An Experimental Study on the Thermal Characteristics of Firestop Systems Depending on Physical Properties of Metallic Pipe Materials

Abstract

We quantitatively analyzed the effects of physical properties of metallic pipe materials on the thermal performance of firestop systems. Fire-resistance tests under realistic fire conditions were conducted for 120 min using five types of metallic pipes—carbon steel, stainless steel, cast iron, copper, and aluminum—under identical firestop material conditions. The temperature distribution at key locations within the slab and average rate of temperature increase over specific time intervals were compared. Materials with higher thermal conductivity and lower wall thickness exhibited faster thermal response characteristics. High-temperature behavior was most pronounced at the pipe surface, where copper and aluminum pipes reached temperatures approximately equal to 200 °C and 190 °C, respectively. During the initial 30 min, the average rates of temperature increase were the highest for aluminum (2.9 °C/min), followed by copper (2.2 °C/min), although the rate of heat transfer gradually decreased subsequently. A correlation analysis between the composite index of thermal conductivity and cross-sectional area, revealed a strong correlation at the pipe’s surface, with a coefficient of determination greater than 0.85. The thermal properties and cross-sectional characteristics of metallic pipes can directly affect the thermal behaviors of firestop systems. The results may serve as a basis for material-informed structural design and performance evaluation criteria.

1. Introduction

Firestop systems are essential structural elements that are used to maintain the fire compartmentation performance of walls and slabs by preventing the spread of flames and smoke through penetrations created by mechanical, electrical, and plumbing installations in buildings [1,2]. Service penetrations represent points where the continuity of a structure is compromised during a fire, making them potential pathways for the propagation of flames, smoke, and hot gases. Therefore, ensuring effective fire-resistive construction at these points is critical. In response, international standards, such as those of the American Society for Testing and Materials (ASTM) E814 [3], Underwriters Laboratories 1479 [4], and European Norms 1366-3 [5] and the Korean standard (KS) F International Standard Compliance (ISO) 10295-1 [6], have been established. Recently, regulatory requirements have been further strengthened through the revised Fire Safety Regulations on Means of Egress and Fire Compartments in Buildings [7], Ministry of the Interior and Safety’s fire prevention guidelines [8], and the technical guidance provided by Promat [9].

Previously conducted research primarily focused on evaluating the fire-resistance performance of firestop materials in terms of their intumescent characteristics, installation methods, and ratio of the annular space to the penetrating item. For example, Sędłak et al. conducted comparative experimental studies on the performance of intumescent firestop materials applied to metallic and nonmetallic pipes [10], whereas Ye et al. proposed a methodology to assess quantitatively the thermophysical properties of firestop sealants [11]. Choi et al. [12] analyzed the effect of applied material volume under identical opening conditions for metallic and nonmetallic pipes, confirming that the cross-sectional ratio of the penetration plays an important role in fire performance. However, most of these studies focused on the quantity or geometry of the applied materials, and the direct quantitative impact of the inherent physical properties of metallic materials on thermal response remains insufficiently explored.

In practice, various metallic pipes—such as carbon steel (CS), copper (Cu), and aluminum (Al) alloys—are commonly used in Korean constructions, and their physical differences (e.g., thermal conductivity, wall thickness, melting point) may induce distinct thermal distributions and heat accumulation behaviors even under identical firestop conditions. According to domestic KSs, such as KS D 3562 [13], KS D 3507 [14], and KS D 6741 [15], the specifications and dimensions of metallic pipes vary depending on the material, which can affect both the heat transfer pathways and structural thermal response. Materials with higher thermal conductivity may experience more rapid temperature increases, and their thermal expansion or melting behavior can further affect the temperature distribution of the firestop system [16]. However, few experimental studies have systematically compared how these material-specific properties impact structural performance during actual fire-resistance tests.

Therefore, in this study, we conducted full-scale fire resistance tests using different metallic pipe materials under consistent firestop conditions and quantitatively analyzed the effects of key physical properties—such as thermal conductivity and cross-sectional area—on thermal response. The goals were to compare and evaluate fire-performance differences based on material-specific thermal characteristics, and to provide foundational data for optimizing test conditions and enhancing regulatory frameworks.

