Strain field measurements around notches using SIFT features and meshless methods

Area of Science:

• Mechanical Engineering
• Computer Vision
• Computational Mechanics

Background:

• Accurate strain measurement is crucial for understanding material behavior, especially in regions with high stress gradients.
• Traditional methods can struggle with precision in complex stress concentration zones.

Purpose of the Study:

• To introduce and validate a novel hybrid experimental-numerical technique for enhanced strain measurement.
• To improve the accuracy of strain analysis in stress concentration regions, such as near notches.

Main Methods:

• The SIFT-meshless technique integrates the Scale Invariant Feature Transform (SIFT) algorithm for feature point tracking with meshless methods based on moving least squares approximation.
• SIFT identifies and matches feature points across multiple images of a deforming specimen to obtain displacement data.
• Meshless methods then interpolate these displacements to generate continuous strain fields.

Main Results:

• The SIFT-meshless method successfully measured strain distribution around a semicircular notch under bending.
• Results showed high accuracy in capturing peak strain values close to the notch, even with steep strain gradients.
• Comparison with Digital Image Correlation (DIC) and Finite Element Analysis (FEA) validated the method's reliability.

Conclusions:

• The proposed SIFT-meshless technique offers a reliable and accurate approach for experimental strain measurement in critical stress concentration areas.
• This hybrid method enhances the capability to analyze localized deformation behavior in materials.
• It provides a valuable tool for researchers and engineers in experimental mechanics and material science.