The medieval walls surrounding historic urban centers embody their cities' cultural and architectural heritage and register the evolution of construction techniques across centuries. Having been erected in multiple phases with heterogeneous masonry materials, these structures are prone to long‐term deterioration and potential collapse over time. Ensuring continuous monitoring of their structural integrity is essential for both conservation efforts and public safety, particularly in seismic regions where climate change can exacerbate environmental degradation and impose dynamic demands beyond the original design scope.

This paper focuses on Gubbio's medieval walls as a case study, analyzing their mechanical response under earthquake‐induced loads, employing nonlinear finite element modeling with the Concrete Damage Plasticity (CDP) model as a nonlinear constitutive model for stone masonry to identify damage mechanisms and vulnerable zones.

The methodology employs a two-phase analysis to evaluate both immediate seismic damage and post-earthquake structural behavior. Phase 1 establishes a 90-day baseline through fully coupled thermal-mechanical analysis, incorporating solar radiation patterns derived from PVGIS weather data and Grasshopper/Ladybug analysis as surface heat flux boundary conditions, with convective and radiative heat exchange. Model predictions will be validated against existing structural health monitoring (SHM) data from monoaxial inclinometers to establish undamaged structural response patterns. Phase 2 extracts temperature and stress fields from Day 60 as initial conditions for Dynamic, Implicit seismic analysis, where real earthquake ground motion records from the Istituto Nazionale di Geofisica e Vulcanologia's (INGV) catalog will be applied at varying intensity levels using calibrated CDP material definitions for damage evolution. The conditions after that will be used to perform post-seismic fully coupled thermal-mechanical analysis for the rest of the analysis period, enabling direct comparison between baseline and post-earthquake structural response characteristics to identify damage-induced modifications.

The baseline thermal-mechanical validation demonstrates strong agreement between simulated and experimental tilt rotation data over representative time windows, with the model capturing the rapid rotational variations induced by diurnal solar cycles. The validated model should reproduce the characteristic daily inclination amplitude observed in SHM data, confirming the dominant role of solar radiation in short-term structural response. The post-seismic thermal analysis is expected to reveal altered diurnal patterns, modified inclination amplitudes, and phase shifts in thermal expansion cycles that correlate with CDP-predicted damage zones. These signature variations in structural response behavior will provide diagnostic indicators for damage detection algorithms and early warning system development for heritage structures.

This work contributes to heritage preservation by establishing a validated framework for interpreting SHM data during seismic events, addressing critical gaps in understanding how thermal pre-conditioning affects earthquake damage patterns in medieval masonry structures. The approach enables the development of automated damage detection algorithms and supports evidence-based conservation interventions targeting identified vulnerability hotspots. However, limitations include computational demands of coupled analyses, challenges in calibrating CDP parameters for historic materials with unknown properties, and the need for extensive sensor networks to validate model predictions across different wall sections. Future research should address uncertainty quantification in material parameters, extend the approach to three-dimensional modeling incorporating more detailed soil-structure interaction, and develop simplified monitoring protocols for heritage sites with limited instrumentation capabilities.