Investigation of temperature effect of stress detection based on Barkhausen noise

Abstract

The temperature of rail tracks differs throughout the year because of seasonal climate change. Exposure to the elements causes thermal stress, which can be measured using magnetic Barkhausen noise (MBN). However, changes in temperature have an effect on the stress and the MBN signal itself. In order to extract feature values which reflect the state of stress, this paper introduces a design of an MBN stress measurement system for use at different temperatures. To analyze the relationship between the signal features and temperature, we measured the MBN signals at different temperatures when the stress was constant in experiment. It is found that within the elastic range of the steel specimen, the average value, RMS value, counts value, peak value, and peak–width ratio value decrease with increasing temperature. However, the variation rate of different feature values with changes in temperature is different. Finally, we found the feature values which have a smaller change with temperature, have larger change with stress. From the theory, we explain the phenomenon that temperature change has effect on the rail stress and analyze the impact of residual stress and uneven distribution of temperature in the detection.