Sie können Operatoren mit Ihrer Suchanfrage kombinieren, um diese noch präziser einzugrenzen. Klicken Sie auf den Suchoperator, um eine Erklärung seiner Funktionsweise anzuzeigen.
Findet Dokumente, in denen beide Begriffe in beliebiger Reihenfolge innerhalb von maximal n Worten zueinander stehen. Empfehlung: Wählen Sie zwischen 15 und 30 als maximale Wortanzahl (z.B. NEAR(hybrid, antrieb, 20)).
Findet Dokumente, in denen der Begriff in Wortvarianten vorkommt, wobei diese VOR, HINTER oder VOR und HINTER dem Suchbegriff anschließen können (z.B., leichtbau*, *leichtbau, *leichtbau*).
Dieser Artikel präsentiert eine akribische dreiphasige Bewertung von Stahlbetonträgern (RC-Trägern), die einem Brand ausgesetzt sind, wobei eine kritische Lücke in der aktuellen Forschung durch Einbeziehung von Vorlade- und Kühlphasen geschlossen wird. Die Studie beginnt mit einer Einführung in die Häufigkeit von Brandereignissen in Gebäuden, wobei die Widerstandsfähigkeit von Beton unter Brandbedingungen und die Notwendigkeit eines gründlichen Verständnisses seines Verhaltens hervorgehoben werden. Das Experimentalprogramm umfasst die Herstellung und Prüfung von zwei RC-Trägern, RCB1 und RCB2, mit detaillierten Abmessungen und Materialeigenschaften. Die Methodik umfasst Vorbrandtests bei Raumtemperatur, Brandexposition und Resttests nach dem Brand, die eine umfassende Beurteilung der strukturellen Reaktion der Balken liefern. Der Artikel geht auf die thermischen und strukturellen Reaktionen der Balken während des Brandtests ein und diskutiert Temperaturgefälle, Dehnungsentwicklung und beobachtete Versagensmodi. Außerdem werden die experimentellen Ergebnisse mit Vorhersagen der codebasierten und Finite-Elemente-Analyse (FEA) verglichen, wodurch die Genauigkeit der FEA-Simulationen bestätigt wird. Die aus der Studie gezogenen Schlussfolgerungen betonen die Praktikabilität des dreiphasigen Bewertungsansatzes, die Widerstandsfähigkeit von Bewehrungsstahl und die signifikanten Auswirkungen der Abkühlphase auf das strukturelle Verhalten nach dem Brand. Dieser Artikel bietet wertvolle Einblicke in die Beurteilung und Sanierung brandgeschädigter Bauwerke und ist daher eine unverzichtbare Lektüre für Fachleute, die ihr Verständnis von feuerfesten Stahlbetonträgern verbessern möchten.
KI-Generiert
Diese Zusammenfassung des Fachinhalts wurde mit Hilfe von KI generiert.
Abstract
This study introduces an improved three-phase evaluation approach to assess the residual strength and structural response of reinforced concrete (RC) beams subjected to fire exposure. Highlighting the potential for RC structures to be utilized again after fire incidents, this research emphasizes the need for comprehensive assessment and rectification to determine their remaining strength. The methodology employs a systematic approach, segmented into three essential phases: pre-fire testing (conducted at room temperature), fire testing including a cooling phase, and post-fire (residual) testing. These evaluations were carried out on two full-scale beam specimens, each 200 mm wide, 300 mm deep, and spanning 2000 mm in length. One RC beam was not exposed to fire, while the other one was subjected to fire on three sides. Each RC beam was then tested under the four-point bending setup until failure to determine the remaining strength. The furnace temperature curve was also compared with BS 476-20 and ASTM E119-00a standard fire curves for reference. The investigation revealed that factors such as thermal degradation of the concrete and pre-loading significantly affect the post-fire performance of the RC beam, leading to irrecoverable plastic deformation. This was evidenced by a 26% reduction in ultimate load and a 51% decrease in secant stiffness for the fire-exposed beam. Additionally, the beam exhibited a 5 mm residual deflection in the unloaded state and a significant 75.8% increase in deflection at ultimate load, rising from 22.3 to 39.2 mm. Further analysis demonstrated the limitations of the 500 °C Isotherm method, which underestimated residual capacity by 16.5%, compared with Finite Element Analysis simulations that closely matched experimental results with only a 1.2% difference. The proposed three-phase evaluation not only deepens the understanding of the structural behavior under fire exposure but also provides a systematic framework for assessing its continued usage throughout its service life.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
1 Introduction
Fire incidents are one of the most severe threats to buildings and other civil related structures, with statistics from the Malaysia Open Data Portal revealed a consistent trend from 2015 to 2021. In 2015, the total number of reported fire incidents was 5609, with residential fires accounting for 3027 cases. By 2021, although the number of reported fire incidents decreased to 3235, residential buildings remained the highest, with 1902 cases reported [1]. This data highlights a critical aspect of fire incidents across various types of buildings where every year, residential buildings consistently represent more than 50% of all cases, making them the highest among the other types of buildings.
Given the widespread use of concrete in construction, it has become critically important to grasp its fire resistance capabilities. The inherent resilience of concrete under fire conditions often makes it possible for fire-damaged buildings to undergo repairs and be put back into service, assuming the structural elements can recover their original strength, stiffness, and ductility [2‐4]. Kodur and Raut [5] also stated that concrete can endure high temperature resulting from fire incident. However, fire exposure instigates chemical and physical changes within the concrete, such as moisture loss, dehydration of cement paste, and aggregate decomposition. These changes undermined the concrete's structural integrity and can significantly reduce its mechanical properties, even without any visible damage [6‐11].
Anzeige
Over the last few decades, extensive investigations have explored the behavior of reinforced concrete (RC) structures under fire conditions, providing valuable insights into their responses to extreme temperatures [12‐21]. However, many of these studies have overlooked the effects of pre-loading and the cooling phase, both critical for accurately replicating practical scenarios. Pre-loading, which simulates in-service conditions, offers a precise evaluation of RC beams' resistance to elevated temperatures while testing them in their most critical state [22]. Fire testing standards, including ISO 834-1 [23], ASTM E119-00a [24], and BS 476-20 [25], emphasize the importance of incorporating pre-loading. Equally vital, the cooling phase significantly affects material recovery and residual behavior, as concrete can continue to lose strength during this stage [26, 27]. Furthermore, the thermal inertia of concrete delays peak temperatures, potentially compromising structural performance during the cooling phase [28‐30].
Addressing this need, the current research introduces an improved three-phase evaluation approach that incorporates both pre-loading and a cooling phase into the experimental framework. This methodology systematically evaluates the residual strength and structural response of RC beams after fire exposure. It encompasses pre-fire testing at room temperature, fire exposure, and post-fire (residual) tests. Through this approach, the study provides a comprehensive understanding of the mechanical property changes in RC beams, offering practical guidelines for assessing and rehabilitating fire-damaged structures to ensure their safety and structural performance in post-fire scenarios.
2 Experimental Program
2.1 Test Specimens
Two rectangular RC beams, named RCB1 and RCB2 with dimensions of 200 mm wide, 300 mm deep and span of 2000 mm, were fabricated at the Structural and Material Laboratory. In addition, the effective span between the support was maintained at 1800 mm. The schematic diagram of the RC beam is shown in Fig. 1 and the detailed dimensions are provided in Table 1. Each RC beam featured a concrete cover of 45 mm with two H12 bars positioned at both the top and bottom section for the compression and tension reinforcements, respectively. In addition, shear links were also provided using R8 reinforcement bars, spaced at 150 mm and bent at 135° into the central part of the RC beams. These RC beams were designed as under-reinforced section to ensure the failure would occur through flexural behaviour. Based on EN 1992-1-2 [31] provisions for flexural design, the calculated design ultimate moment capacity (Mu) of the beams was 30.97 kNm.
