ABSTRACT. Predicting Fire Performance of Reinforced Polymeric Concrete Beams Using Finite Element Modeling Manish Sah1, Venkatesh Kodur1, Ahmed Almaadawy2, and Srishti Banerji2 1Michigan State University 1Utah State University *Correspondence to: kodur@egr.msu.edu

Keywords Polymeric concrete; High-temperature properties; Numerical model; Fire resistance

Introduction Polymeric concrete, a composite material containing aggregates together with full or partial amount of polymeric binder (as replacement to cement) exhibits superior strength, rapid curing, durability, and sustainability properties, as compared to conventional cement-based concrete [1]. Depending on the type and extent of cement replacement, polymeric concretes are generally classified into three categories: polymer concrete (PC), polymer-modified concrete (PMC), and polymer-impregnated concrete (PIC). These materials have gained increasing use in structural applications such as bridges, marine structures, and industrial facilities where long-term durability is essential [2]. Experimental studies at ambient temperature have shown that PC beams can achieve considerably higher flexural strength than normal strength concrete (NSC) beams. However, when used in building applications, polymeric concrete members must satisfy the required fire ratings to comply with building code requirements. Recent studies have shown that despite their superior ambient performance, polymeric concretes exhibit reduced material properties under fire exposure as compared to NSC [3]. This reduction is primarily due to the presence of polymer resin, which undergoes rapid degradation at elevated temperatures. Currently, the fire ratings of reinforced concrete (RC) members specified in building codes are determined using prescriptive methods, where the required fire resistance is ensured by meeting minimum section dimensions and concrete cover thickness to reinforcement. These provisions are based on standard fire tests conducted mainly on NSC members subjected to service loads under standard fire exposure. Due to differences in composition and microstructure, the fire behavior of polymeric concrete members can vary significantly from that of NSC members under realistic fire conditions. To address this knowledge gap, the present study develops a finite element (FE) model in ABAQUS to predict the fire response of RC beams made with PC and PMC.

Numerical Model The developed model incorporates geometric details, element types, material properties, loading, support conditions, and fire exposure. The RC rectangular beam tested by Dwaikat and Kodur (2009) [4] was used to validate the model due to the availability of comprehensive experimental data. The validated model is then applied to predict the thermal and structural response of PC and PMC beams under fire. PC/ PMC Beams Selected for Analysis A simply supported beam, as shown in Figure 1, is considered in this numerical study. Each beam is subjected to two-point loads (P) corresponding to 55% of its room-temperature capacity, while the middle 2.44 m is exposed to the ASTM E119 standard fire. A total of six beams are analyzed, including two polymer concrete (PC) and four polymer-modified concrete (PMC) beams. The PC beams include plain epoxy-based polymer concrete (PC-EP-PLAIN) and epoxy-based polymer concrete with basalt fiber (PC-EP-BF). The PMC beams include plain polymer-modified concrete (PMC-PLAIN), polymer-modified concrete with basalt fiber (PMC-BF), polymer-modified concrete with aluminum hydroxide (PMC-ATH), and polymer-modified concrete with both basalt fiber and aluminum hydroxide (PMC-BF-ATH). Basalt fiber is used to improve high-temperature mechanical stability and crack resistance, and ATH is added as a flame-retardant filler to enhance fire performance.

Figure 1. Elevation and cross-section of the reinforced concrete beam. Mid-Span Deflection Response

The model was validated against the thermal and structural responses of the NSC beam and subsequently applied to predict the corresponding responses of the PC and PMC beams. Figure 2 shows the variation of the predicted mid-span deflection with exposure time for RC beams made with PC, PMC, and NSC under the ASTM E119 standard fire. The NSC beam exhibited a gradual increase in deflection and sustained the applied load for about 185 minutes, whereas the PC-EP-PLAIN and PC-EP-BF beams failed much earlier at around 45 and 23 minutes, respectively, showing rapid deflection. Among the PMC-based beams, PMC-PLAIN had the lowest fire resistance of 115 minutes, while PMC-BF and PMC-BF-ATH showed improved performance with longer failure times (173 and 165 minutes respectively) compared to PMC-ATH, indicating enhanced fire endurance due to basalt fiber and ATH additions.

Figure 1. Elevation and cross-section of the reinforced concrete beam. Detailed Results Detailed thermal and structural response results for PC, PMC, and NSC beams will be presented and discussed in the paper. In addition, the fire resistance of these beams will be evaluated based on the failure limits criteria to provide a comprehensive understanding of their performance under elevated temperatures. References [1] ACI Committee 548. 548.6R-96: Polymer concrete-structural applications. State-of-the-art report. American Concrete Institute; 1996 [p. 1–23]. [2] Taha, M. R., Genedy, M., and Ohama, Y. (2019). Polymer concrete. In Developments in the Formulation and Reinforcement of Concrete (pp. 391-408). Woodhead Publishing [3] Sah, M., Kodur, V., Almaadawy, A., and Banerji, S. (2025). Characterizing the thermal properties of polymer concrete at elevated temperatures. Cement and Concrete Composites, 163, 106175. [4] Dwaikat, M. B., and Kodur, V. K. R. (2009). Response of restrained concrete beams under design fire exposure. Journal of structural engineering, 135(11), 1408-1417.

ABSTRACT. Introduction Bridges are critical components of transportation networks, whose functionality after extreme events is essential for maintaining post-disaster connectivity and overall network resilience. For deck-type bridges - commonly adopted in urban areas - piers play a fundamental role, as they largely govern the seismic vulnerability of the entire system. Owing to their structural function, bridge piers are prone to damage during earthquakes and may subsequently be exposed to additional hazards, such as fire resulting from vehicle collisions or other anthropogenic sources. Consequently, assessing the multi-hazard vulnerability of piers subjected to both seismic and fire actions is crucial for enhancing infra-structure resilience. Importantly, the earthquake–fire combination considered herein does not stem from a direct cause–effect relationship; rather, the focus is on the behaviour of a pier exposed to fire after sustaining light to moderate seismic damage levels compatible with the bridge remaining in service. Recent advances in multi-hazard risk assessment underscore the need for integrated meth-odologies capable of capturing such sequential, non-cascading hazard scenarios, where pre-existing damage may alter the structural response to subsequent events. Methodology This study proposes a multi-hazard vulnerability framework for reinforced concrete (RC) bridge piers subjected to sequential, yet non-casually related, earthquake and fire events. The primary objective is the derivation of multi-hazard fragility curves that quantify the probability of exceeding predefined performance levels as a function of the combined effect of both hazards (Fig.1).

Figure 1. Synopsis of the proposed methodology for multi-hazard earthquake-fire vulnerability assessment Differently from conventional approaches focusing on residual strength or load-bearing capacity, the present methodology defines structural damage through the maximum total displacement demand, computed as the sum of the residual displacement induced by the earthquake and the additional displacement generated during the subsequent fire exposure. The seismic response of the pier is assessed using nonlinear static analysis. From the resulting capacity curve, four conventional dam-age states (DS1-DS4) are identified based on yielding and ultimate displacements [1]. The analysis focuses on DS1 and DS2, corresponding to slight and moderate damage levels as being consistent with post-earthquake serviceability conditions, under which a fire event may realistically occur. The fire scenario considers a vehicle-induced fire beneath the bridge deck, in close proximity to the pier shaft. To analyse this condition, the bridge is modelled in FDS (Fire Dynamics Simulator by NIST [2]), where fluid dynamics simulations of single and multi-vehicle combustion are performed, provid-ing time-temperature histories. These thermal curves are subsequently applied in SAFIR® [3] to the structural model, initially configured in its DS1 and DS2 deformed state. Transient thermomechanical analyses capture temperature-dependent degradation of concrete and reinforcing steel, as well as additional displacements arising from differential thermal expansion and stiffness reduction. The total displacement demand is thus obtained by combining the residual seismic displacement, at either DS1 or DS2, with the maximum displacement induced by the fire. This displacement-based damage metric enables a consistent representation of the cumulative effects of sequential hazards. Displacement capacity is evaluated using two innovative performance levels (PL3 and PL4), defined as functions of pier height and representing increasing thresholds for acceptable lateral displacement prior to substantial functional loss or collapse [4] . Finally, a cloud analysis is conducted to quantify the probabilistic response. The resulting dataset of demand-to-capacity ratios is used to derive multi-hazard fragility curves, expressed in terms of fire intensity measures, namely the peak heat release rate (HRR) and the fire load (Qf). These curves pro-vide the conditional probability of exceeding PL3 and PL4 varying fire intensities, thereby capturing the influence of pre-existing seismic damage on the fire response of the pier. Results The results indicate that even moderate seismic pre-damage (DS1 or DS2) significantly increases the displacement demand during subsequent fire exposure, leading to higher demand-to-capacity ratios across all fire intensity levels. As the HRR and fire load increase, both thermal expansion and tem-perature-induced stiffness degradation further amplify the deformations accumulated during the fire event. The resulting multi-hazard fragility curves exhibit a systematic shift toward higher exceedance probabilities, compared with those obtained from isolated single-hazard analyses. These findings highlight the necessity of explicitly accounting for sequential exceptional events in the safety and performance assessment of bridge infrastructure. The proposed methodology offers an integrated framework capable of capturing the combined effects of seismic and fire actions on RC bridge piers. The main conclusions of the study can be summarized as follows. • The combination of seismic pre-damage and subsequent fire exposure significantly increases structural vulnerability, even for limited initial damage levels, as demonstrated by the D/C values. • Cloud-based D/C analysis proves to be an effective and robust strategy for deriving multi-hazard fragility curves associated with appropriate intensity measures. • The proposed approach advances multi-hazard vulnerability assessment of critical infrastruc-tures, providing a methodological foundation that extends beyond the earthquake-fire sequence. • The framework can be adapted to investigate other combined hazard scenarios, e.g., corrosion-fire or fatigue-fire interactions, and to analyse different pier typologies and more complex struc-tural systems, thereby facilitating the transition from component-level to system-level evaluation.

References [1] Moschonas, I. F., Kappos, A. J., Panetsos, P., Papadopoulos, V., Makarios, T., & Thanopoulos, P. (2009). Seismic fragility curves for Greek bridges: methodology and case studies. Bulletin of Earthquake Engineering, 7(2), 439-468. [2] National Institute of Standard and Technology. Fire Dynamics Simulator. [3] Franssen, J. M. (2005). SAFIR: A thermal and structural program for modeling structures under fire. Engineering Journal, 42(3), 143-158. [4] de Silva, D., Miano, A., De Rosa, G., Di Meglio, F., Prota, A., & Nigro, E. (2025). Analitycal fire fragility assessment for bridges considering fire scenarios variability. Eng. Structures, 325, 119442.

ABSTRACT. Cold-formed steel (CFS) is increasingly used for mid-rise structural systems, yet performance of these systems in real fires and in multi-hazard scenarios remains insufficiently understood. While recent research has advanced knowledge of fire performance of CFS assemblies, data on thermal and structural behavior in CFS buildings exposed to fire are scarce. To support performance-based design methods, a ten-story CFS building was tested on the NHERI LHPOST shake table at UCSD (CFS10). Following a sequence of eighteen multi-directional earthquake motions, two compartment fire tests were conducted in July 2025. One of the key objectives of the program was to advance numerical modeling: finite element (FE) models were used both before the tests to inform the test design and after the tests to compare and improve the models. This paper will present the numerical modeling strategy, the comparison of predicted behavior with the full-scale test data, and parametric studies with the calibrated numerical models. Overall, the integrated modeling-testing program shows the value of predictive numerical modeling and provides unique benchmark data to enhance performance-based structural fire design methods for CFS structures in fire.

ABSTRACT. Large-scale fire tests provide essential insight into the real behaviour of steel structures exposed to fire, particularly where interaction between structural and cladding components influences global stability. This paper presents a full-scale fire test of a single-storey steel portal frame hall with trapezoidal steel cladding subjected to a localized timber crib fire and a simultaneous horizontal mechanical load. Gas and structural temperatures, as well as global and local deformations, were monitored using thermocouples, plate thermometers and displacement transducers. Peak gas temperatures reached approximately 1100 °C in the flame region, while structural temperatures in the rafter above the fire exceeded 900 °C. Despite severe thermal exposure and local damage of secondary members, the primary load-bearing structure maintained global stability, and frame deformations were largely reversible after cooling. A numerical simulation based on the CFAST zone model was developed to simulate temperature distribution. The model showed reasonable agreement with experimental results, although uncertainties in the heat release rate significantly affected prediction accuracy. The findings provide valuable experimental data for validation of advanced numerical models and for improving fire design approaches for steel halls.

ABSTRACT. This study examines the elevated-temperature mechanical properties of seven widely used stainless steel grades and develops improved predictive models for structural fire design. Current standards impose a minimum ultimate strain of 0.02, but test data show that about one-quarter of ultimate strains fall below this value when the temperature is above 850 °C, indicating shortcomings in existing provisions. A new stress–strain formulation is proposed to accommodate both low and high ultimate strains and more accurately capture material behaviour in fire. New predictive equations for elastic modulus, yield stress, ultimate strength, and ultimate strain are developed, providing average and lower-bound values and demonstrating improved agreement with test data.

