Three-Point Inverse and Forward Kinematic Algorithms for Circle Measurement from Distributed Displacement Sensor Network

Abstract

Automatic fitting of an arc center and radius is a quality problem frequently encountered when manufacturing a mechanical component. Due to the complexity of the measurement, validating each manufactured component via inspection is not feasible or economical. This paper introduces a new validation procedure for measuring arcs from distributed sensors. The goal of this proposed measurement process is to improve measurement throughput (i.e., parts measured per unit of time) and reduce measurement errors associated with hardware and algorithms. This proposed model develops a three-point inverse kinematic algorithm (TPIK) accompanied by a calibration master to obtain the relative location of the measurement system by solving a set of six non-linear equations. This technique allows deployment of a high accuracy gauge systems that in general, reduces machine and algorithm errors. The direct fitting is validated by using mathematical, CAD, and experimental models. Furthermore, a modified definition for the roundness index is introduced based on the proposed forward and inverse algorithms. The simulations examine the roundness index in relation to the measurement precision, sampling angle, nominal radius, and part variation. A benefit of this proposed method is accurate and rapid inspection of the radii and elimination of the human error associated with part loading variation during conventional radii measurement. The rapid, accurate inspection and corresponding reduction in human error make this method an excellent process for inspection of large quantities of components.