Type I Portland cement, graded river sand as fine aggregate, and 10 mm crushed granite as coarse aggregate were used in the concrete mix for both RC beams. The concrete composition shown in Table 2 was designed to reach a compressive strength of 30 N/mm2 at 28 days. The actual compressive strength at 28 days for RCB1 and RCB2 specimens based on the average of three 100 mm × 100 mm × 100 mm cube samples were 50.99 N/mm2 and 45.95 N/mm2, respectively. On the other hand, high strength steel and mild steel were used for the main and shear reinforcements, respectively. To confirm this, tensile test was conducted on three samples for each the high strength steel and mild steel, where the average yield strength was 604 N/mm2 and 862 N/mm2, respectively as detailed in Table 3.
Table 2
Concrete mix proportions composition for 1 m3
Mix type
Water-to-cement ratio, w/c
Water (kg/m3)
Cement (kg/m3)
Fine aggregates (kg/m3)
10 mm coarse aggregates (kg/m3)
Normal strength concrete (NSC)
0.52
235
450
795
895
Table 3
Mechanical properties of reinforcement bars
Reinforcement
Diameter, d (mm)
Average Yield Strength, fy (N/mm2)
Average Ultimate Strength, fu (N/mm2)
Average Elastic Modulus (kN/mm2)
H12
11.8
604
684
217
R8
7.17
862
900
214
Anzeige
2.3 Instrumentations
Steel strain gauges, namely SG1 and SG2, were attached to both tension and compression reinforcements to assess the stress–strain behavior under the applied incremental load, as shown in Fig. 2a and b. Additionally, concrete strains were measured using DEMEC discs, which were mounted on the side of the RC beams at six locations as shown in Fig. 4. These DEMEC discs were positioned at a gauge length of 150 mm along its mid-span region.
Fig. 2
Strain gauges and thermocouple locations for: a RCB1, and b RCB2 specimens
For RCB2 specimen, which underwent fire resistance testing, five thermocouples, namely TC1 to TC5, were embedded to measure the temperature of the RC beam along its depth as shown in Fig. 2b. The specific depth locations are as follows: TC1 at the concrete cover, TC2 at the corner tension reinforcement bar, TC3 at 100 mm depth from the soffit of the beam, TC4 at 150 mm depth or the centroid of the beam, and TC5 at the compression reinforcement bar.
2.4 Details of Furnace
The fire resistance test for RCB2 specimen was conducted using a furnace designed with reference to the time–temperature curves outlined in BS 476-20 [25] ISO 834-1 [23], and ASTM E119-00a [24]. This furnace accommodates both static and cyclic loading, which can be applied to the specimen simultaneously during the heating process. It is equipped with a gas burner control unit designed to regulate gas flow, aiming to align the furnace temperature with the standard fire curve. The furnace dimensions are 1500 mm wide × 1500 mm long × 1500 mm high. Its isometric view and the actual furnace in the laboratory are shown in Fig. 3a and b, respectively.
Fig. 3
Furnace setup: a Isometric view, and b Actual furnace in the laboratory
In the first phase testing, a four-point bending test was performed on RCB1 specimen following ASTM C78 [32] procedures. This phase was designed to assess the structural behavior of the RC beam under static loading condition at room temperature. It also aimed to identify the first crack load, which was crucial in establishing the pre-load for the subsequent fire test, as well as its ultimate loading capacity. Being designed as an under-reinforced section will also ensure failure would occur through the yielding of the tension reinforcements prior to concrete crushing. For enhanced crack visibility, both RCB1 and RCB2 specimens were white painted on its side after fabrication, and cracking development was marked using a black marker pen during the test.
To measure deflection, three 100 mm linear variable displacement transducers (LVDTs) were also positioned at mid-span and under the two loading points. The test setup involved placing RCB1 specimen on a testing rig capable of supporting load of up to 500 kN. Additionally, a spreader beam was used to distribute the load into two points arranged symmetrically at 600 mm from both end supports to produce a constant bending moment at the mid-span region. The load was first applied incrementally and in a controlled manner using hydraulic jacks. This load control method involved applying the load at a rate of 1 kN/min. Once the first crack was observed, displacement control was applied at a rate of 0.05 mm/min. Deflection, strain (both steel and concrete), and crack development were continuously monitored throughout the test.
LVDTs and strain gauges were logged into a TDS-530 data logger to collect data for deflection and steel strain. Meanwhile, concrete strain was measured using DEMEC discs mounted at predefined points along the sides of the specimens. A digital extensometer with an accuracy of 0.001 mm was used to record the strains at these points. Bending test was carried out until failure, which was designed by the yielding of the tension reinforcements followed by either concrete crushing or large deflection without any increment of the applied load. Figure 4 shows the schematic drawing of the bending test setup in Phase 1, while Fig. 5 illustrates the actual test carried out in the laboratory. The same test setup was also used in Phase 3 of the post-fire residual test, which is described in Sect. 3.3.
Fig. 4
Schematic drawing of the bending test setup in Phase 1 and Phase 3
In Phase 2, the fire test procedures focused on two main stages, (i) loading stage, and (ii) heating stage. During the loading stage, RCB2 specimen was first positioned as simply supported arrangement inside the furnace. A two-point loading system (using a spreader beam) spaced at 700 mm from each end was then positioned on top of the RC beam specimen. Loading was applied incrementally, using load control, increasing by 2 kN at each step, until it reached the first crack load. The first crack load was determined based on the bending test conducted on RCB1 specimen in Phase 1. This pre-load was then maintained on RCB2 specimen throughout the fire test [22, 33]. Prior to igniting the furnace, the pre-load was maintained for approximately 15 min to allow deflection to stabilize [34, 35], thereby establishing a baseline condition for the subsequent thermal loading.
During the heating stage, the temperature increase was primarily referenced to the BS 476-20 [25] fire curve, with ASTM E119-00a [24] serving as an additional benchmark. The furnace temperature was monitored throughout the test using thermocouples to ensure reliable data collection. RCB2 specimen was exposed to fire on three sides; bottom surface and both side surfaces. Meanwhile, the top surface was protected using a 50 mm thick layer of fiber material insulation, as depicted in Fig. 6. This setup reflects typical condition, where the top surface is protected by concrete slab. In this specific fire test, only the mid-span of RCB2 specimen, spanning 1200 mm, was directly exposed to fire, a setup reflecting typical building section where the supported section remained unexposed. The 90 min (R90) duration of fire resistance was chosen based on the design concrete cover thickness of 45 mm to be in line with EN 1992-1-2 [31] specification. After the heating stage, the furnace was switched off to initiate the cooling phase. During this phase, RCB2 specimen was allowed to cool naturally at room temperature while it remained inside the furnace to prevent thermal shock. Thermal shock can occur when hot structures come into contact with large quantity of cold extinguishing medium, such as water, or from rapid cooling. This can lead to additional damage, thus compromised the structural integrity of the component [12, 36‐38]. The same pre-load applied during the fire exposure was continuously maintained throughout the cooling phase until the specimen returned to ambient temperature [39] and was ready for unloading. Thermocouples embedded within the beam monitored the temperature gradient, capturing the transition of the concrete and reinforcement bars from elevated temperature to ambient condition.
Fig. 6
RCB2 Cross-sectional view of fire-exposed and unexposed surfaces
Due to equipment limitations, LVDTs were not installed during the fire test to capture real-time deflection. Instead, steel strain gauges were installed on the tension and compression reinforcement bars to monitor strain development during fire exposure. While initial readings were successfully recorded, both gauges failed partway through the heating stage as a result of elevated temperature effects. Consequently, complete strain histories and deflection measurements were not available throughout the heating and cooling phases.