ABSTRACT. Introduction Fire poses a significant risk in the case of steel structures. Indeed, the three components contributing to the definition of risk, i.e., vulnerability, hazard, and exposure, may be highly present in steel structures. In particular, vulnerability tends to be higher in buildings lacking fire protection or exhibiting low structural redundancy. Furthermore, large steel structures often store machinery or materials that contribute to an increased fire load, while also hosting substantial numbers of occupants, thereby increasing the hazard and the exposure. Consequently, the potential consequences of fire must be carefully assessed, especially in the case of older and larger steel structures. This paper investigates the structural response of the arch-suspended steel roof of an ice stadium (see Figure 1) during a fire. The stadium was built in Italy and was opened in 1989. It measures 90 m in length, 55 m in width, and 17 m in height, increasing to 25 m if the external arch is included. The investigated steel roof structure is connected to the reinforced concrete perimeter walls. The steel roof and arch systems were designed to withstand significant snow and seismic loads. The study aims to assess the fire response of the structure following a potential renovation of the first floor stands, which involves the installation of plastic seating (see Figure 1 b)). The ignition of a plastic seat and the subsequent fire spread among adjacent seats are simulated with Computational Fluid Dynamics (CFD) analysis using the Fire Dynamics Simulator (FDS) software [1]. Heat transfer and structural analyses are then carried out using SAFIR [2] to capture the complex non-linear behaviour of the structure. A preliminary investigation showed that, despite the low-redundant design, the structure can survive a fire spreading in the changing room (see Figure 1 b)) located at the ground storey. Indeed, the temperatures of the smoke layer impacting the roof remain low, as shown in Figure 2a), having little effect on the structure’s stability. However, the ability of the steel roof to withstand a more severe fire scenario, such as a fire closer to the ceiling, has yet to be demonstrated.

Next Steps The results of the CFD analysis for the plastic seats fire will be integrated into the thermal and structural analyses of the steel roof. The goal is to assess whether the structure can withstand the whole duration of the fire, as required by the Italian Fire Prevention Code [3] for this type of structure. A fire protection measure will be designed in the case that the structure cannot survive the fire, and its effectiveness will be assessed through additional numerical analyses.

References [1] McGrattan, K., Hostikka, S., McDermott, R., Floyd, J., Weinschenk, C., & Overholt, K. (2013). Fire dynamics simulator user’s guide. NIST special publication, 1019(6), 1-339. [2] J.M. Franssen, T. Gernay, Modeling structures in fire with SAFIR: theoretical background and capabilities, J. Struct. Fire Eng. 8 (2017) 300–323, https://doi. org/10.1108/JSFE-07-2017-0023 [3] Approvazione di norme tecniche di prevenzione incendi, ai sensi dell’articolo 15 del decreto legislativo 8 marzo 2006, n. 139, GU n. 192 del 20/8/2015 – S.O. n. 51, 2015 in Italian

ABSTRACT. Research background High-strength steel (HSS) offers superior yield strength, tensile strength, and corrosion resistance compared to mild steel, which is widely used in high-rise and long-span structures [1]. HSSs generally exhibit relatively low ductility and greater susceptibility to fracture, while exposure to fire significantly degrades key mechanical properties (yield strength and fracture ductility) potentially leading to structural fracture and collapse [2–5]. Such failures can result in severe casualties and substantial economic losses. To authors’ knowledge, existing research on fracture performance at elevated temperatures has primarily focused on conventional strength steels [5], whereas systematic investigations into high-strength steels remain limited. Moreover, most existing studies simplify loading conditions to uniaxial tension, whereas actual steel structures experience complex three-dimensional stress states. Furthermore, available models seldom incorporate the coupling effect of temperature and stress for HSSs, and thus fail to adequately capture the influence of multiaxial stress on high-temperature fracture behavior. Therefore, the fracture ductility of HSSs under high temperatures is essential to enable accurate fire safety assessment and collapse prevention in HSS structures. Research content In this study, smooth round bars, notched round bars, double-edge notched plates, and four-side notched plates were designed and tested. The specimens were heated using a split-type atmospheric high-temperature furnace and subjected to tensile tests under elevated temperatures with a universal testing machine. By integrating the digital image correlation (DIC) technique, full-field strain measurements were obtained alongside stress–strain data, thereby elucidating the fracture characteristics of Q460 HSS at high temperatures. Building on the experimental results and grounded in phase-field theory, a phase-field fracture model was innovatively developed specifically for Q460 HSS during fire exposure. Conclusion 1. Affected by the blue brittleness effect, the steel ductility degrades at 200℃, When above 300℃,with the increase in temperature, the fracture displacement of Q460 high-strength steel exhibits a gradual decreasing trend;, 2.For a given Lode angle, the fracture ductility decreases with the increase in stress triaxiality. The fracture ductility decreases with the increase in stress triaxiality and increases with the increase in Lode angle. 3.The fracture prediction model constructed based on phase-field theory, which effectively provides reliable predicting ductile fracture of Q460 HSS under complex stress conditions in high-temperature environments.

ABSTRACT. Unprotected steel elements are often fixed to protected structural steel elements. In the event of a fire, the unprotected element can create a localised hotspot in the primary element, potentially leading to premature failure. To prevent this unwanted heat transfer, a coat-back is applied to the unprotected secondary steel element [1-2]. Coat-backs are typically applied to the unprotected steel elements, extending 450 [3,4] - 500 [5,6] mm in length (at the same level of protection as the primary protected element), to prevent unwanted heat transfer to the primary (protected) steel elements.

Previous research on coat-backs has mainly focused on open-section structural elements and has not addressed their use on hollow profiles. This paper investigates the impact of coat-backs on hollow profiles via heat transfer analysis using both two-dimensional and three-dimensional methods through the Finite Element Method (FEM). It assesses the impact of coat-backs on unprotected hollow steel elements that frame into protected I-section columns which was not covered under the original research carried out by the SCI [3]. The study highlights the necessity to reevaluate current best practices when dealing with unprotected hollow sections that frame into protected primary elements under fire conditions.

ABSTRACT. The fire resistance of columns is essential in the design of multi-story buildings and thus generally well understood. However, the use of columns made from unconventional construction materials render it necessary to use a performance-based design approach. This paper presents investigations carried out to prove the resistance to fire of 60 minutes of massive granite columns. The assessment included initial tests of the thermal behaviour of different granite types subjected to the ISO 834 Standard Fire Curve. All specimens suffered significant spalling and cracking, probably caused by the α-β transition of quartz leading to a high thermal expansion at 573 °C. Therefore, gas temperatures were computed in CFD simulations with the software FDS for further experimental investigations. The envelope gas temperature of the simulated scenarios was later employed for loaded column fire tests conducted at the Bundesanstalt für Materialforschung und -prüfung (BAM) in Berlin. To understand the vertical crack development seen in all specimens, these tests were thermo-mechanically simulated in the finite element software Abaqus (Dassault Systèmes), considering material and geometric nonlinearity.

ABSTRACT. A sequentially coupled CFD-FEM modelling approach consisting of fire modelling, heat transfer analysis and mechanical analysis was adopted to simulate the response of post-tensioned concrete bridge beams exposed to hydrocarbon fire. Firstly, the furnace fire model was built and solved by fire dynamics simulator (FDS). The adiabatic surface temperature obtained from fire modelling was then applied on the surfaces of the finite element model of the structure as temperature boundary conditions for heat transfer analysis. Finally, the temperature-dependent nonlinear mechanical analysis was conducted. Special attention was paid on the modelling of bond and spalling phenomenon. The bonding between the strand and surrounding grout was modelled with adhesive element. The spalling was modelled with user-defined material model UMAT. The numerical approach was validated by fire tests on six scale bridge beams. The simulated thermal and structural responses agreed well with the experimental results, demonstrating the feasibility of the modelling approach.

ABSTRACT. This contribution presents and evaluates the results of spalling tests and four-point bending tests with partial heating on CFRP-reinforced concrete specimens.

ABSTRACT. Tunnel fire resistance is traditionally ensured by imposing temperature limits on the reinforcement and the exposed concrete. Meeting these limits is generally interpreted as proof that the required fire rating is achieved and that the structure remains repairable after a fire. Such temperature-based checks avoid explicit structural analysis and do not provide information on the actual structural damage state, nor the repairability. To move beyond this prescriptive approach, this paper presents a dedicated post-processing tool for SAFIR that quantifies post-fire damage and repairability in reinforced concrete tunnels by combining thermal and mechanically induced damage. The method enables a direct assessment of residual structural performance and supports a more rational evaluation of repairability. The tool is demonstrated through a case study of an existing cut-and-cover tunnel with spalling-sensitive concrete, in which different passive fire protection configurations are analysed. The results show that significant structural damage may develop, even in tunnels equipped with substantial amounts of passive fire protection. Furthermore, the study illustrates how the proposed tool can inform decision-making and contribute to the optimization of passive fire protection strategies for concrete tunnels.

ABSTRACT. Within mass timber connections, small gaps can occur due to construction tolerances or the natural creep of timber products. These gaps may have negative impacts on the fire performance of the connections, enabling fire to spread from floor to floor or loss of structural capacity. To investigate the heat transfer within mass timber connections due to these gaps, three unprotected glulam beam column connections were exposed to a one-hour standard fire (ASTM E119). With gaps sizes from 0 mm to 8 mm, the authors found that gaps sizes up to 3 mm had no substantial impact on char penetration while gaps greater than 5 mm allowed for almost complete charring at the gap interface. Despite this, the charring observed did not reflect a fully exposed surface and current fire design standards may be overconservative when addressing these gaps.

ABSTRACT. Timber elements lose strength under fire due to charring and thermal degradation, and current Eurocode design practice represents this through a char layer and a prescribed zero-strength layer (ZSL). However, recent research suggests that a fixed ZSL thickness does not consistently reflect the loss of capacity across different fire exposures. This study develops a probabilistic framework to quantify how varying the ZSL affects the burnout reliability of timber columns subjected to Eurocode Parametric Fire Curves. High-resolution heat-transfer simulations generate temperature fields for a wide range of opening factors, fuel loads, and column sizes, which are paired with probabilistic models for temperature-dependent timber strength to derive structural resistance. Comparing probabilistic load and resistance distributions yields the probability of failure for different ZSL assumptions. Results across 100 fire scenarios show large variability in required ZSL thickness, where in some severe fire exposures, these targets cannot be met. The findings highlight that fixed ZSL values can lead to inconsistent safety levels and support adopting a reliability-based, performance-oriented definition of the ZSL in timber fire design.

ABSTRACT. This paper presents a numerical investigation of twenty recent fire tests on loaded glulam columns. The tests, which focus on the response until burnout, include eight furnace tests under standard fire followed by controlled cooling, and twelve natural compartment fire tests. Finite element analyses were performed using SAFIR and in accordance with the Eurocode 5 advanced method. The analyses combine two-dimensional heat-transfer analysis with structural analysis with fiber-based beam elements using effective temperature-dependent properties. Model predictions were compared with measured temperatures, displacements, and failure times across a range of fire exposures and column sizes. The analyses provided generally good and conservative agreement with the tests, capturing the phenomenon of thermal wave resulting in delayed failure, although discrepancies were observed in internal temperatures and failure time especially during the decay phase, reflecting limitations of properties calibrated under standard fire. Three experiments on 40x40 cm2 section columns exhibited smoldering-induced failures not captured by the model. The full paper will provide detailed comparative analyses of thermal and structural response across the twenty tests.

ABSTRACT. 1.0 Introduction Testing for a fire resistance rating (FRR) of mass timber frame connections between beam to column or beam to girder has become more relevant in the past decade as mass timber buildings increase in height and complexity. Mass timber buildings utilizing engineered timber such as glulam for the structural frame may require connections with a proven FRR, which requires ad-hoc fire testing as methods are not included in national standards, leading to inconsistencies. 2.0 Structural connection fire resistance The fire resistance of connections for any structural material are not specifically addressed in national standards, such as ASTM E119, EN 1363 & 1365 series, CAN/ULC S101, or AS 1530 pt 4. These standards address floors, walls, beams and columns. Floor or wall connections have their fire resistance proven as part of the tested system, whereas frame connections (i.e. beam to a column) do not have a test method. Tests undertaken and the results are therefore unique. 3.0 Completed full-scale fire resistance testing Fire resistance testing of glulam connectors has primarily been research based, conducted on medium scale timber elements in tension [1]. These typologies are easy to construct, fire test and analyse, though limited in findings as they do not reflect scale, stresses or fire exposure of actual mass timber frame connections. A limited number of glulam beam to column connection FRR tests have been completed that replicate both the size and applied forces of actual building conditions. The first published was by ETH in 2014 [2], testing a range of connectors to 90 mins. In the US, a FRR testing method was developed for high-rise projects, based on the ETH approach and over 30 research and product development FRR tests have been completed [3].

