Mechanical properties of concrete composites subject to elevated temperature
Fire Safety Journal 95 (2018) 66–76
Contents lists available at ScienceDirect
Fire Safety Journal j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / fi r e s a f
Mechanical properties of concrete composites subject to elevated temperature Josef Novak *, Alena Kohoutkova Czech Technical University in Prague, Faculty of Civil Engineering, Department of Concrete and Masonry Structures, Thakurova 7, 166 29, Prague 6, Dejvice, Czechia
A R T I C L E I N F O
A B S T R A C T
Keywords: Fibre reinforced concrete Heat transport test Heat treatment Mechanical properties
Elevated temperature Structural behaviour
Fire resistance represents an important parameter which is necessary to consider during the structural design of buildings. It is deﬁned as an ability of building components to perform their intended load-bearing functions under ﬁre exposure. In terms of ﬁre resistance, the right choice of a construction material plays a key role and can reduce structural damage or even save human lives. The building industry offers a wide range of materials whose structural behaviour is more or less affected by temperature. Recently, concrete has become one of the most utilized materials used for a various kind of buildings. While the knowledge and experience with concrete behaviour under ambient temperature are well-known, the behaviour under elevated temperature has to be deeply investigated. The paper deals with observing the behaviour of concrete composites with addition of ﬁbres under ambient and elevated temperature with the aim to determine the mechanical properties of materials. The experimental tests were conducted on three selected concrete composites which differ in a type and content of ﬁbrous reinforcement used. The experimental work carried out was divided into several phases. First of all it was necessary to leave the produced specimens aging and drying in order to minimize the risk of unexpected damage caused by concrete spalling during heating. Time to time, the specimens were weighted with the aim to determine the loss of weight imposed by drying. Then, a heat transport test was performed on a few reference specimens in order to determine the time required for uniform heating the specimens up to 200 C, 400 C and 600 C. In the last phase, conventional testing methods were undertaken to determine the mechanical properties of concrete composites at ambient and elevated temperature. A compression test and a splitting tensile test were conducted on 150 mm cubes. Based on the results, the peak and residual strength of the materials were determined for various temperature levels. The obtained ﬁndings contribute to improving the knowledge in the ﬁeld of both concrete structures exposed to high temperature and structural behaviour of ﬁbre reinforced concrete. The ﬁndings can be also utilized in case of the structural design of concrete structures with the high risk of ﬁre loading.
1. Introduction Today, a structural design represents a very complex task which includes analyzing many parameters such as ultimate bearing capacity, stability, deﬂection, rigidity of structure etc. The essential part, which has to be paid attention to, is also ﬁre resistance deﬁned as an ability of building components to perform their intended load-bearing functions under ﬁre exposure. Although the issue of ﬁre safety seems to be not so important, the inadequate ﬁre design or a construction material choice can cause structural damage or even loss of human lives. For instance, in 2008 an extensive ﬁre occurred in the 13-story Faculty of Architecture Building at the Delft University of Technology (TUD) in Netherlands .
Although all building occupants evacuated safely, the ﬁre burned uncontrolled for several hours and caused the structural collapse of the major portion of the building with a reinforced concrete structural system. The key cause of the collapse is assumed to be the occurrence of spalling which rapidly and dramatically reduced reinforced concrete member capacities. A massive ﬁre also hit the Windsor Tower in Madrid, Spain, in February 2015 . The structural system of the tower was
constituted of a reinforced concrete core, interior columns, wafﬂe slab and steel exterior columns. The part of building was exposed to the ﬁre for 20 h resulting in the extensive structural damage. As lately analyzed, a large portion of the ﬂoor slabs collapsed during the ﬁre due to the failure of reinforced concrete elements. However, according to other sources 
