Locating Transverse Cracks in Prismatic Beams Using Random Forest Method and The Frequency Drop

Main Article Content

Cristian TUFIȘI
Vasile-Catalin RUSU
Gilbert-Rainer GILLICH
https://orcid.org/0000-0003-4962-2567

Abstract

As the infrastructure ages and nears the end of its estimated operating time, the detection of damages becomes a major issue in structural health monitoring (SHM). In this paper, the authors propose an analytical approach for generating the data needed to train a Random Forest model (RF) that will perform the SHM task, namely, to detect, locate, and assess the severity of transverse cracks in beam-like structures. Using an original method, we calculate the relative frequency shifts (RFS) for different damage scenarios and use the generated data to train the RF model. Subsequently, the validation of the RF model is performed using data obtained from FEM simulations using different mesh sizes, for different damage scenarios on steel beams. The results indicate that the RF model can detect the presence of the defect and find the position and depth of the transverse cracks very precisely if the crack is located in the area where the beam achieves the maximum bending moment.

Downloads

Download data is not yet available.

Article Details

How to Cite
[1]
2022. Locating Transverse Cracks in Prismatic Beams Using Random Forest Method and The Frequency Drop. Romanian Journal of Acoustics and Vibration. 18, 2 (Feb. 2022), 119–125.
Section
Articles

How to Cite

[1]
2022. Locating Transverse Cracks in Prismatic Beams Using Random Forest Method and The Frequency Drop. Romanian Journal of Acoustics and Vibration. 18, 2 (Feb. 2022), 119–125.

Most read articles by the same author(s)