UNI-MB - logo
UMNIK - logo
 
(UM)
  • Global optimization algorithm applied to design gear pair
    Zafošnik, Boštjan ...
    During the gear meshing, gear tooth flanks, slide and roll one over another. Geometrically, sliding and rolling could be described by using the equivalent model of two contacting cylinders that have ... the same radius as the curvature radius of contacting mechanical elements at any point of the real contact. Sliding is described by a parameter known as specific sliding. Specific sliding is an appropriate measure to determine wear of tooth flanks. Specific sliding should be equal and minimal for a tooth tip of pinion (gear) and a tooth root of gear (pinion). In general, high specific sliding occurs generating of the heat, resulting in a substantial increase of the temperature of contacting surfaces. Also, when a specific sliding of tooth root is too high, fracture of the tooth could occur. Due to the rolling of the tooth flanks Hertz pressure is caused as a consequence of the normal load. Consequently, regarding the rolling-sliding circumstances damaging of gear flanks can occur. Typical form of failure caused by the above phenomenon is pitting. Appearance of pitting can severely condition services life of the gear. Both, specific sliding and Hertz pressure depend on the geometry of contacting surfaces which is dominantly influenced by the profile offset. To maximize the service life of a gear pair optimal values or profile offset have to be found in order to satisfy the prescribed criteria for specific sliding and Hertz pressure. The general optimization problem is addressed in the form of nonlinear programming problem. The objective of this approcah is to determine the optimal values of the profile offset to minimize specific sliding, while Hertz pressure remains within the specified values. The global optimization method is used to find a global optimal solution. The Adaptive Grid Refinement algorithm procedure is applied. This algorithm is based on the identification of the feasible points inside each of the iterations, defining the solution set. Points far from the current optimum are pruned from the solution. The algorithm identifies optimal regions, rather than only a single optimal solution.
    Source: Advances in computational engineering & sciences : 2000 (Vol. 2, str. 1592-1597)
    Type of material - article, component part
    Publish date - 2000
    Language - english
    COBISS.SI-ID - 5610006

Shelf entry