the collapse was induced by buckling of steel columns during the ﬁre. Today, there are many regulations and standards which deal with the demands for the durability and ﬁre resistance of concrete structures . The European standard EN 1992-1-2  is a widely-used document which serves for designing concrete structures exposed to high temperature. Recommendations and principles stated not only in this standard are usually based on the extensive experience and experimental investigations in the ﬁeld of concrete composites subject to high temperature. However, the available data does not have to accurately describe the structural behaviour of relatively new materials including ﬁbre reinforced concrete (FRC). While the knowledge and experience with FRC behaviour under ambient temperature are well-known, the behaviour under elevated temperature has to be deeply investigated. FRC is a building material whose utilization in the concrete industry has been rapidly increasing. This development is motivated by its physical and mechanical properties which contribute to traditional concrete elements and structures various economical beneﬁts such as structure subtlety, part or full elimination of conventional reinforcement, enhanced impact resistance, resistance to mechanical loads and environmental loads. Recently, many comprehensive studies have been undertaken with the aim to observe the mechanical behaviour of FRC exposed to elevated temperature. Although concrete is well-known for a high degree of ﬁre resistance, high temperature seriously damages microstructure and mesostructure which results in generalised mechanical decay of a concrete composite . As a consequence, the extensive knowledge of mechanical properties of FRC exposed to elevated temperature seems to be decisive for a wider utilization of the material. There are two fundamental types of methodological procedure used for observing mechanical properties of concrete at elevated temperature. Most of experimental investigations [7–33] are conducted on test specimens at ambient temperature after high temperature exposure and only a few is performed on hot test specimens [34–36]. Such approach of experimental testing is preferred mainly due to a simple way of testing as tests are easier to conduct on test specimens at ambient temperature. However, if results obtained from the tests on specimens after high temperature exposure correspond enough to the mechanical properties of a tested material at a certain temperature level has not been still fully understood. Bamonte and Gambarova belong to authors which very intensively deals with such issue and states in their publications [37,38] that the hot and residual (after high temperature exposure) behaviour in compression are very close; the only difference was observed in case of the peak strain in compression which is larger on heated specimens in comparison with specimens cooled down after temperature exposure. The ﬁre response of concrete composites is closely associated with concrete composition, particularly with a type and content of concrete components used. Generally speaking, concrete made of siliceous aggregates, such as granite, shows unfavourable mechanical properties at high temperature compared to concrete composed of calcareous aggregates such as dolomite and limestone [34,39]. Recently, a lot of interest is being paid on the possible use of metakaolin, ﬂy ash and silica fume as partial cement replacement in concrete subject to high temperature [7,12,13]. Owing to silica fume and ﬂy ash ﬁneness, concrete composites with such additions have denser microstructure and as a consequence their explosive spalling tendency increases . The amount and type of ﬁbres also have an inﬂuence on the structural performance of a composite exposed to high temperature. A number of experimental investigations have been conducted up to date with the aim to observe the ﬁre response of FRC with a various type and amount of ﬁbres. Particularly, the studies are focused on the effect of a type, shape and content of ﬁbres on the mechanical properties of concrete composites, mostly compressive and tensile strength including elastic modulus. Namely, it concerns steel ﬁbres [7,10,11,13–16,19,21,24–27,29,32–35], synthetic ﬁbres [7,9,11–14,16,17,20,22,24,27,32,34,35] and mix of steel and polypropylene ﬁbres [7,11,16,18,24,27,32,34,35] which are widely used in the concrete industry. There also a few investigations which deals with carbon ﬁbres [8,17] and glass ﬁbres .
As the melting point of steel is relatively high in comparison with other materials, the use of steel ﬁbres seems to be beneﬁcial for concrete composites exposed to high temperature. Incorporating steel ﬁbres into concrete composites remains advantageous even when the concrete composites are exposed to high temperature up to temperature up to 1200 C, particularly 1% content has no deleterious effect on heated concrete. In fact, the inclusion of steel ﬁbres in a concrete mix leads to an improvement in both mechanical properties and resistance to heating effects in comparison with unreinforced concrete [10,21,27,28,32]. Some experimental investigations [15,19,26] even demonstrate the compressive strength of steel ﬁbre reinforced reactive powder concrete and geopolymer concrete gradually increases when the material is heated up to 200–300 C, but starts to decrease as temperatures further increase. The compressive strength of reactive powder concrete with 1% steel content is higher between 200 C and 400 C than at room temperature and subsides when temperature exceed 500 C. Furthermore, 2% and 3% steel ﬁbre content signiﬁcantly increase compressive strength from 200 C to 300 C which then gradually decreases as the temperatures reach 400 C and beyond. However, a higher content of steel ﬁbres cannot improve the compressive strength of concrete composites at elevated temperatures . On the contrary, it has been also demonstrated that steel ﬁbres have negligible effect on high temperature compressive strength and only improve tensile