Prediction of load sharing based HCR spur gear stresses and critical loading points using artificial neural networks

Abstract

The prediction of the load shared by a pair of teeth, maximum contact and fillet stresses and the respective location of the critical loading point becomes rather a difficult task in High Contact Ratio HCR gears as the contact ratio exceeds two. As this prediction greatly depends on the gear parameters like pressure angle, addendum factor and teeth number, an attempt has been made to work on this area highlighting these aspects using Finite Element FE Multi Pair Contact Model MPCM. The minimum value of contact ratio under consideration is 2.1. However, the maximum is chosen as 2.9. A new methodology based on Artificial Neural Networks ANNs is proposed for the prediction of Load-Sharing Ratio LSR, maximum fillet and contact stresses and the respective critical loading points. The data set generated from the MPCM has been used to train the networks and, furthermore, its effectiveness is proved by a different data set of HCR gear pairs determined for the randomly selected parameters.