Modellazione numerica di colonne in acciaio in caso di incendio post-sisma

Questo articolo presenta i risultati di calibrazioni di analisi numeriche di incendio post-sisma (FFE) su colonne in acciaio appartenenti a un telaio in acciaio controventato. In particolare, per calibrare i modelli numerici 3D, sono stati utilizzati i dati sperimentali relativi a cinque prove FFE su colonne in acciaio, eseguite presso l'Istituto Federale per la Ricerca e le Prove sui Materiali (BAM) mediante tecnica di prova ibrida. Due test FFE sono stati eseguiti su colonne senza alcun tipo di protezione antincendio, mentre gli altri tre test FFE sono stati eseguiti con diverse soluzioni antincendio: pannelli in silicato di calcio convenzionali e sismo-resistenti e protezione a base di intonaci antincendio. Sono state impiegate diverse strategie di modellazione utilizzando il software SAFIR: elementi di tipo trave e elementi shell. La modellazione ed i risultati della calibrazione numerica degli scenari FFE applicati a colonne in acciaio sono riportati e descritti nell’articolo. I risultati dei modelli numerici hanno mostrato una buona rappresentazione dei dati sperimentali.

After an earthquake, as has happened on many historic occasions, a fire may be triggered by earth- quake-induced rupture of gas piping or failure of electrical systems. In this context, the structural fire performance can deteriorate because the fire acts on a previously damaged structure. Moreover, fire protection elements may have been damaged by the earthquake and the fire can spread more rapidly if compartmentation walls have failed.
This paper presents the results of a numerical Fire following earthquake (FFE) analysis on steel columns belonging to a braced steel frame. In particular, experimental data regarding five FFE tests on steel columns, performed at the Federal Institute for Materials Research and Testing (BAM) by means of hybrid fire testing technique, were used to calibrate 3D numerical models. Two FFE tests were performed without fire protection, whereas the other three FFE tests were performed with different fire protection solutions: conventional and seismic-resistant calcium silicate boards and spray-based fire protection. Different modelling strategies were employed: beam and shell elements using the thermomechanical SAFIR software. The calibration of the column numerical models con- sidered boundary conditions estimated by computing the actual initial rotational stiffness of the end joints and a more representative temperature distribution based on the measurements of the ther- mocouples placed in different cross-section locations. The description and the results of the numerical calibration of modelling FFE scenarios applied to steel columns are thoroughly reported in the paper. The results of the calibrated model showed good agreement with the experimental data.

Numerous historical events highlight that the impacts of fire following an earthquake (FFE) can be notably greater compared to the damages and losses caused only by the earthquake. Major FFE events that occurred in the past include the San Francisco earthquake (1906), the Tokyo earthquake (1923), the Kobe earthquake (1995) and the Tohoku earthquake (2011). An FFE event recently occurred in Wajima (Japan) after the Noto earthquake (2024), in which the fire affected an estimated 200 buildings. Post-earthquake ignition sources identified from past earthquakes are reviewed by Botting and Scawthorn.

In brief, the principal ignition sources are the overturning of electrical appliances, short-circuiting of electrical equipment, gas leakage from damaged equipment and pipework, and leakage of flammable fluids. Damage to gas equipment and pipes can cause sparks and fuel fire propagation, while electrical appliances may ignite sparks with combustible materials. Incidents of leaking gas and damaged electrical appliances causing fires were ob- served after earthquakes such as Kobe and Northridge.

Earthquakes can result in single or multiple building ignitions, compromising the structural fire performance by damaging the fire protection elements and compartment measures, leading to rapid fire spread.

In this context, the structural fire performance can significantly deteriorate because the fire acts on an already damaged structure. Additionally, passive and active fire protections may have been com- promised by the seismic action, and the fire can spread more rapidly if compartmentation measures have failed. Thus, the seismic performance of non-structural components directly affects the fire performance of structural members. Minimizing non-structural damage is crucial in mitigating the potential decline in structural fire performance.