Figure 1a: Completed 120 mins glulam beam to column test with applied shear force, exposed to ASTM E119 Figure 1b: Post-fire test disassembly of a glulam two-part connector (used with permission)

A French research series by CTSB has been on-going nearly 25 years with recent tests replicating shear and bending of building construction [4]. Fire testing of full-scale glulam connections has also occurred in Canada by FPInnovations, National Research Council and Lakehead University. Connection research is ongoing at universities such as Canterbury (NZ), Oregon State and Stellenbosch. Project specific fire resistance testing has occurred in Austria, Germany and Australia, though with few details being publicly available. 4.0 Observations and lessons learnt from connection fire resistance testing A review of the full-scale fire resistance testing of glulam beam to column connections (published and some unpublished) shows that test methods are not consistent in set-up, load-application, allowing for rotation and deflections, pass / fail criteria and test monitoring. The methods are generally reasonable, though some could result in load-sharing and inhibit rotation. The lack of a published test method in guidance or a national standard leads to the ad-hoc nature of test setup, limits comparison of results and restricts improvements to be made. 5.0 The need for fire resistance testing Driven by project demands due to code changes allowing taller and larger mass timber buildings, connection hardware manufacturers are continually innovating, with concealed steel or aluminium connectors that offer increased shear capacity, allow for lateral drift and provide ease of on-site assembly. Thus, testing for FRR of full-scale glulam connectors with combined bending and shear is continually needed for building approval, research on failure modes and improving engineering analysis methods [4]. Some connection hardware manufacturers have seen a competitive advantage by fire testing and making the results available to designers [5]. 6.0 Recommendations for fire resistance testing of beam to column connections A new FRR testing methodology for mass timber frame connections requires a motivated national standard committee who understand the need. The full paper provides a recommended FRR test methodology for mass timber frame connections, based on completed tests and lessons learnt, to assist with future fire testing, provide guidance and inform changes to national standards. The method addresses: Support and reaction frame, furnace size, species of timber being tested, boundary conditions, restraint, secondary effects of exposed mass timber, application of load, reaction points, possibility of load-sharing, allowing for deflection and rotation, deflection and temperature monitoring, fail criteria, post-test analysis. References [1] Maraveas, C., Miamis, K., Mathew, C., 2015 Performance of Timber Connections Exposed to Fire: A Review Fire Technology 51, 1401–1432 [2] Palma, P., Frangi, A., Hugi, E., Cachim P., Cruz H., 2014 Fire Resistance Tests on Beam-to-Column Shear Connections 8th International Conference on Structures in Fire, Shanghai, China [3] Barber D., 2018 Fire Tests to ASTM E119 on Full-Size Glulam Beam to Column Connections 10th International Conference on Structures in Fire, Ulster University, Belfast [4] Audebert M., Dhima D., Bouchaïr A., 2020 Proposal for a New Formula to Predict the Fire Resistance of Timber Connections, Engineering Structures, Volume 204 [5] Simpson Strong-Tie Heavy Seated Plate Connector, 2024 Summary of ASTM E119 Fire Testing www.strongtie.com/facemounthangers_engineeredwoodconnectors/hskp_hanger/p/hskp

ABSTRACT. Concealed connections are increasingly used in North America for architectural purpose and to improve their fire resistance. A review of recent fire performance studies indicates that, in such assemblies, the unprotected gap between connected members is one of the key parameters influencing fire performance. However, the effects of gap size and wood grain orientation on thermal propagation and char penetration in mass timber joints remain insufficiently understood. This study investigates thermal and char penetration within timber joints featuring gap sizes of 3.2, 4.8, and 6.4 mm (1/8”, 3/16” and 1/4") and different grain orientation combinations. Specimens were tested in an intermediate‑scale furnace under standard fire exposure in accordance with CAN/ULC S101. Results show that increasing gap width systematically increases char penetration within the joint. Longitudinal Longitudinal configurations exhibit deeper charring than Transversal Transversal configurations. Comparison with the American charring model detailed in the Fire Design Specification for Wood Construction indicates that its predictions are conservative, aligning more closely with the behaviour observed for 4.8 and 6.4 mm rather than 3.2 mm gaps. These findings provide new insight into char penetration mechanisms in fully concealed mass timber connections and support the development of improved design guidelines improving the fire safety of timber construction.

ABSTRACT. In a fire, the load-bearing capacity of timber structures is influenced not only by charring but also by the elevated temperatures in the wood behind the char layer. In structural load-bearing assessments, this effect can be accounted for using the effective cross-section method. The procedure specified in the next-generation version of Eurocode EN 1995-1-2:2025 [1] applies to timber and cross-laminated timber (CLT) structures exposed to standard fire conditions, as well as to those subjected to natural fire scenarios, provided that the bond line integrity is maintained in fire. For production-related reasons, adhesives that soften before the onset of charring (i.e., where integrity is not maintained) are still commonly used in CLT manufacturing. Such adhesives include many polyurethane-based (PUR) products. The Eurocode standard does not provide guidance for determining the effective charring depth of such CLT structures under natural fire exposure, including the decay phase. This paper presents the implementation and results of five load-bearing fire tests conducted to investigate heat transfer, temperature distribution, and charring behaviour in CLT floor slab elements exposed to standard and natural fire conditions. The natural fire scenarios were modelled using the Eurocode parametric fire model, and the specimens were tested under mechanical loading throughout the tests. The study also investigates how the experimental results compare with those predicted by the Eurocode for CLT structures in which bond line integrity is maintained.

ABSTRACT. This study investigates the residual compressive strength of low-cement concrete (LCC) after heating–cooling fire cycles. A total of 90 elevated-temperature tests were conducted on six mixes: an OPC control, 35% fly ash (FA), and 50% ground granulated blast-furnace slag (GGBS), each with and without polypropylene (PP) fibres. Cylinders (100 × 200 mm) were heated to 400 °C or 600 °C under sustained preload levels of 0% and 30%, then cooled to intermediate hold temperatures or ambient. Residual strength reduction factors (RF) were obtained from post-fire stress–strain response and compared with Eurocode cooling predictions. Peak temperature dominated degradation, with 600 °C giving markedly lower retained capacity than 400 °C across all mixes. The extra strength loss associated with cooling to ambient, relative to higher cooling-hold states, often exceeded the 10% post-peak reduction assumed in Eurocode models and reached approximately 20% in the most severe cases. Sustained preload improved retention, particularly at 600 °C, suggesting restraint limits cooling-stage crack growth. Binder composition affected trends, with SCM-rich mixes showing mix-dependent degradation and PP-fibre effectiveness. Overall, temperature-only code reduction factors may not capture the combined influence of cooling history, sustained loading, and LCC binder chemistry, supporting improved post-fire assessment of low-carbon concrete systems.

ABSTRACT. Reinforced concrete (RC) members may retain sufficient load-bearing capacity during fire, yet undergo substantial thermal degradation that affects their post-fire performance. The cooling phase (full burnout) is recognized as a critical stage influencing residual strength. Meanwhile, prior studies have examined bond behavior between reinforcing steel and concrete at elevated and residual temperatures; however, a gap remains in understanding how bond degradation evolves across the heating, cooling, and post-fire phases, and how these changes influence structural capacity. This study investigates the coupled effects of full burnout and temperature-dependent bond-slip degradation on the residual performance of RC beams. Experimental beam tests from the literature that include heating, cooling, and post-fire loading are used as benchmarks. Published data on high-temperature and residual bond behavior are synthesized to construct a continuous transition model for bond degradation throughout all thermal phases. A finite element framework in Abaqus is developed to explicitly represent rebar-concrete interaction using temperature-adjusted bond stress–slip relationships. Baseline simulations assuming perfect bond establish the reference response against which the impact of bond deterioration is quantified. The findings provide new insight into mechanisms governing post-fire capacity loss and support more reliable assessment methodologies for fire-exposed RC structures.

ABSTRACT. Introduction Reinforced concrete (RC) structures generally exhibit good fire resistance due to the non-combustible nature of concrete and its relatively low thermal conductivity. Nevertheless, severe thermal gradients across the section and material degradation within concrete cover can still induce significant damage and compromise structural integrity. Tunnels represent a particular case in which fire exposure can be especially severe, as the surrounding structure restricts heat dissipation, leading to temperatures often exceeding 1000°C. Although complete or partial collapse of tunnel linings due to fire is rarely observed, the functionality and resilience of these infrastructures are strongly affected [1]. Among the various consequences of tunnel fires, explosive spalling represents one of the most critical dam-age mechanisms, as it can rapidly reduce the load-bearing capacity and durability of concrete linings. To mitigate these effects, the use of fiber-reinforced concrete (FRC) has emerged as an effective mitigating solution. FRC is a composite material that has proven to be a competitive and versatile solution for a wide range of structural applications. In particular, the use of FRC in the production of precast tunnel segments offers several advantages for tunnel linings constructed using tunnel boring machines (TBM) [2]. The incorporation of fibers, especially polypropylene and steel, has been shown to significantly reduce the risk of explosive spalling by improving tensile ductility, controlling crack propagation, and enhancing pore pressure release during heating. In this context, the present study aims to characterize the fire behaviour of FRC through an extensive literature survey, with particular attention to the mechanisms governing explosive spalling. The analysis seeks to (i) identify the key parameters influencing this phenomenon and (ii) outline the current state of knowledge as well as the research gaps relevant to the fire performance of fiber-reinforced concrete in tunnel applications. Spalling in Concrete Recent studies have identified three main mechanisms contributing to explosive spalling, which may act individually or in combination: (i) pore pressure build-up, caused by the transformation of water into vapor within the concrete pores - since permeability is low, the vapor becomes trapped, generat-ing internal pressures that can break the surface layer; (ii) thermal stresses, induced by steep temper-ature gradients and restrained thermal expansion, producing compressive and tensile stresses that may exceed the concrete’s strength; and (iii) combined effects, where the simultaneous action of va-por pressure and thermal stresses increases the likelihood of material detachment. In addition to these mechanisms, load-induced thermal strains (LITS) influence the structural response of concrete at high temperature: in elements subjected to mechanical loads during heating, LITS introduce addi-tional inelastic deformations that modify the internal stress distribution and may intensify the effects of thermally induced stresses, indirectly increasing the susceptibility to spalling. In tunnel structures, the high confinement provided by soil limits thermal expansion, increasing thermal stresses and fur-ther heightening the risk of explosive spalling. When exposed to high temperatures, often exceeding 1000°C, this restriction causes internal pressure to build up, which in turn increases the risk of explo-sive spalling. Figs. 1(b,c) show temperature distribution and stresses in a typological tunnel subject-ed to RWS. Mitigation of Spalling in Concrete through Fibers The performance of FRC strongly depends on the type, geometry, and distribution of fibers within the matrix. Steel, polypropylene, and carbon fibers each provide distinct benefits, such as enhanced tensile strength, ductility, crack control, or impact resistance. Hybridization, combining two different fiber types in a single matrix, has been shown to optimize the overall properties of FRC. Combining steel and polypropylene fibers offers an ideal balance: steel fibers improve tensile and impact re-sistance, while polypropylene fibers reduce the risk of explosive spalling under fire by creating addi-tional porosity as they melt. The evidence indicates that the integrated use of steel and polypropyl-ene fibers in appropriate proportions provides superior structural performance, durability, and fire resilience, making hybrid FRC the preferred solution for tunnel linings and other demanding applica-tions. Several studies have demonstrated that the addition of dispersed fibers in the cement matrix can significantly reduce the depth of spalling in concrete, as shown in Fig. 1(a). In particular, poly-propylene fibers limit spalling to superficial delamination due to their melting at approximately 170 °C, close to the temperature range where spalling typically occurs [3]. During melting, the fibers are par-tially absorbed by the cement matrix, creating pathways that enhance gas migration and relieve inter-nal pore pressure. Experimental investigations on high-strength concrete have shown that these fi-bers increase matrix permeability, facilitating mass transport and reducing pressure peaks associated with spalling. The effectiveness of this mechanism depends on the fiber dosage, which must exceed a critical filtration threshold to form a connected system of preferential pathways for gases and flu-ids. This process underscores the importance of fiber content and distribution in mitigating fire-induced spalling in concrete elements. Beyond their role in spalling prevention, fibers in concrete are increasingly recognized for their structural potential. Research has shown that fibers can act not only as auxiliary reinforcement but also as a partial alternative to steel, contributing to the load-bearing capacity and stability of elements such as tunnel segments and rings under fire conditions [4, 5, 6]. These studies highlight the two-fold function of fibers in enhancing both fire resistance and structural performance.

