PSI - Issue 78

Pasquale Guarino et al. / Procedia Structural Integrity 78 (2026) 1561–1568

1562

Finite Element (FE) models are commonly employed to analyze complex masonry structures; however, discrepancies often arise between numerical predictions and experimental observations, as highlighted by Mottershead et al. (1993) and Zhao et al. (2025), potentially leading to unreliable results. Finite Element Model Updating (FEMU) techniques have emerged as a key solution, calibrating numerical models using data from static and / or dynamic testing to re flect the actual structural behavior. FEMU is typically formulated as an optimization problem, where measurement ill conditioning and the leak of convexity may produce suboptimal solutions, an issue mitigated through regularization methods, as explored by Grip et al. (2017). Particle Swarm Optimization (PSO), introduced by Kennedy et al. (1995), is a nature-inspired algorithm particularly suited to such optimization tasks. It is computationally e ffi cient, derivative free, supports parallel exploration of candidate solutions, and is not sensitive to initial population conditions. Recent studies, including those by Ereiz et al. (2022) and. H. Tran-Ngoc et al. (2018), demonstrate the superior performance of PSO over Genetic Algorithms (GA) for FEMU applications such as damage detection and model selection. Despite these advances, integrating real-time monitoring data with sophisticated optimization algorithms into user-friendly practical tools remains challenging, particularly for historical buildings. Uncertainty in mechanical properties and incomplete documentation pose additional di ffi culties, as emphasized Piselli et al. (2020). While Historical Build ing Information Modeling (HBIM) tools facilitate cultural heritage management, they could better support structural modeling and provide unified, interdisciplinary workspaces Jorda´n Palomar et al. (2020). Vibration-based Structural Health Monitoring (SHM), valued for its minimal intrusiveness and global damage assessment capabilities, has been successfully applied to heritage structures and bridges (Garc´ıa-Mac´ıas et al. (2020), Costa et al. (2013), Sevim et al. (2011) and Meoni et al. (2025)). Revit is one of the most used BIM software for historical buildings. Plug-ins such Pyrevit allows one to create custom function and activities within it’s environment. For example, Meoni et al. (2022) developed the H2BIM Revit (2024) add-in for integrating structural performance data. To support informed decision-making in design, restoration, and maintenance, advanced SHM algorithm ought to be implemented and in tegrated in these enviroments Olivito et al. (2021). needs to be implemented in these environments. To address this gap in the literature, an easy-to-use plug-in named FEMUP (Finite-Element Model Updating Plug-in) is introduced to automatically calibrate numerical models with real monitoring data, improving the accuracy and e ffi ciency of FEMU for historic.

2. FEMUP

2.1. Plug-in architecture

FEMUP is a Python-based plug-in. Its functionalities are realized with the help of pyRevit (2025), a plug-in for Revit software that leverages IronPython, an open-source implementation of the Python programming language. Pyrevit allows the creation of buttons for running Python codes directly inside the GUI of the program. Modal updating is be performed using FE software Abaqus CAE, through a python script that uses Particle Swarm Optimization as the model calibration algorithm. All the scripts require di ff erent python libraries to be installed in order to work properly, including Numpy, Pymoo, Scikit-Learn, Matplotlib and Plotly. To visualize the geometry, four parametric families have been created to display points and reference axes that are imported using the buttons.

2.2. Plug-in description

For the proposed plug-in, a tab containing 4 buttons has been created for 4 di ff erent functionalities as shown in Fig.1, namely:

1. Import mesh from obj file. 2. Assign sensors’ position and direction, material properties. 3. Assign boundary conditions. 4. Define the fitting parameters and run modal updating.

Mesh visualization within Autodesk Revit presents a unique challenge due to Revit’s fundamental operation as a BIM platform, primarily handling parametric, rather than discretized, geometric elements. Since Revit is based on