.14%, 4.13%, and 7.61% compared to the CFM’s fitness function value with the case I, case II, and case IV, respectively. Hence, the 3–4–5 degree polynomial motion of the follower is effective for better performance of the CFM. Also, the average time required for Rao algorithms and their variants is comparatively significantly less than the other algorithms considered. The effectiveness of Rao algorithms and their variants will help the community to improve the optimum solutions to their problems in their respective research field. Also, Rao algorithms will decrease the researchers’ efforts required to tune the algorithm-specific control parameters as these algorithms do not have algorithm-specific control parameters. Furthermore, the Friedman rank test proves the significance of the statistical results acquired using Rao algorithms and their variants.