strength when temperature up to 400 C is considered . The tensile behaviour of SFRC subjected to elevated temperature is more sensitive to the volume fraction and the aspect ratio of the ﬁbre than to its type . SFRC also has the higher toughness after the high-temperature exposures when compared to the initial values of unheated concrete . As temperature increases, SFRC weakens and shows reduced stiffness with the degradation depending on a type, aspect ratio, and volume fraction of the ﬁbre  and the reduction in modulus of elasticity is more pronounced than the reduction in compressive strength for the same heat treatment . Steel ﬁbre reinforced recycled aggregate concrete exhibits the identical behaviour and loses stiffness much faster than strength after exposure to elevated temperature . As a consequence, peak strains gradually increases together with temperature. The increase in peak strains along with steel ﬁbre content does not signiﬁcantly differ between ambient temperature and 200 C. When temperatures exceed 200 C, a higher steel ﬁbre content generally is associated with a higher peak strain . Moreover, the addition of steel ﬁbres does not eliminate the spalling tendency of concrete mixtures . 1.2. Synthetic ﬁbre reinforced concrete In the concrete industry, synthetic ﬁbres, regardless of a type, shape and length, are mostly utilized with the aim to increase concrete spalling resistance when concrete structures exposed to elevated temperature [20,35]. As the melting point of synthetic ﬁbres is relatively low, the presence of ﬁbres in a concrete composite subjected to elevated temperature affects the mechanical properties of the concrete composite, particularly residual compressive strength, modulus of elasticity and splitting tensile strength . While the presence of polypropylene ﬁbres slightly increase ductility and the speciﬁc toughness, deﬁned as the ratio of the area under the stress–strain curve, and the compressive strength of unheated concrete, after heating all the enhanced characteristics are lost [7,27]. Many contributions even demonstrate that polypropylene ﬁbres have negative effect on the residual mechanical properties of polypropylene ﬁbre reinforced concrete (PPFRC) after high-temperature exposure as they signiﬁcantly decrease the residual compressive strength, elastic modulus and tensile strength as well as they increase peak strain [13,20,34]. On the other hand, some experimental investigations show that polypropylene ﬁbres can improve the relative residual compressive strength of a concrete composite after the exposure to ﬁre . While the presence of polypropylene ﬁbres at different 67
J. Novak, A. Kohoutkova
Fire Safety Journal 95 (2018) 66–76
dosages does not affect the residual compressive strength at 200 C and 400 C, it considerably increases the residual compressive strength of concretes after exposure to 600 C . Moreover, some even shows that compressive strength of concrete with polypropylene ﬁbre additions is greater than unreinforced concrete without additions, when subjected to temperature up to 400 C . In comparison with SFRC, polyvinyl alcohol ﬁbres reduce the uniaxial compressive strength but cause no appreciable change in elastic modulus in reactive powder concrete .
2. Experimental investigation Based on the gap in the ﬁeld of FRC performance at high temperature, the presented experimental investigation was developed. The whole work was carried out by the team of research workers at Czech Technical University in Prague for one year. As the research was very extensive, the investigation was divided into several steps in the following order - test specimens production, heat treatment, compression test and splitting tensile test.
1.3. Hybrid (steel þ synthetic) ﬁbre reinforced concrete 2.1. Test specimens The combination of steel and synthetic ﬁbres represents a promising alternative how to ensure good toughness of a concrete composite before heating and improve its residual mechanical behaviour and spalling resistance as well as the ductility after heating . Although a few contributions declare that a ﬁbre cocktail does not have much effect on high temperature compressive strength , the most of presented work afﬁrm that the incorporation of steel ﬁbres can effectively improve the compressive properties of a concrete composite when exposed to elevated temperatures while polypropylene ﬁbres enhance concrete spalling resistance [18,32]. Speciﬁcally, the combination of steel ﬁbres and polypropylene ﬁbres shows positive synergy effect on the post-peak behaviour of concrete composites before and after exposure to high temperature . However, as synthetic ﬁbres have a low melting point and ignition point, only steel ﬁbres provide the stability and enhanced mechanical behaviour to a concrete composite after exposing to elevated temperatures . Considering steel to synthetic ﬁbre content ratio, a concrete composite containing 1% synthetic ﬁbres and 1% steel ﬁbres by volume seems to produce the best results, balancing performance at high temperature with consideration of initial mechanical properties . As a consequence, using hybrid ﬁbre reinforced concrete (HFRC) (steel and synthetic ﬁbres) might provide necessary safe guarantee for the rescue work and structure repair during and after a ﬁre disaster.
