Displacements and strains in Ncorr vs Ncorr_post

Hi,

The strains obtained for using Ncorr_post are 1/10 times smaller than expected.

After checking, I noticed that the Ncorr_post plots displacements 1/10 as Ncorr, resulting in 1/10 smaller strains.

Have you seen this error before? Is there a fix for this?

Answer 1

[Aditya]

Hi Dhanushka,

This discrepancy typically occurs because Ncorr_post applies a scaling factor to convert pixel displacements to physical units (such as millimeters), and if this factor is set incorrectly—often to 0.1 instead of 1—it will cause both displacements and strains to appear 1/10 of their true values. To fix this, check the section of the Ncorr_post script where the scaling or calibration is applied, and ensure it matches your experimental pixel size. For example, if your images are not calibrated to 0.1 mm/pixel, but rather 1 mm/pixel (or you want results in pixels), set the scale factor accordingly. Here is a code snippet showing what to look for and how to correct it in MATLAB:

% Original (incorrect if your pixel size is not 0.1 mm/pixel)
scale_factor = 0.1; % mm/pixel

% Corrected (set to your actual pixel size, or 1 for pixels)
scale_factor = 1; % use 1 if you want results in pixels

% Then, apply scaling to displacements
u_physical = u_pixels * scale_factor;
v_physical = v_pixels * scale_factor;