2. Experimental Plan and Methodology

2.1. Experimental Plan

To verify quantitatively the performance differences in firestop systems based on various metallic pipe materials, fire-resistance tests were conducted under uniform conditions using metallic pipes with an outer diameter of 165 mm installed in openings with a diameter of 237 mm. A total of five types of metallic pipes—CS, Cu, cast iron, stainless steel (SS), and Al—were tested. Their key thermal and physical properties, including thickness, thermal conductivity, and melting point, are summarized in

Table 1. Physical properties of the metallic pipes at room temperature.

Properties	Carbon Steel (CS) Pipe	Copper (Cu) Pipe	Cast Iron (CI) Pipe	Stainless Steel (SS) Pipe	Aluminum (Al) Pipe
Thickness (mm)	3/4.85/7.5	3.56	7.5	3	3.5
Melting point (°C)	1425–1540	1085	1150–1200	1375–1450	660
Thermal conductivity (W/m·K)	48–54	390–400	45–55	15–25	220–235

(room-temperature values in the range of 20–25 °C) [17,18,19,20]. Each pipe was used according to standard thickness specifications, and the differences in their thermal conductivity and melting behavior were noted. For CS pipes, three thicknesses (3 mm, 4.85 mm, and 7.5 mm) were tested to evaluate the impact of wall thickness on fire-resistance performance.

All test specimens were designed based on the configuration shown in Figure 1, incorporating consistent components such as top insulation, bottom insulation, firestop material, and ceramic finishing. The specimens were vertically installed through ALC slabs (150 mm thick). The top insulation comprised 600 mm of chemically crosslinked polyethylene foam (thickness of 40 mm), and the bottom insulation comprised a ceramic fiber blanket with a thickness of 50 mm and density of 96 kg/m³.

The firestop material used was a domestic intumescent product with a density of 1.2 g/cm³ (EZONE Co., Ltd., Hwaseong-si, Republic of Korea) that expands by >30 times its original volume when exposed to fire. The installation consisted of two elements: one piece with dimensions of at least 6.0 mm (thickness), 670 mm (length), 75 mm (width), and 362 g in weight, and another with dimensions of 6.0 mm (thickness), 640 mm (length), 75 mm (width), and 346 g in weight. Both were fixed in place using galvanized steel fasteners (0.5 mm thick). A ceramic finishing material (ceramic fiber blanket, 50 mm thick, 50 mm wide, 96 kg/m³ density) was applied to the top of the firestop with a 28% compression ratio. A silicone-based sealant (F-12.5E grade) was used as a surface finish with a thickness of 3 mm and a 10 mm overlap.

The specimen configuration was designed to reflect real-world installation conditions, and both the geometric specifications and installation positions were kept consistent across all tests.

2.2. Experimental Methods

The experiments were conducted in accordance with the KS F ISO 10295-1:2021 standard [6], with the heating conditions set based on the ISO 834 standard time–temperature curve [21]. Each test was performed individually in a horizontal furnace measuring 3 × 4 m, with heating durations of up to 120 min. Figure 2 presents a comparison between the standard time–temperature curve and the measured furnace temperatures during testing.

Thermocouple mounting, averaging, and logging: Seven bead-type K-type thermocouples (OMEGA Engineering, Inc., Norwalk, CT, USA) (wire Ø 0.5 mm, SLE) were installed per specimen at the locations shown in Figure 1 (boundary of the penetration opening and supporting structure; 25 mm from the bottom and top insulations; 25 mm from the pipe surface). Pipe-surface sensors were fixed at the mid-circumference at 15 ± 2 mm (mean ± SD) from the annular gap. Beads were spot-welded, sealed with a small amount of high-temperature ceramic adhesive, and double-fixed with a 5 mm stainless band and ceramic fiber tape; the band joint was placed opposite the annular gap, and ~2–3 mm ceramic paste filled microvoids. Leads were routed toward the slab edge to avoid interference during the intumescent expansion process. To characterize within-surface dispersion, multiple thermocouples were co-located on each surface; plots show the time-wise means, bars denote the ±1 SD across co-located thermocouples (Type A), and time-series plots omit bands for clarity with dispersion described in the text. Temperatures were recorded every 1 min using a Yokogawa MV2000 data logger (Yokogawa Electric Corporation, Tokyo, Japan) (16-bit, cold-junction compensation, monitoring capacity up to 40 channels) [22]. Cold-junction compensation was enabled, and all channels were zero-checked at ambient before each run; any residual offset was included within the Type B uncertainty. Measurement accuracy values were as follows: combined baseline ≈±1.3 °C; contact/attachment ±2.0–2.5 °C for most periods, and ≤±3.0 °C during 10–30 min.