All experimental tests were carried out on 150 mm cubes which are made of three concrete composites with different composition (Table 1). The investigated concrete composites are composed of easy available and widely used components which include Portland cement 42,5 R characterized by high early strength in accordance with CSN EN 197–1 , water and siliceous aggregate. The workability of fresh concrete mass was maintained by using plasticizer Sika Visco Crete 1035 that also reduces the content of used mixing water. The investigated FRCs also contain two types of ﬁbres. Single hook end steel ﬁbres Dramix. RC-80/60-BN have tensile strength 1225 MPa and served as main reinforcement in the concrete composites. Whereas polypropylene ﬁbres were added in concrete with the aim to reduce the risk of explosive concrete spalling which occurs at higher temperature [41,42]. Totally, it was produced about one hundred specimens which were left aging and drying in standard conditions in a laboratory for one year. During this process the specimens were irregularly weighted in order to determine the loss of weight in time caused mainly by drying. However, after one year the decrease in weight was negligible in a range between 0.5% and 1% of the initial weight (Fig. 1). As expected, the highest rate of weight loss was observed immediately after the production when mixing water is consumed for concrete hardening. Long-term aging in normal conditions did not signiﬁcantly affect the weight of specimens as well as the water content.
2.2. Heat transport test
The state of the art review highlights that most of the experimental investigations have been undertaken on specimens cooled down after a ﬁre exposure to obtain material properties. Such methodological procedure likely considers further concrete deterioration in a cooling phase but is not capable to describe material properties at elevated temperature. Since the main objective of the experimental investigation was to obtain the properties of a material right at time of a ﬁre which will be subsequently used for a numerical simulation of steel-FRC composite columns at a ﬁre, the proposed experiments were conducted on heated specimens. The review also indicates several issues associated with the heat treatment of test specimens. In most cases, test specimens are heated in an electric furnace with an automatic control system which only monitors the temperature in the furnace. Heat treatment procedure, speciﬁcally time required for uniform heating, is based only on previous experience and no details are usually provided about uniform temperature distribution all over the specimen [7–25,27,28,32–35]. However, such approach might be inappropriate when various heating rate and specimen shape are considered because the results obtained from experimental tests on non-uniform heated specimens might be vague. The other issue is connected with limited database of results. While there is extensive experience with conventional plain concrete at high temperature  the effect of ﬁbres on the behaviour of FRC has not been fully understood. Speciﬁcally, very limited data are available when pre-cracking and post-cracking tensile strength are considered. Therefore, it is desired to enhance the database and subsequently validate or derive existing or entirely new stress-strain relations for a particular type of FRC, respectively.
As the compression test and splitting tensile test were conducted on hot specimens, ﬁrst of all it was necessary to determine a time period required for the uniform heating of the specimens as well as to determine the rate of cooling in order to avoid excessive temperature loss during the tests. The heat transport test arranged particularly for this investigation was performed on nine 150 mm cubes; three specimens of each composite were gradually used for the three temperature levels - 200 C, Table 1 Concrete composition. Concrete component
Portland cement 42,5 R Water Fine aggregate 0–4 mm Coarse aggregate 4–8 mm Coarse aggregate 8–16 mm Plasticizer Sika Visco Crete 1035 Steel ﬁbres Dramix RC-80/60-BN Polypropylene ﬁbres Forta Ferro 54 mm
Content [kg/m3] Plain concrete (PC)
Steel ﬁbre reinforced concrete (SFRC)
Hybrid ﬁbre reinforced concrete (HFRC)
J. Novak, A. Kohoutkova
Fire Safety Journal 95 (2018) 66–76
Fig. 1. Percentage loss of weight in time during specimens aging.
