Analysis of the pore characteristics of backfill with an air-entraining agent based on SEM images by using fractal theory

Air-entraining agents can introduce tiny bubbles into a filling slurry, thereby improving its fluidity. However, the introduction of bubbles can change the pore characteristics of backfill materials. Pore characteristics can be visually observed using scanning electron microscopy (SEM). SEM images contain a large amount of microstructural characteristic information. However, the analysis of the image information is superficial. We used SEM images to qualitatively analyse the microstructural characteristics of backfill materials, and morphological operations and digital image processing techniques were used to obtain a binarization image for the semi-quantitative analys is of pore distribution. Two fractal theories: box-counting and slit island methods, were used to quantitatively analyse the binarization image to comprehensively analyse the pore characteristics of the backfill materials. The results show that: (1) the pore shape characteristics are obtained quantitatively (Rr = 4.73–5.48, R = 1.46–1.60), and most pores are regular. (2) The pores of backfill materials have good fractal characteristics (R2 > 0.97), which can be quantitatively analysed and studied using the fractal theory (DI = 1.086–1.160 and DB = 1.50–1.62). (3) The research result suggest that the more regular the pore, the smaller its fractal dimension, which is given quantitatively as (D = a*Rr+b*R+c).