During the fire test, direct visual observation of the RCB2 specimen was not possible due to its high temperature throughout the testing period. As a result, the assessment of crack development and the critical temperature of the reinforcement bars was conducted only after the fire test was completed. The critical temperature, defined as 593 °C, marks the point at which steel reinforcement loses up to 50% of its original strength at room temperature, as outlined by ASTM E119-00a [24]. This significant loss in strength compromises the load-bearing capacity of the reinforcement, potentially matching the applied load and leading to material degradation and the onset of structural instability [40]. It is important to emphasize, however, that reaching this threshold does not imply the beam failed to survive the fire test. Instead, the critical temperature serves as a reference for understanding the extent of material degradation during the fire phase. Figures 7 and 8 provide schematic and actual representations of the fire test setup, respectively.
In Phase 3 of the post-fire condition, after completing the cooling phase in Phase 2, the RCB2 specimen was stored indoors at room temperature for an additional 69 h to allow stabilization of its mechanical properties. This duration was selected to ensure the specimen had fully returned to ambient condition before the unloading process. This was also consistent with recommendations by Agrawal and Kodur [39] noting that fire-exposed concrete members may require 24 h to 72 h to completely cool depending on their thermal mass. This extended storage period allowed the concrete and reinforcement bars to regain thermal equilibrium, ensuring that no residual thermal effects influenced the subsequent post-fire testing [41]. Following this stabilization, the RCB2 specimen was subjected to a four-point bending test, as shown in Fig. 9, employing a similar setup to that used in Phase 1 for the RCB1 specimen. In this phase, however, a displacement-controlled hydraulic actuator (Instron with a capacity of 1000 kN) was used instead of a hydraulic jack to apply the loading increment. The two loading points were arranged symmetrically at 700 mm from each end of the RCB2 specimen.
Before commencing the test, DEMEC discs were installed to measure the concrete strain, positioned at the same locations as in Phase 1 of RCB1 specimen. It is important to note that both steel strain gauges (SG1 and SG2) were damaged during the Phase 2 and could not be used in this phase. This test is aimed to evaluate the structural performance of RCB2 specimen after fire exposure, focusing on its load–deflection behavior, concrete strain, residual deflection, and failure mode. The loading was applied incrementally using load control at a rate of 1.5 kN/min until failure was observed, which was characterized by significant deflection without any further load increase, concrete crushing, or visible extensive cracking width and length.
4 Test Results and Discussions
4.1 Phase 1 (Pre-fire): Room Temperature Test
Figure 10 shows the load–deflection relationship for the RCB1 specimen under the Phase 1 test. A straight-linear line was first observed up to 28.1 kN, reflecting the elastic behavior of RCB1 specimen until the first crack started to develop. This first crack load was also applied as the pre-load value in Phase 2 of the fire test. After the first crack was observed, the linear elastic behavior changed to a curvilinear relationship, indicating that RCB1 specimen behaved plastically until it reached the yielding point of the reinforcement bars at 110.6 kN. The smooth transition from the straight-linear line to the curvilinear line shows the ductility behavior of the RCB1 specimen, or its capacity to withstand bending and deformation. When the applied load reached its ultimate at 138.1 kN, the deflection of RCB1 specimen was recorded at 22.3 mm, demonstrating its ability to absorb energy and deform significantly. This behavior showcases the beam's ductility and capacity to sustain load before experiencing flexural failure.
Further detail of the load–deflection behavior is exhibited in Fig. 11, which shows the load-strain relationship, illustrating the behavior of steel and concrete under tension and compression. Strain gauge SG1, which measures the tension reinforcement bar, shows the initial elastic region up to 28.1 kN with a strain of 119 με, indicating a proportional relationship between the load and strain where the material behaves elastically. In the same tension zone, the concrete strain at DM1, located 25 mm from the soffit of the specimen, also demonstrated elastic behavior up to around 28.1 kN with a strain of 220 με. At this point, the concrete strain in DM1 significantly increases, marking 28.1 kN as the first crack load where RCB1 specimen began to crack. As the load increases to 34.7 kN, DM1 exhibited a strain of 2060 με, confirming that cracking had occurred, and the tensile reinforcement bar began to take over the tensile load from the cracked concrete.
Fig. 11
Load-strain relationship of steel reinforcement and concrete of RCB1 specimen
The yield strain for the tensile reinforcement bar recorded at 110.6 kN was 3380 με, where plastic deformation began, marked by the flattening behavior of the curve. Beyond this point, strain hardening in the tensile reinforcement bar was evident as the load increases slightly to 113.2 kN with a significant steel strain of 25,303 με, indicating the increased strength and hardness of the material as deformation progresses. This behavior highlights the ductility of the steel in tension as it effectively accommodates the transferred load from the cracked concrete. For the compression reinforcement bar, the initial elastic behavior was observed up to 110.6 kN with a small compressive strain of -35 με. This corresponds to the behavior of concrete strain at DM6, located 25 mm from the top surface in the compression zone, which shows an elastic response extending up to just before the ultimate load. As the load increases, the compression reinforcement bar reached its maximum strain of 1344 με at an ultimate load of 138.1 kN, while the concrete strain at DM6 demonstrated a compressive strain of -3320 με. Although six DEMEC gauges were installed on the side of RCB1 specimen, the discussion focused on DM1 and DM6 due to their critical location, representing the concrete's behavior in tension and compression zones, respectively.
Figure 12 illustrates the crack patterns and failure modes observed in the RCB1 specimen. The first crack load was recorded at 28.1 kN, marking the onset of deformation as tensile stresses surpassed the tensile strength of the concrete. Beyond this point, the deflection of the RCB1 specimen increased at an accelerated rate as the applied load continued. Flexural cracks were predominantly observed in the constant moment region, where bending stresses were highest. These cracks initiated at the bottom of the beam and propagated vertically upward, indicating that the tensile capacity of the concrete had been exceeded. The failure mode observed was consistent with the expected behavior of an under-reinforced beam, where yielding of the tension reinforcement occurred first, allowing substantial deformation prior to failure. This was followed by concrete crushing in the compression zone at the top of the beam as the compressive strength of the concrete exceeded. The ultimate load of the beam was recorded at 138.1 kN, with crushing localized in the constant moment region. This crushing marked the beam’s ultimate load capacity, where compressive stresses exceeded the material’s limits, leading to failure in the compression zone.
Fig. 12
Crack patterns and failure modes observed in the RCB1 specimen during and after the bending test
4.2.1 Thermal Response (Time–Temperature Relationship of the Furnace)
Figure 13 presents the time–temperature curve comparison between the furnace's average temperature and the BS 476-20 [25] fire curve, the standard used as the primary reference for the heating stage. During the first 5 min of the heating process, the relationship of the furnace's temperature was closely aligned with BS 476-20 [25] standard fire curve. BS 476-20 [25] standard fire curve prescribed a temperature rate increase of 36.5 °C/min in the first 10 min, showing a typical reflective rapid escalation in the initial stage of a fire test. The furnace, on the other hand, shows a more moderate temperature increment rate at 24.3 °C/min, suggesting a small difference with the standard fire curve in BS 476–20 [25]. This indicates that the early-stage simulation of the fire temperature acceleration in the furnace was less intense than that defined in BS 476-20 [25].
Fig. 13
Furnace temperature against BS 476-20 [25] and ASTM E119-00a [24]
Reaching the time from 10 to 37 min, the temperature increment rate specified by BS 476-20 [25] decreased to 7.2 °C/min, while the furnace recorded a slightly higher rate of 12.4 °C/min during this period. From 37 to 60 min, BS 476-20 [25] prescribes a slower temperature increment rate of 3.1 °C/min, reflecting a gradual phase of fire development. The furnace, on the other hand, shows a more controlled rise at a slower rate of 1.5 °C/min during this interval. Between 60 and 90 min, the furnace temperature began to stabilize, showing a slight reduction in the temperature rate of 4 °C/min, whereas BS 476-20 [25] prescribes a minor increase of 2 °C/min. During the cooling phase, spanning from 90 to 240 min, the furnace temperature progressively decreased at an average rate of 6.5 °C/min, ultimately reaching a residual temperature of 66 °C.