400 C and 600 C. The heat treatment was conducted by using a special system (Fig. 2), which consists of a control machine Mannings HTC 70 kW, ceramic pads and K-type thermocouples measuring temperature in a range from À200 C to 1000 C. First of all, three holes serving for the thermocouples installation were drilled in each cube. Two 75 mm deep holes were positioned 10 mm from the edge, the last one at the centre with the aim to measure temperature in the specimen core (Fig. 3). To ensure correct functioning of the thermocouples it was necessary to prepare the holes of the same diameter as the thermocouples have, otherwise any air void in the holes could affect the test results. Afterwards, two ceramic pads were attached directly to four of six specimen sides. The position of ceramic pads was secured by high temperature resistant glass wool insulation which was wrapped tightly around the pads together with a specimen. The insulation used simulates so called thermo-box which also ensures heat accumulation
during the heat transport test. Subsequently, totally ﬁve thermocouples were installed in their position when three of them were inserted through the glass wool into the prepared holes drilled in the specimen and the other two were placed between the ceramic pads and a specimen to measure the surface temperature of the ceramic pads. The temperature of an unheated side was measured time to time by an infrared thermometer and compared with the temperature distribution at the depth of the specimen. Although the temperature development in a ﬁre deﬁned by the standard temperature-time curve (1) in accordance with CSN EN 1991-12  is signiﬁcantly faster, due to the control machine performance the
θg ¼ 20 þ 345: log10 ð8:t þ 1Þ
initial heat transport tests were conducted with the maximum heating rate 1000 C/h in order to simulate as much as possible the real
Fig. 2. System used for heat transport test (a) control machine Mannings HTC 70 kW (b) K-type thermocouples (c) ceramic pads.
Fig. 3. Test specimen preparation for heat transport test (a) specimen with ceramic pads (b) specimen together with ceramic pads placed into thermo-box (c) specimen during heat transport test. 69
J. Novak, A. Kohoutkova
Fire Safety Journal 95 (2018) 66–76
removed from the thermo-box by using ﬁreproof gloves, wrapped in new glass wool insulation and placed on steel plates. The intention of this process was to simulate the steps required for performing the compression test and splitting tensile test on heated specimens and consequently to precisely validate the rate of heat loss. While the temperature in the core was stabilized, the temperature on the sides of specimens decreased rapidly due to the low temperature of surrounding environment. The rate of heat loss is more signiﬁcant at specimens heated up to higher temperature due to a higher temperature gradient. Based on the obtained data from the heat transport test, it was determined the maximum allowable time period for the compression test and splitting tensile test equal to 10 min. The test specimens used for the heat transport test also served for determining the volumetric mass density of materials at ambient and elevated temperature. Before the heat transport tests, the prepared specimens were weighed and measured in order to determine the material density at ambient temperature. The density ranged from 2350 kg/ m3 to 2400 kg/m3 which correspond to conventional concrete. After the heat transport tests, the whole process was repeated. With the aim to obtain as precise results as possible the holes for thermocouples were also considered during the volume calculation. With increasing temperature the volumetric mass density gradually decreased down to a range 2300 kg/m3 - 2350 kg/m3, 2250 kg/m3 - 2300 kg/m3 and 2200 kg/m3 2250 kg/m3 corresponding to the temperatures 200 C, 400 C and 600 C, respectively. The main cause is dehydration of the ingredients of cement paste, dissociation of calcium hydroxide Ca(OH)2 into CaO and water and free water evaporation . The inﬂuence of melted polypropylene ﬁbres was negligible as their content was low. Although the composition of all three concrete composites differed, the effect of temperature on the volumetric mass density of the materials was almost identical with no signiﬁcant deviations.
