An integrated approach for the analysis of historic timber structures combining enhanced photogrammetry and finite element modeling
Abstract
This study develops an integrated photogrammetry-FEM approach for assessing damaged components in historical timber structures. Using smartphone-captured images and an improved SIFT-WOOD (scale-invariant feature transform for wood) feature extraction algorithm, it achieves robust 3D reconstruction of wood surfaces under challenging lighting and textures via SfM-MVS (structure from motion-multi-view stereo). The reconstructed model is post-processed and incorporated into finite element software, where orthotropic elastic, anisotropic elastoplastic, and viscoelastic models simulate mechanical response and long-term performance. A novel “stress-range volume method” quantifies the correlation between damage volume and load-bearing capacity, with a standards-based classification system. Following minimal intervention principles, controlled simulations identify damage repair priorities and inform a stepwise restoration strategy. The method is validated on two in-service historic timber components, demonstrating its practical applicability. This framework establishes a technical pathway—from geometric digitization to quantitative mechanical evaluation—characterized by low cost and high efficiency, improving the scientific basis of conservation practice.