Following the Guide to the Expression of Uncertainty in Measurement (GUM), we treated the ±1 SD across co-located thermocouples on the same surface as Type A, and the instrument/cold-junction specification together with contact/attachment effects as Type B (as described above). For window-averaged heating rates computed from 1 min differences, the associated uncertainty equals the temperature uncertainty over the averaging window divided by the window length (≤0.3 °C/min for 10 min windows; ≤0.1 °C/min for 30 min windows). Under ISO-834 [21] exposure, because the furnace heat input is prescribed, the corresponding thermal-resistance uncertainty is governed by the temperature difference (ΔT).

The analysis focused on relative temperature comparisons according to the pipe material rather than on absolute insulation performance. Key metrics, such as temperature rise behavior, heat transfer trends, and heat accumulation time, were quantitatively compared according to the material type. This experimental study aimed to identify the influence of firestop systems on different metallic pipe materials and to determine structural impact factors based on the collected thermal response data.

3. Experimental Results

3.1. Analysis of Temperature Rise Behavior According to Metallic Pipe Material

Figure 3 presents the temperature–time curves measured over 120 min at key locations—pipe surface, top and bottom insulation layers, and the interface between the opening and slab—for each metallic pipe specimen. The results demonstrate that the thermal response of the firestop system varies considerably depending on the pipe material, even under identical testing conditions.

Cu pipes, which have the highest thermal conductivity among the tested materials, exhibited rapid heat transfer, exceeding 100 °C at the pipe surface within the first 30 min and reaching approximately 200 °C by the 120 min mark. Notably, the pipe temperature exceeded that of the top insulation layer, suggesting that the high conductivity of Cu facilitated direct heat transfer to the pipe before the insulating material could attenuate it. This indicates that the Cu pipe acted as an efficient thermal conduit at the interface with the heat-exposed zone.

Al pipes, which also possess high thermal conductivity (205–235 W/m·K), exhibited a slower initial temperature rise compared with that of Cu pipes. At 120 min, the pipe surface temperature reached approximately 180 °C. This relatively reduced heat transfer efficiency may be attributed to the low-melting point of Al (660 °C); parts of the pipe may have undergone early thermal deformation or physical changes during heating, affecting the consistency of heat conduction.

Although the nominal thermal conductivity of cast iron (CI) pipes is comparable to that of CS (40–60 W/m·K), the CI specimens displayed the lowest overall temperature distribution among all the test specimens. At 120 min, the pipe temperature remained near 60 °C, and the bottom insulation temperature did not exceed 120 °C. These findings suggest a buffering effect potentially due to the greater wall thickness of the CI pipe (7.5 mm) compared with that of the CS pipe (4.85 mm), which may have contributed to the increased thermal mass. Differences in material composition and microstructure may also have resulted in a discrepancy between the nominal and actual thermal behavior.

SS pipes, with a relatively low-thermal conductivity (15–25 W/m·K), exhibited a gradual temperature rise, with pipe and bottom insulation temperatures reaching approximately 120 °C and ~160 °C, respectively, after 120 min. Compared with CI and CS, SS demonstrated higher thermal response, likely due to its thinner wall thickness (3.0 mm), which allowed heat to transfer more readily.

CS pipes experienced a temperature rise to approximately 100 °C at the pipe’s surface and 140 °C at the bottom insulation layer. Despite its similar thermal conductivity to CI, the thinner wall likely facilitated faster heat transfer within the same time frame.

Notably, in most cases, the highest temperature was recorded at the bottom insulation layer, which was closest to the fire source. However, for highly conductive materials such as Cu and Al, the pipe temperature exceeded that of the adjacent insulation, indicating that these pipes served as the primary heat-transfer pathways and reached high temperatures more rapidly compared with the insulating materials. This pattern is also consistent with previous heat-transfer modeling studies [23], which suggest that highly conductive metals positioned near the heat source can trigger an abrupt temperature rise during the early phase of exposure.

