Automatic Determination of the Small Displacements in a Mechanical Assembly of Two Parts

Herein is proposed an automatic method by which the precision of a joint realized by the mechanical assembly of two parts may be totally determined. This precision depends on the dimensional and geometrical defaults existing in the parts, and is materialized by additional small displacements of the parts superposed with the kinematics of the joint. In a first step, the parts are represented by their geometrical entities, and the possible contacts existing between these entities. By ≪ sliding ≫ the parts around their nominal position, i.e. by opening or closing contacts between surfaces, one defines a new configuration of the assembly. In a second step, we give the conditions that this new configuration must fulfill to be physically acceptable. No parasitical motion is accepted, it must be possible to compute the small displacement torsor between the two configurations. And finally, no geometrical interferences are tolerated between the parts. In a third step, we apply this method on a simple but instructive bidimensional example.