Abstract:

As an important component of the ballastless slab track, cement asphalt mortar is prone to void damage under the combined action of high-frequency train loads and temperature loads. The mortar void can be located quickly and accurately according to the change characteristics of wheelset accelerations. The vertical wheelset acceleration was decomposed into a series of product functions via local mean decomposition improved by adaptive dynamic moving step. Based on the change of kurtosis, the first product function that can best reflect the mortar void characteristics was selected, and its instantaneous energy and standardized instantaneous energy were calculated to enhance the mortar void feature. By analyzing whether there are outliers in the envelope functions of instantaneous energy and standardized instantaneous energy, whether the mortar is damaged can be determined. The results show that when the mortar void length is no more than 0.30 m, there is no outlier in envelop function of instantaneous energy at the mortar void position, resulting in “missing judgment”; and the envelop function of standard instantaneous energy has outliers both at the void and non-void position, resulting in “misjudgment”. When the mortar void length is no less than 0.65 m, the envelope functions of instantaneous energy and standardized instantaneous energy only have outliers at the mortar void position, which can be directly used to locate the mortar void position. Identifying mortar void through responses of vehicle structures can effectively alleviate the pressure of track damage detection and improve vehicle operation efficiency.

Key words:  ballastless slab track, mortar void, damage detection, signal processing