Figure 1. (a) concrete spalling due to fire curve RWS by concrete type (b) temperature and internal forces in a typological tunnel at t=60min (c) temperature and stresses at t=120 min. Next steps The full paper will present a comprehensive literature survey about the fire behavior of FRC, focusing on the mechanisms governing explosive spalling. It will delve into key studies related to the interac-tion between fiber composition (steel, polypropylene, and hybrid fibers) and fire resistance, aiming to identify the critical parameters that influence the performance of FRC under fire conditions. The final paper will also explore various types of fibers and their effects on the fire resistance of concrete, discussing the available experimental evidence and identifying gaps in current research. Finally, the analysis will highlight areas for future research, identifying open issues and proposing new ap-proaches to enhance the fire resistance of FRC, particularly in tunnel applications and other critical infrastructure. References [1] Hua, N., Tessari, A. F., & Elhami-Khorasani, N. (2020). Damage assessment framework for tunnel structures subjected to fire, In. Pr. of SiF 2020. [2] Meng, G., Gao, B., Zhou, J., Cao, G., & Zhang, Q. (2016). Experimental investigation of the me-chanical behavior of the steel fiber reinforced concrete tunnel segment. Con-str.&Build.Mat.,126,98-107. [3] Kalifa, P., Chene, G., & Galle, C. (2001). High-temperature behaviour of HPC with polypropylene fibres: From spalling to microstructure. Cem. & Concr. Res., 31(10), 1487-1499. [4] Rodrigues, J. P. C., Laím, L., & Correia, A. M. (2010). Behaviour of fiber reinforced concrete col-umns in fire. Composite Structures, 92(5), 1263-1268. [5] Yan, Z. G., Shen, Y., Zhu, H. H., Li, X. J., & Lu, Y. (2015). Experimental investigation of reinforced concrete and hybrid fibre reinforced concrete shield tunnel segments subjected to elevated tem-perature. Fire Safety Journal, 71, 86-99.

ABSTRACT. Fire-induced collapse of reinforced concrete (RC) structures poses significant challenges due to complex thermo-mechanical coupling and increased stochasticity of concrete at elevated temperatures. This study develops a temperature-dependent stochastic thermo-mechanical damage model within the framework of continuum damage mechanics. The formulation integrates thermal degradation of elastic properties, transient thermal creep, and meso-scale random fracture statistics, enabling prediction of both nonlinear response and statistical variability. A three-dimensional rate-form constitutive model with consistent tangent stiffness is implemented in Abaqus via UMAT. Validation against uniaxial, biaxial, and cyclic tests for C30–C60 concrete demonstrates accurate reproduction of mean responses and dispersion envelopes. The model is further incorporated into a hybrid fire simulation framework for structural analysis and verified against restrained column and RC frame fire tests. Application to a nine-story RC building under localized fire captures progressive damage, load redistribution, and collapse mechanisms. The proposed framework provides a robust basis for performance-based fire engineering and probabilistic collapse assessment of RC structures.

ABSTRACT. Ultra-High-Performance Concrete (UHPC) is a fibre-reinforced cementitious material characterized by enhanced mechanical performance, including compressive strengths between 120–150 MPa and tensile strengths between 5–12 MPa, combined with low porosity and high impermeability, making it suitable for demanding environmental conditions. Despite its widespread use, the direct tensile characterization of UHPC remains an area of ongoing research, and only recently has a standardized uniaxial tension test method been introduced through AASHTO T397-22. In this study, a modified version of this standard was employed to evaluate the tensile response of UHPC reinforced with steel fibres and a novel synthetic fibre type, polyoxymethylene (POM), under elevated temperatures. Digital Image Correlation (DIC) was used to measure crack widths during the test. The results indicate a complete loss of post-cracking ductility at temperatures above 450 °C for specimens containing steel fibres and above 150 °C for specimens with POM fibres, highlighting the significant influence of fibre type and dosage on the tensile behaviour at high temperatures.

ABSTRACT. The fire vulnerability of flat slabs is a well-recognised issue, extensively documented through both numerical and experimental studies. In particular, the risk of punching shear failure in flat slabs under fire conditions requires careful consideration. A notable limitation of much of the existing experimental and numerical research is the use of slab specimens or FE models with boundary conditions that do not accurately represent those in real structures. While the effects of material degradation at high temperatures are widely examined, a clear gap exists in research concerning the fundamental influence of boundary conditions (BCs) and thermally induced deformations on the thermo-mechanical behaviour of flat slabs. This paper addresses this gap through a thermo-mechanical finite element analysis of the Gretzenbach car park collapse using Abaqus, modelling an axisymmetric slab-column segment that captures smeared reinforcement effects. The heat-transfer analysis, based on a passenger car fire with 5 MW peak and 5000 MJ load, drives rapid heating at the soffit, with temperatures up to 850°C and 358°C in the reinforcement, and steep gradients that induce significant compression near the column-slab connection. Preliminary results up to 1000s of fire reveal 4 MPa top tensile cracking and 35 MPa bottom compression crushing, reducing the effective depth by 27% and resulting in 39-46% punching shear capacity utilisation, with 25% column load amplification, highlighting the significance of thermal deformations and BCs in punching shear failure.

ABSTRACT. Concrete spalling under fire exposure is a critical phenomenon that threatens the safety and durability of civil infrastructure. Spalling refers to the violent detachment of surface layers due to rapid heating, vapor pressure buildup, and restrained thermal expansion. Conventional evaluation methods typically employ cylindrical or rectangular specimens heated in an electric furnace. While these methods allow observation of cracking and surface damage, they often fail to reproduce explosive spalling because the specimens expand freely without restraint. In actual structures, however, thermal expansion is constrained by surrounding components, making spalling more likely and more severe. To address this discrepancy, our research group has proposed the ring‑constrained heating test method (Fig.1), in which concrete is cast inside a steel pipe and heated from the underside of the opening. This setup constrains thermal expansion and deformation, thereby reproducing conditions favorable for spalling. The method also enables simultaneous measurement of internal temperature, restraint stress, and vapor pressure, providing a comprehensive picture of the mechanisms driving explosive damage. The Japan Concrete Institute has adopted this method as a testing standard1), and it has been proposed as part of the RILEM TC‑SPF explosive spalling screening protocol2). The present study investigates the relationship between concrete compressive strength and spalling severity using the ring‑constrained heating test. Specifically, we analyze the maximum explosive depth ratio (defined as maximum spalling depth divided by specimen height) across concretes of different strength levels and fiber reinforcement conditions.

By normalizing spalling depth, we aim to verify the effectiveness of the ring‑constrained method and to clarify the role of fibers in mitigating explosive damage3)-9). Specimen geometry: Steel pipes with a diameter of 300 mm and thickness of 8 mm. Two specimen heights were used: 50 mm and 100 mm. Heating conditions: (1) RABT30 heating curve, representing rapid heating typical of tunnel fires; (2) a unique 1050°C heating curve, designed to impose extreme thermal stress.

Concrete types:Normal strength concrete (≤ 50 MPa), High strength concrete (70–100 MPa), Ultra‑high strength concrete (150–230 MPa), Porosity free concrete (≥ 300 MPa)：designed using close‑packing theory, deaired, and steam‑cured. ・Fiber types: Polypropylene (PP), polyvinyl alcohol (PVA), jute, and steel fibers. Fig.2 illustrates representative damage patterns and the relationship between compressive strength and the dimensionless maximum spalling depth ratio(Fig.3) . The experimental results and discussion are shown below. 1. Normal strength concrete (≤ 50 MPa): Spalling was rare. Surface damage manifested primarily as tortoiseshell‑shaped cracks. Explosive penetration was not observed. 2. High strength concrete (70–100 MPa): No fiber specimens exhibited spalling. Addition of PP fibers or jute fibers significantly reduced or eliminated spalling. The maximum explosive depth rate decreased markedly with fiber reinforcement. 3. Ultra‑high strength concrete (150–230 MPa): Severe explosive spalling occurred in specimens without fiber, often penetrating the full specimen height. Incorporation of steel fibers, PVA fibers, or jute fibers reduced spalling depth and prevented catastrophic penetration. Steel fibers provided mechanical bridging, while PVA and jute fibers enhanced vapor release pathways. 4. Porosity free concrete (≥ 300 MPa): Designed with minimized voids and subjected to steam curing, these specimens exhibited extremely high strength. Specimens without fiber experienced violent explosions sufficient to penetrate the specimen. Incorporation of steel fibers and specially engineered PP fibers mitigated explosive damage, demonstrating the necessity of fiber reinforcement in ultra‑fiber reinforced concretes. In summary, the ring‑constrained heating test provides a reliable and standardized method for reproducing spalling phenomena under realistic restraint conditions. Results demonstrate that: Spalling severity increases with compressive strength. Fiber reinforcement is essential for mitigating explosive damage, particularly in high‑strength and non‑porous concretes. Different fibers contribute through distinct mechanisms: vapor release (PP, PVA, jute) and crack bridging (steel). By normalizing spalling depth through the maximum explosive depth ratio, the method enables quantitative comparison across strength levels and fiber types. This study confirms the effectiveness of the ring‑restrained heating test and highlights its value for developing fire‑resistant concrete materials. The paper will also include a detailed discussion of the internal vapor pressure, restraint stress, and pore pressure developed during the ring-constrained heat test. These parameters are critical in understanding the mechanisms underlying explosive spalling in heated concrete. Relevant findings from previous studies will be referenced to contextualize the observed behaviors and to support the interpretation of experimental results. Reference 1) JCI ：Standard of Test Method for Spalling of Concrete under High Temperature Exposure，JCI-S-014-2024, Japan Concrete Institute，2024 (In Japanese) 2) P. Pimienta et al : Recommendation of RILEM TC 256-SPF on the method of testing concrete spalling due to fire: material screening test, Materials and Structures (2023) 56:164 3) M.Ozawa et al:Spalling of geopolymer concrete in ring-restrained specimens under high temperatures,Proceedings of the 11th International Conference on Structures in Fire SiF 2020，pp.313-321,November, 2020. 4) H.Akasaka et al：Preventive effect on fire spalling of high-strength concrete with Jute fibre in ring-restraint specimen, Proceedings of the 14th International Conference on Concrete Engineering and Technology (CONCET 2018), pp.132-139, Kuala Lumpur, Malaysia, 8-9 August, 2018 5) Y.Kobayashi et al: Evaluation of fire spalling behaviour of ultra-high-strength-PVA fiber-reinforced concrete by ring-restrained heating test , Paper ID: iCFSERP20230065, Proceedings of 1st International Conference on Fire Safety Engineering Research and Practice (iCFSERP2024), 24-27 Nov 2024, Sydney, Australia 6) M.Mizoguchi et al：Fire spalling resistance of fiber reinforced PFC (porosity free concrete), JCI，Proceedings of JCI Annual convention ,pp. 940-945, Vol.44, No.1, 2022 (In Japanese) 7) M. Ozawa et al：Screening-test analysis of fire spalling behavior with various concrete samples, Proceeding of 5th International Workshop on Concrete Spalling due to Fire Exposure, pp.315-326, Boras,Sweden，12-13, October, 2017. 8) S.Yamamoto et al：Comparison of fire resistance test results of full-scale RC segment and heating test results of ring-restrained specimen made with the same concrete mixure，JCI，Proceedings of JCI Annual convention, Vol.39,No.1, pp.1123-1128, 2017.July. (In Japanese) 9)Simamura et al：Fire Spalling evaluation of ultra-high strength fiber reinforced concrete containing natural fibers by ring restraint heating test、JCI，Proceedings of JCI Annual convention,、pp.712-717,Vol.45,2023 (In Japanese)

ABSTRACT. Concrete is known to undergo spalling when exposed to elevated temperatures, particularly during fire events. The ring-constrained heat test has been proposed as a screening method for evaluating the spalling resistance of concrete materials. This method is standardized by the Japan Concrete Institute (JCI) 1）and incorporated into the proposed RILEM TCSPF2) spalling test protocol. However, limited knowledge exists regarding the correlation between spalling depths obtained from ring tests and those observed in component-level heating tests. To address this gap, we conducted a comparative analysis using previously reported data from both ring-constrained and component-level heat tests. The maximum spalling depth in the ring test was nondimensionalized by dividing it by the specimen height. Similarly, spalling depths in component-level specimens were normalized using their respective heights.The concrete compressive strength across the test specimens ranged from 42 MPa to 108 MPa. Polypropylene (PP) and jute fibers were used as spalling suppression agents, with fiber contents varying from 0 to 0.2% by volume. Component-level verification included reinforced concrete (RC) columns, RC slabs, prestressed concrete (PC) beams, and small-scale RC beams. Representative heated surface conditions are shown in Figure 1, comparing ring test specimens with component-level elements such as columns3) and slabs4). The addition of PP fibers was found to effectively suppress spalling, a trend consistently observed across both test methods. Figure 2 presents a comparison of the maximum spalling depth ratios between ring and component specimens. In fiber-free cases, spalling was generally observed, with ring test specimens exhibiting greater spalling depth ratios than component-level specimens—except in the case of RC segments. This discrepancy is attributed to the increased susceptibility of ring specimens to spalling due to circumferential restraint. In fiber-reinforced cases, spalling was absent in several specimens across both test types, confirming the efficacy of fiber addition. However, the influence of concrete compressive strength is also considered significant. Further discussion in this paper will address the effects of moisture content, vapor pressure, and restraining stress as observed in the ring-constrained test.