In order to evaluate the furnace's performance against established benchmarks outlined in Sect. 3.2, the increase in temperature was compared to both the BS 476-20 [25] and ASTM E119-00a [24] fire curves, as shown in Table 4. In the first 10 min of the firing process, the temperature inside the furnace was 28% higher than the limit, which permits only a 15% variation as stipulated in BS 476-20 [25]. Between the 10 min and 30 min interval, this difference was 26.9%, exceeding the 10% acceptable limit in the code. Later and beyond the 30 min interval, the furnace's temperature continued to deviate by 8%, surpassing the 5% acceptable variations. As shown in Fig. 13, when compared with ASTM E119-00a [24], the time–temperature curve of the furnace falls within the acceptable limits. The first 60 min recorded a difference of 8.2%, and from 60 to 120 min was 1%, where both measurements were within the ASTM's allowable percentage limit of 10% and 7.5%, respectively. While the furnace's performance deviated from BS 476-20 [25], it remained compliant with the percentage deviation limits recommended by ASTM E119-00a [24].
Table 4
Percentage differences of the area under the curve for the time–temperature relationships between the experimental and code of practice
For comparison with previous studies employing gas burner methods similar to this research, Fig. 14 illustrates significant deviations in the time–temperature curves relative to standard fire curves. Song et al. [42] presented a curve that generally followed the standard trend but consistently remained about 100 °C lower throughout the exposure period. In contrast, Panedpojaman et al. [43] exhibited a more substantial deviation, with their curve being approximately 200 °C lower than the standard for the entire test duration. Huo et al. [44], in a shorter test lasting around 30 min, displayed a curve that was approximately 130 °C lower than the standard, although it followed a similar trend during this interval.
Fig. 14
Time–temperature curves comparison for gas burner between previous and current studies
The thermal response of the furnace during the fire test demonstrated acceptable alignment with the standard time–temperature profile, despite some deviations. When compared with ASTM E119-00a [24], the furnace's performance remained within the permissible limits, with percentage differences of 10% and 7.5% for specific intervals. However, benchmarking against BS 476-20 [25] revealed deviations exceeding the allowable variations, particularly during the first 30 min of heating. These differences, attributed to the manual operation of the gas burner system, underscore the inherent variability of such setups compared with the precision of electric furnaces. Nevertheless, this study effectively utilized the gas burner system to achieve its primary objective of subjecting beams to fire exposure under pre-loading conditions, enabling critical thermal and structural data collection. Comparisons with prior studies employing similar systems, such as Song et al. [42], Panedpojaman et al. [43], and Huo et al. [44], revealed significant deviations from standard fire profiles. Unlike those studies, which consistently recorded temperature curves well below expected profiles, this research demonstrated notable improvements in alignment with standard fire curves, showcasing the reliability and applicability of the manually controlled gas burner system for consistent fire simulation condition.
4.2.2 Thermal Response (Concrete and Reinforcement Bar)
Figures 15 and 16 illustrate the thermal response of the RCB2 specimen exposed to fire, showing the temperature changes in both the concrete and reinforcement bars at various locations along the cross-section of the beam. The temperature was recorded using five thermocouples placed at different points within RCB2 specimen, as shown previously in Fig. 2b. As the fire progresses, the temperature within RCB2 specimen corresponded closely with the average furnace temperature, suggesting a direct influence of the heating process on its material behavior. A temperature plateau pattern was observed between 110 and 140 °C for both the concrete and reinforcement bars at various locations within the cross-section of RCB2 specimen. This plateau pattern occurred due to the absorption of latent heat by the free capillary water within the RCB2 specimen, transitioning from liquid to vapor [34, 39, 45]. As most of this water evaporated, temperature in the concrete and reinforcement bars increases with the fire temperature during the firing process.
Fig. 15
Measured temperature of the concrete in RCB2 specimen
At 42 min of fire exposure, thermocouple TC1, positioned at 23 mm from the bottom soffit of the beam, recorded a sharp temperature spike at 423 °C, showing a 47% increase from the 287 °C recorded at the 40 min. Another significant spike was observed at fire exposure of 53 min, with the temperature reaching 517 °C, showing a 38% rise from the 376 °C recorded at the 49 min. These spikes are likely influenced by the visible crack observed on the soffit at the middle section of the beam, as depicted in Fig. 17. Such cracks can expedite temperature elevation by creating a direct path for heat to reach the core of the beam. Essentially, this crack enabled heat to bypass the concrete cover, thus accelerating the increase in temperature at TC1. Meanwhile, during the cooling phase, temperatures within the cross-section of RCB2 specimen continued to rise, demonstrating the high thermal inertia of concrete. This effect was more pronounced in the inner sections of the concrete, as indicated by thermocouples TC3 and TC4, compared to that of the outer section, represented by TC1. Maximum temperature was reached during this phase, with TC1, located at the concrete cover, recorded a maximum of 675 °C at the 94 min. In contrast, the inner section, measured by thermocouples TC3 and TC4, peaked at 410 °C and 392 °C at the 189 min and 208 min, respectively. The temperature in the outer section, however, began to decrease soon after the furnace temperature entered the cooling phase, whereas the temperature in the inner section of the concrete took significantly longer to drop. According to Chan et al. [46] and Li and Bu [47], the critical temperature for concrete is 400 °C. Up to this temperature, only a small portion of the original strength was lost, between 1 and 15% for NSC. The most severe compressive strength losses occurred primarily between 400 and 800 °C.
Fig. 17
Thermal cracks observed on RCB2 specimen after the fire test
For the reinforcement bars, thermocouple TC2, attached to the tension reinforcement bar 62 mm from the bottom soffit, recorded a maximum temperature of 479 °C at 134 min. Conversely, thermocouple TC5, attached to the compression reinforcement bar 238 mm from the top surface, recorded a lower maximum temperature of 408 °C at 171 min. Both maximum temperatures were reached during the cooling phase and were influenced by thermal inertia of the concrete. The reinforcement bars retained their strength below 500 °C, consistent with the findings of Neves et al. [48], and the tension reinforcement bar remained below the critical temperature of 593 °C specified by ASTM E119-00a [24]. However, the bond between reinforcement bar and concrete may weaken above 300 °C due to differences in thermal expansion coefficients. Below this temperature, the thermal expansion coefficients of steel and concrete are nearly identical. At temperatures exceeding 300 °C, the mismatch in expansion behavior causes stresses at the reinforcement bar-concrete interface, potentially compromising bond integrity [28].
4.2.3 Structural Response (Reinforcement Bar)
During the loading phase of the RCB2 specimen, simulating the service load condition, the applied pre-load induced an initial steel strain of 511 µε for SG1 and 322 µε for SG2, as shown in Fig. 18. As the specimen transitioned into the heating stage, steel strain SG1, installed on the tension reinforcement bar, increased significantly, reaching approximately 2080 µε, while thermocouple TC2 measured 132 °C at the 25-min mark. Similarly, SG2, positioned on the compression reinforcement bar, exhibited a maximum strain of 1105 µε, with thermocouple TC5 recording 74 °C during the same period. At this point, the strain gauges failed due to excessive heat exposure, showing percentage differences of 79% for SG1 and 88% for SG2 relative to the furnace temperature of 626 °C. These observations highlight the distinct roles and positions of SG1 and SG2. SG1 experienced higher strain due to tensile forces and closer proximity to the heat source, resulting in greater thermal expansion. In contrast, SG2 exhibited lower strain and temperature as it was farther from the heat source and subjected to compressive forces. These differences underscore the thermal gradient within the beam and reflect the distinct thermal and mechanical responses of the reinforcement bars under fire exposure.