Quantitatively, median values were 0.495 °C at the pipe’s surface (90th percentile ≤ 2.192 °C), 3.606 °C at the top insulation layer, and 7.849 °C at the bottom insulation layer. At the opening–slab interface (single channel) the SDs at the same locations were 1.34 °C (median) and 7.9 °C (90 min). Thus, the pipe curves were effectively conduction-dominated with sub-degree scatter, whereas the insulation layers—especially the hot-face bottom—exhibited larger within-surface variability due to sensitivity to contact/adhesion, thickness tolerance, and non-uniform radiative/convective flux as the intumescent layer evolved. Early-time separation for Cu (and to a lesser extent Al) reflects a faster conduction front that outpaces insulation inertia; mid- to late-phase divergence in insulation is amplified by dehydration, binder burn-off, and microdebonding. Together with the large intermaterial differences at 120 min (e.g., pipe Cu ≈ 200 °C vs. CI ≈ 60 °C), this indicates that material, wall thickness, and the resulting k/A pathway—not measurement noise—govern the response.

These findings demonstrate that the thermal response of firestop systems is not solely determined by the properties of the insulation materials. Instead, the material type, wall thickness, and actual thermal behavior of the metallic pipes play an interacting and important role in influencing the overall heat-transfer dynamics.

3.2. Analysis of Temperature Distribution Characteristics by Thermocouple Location and Pipe Material

Figure 4 presents the temperature distribution measured by thermocouples installed at four key locations—(a) bottom insulation layer, (b) top insulation layer, (c) pipe surface, and (d) opening–slab interface—for each metallic pipe material under identical test conditions. These results reveal distinct thermal behaviors depending on the thermal conductivity, wall thickness, heat capacity, and spatial configuration of the material within the firestop structure.

The bottom insulation layer (a), directly exposed to the heated surface, exhibited the fastest temperature rise across all specimens within the first 30 min. Notably, the Al pipe reached ~150 °C at 30 min and ~180 °C at 120 min. The Cu pipe also approached 170 °C by the 120 min mark. These results reflect the direct heat transfer effect due to the high thermal conductivity of these materials, suggesting that the pipe itself functioned as a primary heat-transfer path. This pattern is also supported by previous modeling research findings [23] that reported that highly conductive pipes in close contact with firestop sealants may induce localized thermal concentration during early fire exposure. In contrast, the CI pipe exhibited the lowest temperatures, reaching only ~60 °C at 30 min and ~110 °C at 120 min. This is attributed to the combined effect of a thicker wall and higher heat capacity, which suppressed rapid heat transfer.

At the top insulation layer (b), located farther from the fire-exposed surface, the overall temperature curves are more gradual. However, the thinner SS pipe (SS, 3.0 mm) and CS pipe (CS, 4.85 mm) exhibited rapid heat transfer outcomes, reaching approximately 110–115 °C by the end of the 120 min test. Cu and Al pipes, while they were initially slower to heat and displayed a steady increase after the 60 min mark, they reached temperatures of ~120 °C. Notably, the Cu pipe exhibited a stagnation period between 30 and 60 min, forming an atypical curve that may indicate thermal saturation or local deformation within the insulation, disrupting uniform heat conduction.

At the pipe’s surface (c), the differences in thermal performance among materials were most pronounced. The Cu pipe reached ~200 °C and the Al pipe reached ~195 °C at 120 min. The Al pipe exhibited a rapid temperature spike at temperatures > 160 °C within the first 30 min, followed by an abrupt decline in the rate of increase. This behavior suggests possible deformation or alteration in the heat conduction path near the melting point of Al (660 °C). In contrast, the Cu pipe displayed a more gradual and continuous rise, ultimately reaching the highest temperature at the end of the test. The CI pipe remained below 80 °C throughout the test, indicating that its microstructure, thickness, and thermal mass hindered heat conduction. These results are consistent with those of prior studies showing that highly conductive metals act as dominant heat transfer paths. Kodur and Dwaikat [11] emphasized the role of internal conduction in shaping global thermal gradients in steel structures, while Zhu et al. [24] experimentally validated rapid surface heating of steel pipes under fire exposure conditions.

The opening–slab interface (d), where heat is transferred indirectly through the surrounding structures rather than directly through the pipe, recorded the lowest temperature increases and showed relatively minor differences among materials. However, the Cu pipe still reached the highest temperature at this location (approximately 140 °C at 120 min), likely due to heat accumulation transmitted through its high-conductivity structure. Most other materials remained within the 100–120 °C range.

