PSI - Issue 78

Eugênio Moreira et al. / Procedia Structural Integrity 78 (2026) 1967–1974

1970

4. Methodology

4.1. Structural Health Monitoring and Data Processing

Model validation is based on SHM data from three inclinometers and thermo-hygrometers installed along Gubbio’s medieval wall, recording every 10 minutes since July 2018. Data are wirelessly transmitted and aggregated into 20 minute averages, yielding 1,911 daily files (to end of 2024). Station 02, with over 99% data completeness in 2021, serves as the primary validation subset. Data processing involved consolidating CSV files, standardizing timestamps, and segmenting by station. Quality control included numeric conversion, outlier removal, and temperature compensation. Inclination values were cor rected for temperature e ff ects and normalized to a common baseline, producing the validated dataset for subsequent modeling. Thermal boundary conditions were obtained via solar radiation analysis in Rhino.Inside.Revit with Ladybug, using a custom EPW from PVGIS (The Joint Research Centre: EU Science Hub, 2001). PVGIS supplies hourly radiation data from satellite imagery and high-resolution interpolation. The study focused on Station 02 considering geometry, terrain, and nearby structures for shading e ff ects. Hourly radiation data were extracted at five points: terrain front / back, wall front / back, and top. Boundary conditions included: (i) convective transfer h = 200W / m 2 · ◦ C, (ii) radiative exchange with ε = 0 . 9 for limestone, (iii) distributed solar flux per PVGIS methodology, accounting for orientation and shading. This established thermal pre-conditioning states for subsequent seismic assessment. 4.2. Thermal Loading

4.3. Finite Element Modeling

The multi-leaf masonry was modeled via macromodeling, with distinct limestone outer leaves and lime mortar core, using 2D deformable parts for computational e ffi ciency and accurate out-of-plane response (Fig. 2). Material properties, including thermal parameters, are listed in Table 1.

Fig. 2. Wall section modeling and meshing strategy. Nonlinear mechanical behavior was addressed with the Concrete Damage Plasticity (CDP) model (Lee and Fenves, 1998; Lubliner et al., 1989), tailored for brittle materials exhibiting tension / compression asymmetry. The model im poses a linear elastic regime up to threshold stresses σ t 0 (tension) and σ c 0 (compression), beyond which plastic strains develop to represent irreversible deformation resulting from microcracking and crushing (Eq. 1).