Upon successfully enduring the fire exposure test without structural failure, the residual loading capacity of RCB2 specimen was determined from the four-point bending test using the same setup as in Phase 1. Observations were made on its performance, particularly the load–deflection relationship, concrete strain, residual deflection after unloading, and failure mode. The load–deflection relationship of RCB2 specimen, shown in Fig. 19, clearly outlines three main phases: (i) initial linear behavior, (ii) reinforcement bar yielding, and (iii) final plastic deformation leading to the strain hardening and failure. Comparison was also made with the load–deflection relationship of RCB1 specimen.
Fig. 19
Load–deflection relationships of RCB1 and RCB2 specimens
RCB2 specimen exhibited initial linear behavior, primarily reflecting the elastic response of the material. However, this linear phase was influenced by temperature-induced degradation of both the steel and concrete material [14, 39, 49], as fire exposure weakened the properties and induced pre-existing tensile cracks. This persisted until the reinforcement bar yielded at 87.1 kN after being subjected to a maximum temperature of 479 °C, consistent with the findings that reinforcement bars retained their strength below 500 °C and typically recovered nearly all of their original yield strength upon the cooling phase [48]. Beyond this threshold, however, the mechanical properties of the reinforcing steel deteriorate significantly, with strength reductions of 10–15% at 600 °C and more than 50% between 700 and 950 °C [50‐52]. These changes are linked to microstructural transformations, including reductions in pearlite and cementite and increases in ferrite between 500 and 700 °C, which compromise the steel’s integrity [53]. At around 727 °C, pearlite transitions into austenite, a softer phase, leading to substantial reductions in both the yield and ultimate strength [54].
Following the observed behavior of reinforcing steel under elevated temperatures, the analysis transitioned to the concrete strain and stiffness characteristics of the RCB2 specimen. The yield point was determined from the concrete strain measurements using DEMEC strain gauges, as the steel strain gauges were damaged during Phase 2, as discussed in Sects. 3.3 and 4.2.3. Similar to Phase 1, the focus is on DM1 and DM6 due to their critical locations, representing the concrete's behavior in tension and compression zones, respectively. The concrete load-strain relationship, shown in Fig. 20, indicates a distinct inflection point at approximately 5513 µε, corresponding to the yield load of 87.1 kN. The ultimate load of the RCB2 specimen was 102.1 kN, reflecting a 26% reduction compared to the RCB1 specimen at 138.1 kN, with a corresponding concrete strain at failure of approximately − 6920 µε, indicating concrete crushing. Additionally, there was a 51% decrease in secant stiffness, from 18.1 kN/mm in the RCB1 specimen to 8.8 kN/mm in the RCB2 specimen. This reduction resulted from the thermal degradation of concrete, which exceeded the critical temperature of 400 °C [46, 47], as illustrated in Fig. 15. Furthermore, the weakening of the bond between the reinforcement bar and concrete, caused by thermal shrinkage and mismatched thermal expansion coefficients above 300 °C, contributed to this declining relationship [55‐57]. These combined effects, including internal stresses generated by thermal gradients, vaporization of water within the concrete's pores, and weakened cohesive forces in calcium-silicate-hydrate (C–S–H) gel layers, resulted in slippage and a significant reduction in the load-bearing capacity [58].
Fig. 20
Load-strain relationships of the concrete of RCB1 and RCB2 specimens
The interaction between fire exposure and the application of pre-loading in Phase 2 test procedures resulted in permanent alteration to the concrete's thermal and mechanical characteristics, significantly changing the behavior of the RCB2 specimen upon cooling to room temperature. These changes were evidenced by the consistent residual deflection, demonstrating permanent damage caused by elevated temperature and mechanical stress. As highlighted in Fig. 21, the observed 5 mm residual deflection in the unloaded state for the RCB2 specimen, along with the substantial increase in maximum deflection from 22.3 mm in the RCB1 specimen to 39.2 mm in the RCB2 specimen, represents a 75.8% increase. The reduction in ductility ratio from 3.60 to 2.65 (a 26.39% decrease) can be attributed to thermal softening of the concrete matrix and a potential reduction in bond strength between the reinforcement bar and concrete. Although direct evidence of pre-loading effect is limited in this study, the observed reduction in ductility ratio may be linked to pre-loading, which intensifies the temperature distribution within the RC beam, as suggested by Ryu et al. [59]. This exacerbates thermal softening and amplifies structural vulnerabilities after fire exposure. These findings, combined with the enduring thermal damage to the concrete, reinforcement bar, and sustained plastic stresses and strains [60], offer a comprehensive view of the structural changes resulting from the combined effects of fire and mechanical stresses.
Fig. 21
Residual deflection of RCB2 specimen after the fire test
Figure 22 shows the general view of the failure section, revealing vertical flexural cracks in the constant moment region and compression failure of the concrete. The presence of thermal cracks on the surface of RCB2 specimen can be attributed to the decomposition of the C–S–H gel as the temperature increases, which reduced the concrete's brittleness and leads to micro-cracks between the aggregate and cement paste, gradually extending to the surface [45, 61]. Additionally, stresses generated by thermal gradients exceeded the tensile strength of the concrete at elevated temperature, resulting in further cracking. In addition, the heating of the concrete generated self-equilibrating stresses, with compressive stress near the heated surface and tensile stress internally. As RCB2 specimen was left to cool down at room temperature, these stresses reverse, and micro-cracks emerged on the surface as the tensile stresses induced by thermal gradients overcome the tensile strength of the concrete [62, 63].
Fig. 22
Concrete crushing, flexural and thermal cracks of RCB2 specimen
4.4 Experimental, Code-Based, and Finite Element Analysis (FEA) Comparison
4.4.1 500 °C Isotherm Method
The 500 °C Isotherm method, as introduced in Appendix B (informative) of EN 1992-1-2 [31], is a practical tool for estimating the residual capacity of fire-exposed RC beams. This method simplifies the assessment by focusing on the effective cross-section below the 500 °C Isotherm, assuming that concrete exposed to temperatures above 500 °C loses its structural integrity, while concrete below this threshold retains its original mechanical properties [64]. Figure 23 illustrates the effective cross-section, highlighting the heat-damaged zone excluded from the calculations. The thickness of this damaged layer, denoted as a500, is determined as the average distance of the 500 °C Isotherm from the cross-section edge in the compression zone. For tensile reinforcement, the reduction in strength is accounted for using the temperature-dependent factor ks(θ), as outlined in EN 1992-1-2 [31], based on the temperature of the reinforcement. These parameters are integrated into standard beam formulas to calculate the residual moment capacity, factoring in the reduced cross-section and degraded reinforcement strength [31].
Fig. 23
Fire exposure on three sides with the tension zone exposed
The 500 °C Isotherm method is widely used due to its simplicity and reliance on pre-established temperature profiles provided in EN 1992-1-2 [31], making it a convenient approach for evaluating the load-bearing capacity of fire-exposed cross-section. However, the method has notable limitations. It is applicable only to standard fire exposure [31] and is based solely on the heating phase of the standard fire curve, without considering the cooling phase. This omission is critical, as material properties, particularly those of reinforcement, can recover during cooling, influencing the residual capacity. Additionally, the method does not account for stress redistribution within the structure, which is essential for maintaining load-bearing capacity as material degradation progresses. Furthermore, the method is limited to evaluating cross-section and does not capture the behavior of the entire structural member under fire exposure, restricting its application in more complex scenario [65].
