1. Introduction
Since Reinforced Concrete (RC) is well-known thermal resistant materials, studies for the thermal and structural behaviors of fire-damaged concrete structures have been limitedly conducted. Moreover, most of the existing studies have performed fire tests on concrete columns by heating all four sides consistently. However, concrete column is commonly damaged partially and inconstantly due to fire condition and floor plan when a fire occurs, which may reduce the residual strength of fire-damaged columns additionally.
Recently, the performances of fire-damaged concrete columns have been actively studied by many researchers (Kodur et al., 2004; Han et al., 2009; Yaqub and Bailey, 2011; Shah and Sharma, 2017). Jau and Huang (2008) investigated the behavior of corner columns under axial loading, biaxial bending and asymmetric fire loading. Chen et al. (2009) reported an experimental research into the effect of fire exposure time on the post-fire behavior of RC columns. After fire tests, the specimens were tested in axial load combined with uniaxial or biaxial bending. The test results indicated that the residual load-bearing capacity decreased with increase in fire exposure time. In addition, the reduction rate of stiffness was higher than that of ultimate load. In the research (Choe et al., 2015), material properties of Ultra-High-Strength Concrete (UHSC) and residual load of RC column at high temperature was evaluated. RC columns were fabricated with 200 MPa concrete and were heated by ISO-834 standard fire curve for 3 hours. From the results, the rate of compressive strength degradation of UHSC subjected to elevated temperature was higher than that of NSC. Li et al. (2017) tested to understand the cyclic behavior of damaged RC columns repaired with a high-performance fiber-reinforced cementitious composite with high-volume fly ash and to evaluate the effects of axial load and repair height. However, limited research has been conducted on the thermal and structural behaviors of the RC columns depending on heated area.
Therefore, this study aimed to investigate the effect of various heated areas on the thermal and structural behaviors of RC columns under and after the fire. Toward the goal, fire tests and loading tests were conducted to measure temperatures distributions and residual strengths of RC columns exposed to fire on different number of faces.
2. Experimental and Analytical Approaches
2.1 Test Specimens and Parameters
For the tests, five columns were fabricated with the dimensions of 350 × 350 × 1,500 mm (width × depth × height), as shown in Fig. 1. Dimensions were determined with consideration of heating chamber sizes and to exclude the effect of buckling. For the longitudinal and transverse reinforcement, D22 and D10 (diameter = 22.2 and 9.53 mm, respectively) were used, respectively. Stirrups were arranged with intervals of 300 mm at center and 50 mm at the edge of column. The first tie was placed at a height of 50 mm from the column base.
The sheath temperatures inside the RC columns were measured by five thermocouples from measuring points of C1, C3, and C5 along the concrete cover, C4 at the middle of the cross section, and C2 in the mid of C1 and C4, as shown in Fig. 1. The reinforcing bars and thermocouples were placed inside the formworks, before fresh concrete was poured into the mold. All the specimens were cured for three months at ambient temperature to minimize moisture effect during the fire tests.
The variable of the experiment was heated area, which was varied from none to four side surfaces as listed in Table 1. The variable was determined to consider that columns can be partially exposed to fire depending on fire conditions and floor plans. In order to apply heat on the designated surface of the columns, insulations are attached on the surfaces to be unheated during the fire tests.
)
2.2 Materials
The the mix proportions of the tested concrete are designed as listed in Table 2 to satisfy design strength of concrete as 60 MPa. For the material strength tests, the cylindrical specimens were fabricated with 100 mm of diameter and 200 mm of height according to KS F 2403 and subjected to loading after 28 days of curing. The averaged compressive strength of concrete was obtained as 57.8 MPa. The average yield strength and elastic modulus of reinforcing bars were also obtained from the material strength tests as listed in Table 3.
2.3 Experimental Approach
2.3.1 Fire Test
The specimens were tested in the heating chamber as shown in Fig. 2. Top and bottom parts of the specimens were covered by insulations to protect those parts from being damaged during the heating so that the load and support can be applied easily during the loading tests. Specimens were heated for 2 hours according to ISO-834 standard time-temperature curve without axial loading. Nine thermocouples were placed inside the heating chamber to measure and control the temperature of the heating chamber to follow the ISO-834 standard time-temperature curve (ISO 1999) during the tests.
