Efficient buffeting analysis of long-span bridges under non-stationary winds: A 2D interpolation enhanced approach

Abstract

The buffeting analysis of long-span bridges under non-stationary winds is a crucial consideration in the context of exceptional winds recurring in frequent and violent patterns. The time–frequency representation is widely used for the solution of structural non-stationary buffeting responses due to its popularity and simplicity. However, the traditional approach consumes considerable computational resources to achieve high accuracy, which renders the approach computationally inefficient. In such a case, a 2D interpolation enhanced approach is developed for the efficient buffeting analysis of long-span bridges under non-stationary wind actions. Central to this approach is the 2D interpolation applied to the reconstruction of structural buffeting responses, accompanied by the well-designed non-uniformly distributed interpolation knots. For the enhanced approach, the computations in solving governing equations are only required at limited interpolation knots instead of at each frequency and time instant, which yields a dramatic reduction of computational resources and efforts. By taking a long-span suspension bridge as an example, a parametric analysis is conducted to investigate the effect of interpolation intervals on the simulation efficiency and accuracy, and the preferred suggestions are made accordingly. Then, the effectiveness of the enhanced approach is further verified via a case study by focusing on the evolutionary spectra and non-stationary buffeting responses. The analytical results indicate that the enhanced approach is particularly efficient in simulating non-stationary buffeting responses of long-span bridges with few losses of accuracy.