Development of a marker-based displacement tracking technique for tensile testing of geosynthetics using computer vision

Date

2023-7-5

Author

Türkezer, Mehmet

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

216
views

177
downloads

Cite This

Geosynthetics are the manufactured products generally used within soils to improve their weak characteristics. Before using geosynthetics in the field, their tensile strength and deformation characteristics (i.e., strain capacity) should be tested in laboratory conditions. For this purpose, as part of tensile testing, various types of extensometers are used to measure the deformations accurately, which are generally expensive and take time to implement during testing. Therefore, there is a need to use new technologies, such as imaging, to replace extensometers to ease and accelerate testing without losing accuracy. In this sense, computer vision facilitates the quantification of strains based on markers drawn on specimens. This study presents a non-contact strain measurement technique called "Marker-Based Displacement Tracking (MBDT)" developed using Python programming, incorporating OpenCV, an open-source vision-based code platform. Markers designating a gauge length are drawn on specimens and identified using image segmentation techniques. The accuracy of MBDT is validated by comparing the strains calculated by Crosshead Displacement and Digital Image Correlation techniques on geomembrane, geotextile, and geosynthetic clay liner samples, tested by different Universal Testing Machines and conventional video cameras. The feasibility of using a camera with an ultra-wide-angle (fisheye) lens as the capturing device is also assessed. The results obtained with 36 experiments proved the success of MBDT, implying its application to various other types of geosynthetics.

Subject Keywords

Geosynthetics, Tensile testing, Strain measurement, Video extensometer, Digital image correlation

URI

https://hdl.handle.net/11511/104763

Collections

Graduate School of Natural and Applied Sciences, Thesis