Optimizing Gearbox Lubrication With a Fully Integrated PLM Process, Using GPU Based Lattice Boltzmann and Multi-physics Solvers

Abstract

A gearbox is at the center of most power machinery and vehicles. Because of its place in the powertrain, the gearbox efficiency directly affects the device’s overall energy consumption, and reliability. Be it in a car, truck, heavy-duty equipment, wind-turbine or even helicopter, gearboxes failure can lead to a complete stop; lubrication issues account for 54% of failures, inducing expensive maintenance costs or insurance claims. A well-optimized lubrication design can help reduce your insurance and repair costs by half, and lower your vehicle consumption by more than 10%. OEMS are being pushed to evaluate new development methodologies to design more efficient and reliable products faster. Verifying the correct lubrication of powertrain components, such as gearboxes, has always proven a difficult experimental task. Translucent acrylic parts would need to be built specifically for testing, and quantitative data proved hard to come by. With Lattice Boltzmann based fluid simulation solutions however, accurate quantitative data about the lubrication of complex rotating geometry can easily be obtained. In this presentation, we will first show how the lubrication of a gearbox can be accurately estimated with multi-phase simulation using actual design and motion. Not having any need for modeling, on top of the use of multi-phase, is necessary to get validated results for low to high gear rotation speeds. Also, the use of GPU based solver will lower the simulation cost and help replacing testing. Then, we will show how the gearbox can be improved by enabling designers to implement design changes directly from a model based Product Lifecycle Management system. These improvements will focus on assessing wetted surface and churning loss to determine how product alteration affects reliability and efficiency of the gearbox system. Design cycles are getting shorter and shorter; to match this demand we will also demonstrate how a fully integrated workflow can drastically reduce turnaround time for lubrication simulation. Lastly, we will show that the multi-physics performance of the gearbox can be further improved by using automated optimization tools, and combing them with multi-body dynamics solutions.