Title:SFFT-based Homogenization: Using Tensor Trains to Enhance FFT-Based Homogenization

Abstract:Homogenization is a key technique for approximating the macroscopic properties of materials with microscale heterogeneity. The FFT-based Homogenization method has gained widespread usage due to its computational efficiency and accuracy in handling complex microstructures. However, despite its advantages, the method is limited by speed and memory constraints, particularly when applied to high-resolution discretizations. These limitations affect its scalability and efficiency, especially in large-scale simulations or when dealing with highly detailed microstructures. These challenges arise from the fundamental reliance on the Fast Fourier Transform, which imposes inherent restrictions on further advancements. In this paper, we propose a novel SFFT-based Homogenization algorithm that utilizes a Quantized Tensor Train variant of the Quantum Fourier Transform. This method is tailored to the geometry under consideration and offers significant improvements in time complexity and memory efficiency compared to the traditional FFT-based approach while remaining executable on classical hardware. The method is applicable only if a suitable Quantized Tensor Train representation exists for the stiffness operator associated with the underlying geometry.