An Innovative Approach to Geometric Characterization in Profile Tolerancing

Profile tolerances are used to control form or combinations of size, form, orientation and position of free form features (lines or surfaces) usually described by a CAD 2D or 3D model. The procedures reported in the standard ISO/TR 5460 do not address the problem of identifying and separating the size error contribution in profi le error assessment. This issue has become an industrial need in years because complexity and precision requirements of parts functional surfaces are increasing. In order to improve geometric tolerance verification, ISO 14660 and ISO 17450 Standards have been developed, where the “association” operation is introduced. For standard geometries the association operation “is used to fit ideal feature(s) to non-ideal feature(s) according to specific rules which are called criteria” [1]. The associated feature has the same shape of the nominal model and differs from it in size, orientation and position. Following the diffusion of measuring instruments, as mechanical or optical coordinate measuring machines, and their integration with CAD systems, the geometric characterization of a free form profi le is currently obtained by fi tting algorithms that minimize the distance among the measured points and the virtual model. This method doesn’t allow the separation of size errors from geometric errors in profi le error assessment. However this information could be very useful to adjust the manufacturing process, thus increasing the possibility of manufacturing good parts. As an example, it could be possible to correct size errors resulting from wrong tool’s set-up in cutting operation, or from an overestimation or underestimation of material’s shrinkage in extrusion operations, or to regulate the gap in a mechanical system, by mean of correcting errors due to wear phenomena. In this paper a new method for geometric characterization of free-form profi le that allows the separation of size contributions from form, orientation and position errors is presented. The algorithm consist in an iterative process where at every cycle a resizing (offset) of the nominal model and a repositioning (fitting) of the points measures are calculated. The method has been validated with several simulated test-case and the results confirm both the correctness of the proposed approach and its utility in conformity assessment operations.