Contour of Strain of concrete plate

Hi,

I have around 100 pictures with high quality of a concrete plate. I am going to create a contour of strain by these pictures. Is it possible to do this job by Matlab software? Please help me.

My Best Regards Hadi

Answer 1

[Walter Roberson]

Yes, if there is a way to convert the picture elements into strain. For example is the color related to strain?

[Hadi]

Sorry, but this is answer or question?

[Walter Roberson]

When you examine your images, at any given point, how do you evaluate the strain?

If you just give me a non-coherent visible-light of a piece of concrete, such as taken with a smart phone, then it is not going to be possible to figure out the strain at any location. With a coherent visible light (e.g. laser) picture, it might be possible to determine something about strain very near the surface. With a coherent source on other wavelengths, possibly more could be deduced. For example you might get some interesting results from polarized terahertz radiation.

Answer 2

[Sean de Wolski]

Yes.

I had a chapter on this in my Master's Thesis :)

[Hadi]

So Please can you help me? Is it possible for you to send me that chapter?

[Sean de Wolski]

Not easily. It's published through my University. Besides, I don't know if the technique I used is applicable to your images.

• What are you using to measure displacement between the images?
• Is it color as Walter has suggested?
• Is it displacement calculated using digital image correlation (This is what I had)?
• Is there cracking? Strain is not defined across a crack.

[Walter Roberson]

Ah, I had assumed that strain was to be calculated for each individual image. You were able to calculate it from successive images in a time sequence? What kind of light source?

[Sean de Wolski]

For my research we were using the APS at Argonne National Lab to take micro-CT scans so we actually had full three dimensional maps of the concrete specimen. We'd apply a load to it while is was inside of the scanner so that it was actually under load for successive scans and then I did non-rigid digital volume correlation to track every voxel's motion between the scans. With these dense displacement fields, I could calculate strain.

Others in my lab were using a regular high resolution camera to image the surface of a beam under a three-point bend test and from there could calculate surface cracking between scans. Identifying cracking is much easier than strain, it's just simple image processing. They did try calculating strain by painting dots on the concrete specimen but I don't recall it working.