1) Condition Based Monitoring: The primary application of this work for NASA, is as a tool that can be used for creating, testing, and fine-tuning condition-monitoring strategies for rotor craft drive systems. The distributed contact analysis will enable dynamic analysis of full drive system models, both in a healthy state as well as with various kinds of damage. Both surface damage as well cracks can be studied. 2) Life Estimation: Current component life prediction tools are constrained by the limited accuracy of simplified dynamic stress prediction methods. The proposed work will, make it possible run very accurate simulations under dynamic conditions. 3) Dynamic Factors: The proposed work will enable NASA to compute accurate dynamic factors for use during the design evaluation stage of gear boxes. These dynamic factors can be used to account for steady state dynamics, as well as for transients caused by short duration events.
1) Vibration Prediction in Time-Domain: To date, only frequency domain based vibration calculations with linear models have been commercially available. But Time-domain models are necessary to correctly include contact and kinematics induced non-linearities. Having the fast contact solver will allow very realistic drive system dynamic models, to run in the time domain. 2) Impact Dynamics: It will be possible to make predictions for survivability of drive systems subjected to transients caused by short duration events such as a load spikes. This is an important consideration in the wind-turbine and off-highway equipment industries. Modeling these transient dynamics can only be done in the time-domain. A fast contact solver will allow realistic prediction of these effects. 3) Automatic optimization: Access to a very fast solver will make it possible to run fast static analyses inside the optimization loop of a commercial general-purpose optimizer. It will be possible to optimize metrics such as gear contact patterns, transmission error, and stress while automatically varying the surface modifications and other design parameters. 4) Manufacturing Error Studies: Each manufacturing error has a unique probability distribution. A very fast solver will enable Monte Carlo type studies of manufacturing errors with realistic random distributions. The output will be the probability distribution functions of performance and failure metrics for the drive system.