Within a given material, the curves exhibit hot-face (bottom)—top insulation—pipe surface—opening–slab interface sequences with a lag in the order of tens of minutes. For high-conductivity metals (Cu, Al), an early-time crossover appears in which the pipe is briefly hotter than the adjacent insulation, indicating a direct conductive shortcut; during the period of 30–60 min, the insulation often exhibits peaks or mild plateaus, consistent with sensitivity to hot-face flux nonuniformity, evolving contact, and moisture/binder transitions. By contrast, CI exhibits the strongest spatial attenuation and the longest lag owing to its thicker wall and larger heat capacity.

Figure 5 quantitatively compares the average temperature rise rates (°C/min) for each material at each thermocouple location across four time intervals: 0–30, 30–60, 60–90, and 90–120 min.

The window-averaged rates exhibit three salient features. First, at the pipe’s surface, Al and Cu dominate early heating and then decay, consistent with a conduction-dominated pathway and very small error bars (typically ≤ ~0.05 °C/min). Second, in the insulation layers, rates typically peak during the 30–60 min period, and occasional plateaus or slight reversals emerge around 60–90 min (notably at the top Cu/Al layers), reflecting time-varying contact/adhesion, thickness tolerance, and non-uniform hot-face flux. Third, the dispersion structure itself differentiates locations: tiny bands at the pipe versus markedly larger mid/late-window bands at the top layer (up to ~0.94 °C/min). The rate hierarchy at the pipe surface largely tracks k/A, directly expressing material and wall-thickness effects. In contrast, the larger SDs in the insulation indicate higher sensitivity to field variables (contact pressure, surface roughness, moisture/binder transitions). CI exhibits persistently low rates, consistent with a buffering effect from its thicker wall and larger thermal mass. Hence, Figure 5 provides a complementary view: the pipe surface acts as a robust material index, whereas the insulation layers probe contact and installation sensitivity.

For the bottom insulation layer (a), the highest rise rates occurred in the 0–30 min interval. The approximate recorded rates for Cu, Al, SS, CS, and CI were 1.1, 0.9, 1.0, 0.8, and 0.6 °C/min, respectively. After this initial phase, all materials exhibited abrupt decreases in rise rates, indicating thermal saturation and heat accumulation effects within the insulation layers.

At the top insulation layers, the rise rates were typically lower. The SS pipe showed the highest initial rate at ~1.2 °C/min, followed by CS at 1.1 °C/min. Al and Cu exhibited much lower initial rise rates approximately equal 0.3 °C/min. Notably, the Cu pipe recorded a negative rise rate (−0.1 °C/min) during the 30–60 min interval, suggesting localized melting or thermal distortion within the insulation that disrupted consistent heat transfer.

The pipe surface was most sensitive to differences in thermal behavior. The Al pipe recorded the highest rise rate of ~2.9 °C/min during the initial 30 min, followed by gradual decreases to 1.5, 0.8, and 0.4 °C/min in subsequent intervals. In contrast, the Cu pipe exhibited a delayed peak, with the highest rise rate (2.5 °C/min) during the 30–60 min interval, and sustained elevated rates of 1.8 and 1.2 °C/min in subsequent intervals. This suggests that Al exhibits a very rapid early thermal response, followed by reduced conductivity due to deformation, whereas Cu maintains consistent heat transfer performance throughout the test.

Regarding the 0–30 min window, we cautiously note that the differences may be associated with time-dependent interface changes under heating conditions (e.g., pipe thermal dilation/ovalization, subtle breathing of the annular gap, early changes in pressing state), which could modulate local heat transfer and contact. At this stage we simply note this possibility; subsequent studies will broaden the interpretation from multiple perspectives [25,26].

At the opening–slab interface (d), all materials exhibited low temperature rise rates, remaining below 1.0 °C/min. Specifically, the recorded rise rates in the initial phase for Cu, Al, SS, CS, and CI were 0.6, 0.4, 0.5, 0.5, and 0.2 °C/min, respectively. These results reflect the location’s indirect exposure to heat and the gradual diffusion of thermal energy through the surrounding structure.

