Novel Iterative Methodology for Modifications in Planar Straight Line Crank-Rocker Synthesis

Popular and widely referred text on graphical synthesis of crank-rockers invariably mentions Hrones and Nelson Atlas of the coupler curves. The studied coupler points, situated in a limited coupler space, and at the nodal points of the superimposed grid trace coupler curves with “granularity”. The performance of coupler points lying between the two adjacent coupler points is left to the user’s imagination and intuition. This method of providing gross information is suitable only for the initial selection. It can be modified further to meet the targeted output. Otherwise, it remains a solution with compromise. This paper demonstrates a graphic-numeric methodology for modifying/refining the initial selection by effective iterations. Using instantaneous Inflection Circle properties, three different approaches for iteration are demonstrated. Inflection Circle is a locus of infinite possible coupler points tracing some approximate straight path. In this study initial graphical analysis is completed using GIM software. Neural Networks are trained to reduce the time required for completing remaining iterative synthesis. One crank rocker is randomly selected from the Atlas and three routes for modifications using the mechanism specific trained Neural Network are demonstrated for modifications in the length of straight paths.