Generation of SWIFT Models Virtually Using FE Analysis: Application to Cleat Simulations

ABSTRACT Parameterization methods for two “commercial tire models—short wavelength intermediate frequency tire (SWIFT) model and flexible ring tire model—are developed around existing testing protocols. As an example, cleat tests are used for estimating belt mass and stiffness and enveloping tests are used for estimating belt bending and contact stiffness parameters in the respective tire models. This is the only way when commercial tire models are parameterized and used mainly by vehicle original equipment manufacturer companies. At present, tire suppliers are stepping up to supply commercial tire models as a part of virtual tire submissions and are virtually simulating standard testing protocols. It is envisioned that with a proper fundamental understanding of commercial tire model parameters and their modeling approaches to capture the respective tire dynamics, we can develop simple finite element (FE) techniques to estimate respective tire model parameters, thereby avoiding the simulation of cleat and other dynamic tests by using FE. With that motivation, previous work has already shown the estimation of belt mass and bending properties from FE part separation technique. In this work, with a fundamental understanding that the front cam and rear cam models in SWIFT are mathematically modeling the curvature of a loaded tire just outside of the footprint, we show that by fitting the tandem model to the loaded FE model deformed co-ordinates length, height, order of cam, and distance between the cams can be estimated easily. This simple loaded FE model-fitting technique is combined with other computationally simple FE static stiffness, footprint, and modal analyses to estimate other SWIFT parameters. Finally, SWIFT models from the above-mentioned FE techniques are developed for several tire designs and validated against enveloping and dynamic in-plane cleat test data. The variations in enveloping and other ride metrics from simulations are inline with testing data.