3.3. Influence of Thermal Conductivity and Cross-Sectional Area on Temperature Response

This section presents a quantitative analysis of the correlation between the thermal properties of metallic pipes and measured temperatures using Figure 6, Figure 7 and Figure 8. Figure 6 analyzes the thermal conductivity (k, W/m·K) of each pipe material; Figure 7 focuses on the cross-sectional area (A, cm²) of CS pipes with identical outer diameters but varying wall thicknesses, and Figure 8 introduces a composite index (k/A, W/m·K·cm^{2 −1}), combining thermal conductivity and cross-sectional area. The analysis evaluates correlations with temperatures at the bottom insulation (a) and pipe surface (b) positions, using the coefficient of determination (R²) as a metric. For interpretation purposes, the early-time heating of the pipe wall follows a simple conduction–capacity balance, yielding dT/dt ∝ k/(ρ·c·A), that is, the ratio of the conduction potential k and wall thermal mass product ρ·c·A. Because ρ and c vary modestly across the tested metals, while A changes appreciably (via wall thickness), the compact, geometry-aware proxy k/A is informative. For completeness, we also examined the product k·ρ·c (thermal inertia) and found that it was comparable only during the period of 0–30 min; this product did not outperform k/A during the period of 30–60 min [27,28].

In Figure 6 (thermal conductivity, k), a positive correlation was observed at the pipe surface (b), particularly during the initial 0–30 min interval, where the R² value approached 0.7. This suggests that materials with higher thermal conductivity facilitated faster initial heat transfer, resulting in more rapid temperature increases. However, the correlation decreased over time because of the effects of heat capacity and external heat dissipation. Within the 20–200 °C window relevant to this study, the thermal conductivity of metals changes only modestly and in a material-dependent manner—slight decreases for Cu and Al, and a slight increase for austenitic stainless steels, as indicated by previously published studies. Accordingly, the k values (at room temperature) used herein preserve the observed relative ranking among materials, although late-time magnitudes may shift slightly [29,30,31,32].

In Figure 7 (cross-sectional area, A), CS pipes with identical outer diameters but differing thicknesses exhibited a strong negative correlation between cross-sectional area and temperature rise. During the 30–60 min interval, the R² values reached 0.9888, indicating that thicker pipes with larger cross-sectional areas suppressed mid-stage heat transfer considerably. This is attributable to increased thermal mass and structural thermal resistance.

Figure 8 presents a composite index (k/A), which normalizes thermal conductivity by cross-sectional area to express the material’s effective thermal transmissivity within a structural context. This index exhibited the strongest overall correlation at the pipe’s surface where the R² value exceeded 0.85, particularly during the 30–60 min interval. The results demonstrate that the composite index offers a more accurate explanation of thermal behavior than thermal conductivity or cross-sectional area alone.

This interpretation is supported by previous studies. Kodur and Dwaikat [23] modeled heat transfer within complex structural cross-sections, emphasizing that thermal gradients in fire-exposed assemblies cannot be fully explained by thermal conductivity alone. Instead, they highlighted the importance of combined parameters reflecting material configuration. Similarly, Abu-Eishah [29] demonstrated that thermal conductivity varies nonlinearly with temperature, particularly in high-conductivity metals, underscoring the need for composite indicators under elevated thermal conditions. These findings support the analytical validity of the composite index (k/A), which effectively captures both the heat transfer potential and moderating effects of thermal mass in firestop systems.

At the bottom insulation layer, correlations were typically lower. However, during the initial 0–30 min interval, some variables exhibited moderate R² values approximately equal to 0.6 and 0.8, suggesting partial consistency. This behavior is likely due to the direct exposure of the bottom insulation layer to the heat source, where the initial thermal response was strongly influenced. In the subsequent intervals (beyond 60 min), the R² values dropped below 0.4, possibly owing to indirect heat-transfer effects, the effects of intumescent material expansion, variations in contact resistance, and minor differences in thermocouple placement—all of which may have introduced experimental variability and reduced the correlation consistency over time.

In contrast, the pipe surface (b) consistently exhibited higher R² values across all time intervals. The consistent material properties, stable geometry, and reliable thermocouple attachment on the metal surface contributed to robust data reliability. These results indicate the predictive validity of thermal conductivity and cross-sectional area as key factors governing temperature behavior.

In conclusion, while the temperature response at the pipe’s surface is well explained by individual variables, such as thermal conductivity and cross-sectional area, it is even more effectively explained by their composite index (k/A). The influence of these parameters varied as a function of the time interval, with the 30–60 min range demonstrating the clearest correlations. This interval can therefore be considered as a key analytical window for evaluating thermal conduction characteristics in firestop systems.

