Fold analysis of crumpled sheet using micro computed tomography

TL;DR

This study uses micro-CT to analyze the internal structure of crumpled sheets, revealing fractal dimensions and self-similarity across different materials and sizes, and examining the structure of lines on crumpled sheets.

Contribution

It provides the first detailed 3D analysis of crumpled sheets' internal structure using micro-CT, identifying key fractal properties and self-similarity across various materials.

Findings

Mass fractal dimension D_M ≈ 2.7 for crumpled paper balls

Fractal dimension d_f between 2.5 and 2.8 for internal structure

Hurst exponent H ≈ 0.9 for lines on crumpled sheets

Abstract

Hand crumpled paper balls involve intricate structure with a network of creases and vertices, yet show simple scaling properties, which suggests self-similarity of the structure. We investigate the internal structure of crumpled papers by the micro computed tomography (micro-CT) without destroying or unfolding them. From the reconstructed three dimensional data, we examine several power laws for the crumpled square sheets of paper of the sizes $L=50\sim 300$ mm, and obtain the mass fractal dimension $D_M = 2.7\pm 0.1$ by the relation between the mass and the radius of gyration of the balls, and the fractal dimension $2.5\lesssim d_f \lesssim 2.8$ for the internal structure of each crumpled paper ball by the box counting method in the real space and the structure factors in the Fourier space; The data for the paper sheets are consistent with $D_M = d_f$, suggesting that the self-similarity in the structure of each crumpled ball gives rise to the similarity among the balls with different sizes. We also examine the cellophane sheets and the aluminium foils of the size $L=200$ mm and obtain $2.6\lesssim d_f\lesssim 2.8$ for both of them. The micro-CT also allows us to reconstruct 3-d structure of a line drawn on the crumpled sheets of paper. The Hurst exponent for the root mean square displacement along the line is estimated as $H\approx 0.9$ for the length scale shorter than the scale of the radius of gyration, beyond which the line structure becomes more random with $H\sim 0.5$.