4.4.2 FEA
Numerical simulations of the fire-exposed RC beam were conducted using FEA software ABAQUS, employing two sub-models: a thermal model for heat transfer analysis and a structural model for evaluating thermal–mechanical analysis and post-fire behavior. The thermal model discretized the beam using eight-node linear brick elements (DC3D8) for concrete and two-node link elements (DC1D2) for reinforcement, with nodal temperature (NT11) as the active degree of freedom. An initial ambient temperature of 20 °C was applied uniformly across all surfaces, while simulated fire condition replicating Phase 2 experimental furnace temperature was imposed on the exposed surface of the RC beam. Heat transfer occurred through convection and radiation, with parameters such as the Stefan-Boltzmann constant, convection coefficients, and emissivity values as summarized in Table 5. The thermal properties of concrete and reinforcement were derived from EN 1992-1-2 [31] and EN 1993-1-2 [66], respectively, to ensure realistic material behavior under fire condition. To ensure consistency and computational accuracy, the mesh size was maintained across both the thermal and structural models.
Table 5
Heat transfer analysis input parameters
No
Parameter
Value
1
Stephan-Boltzmann constant, σ
5.67 × 10 W/m2K4
2
Absolute zero temperature, Tz
− 273.15
3
Initial temperature, T0
20 °C
4
Coefficient of convection heat transfer at the exposed surface, hc
25 W/m2K
5
Coefficient of convection heat transfer at the unexposed surface, hc
9 W/m2K
6
Emissivity of the exposed surface, εm
0.7
The structural model discretized the concrete using eight-node continuum elements (C3D8) and the reinforcement using two-node truss elements (T3D2), ensuring compatibility with the thermal model. Reinforcement was explicitly modeled to align with the concrete nodes, enabling accurate representation of the composite action between the materials. To replicate experimental condition, “Tie Constraint” was used to attach steel plates to the beam, while the plates were modeled as rigid bodies using the “Rigid Body Constraint” to minimize local stress concentration. The interaction between the concrete and reinforcement was managed through the “Embedded Region Constraint,” embedding the reinforcement within the concrete matrix to simulate realistic load transfer. To ensure consistency and computational accuracy, the mesh size was consistently maintained across both the thermal and structural models. Figure 24 illustrates the three-dimensional (3D) finite element model, highlighting the discretization and material assignments of both thermal (used in Phase 2) and structural elements (used across all phases). Similarly, Fig. 25 presents the thermo-mechanical analysis results of the RC beam during fire exposure at 90 min (Phase 2).
Fig. 24
3D finite element model with discretization and material assignments
Additionally, the Concrete Damaged Plasticity (CDP) model in ABAQUS was used to simulate the inelastic behavior of concrete, capturing both compressive failure and tensile cracking [67]. The CDP model translates uniaxial stress–strain data into stress-equivalent plastic strain curves, evaluating the yield surface based on equivalent plastic strains in tension and compression. It employs the Drucker-Prager yield criterion, which considers both normal and shear stresses, with a non-associative flow rule defined by the Drucker-Prager hyperbolic function for the flow potential. Key parameters in the CDP model include a dilation angle (ψ) of 31°, governing plastic volumetric strain during shear, and a flow potential eccentricity (ϵ) of 0.1, which influences the flow potential function in the meridian plane. The ratio of equibiaxial to uniaxial compressive yield stress (σbo/σco) is set to 1.16, indicating the relative strength of concrete under biaxial compression, while the second stress invariant ratio (Kc) on the tensile meridian is set to 0.6667, ensuring consistency with the Drucker-Prager yield surface. To prevent numerical instability, a viscosity parameter (μ) of 0.0001 was employed for viscoplastic regularization, allowing rate-independent analysis.
4.4.3 Model Validation
Figure 26 compares the experimental and FEA simulation load–deflection responses of RCB1 specimen. Both curves display a linear trend in the elastic phase, indicating alignment in initial stiffness between the experimental specimen and the FEA simulation. The experimental first crack load of 28.1 kN was 28.5% lower than the FEA simulation result of 36.1 kN. This overestimation likely stems from idealized material behavior assumptions in the simulation, which neglect inherent material imperfections and micro-crack initiation observed experimentally. Post-cracking, experimental reinforcement yielding occurred at 110.6 kN, which was 14.1% lower than the FEA simulation prediction of 126.2 kN. This difference may arise from variations in material property calibration and boundary condition simplifications. Notably, the ultimate load capacity aligned closely with the experimental of 138.1 kN differing by only 0.7% from the FEA simulation result of 139.1 kN. However, the FEA simulation underestimated the maximum deflection by 19.7%, with experimental deflection reaching 22.3 mm compared to the simulated 17.9 mm. This underestimation is attributed to limitations in modeling material non-linearities and post-cracking behavior.
Fig. 26
Load–deflection prediction from FEA for RCB1 specimen
Figure 27 illustrates the FEA-predicted temperature profiles of concrete and reinforcement bars for RCB2 specimen during fire exposure and cooling, along with experimental measurement points. The simulated measurement points (MP1–MP5) correspond to thermocouple locations shown in Fig. 2b, with MP1, MP3, and MP4 embedded in concrete, and MP2 and MP5 positioned on the tension and compression reinforcement bars, respectively.
Fig. 27
Temperature profile prediction from FEA for concrete and reinforcement bars in RCB2 specimen during fire exposure and cooling, along with corresponding measurement points
The FEA simulation results showed strong agreement with experimental temperature data. Maximum temperature deviations between simulation and experiment were 5.93% lower for MP1, 6.05% higher for MP2, 5.37% higher for MP3, 3.32% higher for MP4, and 5.39% higher for MP5. These minor discrepancies validate the FEA simulation’s ability to replicate heat ingress accurately. The concrete core exhibited slower temperature rise due to its low thermal conductivity and high thermal inertia, delaying heat propagation. In contrast, tension reinforcement at MP2 experienced faster heating, attributable to multidirectional heat exposure from the beam’s bottom and sides, while compression reinforcement at MP5 was heated primarily from the side surfaces. Peak temperatures occurred during cooling, reflecting delayed heat dissipation. For instance, in the experimental test, thermocouples TC3, TC4, and TC5 recorded peak temperatures at 99 min, 118 min, and 81 min, respectively, after the 90-min heating phase. The FEA simulation recorded peak temperatures at 100 min, 114 min, and 60 min for MP3, MP4, and MP5, respectively.
Figure 28 depicts the strain–time prediction for the reinforcement bars during the heating phase of the RCB2 specimen. Both the experimental and FEA results exhibit a similar trend, with strain increasing rapidly as the temperature rises. The initial strains for the RCB2 specimen were recorded at 511 µε for SG1 (tension reinforcement) and 332 µε for SG2 (compression reinforcement). In comparison, the FEA simulation predicted significantly lower initial strains of 55 µε for SG1 and 54 µε for SG2. This discrepancy can be attributed to the pre-loading applied to the RCB2 specimen, which caused tension in the reinforcement bars, resulting in initial elongation before the heating phase.
Fig. 28
Reinforcement bars strain prediction from FEA for RCB2 specimen during fire exposure
As the temperature increased, the strain in both the tension and compression reinforcement bars rose significantly. At 25 min into the heating phase, when the furnace temperature reached 626 °C, the strain in SG1 of the RCB2 specimen reached 2080 µε, while the FEA model predicted 1774 µε, showing a − 14.7% difference. In contrast, the strain in SG2 measured at 1105 µε in the RCB2 specimen, while the FEA model predicted 682 µε, showing a + 38.3% difference. Despite the strain gauges failing after 25 min, the FEA simulation was able to closely match the experimental results up until that point, confirming the accuracy of the simulation in capturing the strain response during the heating phase.
4.4.4 Comparison Results
The 500 °C Isotherm method predicted a residual capacity of 85.2 kN, which was 16.5% lower than the experimental result of RCB2 specimen at 102.1 kN, yielding a capacity ratio of 0.83. These comparisons are summarized in Table 6. The conservative prediction of 500 °C Isotherm arises from the assumption that concrete beyond the 500 °C boundary contributes no residual strength. This simplification excludes contribution from the heat-damaged zone, leading to an underestimation of residual capacity. However, experimental evidence indicates that concrete exposed to high temperature retained some degree of partial residual strength after cooling, despite being significantly degraded [68]. Additionally, the absence of pre-load effect during fire exposure contributes further to the discrepancy. Previous research shows that concrete subjected to simultaneous high temperature and compressive stress (pre-load) loses strength at a slower rate compared with concrete heated without any applied load [69]. This occurs because compressive stress mitigates crack propagation and maintains material engagement, preserving the concrete's load-bearing capacity during fire exposure. By neglecting this interaction, the Isotherm method fails to account for the additional strength retention observed in the experimental test, resulting in further underestimation of the residual capacity.
