Abstract: This study presents the development of a non-contact displacement detection system for determining stress-strain curves in polymeric materials, overcoming the limitations of conventional mechanical extensometers while maintaining measurement accuracy. The proposed method employs a camera integrated with the YOLOv8 deep learning algorithm to track real-time positional changes of red marker lines on specimens during tensile testing, enabling accurate strain measurement. This contact-free approach eliminates potential specimen interference caused by mechanical extensometers. Comparative experiments were conducted by identifying the grips of mechanical extensometers, confirming the consistency between both methods in stress-strain curve determination. The results demonstrate remarkable similarity between stress-strain curves obtained from the non-contact system and traditional mechanical extensometers, validating the reliability and potential applications of this method for evaluating mechanical properties of polymeric materials. This digital innovation enhances experimental flexibility and offers an alternative solution for mechanical characterization of polymer materials.

Key words: Polymer science experiment, Stress-strain curve, Non-contact displacement detector, Extensometer