Reference 1) JCI standard：Test Method for Spalling of Concrete under High Temperature Exposure，JCI-S-014-2024, Japan Concrete Institute，2024 (In Japanese) 2) P. Pimienta et al M.Ozawa :Recommendation of RILEM TC 256-SPF on the method of testing concrete spalling due to fire: material screening test, Materials and Structures (2023) 56:164 3) M. Ozawa et al：Screening-test analysis of fire spalling behavior with various concrete samples, Proceeding of 5th International Workshop on Concrete Spalling due to Fire Exposure, pp.315-326, Boras,Sweden，12-13,October, 2017. 4) S.Yamamoto et al：Comparison of fire resistance test results of full-scale RC segment and heating test results of ring-restrained specimen made with the same concrete mixure，JCI，Proceedings of JCI Annual convention, Vol.39,No.1, pp.1123-1128, 2017.July. (In Japanese)

ABSTRACT. Flat slabs supported on column grids, which are widely used in car parks and open-plan structures, are highly vulnerable to punching shear failure during fire as evidenced by the 2004 Gretzenbach car park collapse in Switzerland. Previous whole-structure fire simulations captured overall behaviour but did not model the local slab-column stress concentrations that critically govern punching shear. Building on existing nonlinear thermo-mechanical plate and shell research, the present work develops a closed-form thermo-elastic model for the axisymmetric slab-column tributary region. The model quantifies localised membrane stresses and bending moments arising from combined mechanical loads (gravity and perimeter column reaction) and through-thickness thermal gradients. Using von Kármán plate theory, the study formulates coupled nonlinear differential equations for the Airy stress function and transverse deflection. Two boundary-restraint models (free edge vs. radially restrained) are examined to reflect varying column head stiffness. The results identify parameter regimes in which these thermally-induced stress concentrations become strongly amplified, suggesting a previously unrecognised mechanism that may critically exacerbate punching shear vulnerability in flat slabs during fire.

ABSTRACT. Concrete-filled steel tube (CFT) columns are widely used in high-rise and large-scale buildings due to their high strength and ductility. In Korea, several sectional configurations, including internally protruded and externally thickened plate types, have been developed; however, domestic fire design remains prescriptive, requiring fire protection regardless of sectional characteristics. To address the limited research on protected and unprotected CFT columns, this study conducted non-loaded and loaded fire tests on representative Korean sections and carried out sequential thermal and thermal-stress analyses to evaluate their fire behavior. Non-loaded fire tests investigated the influence of sectional shape, size, steel tube thickness, and fire protection type under a 3-hour ISO 834-1 standard fire. Loaded fire tests applied a 40% axial load ratio to examine the performance of columns protected with 16 mm SFRM and 1.9 mm intumescent coating. A two-dimensional transient heat transfer model was developed, and the predicted temperatures agreed with measurements within approximately 10%. The computed temperature histories were subsequently used as input for a finite element thermal-stress analysis incorporating temperature-dependent strength reduction factors from EC3 and EC4, with predicted deformation and failure time matching experimental results within about 10%. Results showed that larger cross-sectional sizes reduced steel tube temperatures by 7–10%, while smaller width-to-thickness ratios achieved an additional 3–4% reduction. Unprotected columns exhibited thermal expansion, load transfer, and failure, whereas SFRM-protected columns sustained continuous expansion for nearly 150 minutes without transitioning to failure.

ABSTRACT. Most experimental research on explosive spalling of UHPC involves small scale specimen testing in the form of cubes, prisms and cylinders, where size effect has been shown to be a factor on spalling behavior. Fire tests on larger UHPC beam specimens available in literature have only subjected beams to standard fire curves, ASTM E-119 and ISO 834. As standard fire curves do not accurately reproduce typical temperatures seen during fire events spalling behaviour may be impacted by the use of localized pool fires. Use of the standard fire also typically focuses on fire rating which neglects the impact post-fire cooling can have on structural elements which may not fail during the duration of a fire event. To further investigate how explosive spalling and post-fire cooling could impact the structural response of larger structural elements four reinforced concrete beams incorporating a steel fibre reinforced commercially available UHPC were cast with different reinforcement configurations. This being an effort to holistically consider novel reinforcement configurations of shape memory bars in reinforced concrete, including UHPC. Initial results demonstrated that exposure to fire reduced the ascending stiffness of beams during flexural testing but had no significant impact on load capacity. All beams were observed to recover the downward thermal induced deflection during fire testing and develop an upward deflection by the end of the cooling phase, the presence of cracks increasing this net positive deflection. Explosive spalling of UHPC beams was only seen in uncracked specimens with cracked specimens allowing for water and vapour migration through micro and macro cracks in the concrete matrix.

ABSTRACT. Mechanically bolted (shear lock) reinforcement couplers offer a valuable alternative to traditional lap splicing, particularly in congested structural zones. However, their performance under fire conditions is not well documented. This study experimentally investigated the thermal and mechanical behaviour of an MBT ET-12 shear lock coupler-bar assembly using 12-mm B500B/C reinforcement. Three test regimes were conducted. Ambient temperature, sustained load (45.2 kN) with transient heating (10 °C/min to 600 °C), and sustained load with transient heating plus an imposed thermal gradient. Full-field strain was measured using Digital Image Correlation (DIC). Under ambient conditions, all assemblies failed by bar necking outside the coupler, as expected. In the transient elevated temperature experiments, failure also occurred in the bar, with thermal strain dominating the total deformation and causing fracture at reduced temperatures. When a thermal gradient was applied by actively cooling the bar, the failure mode shifted to the first shear lock bolt position inside the sleeve, with indications of creep deformation leading to failure in one sample under sustained load and temperature alone. The results demonstrate that while the coupler system effectively maintains the load path during heating, its failure mechanism can be critically altered by thermal gradients and creep, highlighting the need for further research in this area.

ABSTRACT. The Eurocode Parametric Fire Curve (EPFC) is widely used in structural fire engineering to represent natural fire exposure, yet its decay and cooling phases still rely on simplified assumptions (most notably a linear temperature decrease inherited from the 1975 ISO 834 standard). Moreover, the EPFC is not fully aligned with the heat release rate (HRR) history described in Annex E of EN 1991-1-2.

This study introduces a revised, probabilistic formulation for the EPFC decay and cooling phases. To ensure consistency with Annex E, the model assumes that only 70% of the total fuel load is consumed during the fully-developed phase, with the remaining 30% burning during a linear fire decay phase.

Based on a set of simplified compartment-fire assumptions (e.g., well-stirred reactor behaviour), the evolution of the adiabatic surface temperature is modelled, and the resulting temperature–time curve is approximated using an exponential decay function defined by a characteristic time constant, τ. A key predictor of τ is the thermal penetration speed, which depends on lining diffusivity and the time of peak temperature.

More than 4000 simulated ventilation-controlled fire scenarios were analysed to calibrate the model. The resulting τ values follow a normal distribution, enabling a probabilistic representation of the EPFC’s descending branch. Overall, the study enhances the physical basis of the EPFC, particularly for the decay and cooling phases, and provides a probabilistic framework suitable for reliability-based structural fire assessments.

ABSTRACT. Fire is an action that can severely damage bridge structures, which are not generally designed with fire resistance criteria. This study presents a methodology for deriving fragility curves of bridges subjected to fire scenarios, explicitly accounting for key sources of uncertainty. Two representative typologies are analysed: a simply supported steel-concrete composite bridge and a steel arch bridge, selected as common configurations in transportation networks. Fire scenarios are defined through a statistical analysis of traffic data from an Italian highway to ensure realistic fire input parameters. The longitudinal fire location is treated as a variable along the span, enabling location-independent fragility assessment. A sequential thermo-mechanical framework is adopted, combining CFD fire simulations with heat transfer and structural analyses incorporating temperature-dependent material degradation. The results show that the governing response parameter differs by typology. For the composite bridge, performance is controlled by the maximum vertical displacement during fire exposure. For the steel arch bridge, the residual vertical displacement governs structural performance. Fragility curves are derived considering peak heat release rate (HRR) as the intensity measure.

ABSTRACT. Localised fires from passenger vehicles, including internal combustion engine vehicles (ICEVs) and electric vehicles (EVs), generate highly non-uniform flames that differ substantially from conventional compartment fires represented by uniform gas-temperature curves such as ISO 834. This study compiles a comprehensive dataset of full-scale vehicle fire experiments from 1994 to 2025 and performs statistical analysis to derive representative mean and 80th‑percentile heat release rate (HRR) design fire curves for ICEVs and EVs. These percentile-based HRR profiles, including typical, upper-bound, and lower-bound scenarios, are then coupled with simplified conical and cylindrical flame models to estimate radiative heat flux and temperature distributions around burning vehicles. The conical model captures natural flame widening and entrainment effects, while the cylindrical model represents relatively steady burning. Using configuration factors and radiation-dominant heat-transfer formulations, the study evaluates how flame-geometry assumptions influence predicted thermal actions at various distances and elevations. The results provide a first-step methodology for generating realistic, vehicle-specific design fires for structural fire engineering applications in car parks and other localised fire scenarios.

ABSTRACT. Adoption of electric buses in public transportation is accelerating yet the risks tied to their lithium-ion battery systems are still not well characterized in the context of structural fire engineering. Available test data and properties appropriate for computational modeling come from studies focused on passenger Electric Vehicles (EV) or simplified battery setups, which do not capture the size, battery layout, or fire-growth behavior unique to electric buses. What further exacerbates the situation is that public transportation buses require large storage and charging garages/stations, often enclosed in urban areas, and where the parking arrangement density magnifies the hazards.

This study develops a framework for analyzing the potential deflagration and fire loads arising from the failure of lithium-ion battery systems undergoing thermal runaway. Under thermal runaway, electric batteries generate significant heat, produce large quantities of toxic and flammable gases and once ignited, generate deflagration pressures and undergo rapid fire growth. It is essential to characterize these demands in a way that reflects the unique design and battery architecture of modern electric buses, i.e. electric buses introduce new loads from explosions (delayed ignition of the flammable gas cloud) and fire demands (exceeding standard code fire design curves), both of which need to be considered in the assessment of the structure’s response to a thermal-runaway event.

The work identifies representative electric-bus configurations by reviewing common battery capacities, on-board placement, and cathode chemistries in lithium-ion batteries. The methodology uses a bottom-up approach in which the contributions from the individual electric bus components are superimposed for a particular bus makeup and battery configuration. Heat-release rate over time (HRR) and gas mass release profiles are developed to include internal bus materials (common to electric and internal combustion buses), battery capacities and distribution, and battery states (e.g., state of charge, cathode chemistry, total capacity etc.). This study outlines the development of a HRR curve, from initiation to propagation of battery thermal runaway, the off-gassing of batteries, to the eventual deflagration and fire growth, all through the cooling phase. Quantifying the new demands from electric vehicles to structures is a complex engineering problem that needs to consider overpressure developed from potential explosions as well as thermal loads from fire.

A complementary objective is to examine how sprinkler systems influence fire events. A simplified methodology is developed from experimental data available in the literature to quantify the effects of sprinklers on the heat released by an electric bus fire. The methodology focuses on the estimate of heat release rate reduction of the fire as a function of the sprinkler discharge density, i.e. gpm/ft2.

The methodology framework for fire analysis on structures is presented with a case study of a long-range electric bus undergoing thermal runaway in a typical parking garage with close proximity to other buses. The result of this case study, shown in Figure 11, demonstrates its applicability to quantify the thermal effects to structural elements (i.e. beams, columns, deck), effects to adjacent buses to understand bus-to-bus fire propagation potential, as well as air temperature impacting occupants and first respondents. Sprinkler’s effects to heat release rate are considered for this case example following the proposed methodology.

Overall, this work provides a methodology to quantify the electric-bus fire initiated by a battery thermal-runaway event, the time-dependent heat release curve from this event, and compares it against the UL-263 and UL-1709 standard fires (e.g. peak temperatures, fire duration, cooling phase etc.). The proposed method to quantify the effect of sprinklers provides a pathway to assess structural performance under these new demands. It highlights the unique hazards introduced by lithium-ion battery failure in electric buses and sets the objectives for future computational and experimental studies that focus on fire, deflagration, flammable gas-cloud toxins, visibility, and fire suppression.