Table 6
Residual capacity comparison
No
Method
Residual capacity (kN)
Difference (%)
Capacity ratio
1
Experimental
102.1
–
1.00
2
500 °C Isotherm
85.2
16.5
0.83
3
FEA Simulations
103.3
1.2
1.01
The 500 °C Isotherm method also overlooks stress redistribution mechanisms, which are fundamental to structural behavior under load. In the experimental condition, concrete and reinforcement interact dynamically, redistributing stresses across the section as degradation progresses. This redistribution enables the structure to maintain load-bearing capacity even as localized failure occur. However, the method relies solely on predefined reduction factors provided in EN 1992-1-2 [31], which do not capture the dynamic interaction observed experimentally. Together, the exclusion of localized effect, pre-load interaction, and stress redistribution mechanism underscores the limitations of the Isotherm method.
On the other hand, the FEA simulations predicted a residual deflection of 3.59 mm, which was 28.2% lower than the experimental value of 5 mm, as shown in Fig. 29. This discrepancy can be attributed to the model’s assumptions about uniform material degradation and linear plastic deformation behavior. Experimental finding indicates that residual deformation results from irreversible temperature-induced damage in concrete, which does not recover its strength or stiffness after cooling to ambient condition. In the cooling phase, the reinforcing, which remained below 500 °C during fire exposure, fully regained its strength and material properties [48]. This recovery partially counteracts the effects of concrete damage by redistributing stresses; however, it does not fully compensate for the observed permanent deformation. The interplay between the full recovery of reinforcement during cooling and the irreversible damage in concrete explains the differences between the simulated and observed deflection. This residual deformation forms the initial condition for evaluating the post-fire residual strength (Phase 3).
As shown in Fig. 29 and Table 6, the FEA simulation predicted an ultimate load of 103.3 kN, closely matching the experimental result of RCB2 specimen at 102.1 kN, with a percentage difference of 1.2% and a capacity ratio of 1.01. This agreement demonstrates the accuracy of the FEA model in replicating the load-bearing capacity of fire-exposed RC beam. Both experimental and FEA load–deflection curves initially exhibit a linear elastic phase, reflecting the beam’s stiffness under the applied load. However, pre-existing cracks from fire exposure eliminated a distinct first crack load, reducing the beam’s initial stiffness. The curves transition into a non-linear phase, representing plastic deformation leading to failure. While both curves show similar trends, the FEA model predicted a maximum deflection at ultimate of 57.0 mm, overestimating the experimental value of 39.2 mm by 45.4%. This overestimation can be attributed to assumptions about material deformation behavior and the simplified representation of localized damage effects, such as concrete crushing and flexural cracking. These effects are illustrated in Fig. 30, which compares the high-stress zones predicted by FEA with the experimentally observed cracking patterns.
Fig. 30
FEA prediction of high-stress zones, including concrete crushing in the compression zone and flexural cracks in the tension zone, validated against experimental cracking patterns observed during Phase 3
The FEA model’s accuracy in predicting ultimate load capacity can be attributed to multiple factors, including its use of the CDP model, which effectively captures compressive failure, tensile cracking, and temperature-dependent material degradation. The FEA also incorporates stress redistribution mechanism and pre-loading effect, enabling it to reflect the interaction between the applied load and temperature-induced damage during fire exposure and subsequent cooling phases. These features highlight the advantages of FEA over the 500 °C Isotherm method, which does not account for pre-loading effect and relies on predefined reduction factors.
The 500 °C Isotherm method, while practical for baseline evaluations, demonstrates limitations in predicting residual capacity due to its simplified assumptions. Its conservative approach, excluding critical factors like pre-loading, stress redistribution, and the cooling phase, results in underestimation when compared with experimental findings. These limitations underscore the necessity of advanced numerical approaches like FEA, which provide more accurate and detailed assessments of post-fire structural behavior.
5 Conclusions
The following conclusions can be drawn from this study:
1)
The three-phase evaluation methodology proved to be a practical approach for assessing the residual strength and structural response of RC beams after fire exposure;
2)
After 90 min of fire exposure and a 69-h cooling period, RCB2 specimen retained 74% of its original ultimate load capacity, with a reduction from 138.1 to 102.1 kN. This performance reflects the resilience of the reinforcing steel, which remained below its critical temperature threshold of 500 °C (maximum of 479 °C), despite the concrete surpassing its critical temperature of 400 °C, resulting in significant thermal degradation;
3)
The combined effects of fire exposure and pre-loading induced irreversible plastic deformation in the RCB2 specimen, as evidenced by a 5 mm residual deflection in the unloaded state and a 75.8% increase in maximum deflection compared to RCB1;
4)
The cooling phase significantly influenced the thermal behavior of the RCB2 specimen, with temperatures within the cross-section continuing to rise and reaching their maximum during this stage. This highlights the critical role of the cooling phase in post-fire thermal response and structural assessment;
5)
Fire significantly influenced crack initiation and propagation due to the formation of thermal-induced micro-cracks, which exacerbated the vulnerability of concrete to structural weakening;
6)
The effects of pre-loading on post-fire performance, while not directly evidenced, are suggested by the observed 26.39% reduction in ductility ratio, decreasing from 3.60 to 2.65. Pre-loading may have intensified temperature distribution within the beam, contributing to the observed changes in structural behavior and performance;
7)
The 16.5% and 1.2% residual capacity differences predicted by the 500 °C Isotherm method and FEA simulation, respectively, compared with the experimental result, highlight the limitations of the simplified approach such as the Isotherm method. While conservative, the Isotherm method does not account for critical factors such as pre-loading, cooling phase, and stress redistribution. In contrast, FEA simulation, with their advanced modeling capabilities, deliver significantly more accurate assessments of residual capacity and post-fire behavior, reinforcing their importance in reliable structural evaluations.
Acknowledgements
The research presented in this paper is supported by Universiti Teknologi Malaysia and Universiti Kebangsaan Malaysia. Any opinions, findings, and conclusions or recommendations expressed in this paper are those of the authors and do not necessarily reflect the views of the sponsors.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Poon CS, Shui ZH, Lam L (2004) Compressive behavior of fiber reinforced high-performance concrete subjected to elevated temperatures. Cem Concr Res 34(12):2215–2222. https://doi.org/10.1016/j.cemconres.2004.02.011CrossRef
5.
Kodur VR, Raut N (2010) Performance of concrete structures under fire hazard: emerging trends. Indian Concr J 84(2):23–31
Peng GF, Yang WW, Zhao J, Liu YF, Bian SH, Zhao LH (2006) Explosive spalling and residual mechanical properties of fiber-toughened high-performance concrete subjected to high temperatures. Cem Concr Res 36(4):723–727. https://doi.org/10.1016/j.cemconres.2005.12.014CrossRef
Luo X, Sun W, Chan YN (2000) Residual compressive strength and microstructure of high performance concrete after exposure to high temperature. Mater Struct 33:294–298. https://doi.org/10.1007/BF02479699CrossRef
12.
Kumar SV, Suresh N (2021) Study on the residual performance of RC beams exposed to processed temperatures and fire. J Struct Fire Eng 13(3):269–292. https://doi.org/10.1108/JSFE-08-2021-0051CrossRef
13.