conditions of a ﬁre (Fig. 4). While the specimens of HFRC did not exhibit any defects in response to the high temperature gradient, concrete spalling occurred at the specimens made of SFRC and PC in the 26th minute of the heat treatment. Concrete spalling was imposed by both evaporable and nonevaporable water chemically bound with cement paste which causes the pore pressure buildup and subsequently spalling in concrete . At the time of concrete spalling, the temperature on the heated sides of specimens and in a core reached roughly 450 C and 60 C, respectively. While a massive explosion and extensive spalling of a whole PC specimen happened, in case of a SFRC specimen only small pieces of the size in millimetres were released owing to the presence of steel ﬁbres in the concrete composite. The presence of polypropylene ﬁbres in HFRC lowered the risk of concrete spalling as the ﬁbres melted with increasing temperature due to low melting point. The melted polypropylene ﬁbres increased the concrete composite permeability witch it positively affected the pore pressure and subsequently decreased spalling in concrete. As a consequence, the specimens were eventually heated up gradually step by step when temperature was increased depending on the amount of vapour visibly bursting out of the specimens with the aim to avoid any damage caused by extensive vapour expansion in SFRC and PC specimens. When no or little vapour was spotted, the temperature was repeatedly increased by 50 C until a particular temperature level was reached. Such approach was used for the ﬁrst PC cube heated up to 600 C and as no concrete spalling occurred the identical heat treatment pattern was applied for the other specimens. On average, such heat treatment corresponds roughly to the heating rate 200 C/h. The heat transport tests were performed to determine the time period for the uniform heating up to 200 , 400 C and 600 C at all three concrete composites (Fig. 5). The temperature at the interface between the ceramic pads and a concrete sample was increased step by step following always the same pattern until 250 C, 450 C or 650 , respectively, were reached. As the test continued, the interface temperature was kept at the same level until the temperature in the specimen core reached 200 C, 400 C and 600 C respectively. The increase in temperature on the heated sides of the specimen (at the 75 mm depth) was more rapid in comparison with the specimen core which is placed further from the source of heat. The temperature distribution at zero depth on the top unheated side observed using the infrared thermometer differed from the one obtained at the 75 mm depth. However, as the temperature gradient decreased during the heat transport test, the difference in the temperature distribution at both depths ranged from 0 C to 8 C and consequently was negligible. The uniform heating of the specimens up to 200 C, 400 C and 600 C took approximately two, four and 6 h respectively regardless of the concrete composition. When the intended temperature levels were reached, the heat transport system was turned off in order to observe specimen cooling in detail. First, the specimens were immediately
2.3. Compression test The concrete compression test was conducted on the 150 mm cubes under 20 C, 200 C, 400 C and 600 C in accordance with CSN EN 12390–3 . First, the whole set of specimens was divided into four groups according to the intended temperature level during testing. The specimens tested under elevated temperature were heated up to the intended temperature following the procedure used during the heat transport test. The specimens together with two ceramic pads were inserted into the thermo-box of glass wool (Fig. 6) and then two thermo-couples were placed between the specimen and ceramic pads with the aim to control and monitor the temperature of the ceramic pads. The time and heat treatment were identical to those obtained from the heat transport test. Heating up to 200 C, 400 C and 600 C took 113 min, 235 min and 370 min, respectively.
Fig. 4. Nominal ﬁre curve and heating rates used for the heat transport test. 70
J. Novak, A. Kohoutkova
Fire Safety Journal 95 (2018) 66–76
Fig. 5. Heat transport diagrams (a) legend of thermocouples, (b) heating up to 600 C, (c) cooling of specimens heated up to 600 C, (d) heating up to 400 C, (e) cooling of specimens heated up to 400 C, (f) heating up to 200 C, (g) cooling of specimens heated up to 200 C.
Fig. 6. Test specimens during heating.
deformation 0.02 mm/s in the range of 0–5 mm and 0.1 mm/s for deformation greater than 5 mm. The obtained data from the test were used for generating compressive strength-displacement diagrams which describe the material behaviour at various temperature levels.
After the intended temperature in the specimen core was reached the compression test was performed in a testing machine Inova 200F. The time of compression test of each sample had to be kept under 10 min in order to avoid excessive temperature loss. The test was driven by
J. Novak, A. Kohoutkova
Fire Safety Journal 95 (2018) 66–76
The presented results do no exhibit any signiﬁcant deviations caused by either incorrect conducting the experimental tests or poor-quality manufacturing technology. The diagrams of the investigated concrete composites regardless of the amount and type of used ﬁbres correspond
The compressive strength-displacement diagrams (Fig. 7) show the structural behaviour of tested specimens at different temperature levels, mainly the peak and residual (post-peak) strength of the investigated materials.