4. Conclusions

This study quantitatively analyzed the effects of pipe material, thermal conductivity, and cross-sectional characteristics on the thermal performance of firestop systems at differing metallic penetrations. Under consistent firestop material and testing conditions, the temperature responses at key locations within the slab were experimentally compared based on pipe material and geometry. The structural influence factors were comprehensively evaluated using time–temperature curves, average temperature rise rates, and coefficient of determination (R²) analyses based on thermal conductivity.

First, the time–temperature curve analysis (Figure 4) showed that Cu and Al pipes, which had high thermal conductivities, exhibited the most rapid temperature rise during the early stages of the test. In particular, Cu consistently maintained temperatures near 200 °C throughout the 120 min test. This demonstrates that thermal conductivity plays a major role in shaping the internal heat transfer pathways within firestop systems.

Second, the analysis of the average temperature rise rates by time interval (Figure 5) revealed distinct thermal response speed outcomes according to material. The Al pipe exhibited the highest rise rate of 2.9 °C/min during the 0–30 min interval, followed by a gradual decline. In contrast, the Cu pipe displayed a nonlinear trend with increasing rise rates over time. These patterns are interpreted as the combined result of the inherent material properties and structural response.

Third, the R² analyses (Figure 6, Figure 7 and Figure 8) confirmed that both thermal conductivity and cross-sectional area considerably influenced the temperature at the pipe surface (b), with the maximum coefficient reaching 0.9988 for the cross-sectional area, and 0.8546 for the composite index (k/A), confirming their high explanatory power. A positive correlation was observed, with higher thermal conductivity corresponding to greater temperature increases. The CS pipe specimens with smaller cross-sectional areas exhibited greater thermal responses, indicating that for materials with equivalent conductivity, narrower cross-sections led to increased heat flux per unit area.

Based on a comparative analysis of thermal behavior under practical firestop installation conditions using actual metallic pipe products, we provide foundational data for performance-based fire protection design and regulatory refinement. This study intentionally focused on metallic penetrants and interpreted the data based on a negligible-deformation assumption with a stable annulus; we did not apply mechanical deformation, imposed displacement/vibration, internal pressure, or movement to the pipe–sleeve assemblies, and we did not measure smoke/air leakage (L-rating). Consequently, thermo-mechanically induced effects—such as axial growth, ovalization, restraint-related slip, gasket relaxation, and cracking or detachment of firestopping components (“secondary penetration”)—were outside the scope of this work and were not justified by the reported temperature histories. The reported correlations and the k/A-based interpretation therefore apply to conduction-dominated, low-deformation regimes comparable to the tested configuration. Because commercial products were used, the outer diameter and wall thickness could not be fully controlled; future work will include additional tests under standardized-thickness conditions, together with coupled thermo-mechanical protocols (controlled thermal expansion, internal pressure, and imposed displacement/vibration) and complementary modeling. Nonmetallic penetrants—subject to high-temperature material loss and time-dependent annulus evolution—will be examined in a separate, dedicated study using a distinct experimental and analytical framework. Temperature-dependent property datasets (approximately ≤200 °C) will be explicitly incorporated to refine and validate these correlations. In addition to the standardized-thickness and coupled thermo-mechanical tests outlined above, future work will include scaling studies across pipe diameters/wall-thicknesses and slab depths, and field-oriented validation through inter-laboratory replication and instrumented site trials to compare ISO-834 [21] furnace histories with in situ conditions.

For metallic penetrants, we interpret the surface temperature response using thermal conductivity (k) together with cross-sectional area (A) as organizing variables; a concise screening relation based on these variables will be developed and validated in subsequent work before any normative use in building-safety codes.

While further experimental validation is warranted, the findings of this study suggest that thermal performance can be partially inferred based on the intrinsic properties of metallic pipes. This insight may help industry stakeholders to define more material-specific test protocols and refine design conditions aligned with practical applications. In the long term, predictive modeling based on material indices may enable preliminary assessment of fire resistance performance under complex or nonstandard conditions. These implications support a shift toward performance-based, material-informed, fire safety strategies, and contribute to the modernization of related regulations and certification systems.

Ultimately, by clarifying how fundamental material properties influence the thermal behavior of firestop systems, this study offers practical insights that can bridge the gap between laboratory testing and real-world application—providing a meaningful step toward safer, more predictable building fire protection strategies.