Rehman A, Masood A, Akhtar S, Ibrahim SM, Shariq M (2021) Experimental and numerical investigation into flexural bond strength of RC beams exposed to elevated temperature. Constr Build Mater 282:122630. https://doi.org/10.1016/j.conbuildmat.2021.122630CrossRef
Suryanita R, Maizir H, Andriano R, Arditama R (2019) Experimental study and numerical analysis of flexural behaviour of post-fire reinforced concrete beam. IOP Conf Ser: Mater Sci Eng 615(1):012090. https://doi.org/10.1088/1757-899X/615/1/012090CrossRef
17.
Reza W, Subramanian SV (2017) Experimental investigation on RC beam subjected to elevated temperature. Int J Civ Eng Technol (IJCIET) 8(3):220–231
Suhaimi A, Ibrahim IS, Ab Kadir MA (2024) The effects of pre-loading on structural behavior of reinforced concrete beams under fire condition: a review. J Struct Fire Eng. https://doi.org/10.1108/JSFE-11-2023-0041CrossRef
23.
ISO 834-1 (1999) Fire resistance tests - elements of building construction. Part 1: general requirement. International Organization for Standardization, Geneva
24.
ASTM E119-00a (2000) Standard test methods for fire tests of building construction and materials. ASTM International, West Conshohocken
25.
BS 476-20 (1987) Fire tests on building materials and structures. Part 20: methods for determination of the fire resistance of elements of construction (general principles). British Standards Institute, Milton Keynes
Bamonte P, Kalaba N, Maclean J, Bisby L (2020) Numerical investigation of the structural response of eccentrically loaded reinforced concrete columns exposed to non-uniform heating and cooling. In: Proceedings of the 11th international conference on structures in fire (SiF2020), The University of Queensland, 271–281 2020
Salah Dimia M, Guenfoud M, Gernay T, Franssen JM (2011) Collapse of concrete columns during and after the cooling phase of a fire. J Fire Prot Eng 21(4):245–263. https://doi.org/10.1177/104239151142345CrossRef
30.
Gernay T, Franssen JM, Robert F, McNamee R, Felicetti R, Bamonte P, Brunkhorst S, Mohaine S, Zehfuß J (2022) Experimental investigation of structural failure during the cooling phase of a fire: concrete columns. Fire Saf J 134:103691. https://doi.org/10.1016/j.firesaf.2022.103691CrossRef
31.
EN 1992 (2005) Eurocode 2: design of concrete structures - part 1–2: general rules - structural fire design, British Standards Institute
32.
ASTM C78 / C78M-18 (2018) Standard test method for flexural strength of concrete (using simple beam with third-point loading). ASTM International, West Conshohocken
33.
Heiza KM, Taha MS, Soliman MH (2016) Performance of reinforced concrete beam column connections exposed to fire under cyclic loading. In: The international conference on civil and architecture engineering, Military Technical College, 1–23 April 2016. https://doi.org/10.21608/ICCAE.2016.43452
Thongchom C, Lenwari A, Aboutaha RS (2019) Effect of sustained service loading on post-fire flexural response of reinforced concrete t-beams. ACI Struct J. https://doi.org/10.14359/51714477CrossRef
Agrawal A, Kodur VKR (2020) A novel experimental approach for evaluating residual capacity of fire damaged concrete members. Fire Technol 56:715–735. https://doi.org/10.1007/s10694-019-00900-1CrossRef
40.
Rozsypalova I, Schmid P, Danek P (2017) Determining the condition of reinforced and prestressed concrete structures damaged by elevated temperatures. Proc Eng 195:120–126. https://doi.org/10.1016/j.proeng.2017.04.533CrossRef
Song Y, Fu C, Liang S, Li D, Dang L, Sun C, Kong W (2020) Residual shear capacity of reinforced concrete beams after fire exposure. KSCE J Civ Eng 24(11):3330–3341. https://doi.org/10.1007/s12205-020-1758-7CrossRef
43.
Panedpojaman P, Jina P, Limkatanyu S (2016) Moment capacity and fire protection of the welded plate joint for precast members. Arch Civ Mech Eng 16:753–766. https://doi.org/10.1016/j.acme.2016.04.017CrossRef
Liu C, Wang P, Lu X, Miao J, Wang X (2023) Residual flexural behaviour comparison between composite and monolithic beams after fire exposure. J Build Eng 64:105584. https://doi.org/10.1016/j.jobe.2022.105584CrossRef
46.
Chan YN, Luo X, Sun W (2000) Compressive strength and pore structure of high-performance concrete after exposure to high temperature up to 800 C. Cem Concr Res 30(2):247–251. https://doi.org/10.1016/S0008-8846(99)00240-9CrossRef
Fletcher IA, Borg A, Hitchen N, Welch S (2006) Performance of concrete in fire: a review of the state of the art, with a case study of the Windsor tower fire. http://hdl.handle.net/1842/1987
Elghazouli AY, Cashell KA, Izzuddin BA (2009) Experimental evaluation of the mechanical properties of steel reinforcement at elevated temperature. Fire Saf J 44(6):909–919. https://doi.org/10.1016/j.firesaf.2009.05.004CrossRef
Quiel SE, Irwin CH, Naito CJ, Vermaak N (2020) Mechanical characterization of normal and high-strength steel bars in reinforced concrete members under fire. J Struct Eng 146(7):04020110. https://doi.org/10.1061/(ASCE)ST.1943-541X.0002644CrossRef
54.
Nikolaou J, Papadimitriou GD (2004) Microstructures and mechanical properties after heating of reinforcing 500 MPa class weldable steels produced by various processes (tempcore, microalloyed with vanadium and work-hardened). Constr Build Mater 18(4):243–254. https://doi.org/10.1016/j.conbuildmat.2004.01.001CrossRef
55.
Poon CS, Azhar S, Anson M, Wong YL (2001) Comparison of the strength and durability performance of normal-and high-strength pozzolanic concretes at elevated temperatures. Cem Concr Res 31(9):1291–1300. https://doi.org/10.1016/S0008-8846(01)00580-4CrossRef
56.
Elleithy WM, Al-Amoudi OSB, Sharif AM, Maslehuddin M (1998) Effect of thermal variations on bond strength of fusion-bonded epoxy-coated bars. Cem Concr Aggreg 20(1):163–168. https://doi.org/10.1520/CCA10450JCrossRef
Esfahani M, Hoseinzade M, Shakiba M, Arbab F, Yekrangnia M, Pachideh G (2021) Experimental investigation of residual flexural capacity of damaged reinforced concrete beams exposed to elevated temperatures. Eng Struct 240:112388. https://doi.org/10.1016/j.engstruct.2021.112388CrossRef
59.
Ryu E, Shin Y, Kim H (2018) Effect of loading and beam sizes on the structural behaviors of reinforced concrete beams under and after fire. Int J Concr Struct Mater 12(1):54. https://doi.org/10.1186/s40069-018-0280-5CrossRef
60.
Banerji S, Kodur V (2023) Numerical model for tracing the response of ultra-high performance concrete beams exposed to fire. Fire Mater 47(3):322–340. https://doi.org/10.1002/fam.3099CrossRef
Jaszczak B, Kuczma M, Szymkuć W (2021) Comparison of the load-bearing capacity of reinforced concrete columns under fire conditions using the method A, zone method and isotherm 500 method. Fire Saf J 124:103396. https://doi.org/10.1016/j.firesaf.2021.103396CrossRef
Kowalski R, Król P (2010) Experimental examination of residual load bearing capacity of RC beams heated up to high temperature. In: Kodur VKR, Fransen JM (eds) Sixth international conference structures in fire, Michigan state university, East Lansing, Michigan, USA, DEStech Publications Inc, pp 254–261
69.
Głowacki M, Kowalski R (2020) An experimental approach to the estimation of stiffness changes in RC elements exposed to bending and high temperature. Eng Struct 217:110720. https://doi.org/10.1016/j.engstruct.2020.110720CrossRef
