AGMA 10FTM07

Load-sharing among the teeth of involute splines is littleunderstood. Designers typically assume only a fraction of the teethare engaged and distribute the load uniformly over the assumednumber of engaged teeth. This procedure can widely over- orunderestimate tooth loads. A new statistical model for involutespline tooth engagement has been developed and presented earlier,which takes into account the random variation of gear manufacturingprocesses. It predicts the number of teeth engaged and percent ofload carried by each tooth pair. Tooth-to-tooth variations causethe clearance between each pair of mating teeth to vary randomly,resulting in a sequential, rather than simultaneous toothengagement. The sequence begins with the tooth pair with thesmallest clearance and proceeds to pick up additional teeth as theload is increased to the maximum applied load. The new model canpredict the number of teeth in contact and the load share for eachat any load increment. This report presents an extension of the newsequential engagement model, which more completely predicts thevariations in the engagement sequence for a set of splineassemblies. A statistical distribution is derived for each tooth inthe sequence, along with its mean, standard deviation and skewness.Innovative techniques for determining the resulting statisticaldistributions are described. The results of an in-depth study arealso presented, which verify the new statistical model. Monte CarloSimulation of spline assemblies with random errors was performedand the results compared to the closed-form solution. Extremelyclose agreement was found. The new approach shows promise forproviding keener insights into the performance of spline couplingsand will serve as an effective tool in the design of powertransmission systems.

Edition:

10

Published:

10/01/2010

Number of Pages:

17

File Size:

1 file , 810 KB