ABSTRACT. Wildland–urban interface (WUI) fires subject residential buildings to severe thermal exposures that can transition rapidly from compartment ignition to external flame jets, full burnout, and structural collapse. While ignition pathways in the WUI have been extensively studied, less attention has been given to how the combustibility of structural systems influences fire intensity, external radiant exposure, and neighborhood-scale vulnerability. This study presents a comparative structural fire engineering assessment of structurally equivalent timber and cold-formed steel (CFS) residential systems to quantify how differences in structural fuel load affect both building-scale fire behavior and cluster-scale risk.

Representative Type V-B archetypes, consistent with parcel geometries in Pacific Palisades, California, are analyzed to determine structural wood mass and associated fuel loads. Equivalent CFS versions of the same plans are developed to ensure comparable geometry and ambient structural performance. Compartment-fire models provide thermal exposure histories for both systems, and a thermo-mechanical analysis of CFS walls and floors captures degradation mechanisms including stiffness loss, thermal bowing, and local instability. These structural outcomes are used to adjust external heat-release and exposure characteristics for CFS buildings, reflecting differences in structural survivability and envelope breach.

A simplified spatial fire-risk model then propagates these exposure characteristics across inter-building distances to quantify damage potential for timber-dominated and CFS-dominated clusters. Preliminary results indicate that structural fuel load strongly influences burning duration and external radiant exposure, with timber systems producing substantially higher cluster-scale vulnerability. Ongoing work evaluates the relative influence of structural versus contents fuel load and integrates these findings into a risk-adjusted carbon-footprint comparison.

ABSTRACT. Standardised furnace testing provides fire resistance ratings for regulatory compliance but typically offers limited insight into real-fire performance and associated uncertainties. Using a simple, one-way spanning reinforced concrete slab example, we show that furnace test results contain quantitative information that can be extracted to reduce uncertainty in predicting structural performance under natural fire exposures. We highlight the importance of using all test outcomes (including those that do not reach the target resistance) to avoid bias, the benefits of running tests to structural failure, and how furnace testing campaigns can be optimised through pre-posterior analysis. We propose principles for future fire-safety regulations that treat furnace tests as controlled information-generating experiments to quantify reliability and guide performance-based design, rather than solely as demonstrations of compliance with prescriptive guidance.

ABSTRACT. Steel tube confined reinforced concrete (STCRC) columns are innovative steel-concrete composite structural elements, characterized by intentional discontinuity of the steel tube in beam-to-column joint regions. In this system, the steel tube is primarily designed to provide confinement to the concrete core, rather than to directly bear axial loads. Over the past decade, STCRC columns have been increasingly adopted in high-rise buildings and large-span structures, where fire safety constitutes a key design consideration. To date, the fire performance of STCRC columns under idealized uniform heating has been examined, leading to the development of practical fire performance design methods. It is noteworthy that, despite similarities in external geometry, STCRC columns exhibit distinct deformation behaviour and failure modes in fire, when compared with the concrete-filled steel tubular (CFST) columns. In real buildings, columns may be subjected to non-uniform fire exposure due to the presence of adjacent walls. Such scenarios would result in asymmetric temperature distributions, inducing differential thermal expansion, non-uniform material degradation, and differential thermal stresses. Previous investigations into the fire performance of CFST columns have considered idealized non-uniform thermal boundaries (e.g., one-, two-, or three-sided exposure). In practice, however, the presence of adjacent walls can exacerbate effects such as non-uniform confinement, additional lateral deflections, and additional eccentricity. There remains a paucity of data on the temperature development within and structural performance of STCRC columns under non-uniform heating, with or without consideration of adjacent walls. Moreover, no design guidelines exist which specifically address the structural fire performance of STCRC columns under these conditions. This study first presents an experimental investigation into the fire performance of square STCRC columns under three-sided heating, considering both concentric and eccentric loading. Numerical analyses are then presented to examine the influence of parameters including the presence of connecting walls, number of exposed faces, and load ratio on the temperature development and structural fire performance of STCRC columns under non-uniform heating. Finally, design recommendations are proposed for the fire safety design of square STCRC columns under non-uniform heating conditions.

ABSTRACT. This study investigates explosive spalling in high-strength concrete by focusing on the thermal stress hypothesis. Material properties—including compressive strength, Young’s modulus, free thermal strain, and transient strain—were compared between concretes that exhibit spalling and those that do not. High-temperature compression tests and overall strain tests were conducted on cylindrical specimens, and the obtained data were used in a thermal stress analysis with finite-element modeling. Results showed that, compared with non-spalling concrete, the spalling concrete had a higher residual Young’s modulus, smaller transient strain, and lower residual compressive strength. Thermal stress analysis revealed that the spalling concrete developed significantly higher normalized thermal stresses during heating, particularly at the onset of temperature rise, indicating a strong association between thermal stress development and explosive spalling.

ABSTRACT. The thermo–mechanical response of composite beams in steel framed composite construction in fire is strongly governed by the nonuniform temperature field that develops in the composite slab and steel beam cross-sections. As heating progresses, the top of the slab remains relatively cool while the soffit and the steel beam heat rapidly, producing significant thermal gradients that govern deformations which determine the forces and moments in the composites section based on the boundary restraint conditions. Previous work has shown that accurately capturing these temperature gradients is essential for understanding the transition between different response mechanisms in composite beams subjected to various fire exposure scenarios. This paper provides the inspiration for developing a database containing a large range of temperature distributions for a full range of composite steel-concrete structural cross-sections such as universal beam, universal column and square hollow sections composite with a range of concrete slab thicknesses under a reasonably comprehensive range for fire scenarios. Such a database will enable instantly resolving thermal gradients in arbitrary composite cross-sections without conducting a full heat transfer analysis every time using machine learning.

ABSTRACT. Steel-framed structures are extensively used in high occupancy buildings, where fire safety is a critical concern. While steel offers high strength and ductility under normal conditions, its mechanical properties degrade significantly at elevated temperatures, increasing the risk of collapse during fire incidents. To mitigate this risk, passive fire protection measures are widely applied to steel components. This helps delay the temperature rise of steel members during fire exposure, maintaining their load-bearing capacity for a longer duration. This paper presents a progressive framework for performance-based structural fire design that balances practicality with engineering rigour. The first stage introduces a streamlined performance-based structural fire engineering (PBSFE) approach that preserves the essential modelling of structural response while avoiding the computational demand of fully nonlinear finite element analyses with coupled heat transfer. This enables efficient, physics-informed determination of fire protection thickness for real building frames and makes performance-based design more accessible for routine engineering practice. Building on this foundation, we further examine whether fire protection requirements can be predicted without any structural simulation. To this end, we propose a Graph Neural Network (GNN) model that learns to map structural topology and member properties directly to fire protection thickness. GNNs operate on graph-structured data, which matches the representation of building frames: columns can be treated as nodes and beams as edges linking them.

ABSTRACT. Introduction The European built environment is largely affected by structural obsolescence, as the majority of existing reinforced concrete (RC) buildings were constructed before the 1980s and no longer meet current structural and safety requirements [1]. Similarly, most transport infrastructures date back to the same construction period, leading current public investments to be primarily devoted to their maintenance and strengthening [2]. Consequently, in recent decades, both researchers and practi-tioners have been challenged to identify effective repair and strengthening techniques for existing RC structures, with the aim of (i) enhancing their bearing capacity under gravity loads and (ii) mitigating their vulnerability to accidental actions, such as, e.g., earthquakes. Among the wide range of retrofit-ting strategies—such as bonding or jacketing with composite materials, modifying the structural sys-tem through the addition of RC shear walls or bracings, or installing special devices—the use of steel elements has proven to be a valuable and versatile solution. Steel-based interventions can be easily adapted to address various structural deficiencies while ensuring simplicity in design, implementa-tion, and detailing. However, steel is highly vulnerable to fire, as it experiences a significant loss of stiffness and strength when its temperature exceeds approximately 400°C. This raises a critical issue regarding the fire performance of RC members strengthened with externally bonded or anchored steel plates, since both steel and its connection system are directly exposed to high temperatures. This paper presents the results of an extensive literature survey focused on (i) the current knowledge about the behaviour of steel-strengthened RC structural members under fire conditions, aiming to identify their governing factors, and (ii) how codes and standards, worldwide, deal with the issues associated with the mechanical decay affecting the strengthening system in case of fire. RC members strengthened with external steel plates The use of external steel plates is a widespread technique to strengthen existing RC members suffer-ing from structural deficiencies, ranging from inadequate flexural or shear capacity to poor ductility. It can be effectively applied to both beams and columns. Relevant to beams, externally bonded steel plates are mainly used to enhance flexural and shear capacity. In the literature, early research focuses on the response of plates bonded to either tension or vertical faces, to overcome either flexural or shear deficiency, respectively. Later studies, instead, address issues mainly referring to the capacity of the strengthened element to fully deploy its en-hanced capacity. Such condition is ensured only if the mechanical connection between concrete sub-strate and strengthening plate is designed to prevent early interface failure mechanisms, such as end/intermediate debonding, loss of anchorage system’s capability of transferring shear stresses and loss of full interaction between existing concrete and installed plates. Looking at existing RC columns, external plates form the basis of steel jacketing systems, where continuous or discrete plates welded to corner angles provide (i) confinement, (ii) increase shear and axial capacity, and (iii) ductility improvement. The solutions for plated RC elements fall into two main categories: externally bonded (EB) plates and mechanically bonded (MB) plates, with EB systems often combined with mechanical steel anchors to enhance adhesion and the interaction between plates and concrete. The bibliographic review carried out by the authors indicates that strengthening systems are generally designed to address static or seismic deficiencies, while their vulnerability under fire conditions is rarely considered—despite the pronounced degradation of steel properties at elevated temperatures, which may critically affect both design and safety verification. Response in fire conditions The full analysis conducted by the authors confirms that the fire behaviour of steel-reinforced con-crete elements is still only partially understood, with a significant lack of experimental and numerical studies on their global response at elevated temperatures, as illustrated in Figure 1b. To address this gap, the authors present a comprehensive literature survey that examines both the thermo-mechanical behaviour of the individual components of the steel–concrete–anchorage system and the overall response of the strengthened element, providing an integrated understanding of its performance under high-temperature conditions. Steel experiences a marked reduction in strength and stiffness at temperatures above 400°C, while the low thermal conductivity of concrete generates significant tem-perature gradients within the section. Connection systems, such as epoxy resin and mechanical con-nectors, also degrade under elevated temperatures: resin-bonded reinforcements are prone to delam-ination once the adhesive approaches its glass transition temperature, often below 100°C, while me-chanical connectors lose strength due to both thermal softening and differential expansion between steel and concrete. Additionally, the thickness and type of mortars or plasters covering the rein-forcement can influence the overall fire resistance of the element.

(a)

(b) Figure 1(a) Pie chart showing the distribution of studies considered in the review, (b) Methodology for the case study, illustrating key steps in the analysis. Due to the limited number of studies on this topic, a representative case study was selected to help bridge this gap through a series of detailed thermo-mechanical analyses. As illustrated in Figure 1b, the methodology involved the selection of an existing building representative of typical Italian con-structions, chosen based on its year of construction to reflect the large stock of buildings designed without adequate consideration of seismic actions. The building was subjected to a seismic vulnera-bility assessment to identify deficient elements, specifically the beams. Steel plate reinforcements were then designed for these elements, addressing both bending and shear. The strengthened ele-ments were subsequently evaluated under fire conditions using both prescriptive and performance-based approaches. The results of these analyses are currently being processed and will be presented in the final paper. References [1] A. Gevorgian, S. Pezzuto, S. Zambotti, S. Croce, U. F. Oberegger, R. Lollini, L. Kranzl, A. Müller, European building stock analysis, Eurac Research (2021) [2] A. Del Grosso, D. Inaudi, L. Pardi, Overview of European activities in the health monitoring of bridges, in: Pr. of the First International Conference on Bridge Maintenance, Safety and Manage-ment, Barcelona (Spain), 2002 [3] Annalisa Napoli, Enzo Martinelli, Maria Rosaria Pecce, RC beams externally strengthened by steel plates: experimental database and preliminary analysis, Procedia Structural Integrity, Volume 64, 2024, Pages 975-982, ISSN 2452-3216.

ABSTRACT. It is widely accepted that the porosity distribution of concrete significantly influences its strength. When concrete is exposed to elevated temperatures during a fire, its porosity undergoes notable changes due to the evaporation of bound water. Mercury Intrusion Porosimetry (MIP) is a commonly used technique for quantifying porosity and evaluating the relationship between pore volume and pore size. This study examines Normal Strength Concrete (NSC) and High Strength Concrete (HSC), with specimens exposed to six target temperatures ranging from room temperature (24 ± 2 °C) up to 1000 °C. MIP results from these specimens are used to establish the relationship between intruded pore volume and pore size distribution. Parameters governing the pore size distribution (PSD)—including the mean pore radius, coefficient of pore dispersion, and permeable porosity—are determined using the Morgan–Mercer–Flodin model. The influence of temperature-induced variation in PSD parameters on concrete strength is then assessed. The applicability of PSD is demonstrated by estimating the compressive strength of concrete using established relationships reported in the literature.