The loss of fire protection is particularly dangerous for steel structures due to their high thermal conductivity and thin profiles, leading to a rapid rise in temperature and a consequent quick loss of strength and stiffness.
Most of the literature focuses on numerical simulations of steel moment-resisting frames, with few studies dedicated to buckling-restrained and conventional brace systems. These studies have developed frameworks for evaluating the post-earthquake performance of steel structures in a multi-hazard context, incorporating tools for probabilistic struc- tural analyses under fire and seismic loads. Experimental studies have been conducted on single elements, steel–concrete composite beam-to-column joints, and full-scale reinforced concrete frames.

The literature review reveals that many numerical studies on the post-earthquake fire behaviour of structural components lack comprehensive experimental research support, and research on non-structural components is also limited. Based on these premises, the European project EQUFIRE was funded.

Experimental Test

The experimental tests at BAM were performed using a hybrid simulation technique as shown in Fig. 1a, in which the physical column (HEB 220 - S355) was firstly subjected to the horizontal and vertical displacement time-histories representing the seismic action and computed through numerical modelling in OpenSees. Then, the column was heated by the ISO 834 standard heating curve  and a constant numerical axial stiffness representative of the surrounding structure (a four-storey three-bay structure with concentric bracings in the central bay, as shown in Fig. 1b) was applied as boundary condition at the top of the physical column.

Numerous historical events highlight that the impacts of fire following an earthquake (FFE) can be notably greater compared to the damages and losses caused only by the earthquake. Major FFE events that occurred in the past include the San Francisco earthquake (1906), the Tokyo earthquake (1923), the Kobe earthquake (1995) and the Tohoku earthquake (2011). An FFE event re- cently occurred in Wajima (Japan) after the Noto earthquake (2024), in which the fire affected an estimated 200 buildings.

Post-earthquake ignition sources identified from past earthquakes are reviewed by Botting and Scawthorn. In brief, the principal ignition sources are the overturning of electrical appliances, short-circuiting of electrical equipment, gas leakage from damaged equipment and pipework, and leakage of flammable fluids. Damage to gas equipment and pipes can cause sparks and fuel fire propagation, while electrical appliances may ignite sparks with combus- tible materials. Incidents of leaking gas and damaged electrical appliances causing fires were ob- served after earthquakes such as Kobe and Northridge. Earthquakes can result in single or multiple building ignitions, compromising the structural fire performance by damaging the fire protection elements and compartment measures, leading to rapid fire spread.
In this context, the structural fire performance can significantly deteriorate because the fire acts on an already damaged structure. Additionally, passive and active fire protections may have been compromised by the seismic action, and the fire can spread more rapidly if compartmentation measures have failed. Thus, the seismic performance of non-structural components directly affects the fire performance of structural members. Minimizing non-structural damage is crucial in mitigating the potential decline in structural fire performance.

The loss of fire protection is particularly dangerous for steel structures due to their high thermal conductivity and thin profiles, leading to a rapid rise in temperature and a consequent quick loss of strength and stiffness.
Most of the literature focuses on numerical simulations of steel moment-resisting frames, with few studies dedicated to buckling-restrained and conventional brace systems. These studies have developed frameworks for evaluating the post-earthquake performance of steel structures in a multi-hazard context, incorporating tools for probabilistic struc- tural analyses under fire and seismic loads.

Experimental studies have been conducted on single elements, steel–concrete composite beam-to-column joints, and full-scale reinforced concrete frames. The literature review reveals that many numerical studies on the post-earthquake fire behaviour of structural components lack comprehensive experimental research support, and research on non-structural components is also limited. Based on these premises, the European project EQUFIRE was funded.

...Continua a leggere nel PDF in allegato.

L'intera memoria è in lingua inglese.

La presente relazione è stata presentata in occasione del XXIX Congresso CTA, svoltosi a Milano il 26 e 27 settembre 2024.