Three-dimensional Shape Reconstruction of Aggregate in Concrete Using Two-Dimensional Periodic Wavelet Series

A novel two-dimensional (2-D) periodic wavelet series (PWS) method was proposed in this research. It creatively used the 2-D compactly supported periodic wavelet basis functions to reconstruct the 3-D shape of the aggregates with numerous sharp angularities (e.g., the manufactured aggregates). The detailed reconstruction procedure was given, including the acquisition of the shape functions for reconstruction, the solution of 2-D PWS coefficients, the fitting of particle surface functions and the assessment of reconstruction accuracy. To validate the feasibility of the proposed method, four manufactured sand particles were selected and reconstructed using the 2-D PWS method, then the proper reconstruction parameters (i.e., the wavelet expanding level, wavelet type) were discussed for the balance of computational efficiency and reconstruction accuracy. Finally, compared with the commonly used spherical harmonic reconstruction method, the advantages of local variation capacity of 2-D PWS method and its potential applications in recycled aggregates were pointed out.