ABSTRACT. The rapid population expansion in the 21st century has resulted in a significant increase in high-rise constructions. Concrete has long served as a primary material in construction which in turn has raised critical concerns about fire safety in high-rise constructions. Studies on the fire performance of concrete shows that the mechanical properties of concrete, particularly compressive strength that begin to deteriorate markedly beyond 150 °C, primarily due to the loss of chemically bound water, increased pore pressure, and the potential for explosive spalling. Such failure mechanisms pose severe risks in high-rise structures where evacuation routes are constrained. Existing passive fire insulation solutions such as vermiculite and gypsum boards are often hindered by high cost, installation complexity, limited durability, and reduced scalability, restricting their practicality in large-scale or cost-sensitive projects. Consequently, recent research has shifted towards developing embedded fire-mitigation systems within the concrete matrix itself. This study presents an innovative embedded insulation solution using phase change materials (PCM) to enhance the thermal and mechanical performance of concrete under elevated temperatures. PCM exhibit the ability to absorb and release latent heat through reversible solid–liquid phase transitions, offering a promising mechanism for attenuating internal temperature rise during fire exposure. Much of the existing research focuses on PCM for thermal comfort applications in building envelopes using organic PCM with low phase-transition temperatures. And limited attention has been given to inorganic and eutectic PCM that possess higher transition temperatures, superior thermal stability, and properties relevant to fire scenarios. This gap in knowledge highlights the need for systematic investigation of PCM-integrated concrete for fire-resistant structural applications. The study presented herein investigate the influence of two types of PCM: organic and inorganic which directly incorporated into concrete using a volumetric replacement method. The experimental program was structured into three phases. Phase I involved the systematic evaluation of mechanical, physical, microstructural, and thermal properties of PCM concrete to identify the optimal mix design. PCM was introduced at 5%, 10%,15% by volume and compared with a standard 25MPa concrete control mixture. Tests included compressive strength, density, Poisson’s ratio, thermal conductivity, and Scanning Electron Microscopy (SEM) to assess PCM–matrix interaction. Phase II examined transient heat-transfer behaviour under uniaxial heating using a controlled muffle furnace, simulating one-dimensional heat flow representative of fire exposure. Internal temperatures were recorded with embedded thermocouples to evaluate heat-transfer delay. Phase III developed a finite element (FE) model in Abaqus using experimentally derived material properties, enabling calibration and validation against measured thermal and mechanical responses. The results indicate that the incorporation of inorganic PCM significantly improves workability while maintaining both compressive and tensile strength, attributed to its uniform dispersion and lubricating characteristics. In contrast, organic PCM led to reduced workability and lower strength, due to non-uniform distribution and poor interfacial bonding. Heat-transfer analysis showed that concrete containing inorganic PCM provided approximately a 10% delay in heat propagation and a 5% reduction in peak internal temperature compared to organic PCM mixes, demonstrating its superior thermal buffering capacity. The FE simulations exhibited strong agreement with experimental results, with minimal deviation in thermal profiles and predicted failure patterns. This confirms the reliability of the numerical framework and its applicability to modelling PCM-integrated concrete under structural fire conditions. Collectively, this study advances the understanding of PCM-modified concrete composites and demonstrates their potential to bridge the gap between thermal-energy-storage technologies and performance-based structural fire engineering, offering a viable pathway toward next-generation fire-resistant concrete systems. Furthermore, the proposed PCM-integrated concrete system contributes directly to the United Nations Sustainable Development Goals (UNSDGs), particularly SDG 9 (Industry, Innovation and Infrastructure) and SDG 11 (Sustainable Cities and Communities), by enhancing structural fire resilience, reducing material usage in passive fire protection, and improving community safety in high-density urban environments

ABSTRACT. This study performed thermal stress analyses on high-strength concrete columns while incorporating the effects of vapor pressure, with the aim of clarifying stress development during heating. The results demonstrated that transient thermal strain and free thermal expansion strain have a substantial influence on the magnitude of thermal stress, and that variations in these parameters can lead to pronounced increases in stress. When vapor pressure was considered, tensile stress developed in the radial direction, causing the stress state to move closer to the failure surface of concrete defined by the Drucker–Prager criterion. In contrast, the stress state in the model without vapor pressure remained well below the failure surface. These findings indicate that early-stage spalling may be assessed by examining the multiaxial failure behavior of concrete while accounting for vapor pressure effects.

ABSTRACT. Partially encased steel-concrete composite (PEC) columns, formed by infilling concrete between the flanges and web of a steel section, exhibit high load-bearing capacity and ease of construction. However, the exposed steel flanges heat up rapidly under fire exposure, which may lead to premature failure. To date, research on the fire performance of protected PEC columns remains limited. This paper presents an experimental investigation into the fire resistance of PEC columns with intumescent coating, non-intumescent coating, and without fire protection. The results show that, compared to unprotected columns, the application of intumescent and non-intumescent coatings enhanced fire resistance by 107.7% and 121.8%, respectively. Failure of the specimens was characterized by spalling of the concrete cover, crushing of concrete in the compression zone, and eventual overall instability about the weak axis. Parametric analysis reveals that load ratio, slenderness ratio, and cross-sectional dimensions considerably influence the fire resistance time. Considering both fire performance and economic efficiency, an extension length of 30 mm for the fire protection is recommended. Simplified calculation methods are proposed for determining the fire resistance capacity of both unprotected and protected PEC columns under axial compression, providing a practical foundation for the fire-resistant design of PEC columns.

ABSTRACT. The parametric fire is a simple fire model that can be considered a first step towards real fire calculations. It is capable of reflecting the influence of a number of physical parameters, both during the heating and cooling phases. Unfortunately, there is an inconsistency in the current EN model, which has also been copied in the formal vote of the upcoming Eurocode 1-1-2. The already known problem is briefly described, as are the already known solution and the consequences of the prescribed imperfections. Simulations are then carried out for different occupancies (which are reflected in fire loads) and the typical dimensions sometimes associated with them, using the proposed model, and compared with the results obtained from a two-zone model (Ozone). Finally, based on this higher-order model, the preconditions are outlined within which safe approximations can be obtained with the simple parametric model.

ABSTRACT. Large-scale modular steel compartments contain wide internal spaces and substantial combustible fuel loads, promoting severe thermal exposure during real fires. Concrete-Filled Tube (CFT) columns are commonly used in these structures due to their high compressive capacity, yet their performance can deteriorate significantly under fire-induced heating. While realistic thermal boundary conditions are essential for evaluating their response, direct implementation of experimental temperature data into FEM-based heat-transfer analysis remains challenging. This study establishes a coupled simulation framework that integrates real-scale compartment fire data with numerical analysis by linking Fire Dynamics Simulator (FDS) outputs to ABAQUS thermal modelling. A full-scale modular fire test was conducted in accordance with KS F ISO 9705-1 to obtain authentic fire temperature development, and Adiabatic Surface Temperature (AST) was extracted using BNDF settings to formulate the thermal boundary conditions. The extracted thermal field was subsequently mapped to CFT column surfaces within ABAQUS, enabling heat-transfer analysis that accurately reflects experimentally observed exposure. Verification of the temperature-mapping process demonstrated strong agreement between BNDF-derived and ABAQUS-applied temperatures, confirming the reliability and fidelity of the coupled FDS–ABAQUS methodology for assessing the thermal response of CFT columns in modular compartment fires

ABSTRACT. Structural response predictions during fire increasingly rely on numerical models. All numerical models have a certain modelling uncertainty, as they cannot represent the physical phenomenon perfectly. Because model uncertainty can significantly affect the predicted response, quantifying and propagating model uncertainty is essential for reliable structural fire performance assessments. Several methods have been used in fire engineering to quantify the model uncertainty. However, these methods are not suitable for time-dependent responses, where each time step represents a different state and each state is highly correlated to the others. This high correlation results in redundant information and inflated dimensionality, leading to lower computational efficiency and reduced interpretability. Principal components analysis (PCA) can be applied to effectively account for this correlation. This paper applies PCA to quantify the model uncertainty of a fibre model for time-dependent thermo-mechanical analysis, providing a systematic technique to characterise time-dependent model uncertainty in structural fire engineering and improve predictive capability within probabilistic performance-based design.

ABSTRACT. Fire-induced concrete spalling accelerates heat penetration and can significantly alter the damage state of reinforced concrete members. The spalling process itself is complex, and existing finite element tools do not capture spalling progression within the section. To address this limitation, the research team previously developed and implemented a simplified spalling model in finite element platforms. In this model, heat-induced spalling is idealized as a gradual and uniform process governed by three parameters (spalling initiation time, a constant spalling rate, and a termination criterion), determined from available experimental data. This study collects and analyzes a comprehensive dataset to develop probabilistic models for spalling initiation and spalling rate, incorporating variations in concrete strength and fire exposure type (ISO 834 and hydrocarbon-based curves). Several statistical distributions were fitted to medium- and large-scale test data and incorporated into heat-transfer analyses to assess temperature evolution and corresponding fire-induced damage classes. The Weibull distribution provided a suitable representation of both spalling parameters. Application of the probabilistic parameters to an ACI-designed slab illustrates the influence of uncertainty: using 90th-percentile spalling parameters results in rebar temperatures approximately 15–20 °C higher than median-based predictions at both 30 and 60 minutes of ISO 834 exposure, while the predicted damage class remains unchanged. Overall, this study provides a systematic quantification of spalling parameters for simplified modeling, with the goal of supporting the development of bounding estimates of fire-induced damage in concrete sections.

ABSTRACT. Although structural fire design is still predominantly based on prescriptive approaches, performance-based methods are gaining increasing attention due to their ability to account for realistic fire conditions and thus to support more effective and economical design solutions. In this framework, fragility curves — a well-established tool in seismic engineering [1] — have recently been introduced for probabilistic fire structural assessments [2], [3]. Their use enables engineers to estimate the probability that a structure will exceed specific damage states as a function of selected intensity measures (IMs), providing a probabilistic viewpoint that complements traditional deterministic fire assessments. The methodology adopted in this study consists of four main stages, summarised in Figure 1. First, a representative building typology is selected, and the required performance levels (PLs) are defined according to the Italian national fire code. Subsequently, realistic fire scenarios are identified, and corresponding thermal analyses are carried out. In the third stage, the thermal results are applied to the global or sub-structural model to perform advanced thermomechanical analyses and evaluate the structural response. Finally, fire fragility curves are developed based on the demand-capacity ratios (DCRs) obtained from the full set of thermomechanical simulations.

Figure 1. Flow chart of the methodology

The analysed structure is a six-storey reinforced concrete building with a rectangular plan (21.05 m × 11.45 m) and a total height of 19.2 m. The structural elements are made of C28/35 concrete and B450C reinforcing steel, and the loads are defined according to the residential occupancy of the building. For the fire scenario definition, a representative apartment within the reinforced concrete residential building was analysed. More than one hundred fire scenarios were considered, varying ignition location, window ventilation conditions, fire load [4], peak heat release rate (RHRf) [5], and auto-ignition temperature for travelling fire conditions (Figure 2a). In this application, simulations were performed using the zone-model software CFAST, which enabled a simplified yet effective analysis of the temperature evolution over time during the thermal transient (Figure 2b). Thermomechanical analyses were conducted in SAFIR to evaluate the structural response under fire exposure. The thermal analysis assessed temperature histories within the structural components based on the scenarios previously defined with CFAST. The subsequent thermomechanical analysis, performed by applying the temperature fields extracted from the thermal simulations, evaluated the building’s structural behaviour under combined thermal and mechanical actions, accounting for nonlinearity. Specifically, the analyses were conducted on a representative substructure of the building—previously validated against the global model and subjected to the same fire input. The substructure includes a fire-exposed floor and the supporting columns carrying the loads of the upper storeys (Figure 3). The natural fire curves were generated for all the selected scenarios, and the subsequent thermomechanical analyses produced preliminary results in terms of structural deformations and displacements (Figure 4). These outcomes constitute the basis for the ongoing development and calibration of the fire fragility curves. In the extended version of the paper, fragility curves will be presented for different intensity measures and multiple performance levels. The resulting curves are intended to serve as an effective probabilistic tool, supporting designers and decision-makers in the definition of reliable and performance-oriented fire safety strategies.

Figure 2. CFAST: a) Travelling fire in CFAST; b) CFAST output fire curves

Figure 3. a) plane view of the global structure model (yz); b) plane view of the sub-structure model (yz); c) 3D view of the sub-structure model

Figure 4. a) Structural displacement; b) slab displacement; c) column displacement.

