Abstract
Water film depth (WFD) on runways is a key factor contributing to aircraft hydroplaning during takeoff and landing. Thus, the early measurement or prediction of WFD during rain is critical for reducing accidents. Most existing WFD prediction models are derived from experiments conducted on road surfaces. However, an accurate prediction of WFD on runways and reduced hydroplaning risk require a precise empirical prediction model. This study conducted experiments involving four parameters-rainfall intensity, pavement mean texture depth, drainage length, and transverse slope-to develop a WFD dataset specific to different runway conditions. The multiple linear regression method is employed to establish a model for WFD predictions. The proposed National Taiwan University (NTU) model's predictability is compared with three existing empirical models using NTU and Gallaway datasets. The results clearly demonstrate the superior accuracy and robustness of the NTU model compared to the other evaluated models. The NTU model offers a precise and practical predictive formula, making it highly suitable for integration into contaminated runway warning and management systems. This study employed a laser displacement sensor and a programmable logic controller to obtain high-accuracy, high-sampling-rate WFD data. Modern automated data acquisition enables simultaneous measurement at multiple points and captures the complete WFD curve from zero to a stable depth, which was previously difficult to obtain.