2.3.2 Axial Loading Test
After cooled down in the air for a week, the fire-damaged columns were subjected to axial loading as shown in Fig. 3. A compressive force was applied to the top surface of specimen using a hydraulic jack. Linear Variable Differential Transformers (LVDTs) were installed at each side surface of the specimens to measure displacement of the columns during the loading.
3. Experimental Results
3.1 Experimental Results from Fire Test
3.1.1 Observation of the Fire-damaged Columns
As shown in Figs. 4(a) and 4(b), white and black stains were visible on the non-heated surfaces due to evaporation of moisture during heating. In addition, post-cooling spalling of specimens was observed from surfaces of the cooled down specimens. This was observed not during the heating, but cooling. The post-cooling spalling occurred mainly on the heated surfaces with depth less than 3 mm, and some of the columns show relatively large post-cooling spalling at corners with maximum depth of 15 mm. Since the post-cooling spalling occur during cooling, it is not possible to be explained by neither thermal stress spalling theory nor pore pressure spalling theory. Rather, it has been known that the spalling can occur from the concrete mixed with calcareous aggregate (Khoury, 2003, 2008), caused by over 40% of volume expansion resulting from rehydration of CaO to Ca (OH)2 during cooling and then the expansion leads to make severe internal cracks (Annerel and Taerwe, 2007, 2009). In other studies (Xing et al., 2011; Klingsch, 2014), post-cooling spalling was reported from the heated specimens during cooling down in ambient temperature.
3.1.2 Temperature Distribution
Figs. 5(a) and 5(b) show the time–temperature curves of columns measured at different depths along the thickness. The temperatures measured at 40 mm from the heated surface increased rapidly in the beginning of the fire test and then slowly increased in time. The temperatures at 100 and 175 mm from the heated surface reached 100 °C and stayed at this temperature longer than those at other locations because the moisture moved from the single face exposed to fire to the unexposed surface. Time-temperature curves of all tested specimens tended to be similar in other specimens. In case of H1, the temperatures at the C1 and C5 located in the concrete cover were similar and lower than other specimens because H1 was heated on one side. On the other hand, there was no significant difference of C2, C3, and C5 among heated specimens regardless of heated area.
3.2 Experimental Results from Axial Loading Test
3.2.1 Failure Modes
Figs. 6(a)~6(d) show failure modes of the fire-damaged specimens after the loading tests. Severe damages such as concrete crushing and buckling of longitudinal bars were observed. In the loading tests, the specimen exposed to fire was observed to be more brittle than the control specimen. Moreover, increased concrete crushing was observed from the heated surface compared to the non-heated surface. This phenomenon became obvious in H3 and H4 due to the relatively large heated area.
3.2.2 Residual Strength
The experimental results of residual strength for the specimens were obtained from the axial loading test as listed in Table 4. It was obvious that the residual strength of the specimen decreased as the heated area increased. It was interesting to note that the residual strengths of the partially heated specimens tended to decrease slightly larger than numerical results obtained by the 700 °C isotherm (Ryu, 2020). This was because the asymmetric cross sectional properties of the partially heated columns worked as if the column was subjected to eccentric loading. Therefore, it can be said that additional decrease of residual strengths can occur from the partially heated specimens because of the eccentric loading effect. This may become more obvious when the longer columns are used in real building.
Table 4
Residual Strength of RC Columns
| Specimen | Experiment | Analysis (Ryu, 2020) | ||
|---|---|---|---|---|
| Residual Strength (kN) | Ratio | Residual Strength (kN) | Ratio | |
| C0 | 5,931.4 | 1.000 | 6,096.7 | 1.000 |
| H1 | 5,460.3 | 0.921 | 5,629.7 | 0.923 |
| H2 | 4,810.8 | 0.811 | 5,168.1 | 0.848 |
| H3 | 4,464.5 | 0.753 | 4,691.6 | 0.770 |
| H4 | 3,991.9 | 0.673 | 4,128.4 | 0.677 |
4. Conclusions
In this paper, the thermal and structural behaviors of RC columns under and after fire were investigated according to heated area. The following conclusions can be drawn from the results.
(1) Post-cooling spalling of specimens were observed on all parts exposed to high temperature not during the heating but after cooling down to ambient temperature. But there was no noticeable differences of crack patterns and temperature distributions among the heated specimens.
(2) From axial loading test results, the residual strength of the columns decreased as the heated area increased. In addition, there was eccentric loading effect of the columns heated on one ~ three side surfaces, which causes additional decrease of residual strength.
PDF Links
PubReader
ePub Link
Full text via DOI
Download Citation
Print