References [1] Jalayer F, Ebrahimian H, Miano A, Manfredi G, Sezen H. Analytical fragility assessment using unscaled ground motion records. Earthq Eng Struct Dyn 2017;46 (15):2639–63. [2] Possidente L., Randaxhe J., Tondini N. (2023). Fire Fragility Curves for Industrial Steel Pipe-Racks Integrating Demand and Capacity Uncertainties. Fire Tech.59(3) doi:10.1007/s10694-022-01358-4 [3] De Silva, D., Miano, A., De Rosa, G., Di Meglio, F., Prota, A., Nigro, E. (2025). Analytical fire fragility assessment for bridges considering fire scenarios variability. Engineering Structures. https://doi.org/10.1016/j.engstruct.2024.119442 [4] Hopkin, C., Spearpoint, M., Wang, Y. et al. (2020) Design fire characteristics for probabilistic assessments of dwellings in England. Fire Tech., 56, 1179–1196, doi: 10.1007/s10694-019-00925-6 [5] Dundar, U., Selamet, S. (2023). Fire load and fire growth characteristics in modern high-rise buildings. Fire Safety Journal, Volume 135. https://doi.org/10.1016/j.firesaf.2022.103710.

ABSTRACT. Introduction Growing attention toward the safety of existing reinforced concrete (RC) buildings, particularly those of strategic importance or characterized by high occupancy, has highlighted the need for advanced tools to investigate structural performance under multiple hazard scenarios. Among the possible haz-ard combinations, the fire–earthquake scenario represents a particularly critical case. Although the two hazards are not causally correlated, their consecutive occurrence can significantly influence the structural response, since fire-induced degradation may compromise the residual load-bearing capac-ity before the onset of seismic loading. In this context, the present study proposes an integrated methodology for evaluating structural vulnerability under such non-simultaneous, multi-hazard loading sequences. Methodology The proposed methodology is structured into a sequence of progressive and interconnected phases aimed at realistically assessing the response of existing RC structures subjected first to a fire event and then to seismic action (Fig.1). The process begins with the definition of a RC structural archetype representative of the considered building use (e.g. strategic or ordinary buildings).

Figure 1: Flowchart of the proposed methodology Fire scenarios are identified on the basis of compartmental functional use and associated fire load density, from which the heat release rate (HRR) curves for the thermal analyses are derived. These scenarios serve as input for advanced thermo-mechanical analyses, performed using dedicated nu-merical tools [1], enabling the determination of temperature histories within structural sections, to-gether with the resulting degradation of material properties under fire exposure. In parallel, the seismic performance of the undamaged structure is assessed through non-linear static analysis, in order to evaluate some performance parameters, including maximum capacity, ductility and collapse mechanisms. The thermal analyses enable the estimation of residual material strength after fire exposure by corre-lating peak temperatures with appropriate reduction curves. For concrete, empirical relationships cod-ified in the Eurocodes [2] are adopted to express the reduction in strength and stiffness, as a func-tion of temperature. For steel reinforcement, reduction curves commonly referenced in the technical literature and supported by experimental studies are employed [3]. These procedures yield the reduc-tion coefficients necessary to characterize the post-fire mechanical properties of structural elements. The implementation of the degraded properties into the numerical model enables a new seismic vul-nerability assessment of the structure in its post-fire condition. The comparison of the seismic per-formance in the pre- and post-fire scenario allows to quantify the impact of fire events on the loading bearing capacity of the archetype structure both in terms of gravity and horizontal loads. The methodological framework ends with the development of seismic fragility curves that explicitly account for pre-existing fire-induced damage. This allows the effects of thermal degradation to be integrated into seismic risk assessments and enables a probabilistic evaluation of the likelihood of reaching different limit states as a function of both seismic intensity and prior material degradation. Proof-of-Concept application The methodology was initially validated through a simplified single-span frame model, which proved effective for verifying the consistency of the computational workflow and the integration between thermal and nonlinear structural analyses (Fig.2).

Figure 2: (a) simplified single-span frame model; (b) ISO 834 - fire thermal input; (c) fire thermal output; (d) Comparision between the pushover curves.

The comparison between the pre- and post-fire event pushover curves demonstrated the significant impact of material degradation due to fire on the seismic performance both in terms of strength and ductility. Following the validation phase, the procedure was applied to a real case study involving a strategical structure in Italy. For this structure, fire scenarios consistent with the functional use of each com-partment were defined, and advanced thermal analyses were carried out on the sections of the struc-tural elements. Then, nonlinear static analyses were caried out to investigate the seismic performance in the pre- and post-fire conditions. References [1] Jalayer, F., Ebrahimian, H., Miano, A., Manfredi, G., and Sezen, H. (2017), Analytical fragility as-sessment using unscaled ground motion records. Earthquake Engineering and Structural Dynam-ics. https://doi.org/10.1002/eqe.2922 [2] Ma, C., Van Coile, R., and Gernay, T. (2024), Fire protection costs in composite buildings for cost-benefit analysis of fire designs. Journal of Constructional Steel Research. https:// doi.org/10.1016/j.jcsr.2024.108517 [3] Miano, A., Sezen, H., Jalayer, F., and Prota, A. (2019), Performance-based assessment method-ology for retrofit of buildings. Journal of Structural Engineering. https://doi.org/ 10.1061/(ASCE)ST.1943-541X.0002419 [4] National Institute of Building Science (NIBS). (2003), HAZUS-MH MR1 Technical Manual, devel-oped by the Federal Emergency Management Agency, Washington, D.C. [5] Hicks, H. L., and Liebermann, R. R. (1979), Study of Indirect Fire Losses in Non-Residential Prop-erties. FoU-brand, 8-14, January 1979.

ABSTRACT. Concrete-filled steel tube (CFT) columns are known to have high fire resistance due to the confinement provided by the steel tube and the thermal inertia of the concrete core. Accordingly, numerous studies have been conducted to investigate the structural behavior and fire resistance of CFT columns exposed to elevated temperatures. Early research primarily focused on experimental parametric studies of circular and square CFT columns, examining the influence of section geometry, load ratio, concrete strength, and slenderness on fire resistance. With the development of numerical techniques, thermo-mechanical finite element analysis (FEA) has subsequently been employed to simulate fire tests on CFT columns. These studies consistently reported that fire resistance tends to decrease in the order of circular, elliptical, square, and rectangular sections. Despite the diffences in fire behavior and structural response with cross-sectional shape, most existing fire tests have been concentrated on circular and square CFT columns. As a result, the fire resistance prediction equation in AISC 360-22, which was developed mainly from circular and square data, and considers rectangular sections only by reflecting the short side, may not fully reflect the behavior of rectangular CFT columns. To address this limitation, the present study conducted thermo-mechanical finite element analyses to investigate the structural fire behavior of rectangular CFT columns at elevated temperatures. The numerical model was first validated using available fire test data for both square and rectangular CFT columns to ensure model reliability. During this validation process, a detailed examination of the concrete core stress distribution revealed clear differences between the two sections at the final failure stage. In square columns, stresses were concentrated toward the center region as thermal degradation progressed, indicating that the concrete core played a significant role in carrying axial loads under global buckling. However, in rectangular columns, minor-axis buckling governed the structural response, and the deformation pattern of the steel tube altered the internal stress transfer paths, leaving the concrete core carrying only limited stress. These observations suggested that the structural fire behavior of CFT columns is strongly influenced by the depth-to-width aspect ratio. Therefore, an extensive parametric study was performed considering concrete strength, load ratio, section dimensions, aspect ratio, and reinforcement ratio. The results confirmed that as the aspect ratio increased beyond 1.0, fire resistance rapidly decreased, and the contribution of the concrete core diminished due to an increased tendency toward minor-axis buckling. This trend was consistently observed in both the stress distribution analysis and the final fire resistance performance obtained from the parametric investigation. Finally, the outcomes from the parametric study were compared with the current AISC 360-22 Appendix 4 prediction approach. The comparison revealed that the existing AISC formulation does not explicitly account for aspect ratio and consequently fails to represent the reduced concrete contribution and increased buckling sensitivity in rectangular sections. Based on these findings, this study emphasizes the need to incorporate the aspect ratio as a modifying factor in the AISC fire resistance evaluation to more accurately reflect the structural behavior of rectangular CFT columns under fire conditions.

ABSTRACT. Deep tunnels operating under high geostatic confinement face severe deformation and degradation when exposed to intense fire scenarios. Predicting their structural response requires advanced numerical tools capable of integrating pre-fire stress history, soil–lining interaction, and temperature-dependent material behavior. This study presents a high-fidelity, explicitly staged 2D finite element model that simulates geostatic loading, excavation, lining installation, and long-duration fire exposure. The model incorporates Eurocode-based thermo-mechanical degradation of concrete and reinforcement, transient heat transfer from severe fire curves, and coupling between thermal expansion and mechanical restraint. A comprehensive parametric study evaluates the influence of soil stiffness, lining thickness, concrete type, fire curve severity, and geostatic stress ratio on internal forces, deformation patterns, and thermal damage progression. The results highlight the critical role of excavation-induced stress redistribution and soil confinement in amplifying fire-induced bending moments, axial forces, and crack formation. The model provides a realistic and robust framework for performance-based fire design of deep tunnels.

ABSTRACT. Jet fires, resulting from the ignited release of high-pressure flammable fluids, are among the recognised industrial accidents [1], [2]. Nevertheless, their effects on structural members or assemblies have not yet been thoroughly investigated, apart from synthetic formulations of time-to-failure, typically associated with specific industrial equipment. Fire Dynamic Simulator (FDS) software [3] was used to simulate two cases: a subsonic (at 1.3 bar pressure) and a supersonic (at 3 bar pressure) jet flames. Although FDS employs low Mach number combustion equations, which are distinctive to subsonic flows, the model was applied beyond its intended regime to also encompass the supersonic case. The process involves a spray of cryogenic methane droplets with high injection velocity and a constrained dispersion cone, which is subsequently ignited [4], [5]. A droplet size of 500 μm [6] and a nozzle diameter of 40 mm were considered. Each presented case was subdivided to analyse three distinct distances between the jet flame and the concrete column, representing far-field, near-field and engulfing scenarios, while the vertical distance of the nozzle to the beam was kept constant at 4 meters. Figure 1 shows pictures taken from the simulations of the above-cited analysis supersonic cases, only. The simulation outputs include the Rate of Heat Release (RHR), along with temperature and Incident Heat Flux (IHF) data, the latter being measured by sensors located at 1-meter intervals along the beam length (15 m) and column height (5 m). The plateau IHF for the far-field, near-field and engulfing subcases are plotted in Figure 2. It can be seen how moving the nozzle towards the column, the IHF peak over the beam is reached at a different x-coordinate (left figure), with decreasing values. This can be explained by the impingement and engulfment of the column (see the right figure – IHF on the z-axis). Much higher variations are observed for the column element, where the peak value increases from 12.5 kW/m2 (case a) to 63.5 (case b), reaching its maximum for the engulfing case with 191.8 kW/m2. The six cases identified, and the respective irradiance time histories, have been implemented in SAFIR [7] for a thermomechanical analysis. The results of the thermal analysis (Figure 3) highlight that under both far- and near-field jet fires, assumed a duration of 400 s, the thermal effect is limited to the very external concrete cover, whilst it may endanger the structural safety in case of engulfment, exposing three column sides. The paper also discusses the influence of the exposure time to the jet fire up to the closure of the safety valve over the structural performance of the frame.

References [1] A. Palacios, M. Muñoz, and J. Casal, ‘Jet fires: An experimental study of the main geometrical features of the flame in subsonic and sonic regimes’, AIChE J., vol. 55, no. 1, pp. 256–263, Jan. 2009, doi: 10.1002/aic.11653. [2] Joaquim Casal, M. Gomez-Mares, M. Muñoz, and A. Palacios, Eds, ‘Jet Fires: a “Minor” Fire Hazard?’, Jet Fires “Minor” Fire Hazard, vol. Chemical Engineering Transactions, no. 26, pp. 13–20, 2012, doi: 10.3303/CET1226003. [3] McGrattan K., McDermott R., Hostikka S., Floyd J., Vanella M., Weinschenk C., Overholt K, FDS - Fire Dynamics Simulator. (2017). NIST Special Publication 1019. [4] A. Palacios and B. Rengel, ‘Computational analysis of vertical and horizontal jet fires’, J. Loss Prev. Process Ind., vol. 65, p. 104096, May 2020, doi: 10.1016/j.jlp.2020.104096. [5] K. McGrattan, ‘MODELING LARGE LIQUEFIED NATURAL GAS FIRES’, presented at the Fire and Evacuation Modeling Technical Conference (FEMTC), Brno, Czech Republic, 2022. [6] A. Àgueda, V. Valdeabella, X. Molina, V. Foroughi, and E. Pastor, ‘Parametric study of jet fires simulation in FDS using a spray-of-particles approach’, J. Phys. Conf. Ser., vol. 3121, no. 1, p. 012046, Sept. 2025, doi: 10.1088/1742-6596/3121/1/012046. [7] J.-M. Franssen, ‘SAFIR: A Thermal and Structural Program for Modeling Structures Under Fire’, Eng. J., vol. 42, no. 3, pp. 143–158, Sept. 2005, doi: 10.62913/engj.v42i3.856.