Fig. 7. Compressive strength-displacement diagram of the concrete composites. 72
J. Novak, A. Kohoutkova
Fire Safety Journal 95 (2018) 66–76
200 C, 400 C and 600 C in accordance with CSN EN 12390–6 . The preparation phase was identical to that used for the compression test. The specimens were divided into four groups according to the intended temperature during testing, the specimens tested under elevated temperature were inserted into thermo-boxes together with two ceramic pads and heated up for the speciﬁed time period to the intended temperature. After the intended temperature in the specimen core was reached, the splitting tensile test was performed in a testing machine Inova 200 F (Fig. 9). The time of the splitting tensile test of each sample had to be kept under 10 min in order to avoid excessive temperature loss. The test was driven by deformation 0.02 mm/s in the range of 0–5 mm and 0.1 mm/s for deformation greater than 5 mm. Based on the obtained data from the tests, splitting tensile strength-displacement diagrams describing the material behaviour at various temperature levels were generated. The presented diagrams (Fig. 10) show the tensile behaviour of the tested concrete composites at various temperature levels, mainly the peak and residual (post-peak) splitting tensile strength. The obtained results are with no signiﬁcant deviations imposed by mistakes made during testing or producing experimental specimens. The mechanical behaviour described by the presented diagrams corresponds to the typical structural behaviour of either plain concrete or FRC. In the opening phase the splitting tensile strength raises almost linearly until the peak strength is reached. While a sudden fall in strength occurs at PC at this point, HFRC and SFRC exhibit residual strength. Such phenomenon results from the absence of ﬁbres in PC which is then weak in tension. As consequence, a brittle failure occurs at specimens made of PC in comparison with other two concrete composites. HFRC and SFRC are characterized as FRC with a tension softening curve  as their residual strength decrease after the ﬁrst macro-crack occurs. The regions with short-term increase in residual strength result usually from ﬁbres which are gradually activated with increasing loading until they are fully activated. Comparing to the composition of SFRC, HFRC contains extra 3 kg/m3 polypropylene ﬁbres Forta Ferro 54 mm (Table 1) and as a consequence its residual tensile strength is higher at ambient temperature when the polypropylene ﬁbres act in the concrete composite together with the steel ﬁbres. Subsequently, the polypropylene ﬁbres in HFRC burn out with increasing temperature and their inﬂuence on the concrete composite strength is negligible and as a consequence the peak and residual strength of HFRC and SFRC are almost identical. The most interesting ﬁnding relates to the peak splitting tensile strength (when the ﬁrst macro-crack occurs) of the investigated concrete composites. The dependence of peak splitting tensile strength on ascending temperature at both PC and SFRC with no polypropylene ﬁbres signiﬁcantly differs in comparison with HFRC (Fig. 11). The peak strength of HFRC gradually decreases with increasing temperature, whereas the ultimate strength of both SFRC and PC at 200 C is lower than the strength at 400 C. This phenomenon is caused by the testing methodology when several types of stresses the hot specimens had to withstand during the splitting tensile test to. Namely, it concerns the stresses from static loading and pore pressure tending to tear the specimens. Moreover, the material properties worsen with increasing
to the typical behaviour of concrete in compression. In the opening phase, the compressive strength of the materials increases linearly until the concrete composites start yielding and ultimate compressive strength is reached. The modulus of elasticity deﬁned by the slope of the linear part of the curves signiﬁcantly decreases at all concrete composites with increasing temperature. This phenomenon is caused by the specimen structure signiﬁcantly damaged by cracks resulting from both the high temperatures and pore pressure. In case of the HFRC, the material structure at elevated temperature is also more porous as the polypropylene ﬁbres melts due to the low melting point of the material . Consequently, the presence of polypropylene ﬁbres in a concrete composite also affects the peak compressive strength of material (Fig. 8). Moreover, the specimens, regardless of the type of a concrete composite, exhibit enhanced ductility with increasing temperature. While the effect of temperature on the compressive strength of PC and SFRC with no polypropylene ﬁbres is noticeable at the temperature higher than 400 C, the compressive strength of HFRC signiﬁcantly starts decreasing already at lower temperature due to the porous structure of specimens caused by burned polypropylene ﬁbres. This phenomenon is particularly visible at SFRC whose compressive strength at 200 C, 400 C and 600 C corresponds to 99%, 99% and 55%, respectively, of its initial strength at ambient temperature and HFRC whose compressive strength at the same temperature levels represents 98%, 88% and 45%, respectively, of its initial strength. The compressive strength of PC at 200 C, 400 C and 600 C is 94%, 94% and 51%, respectively, of the initial strength at ambient temperature. In comparison with CSN EN 1992-1-2  recommending to use for plain concrete with siliceous aggregates 95%, 75% and 45% of the initial compressive strength, respectively, the obtained results from the experimental tests differ signiﬁcantly. The effect of steel ﬁbres in SFRC and HFRC on the behaviour and compressive strength of the materials at ambient and elevated temperature is negligible. 2.4. Splitting tensile test Splitting tensile test was performed on 150 mm cubes under 20 C,
Fig. 8. Peak compressive strength of the concrete composites at ambient and elevated temperature.
Fig. 9. Testing machine Inova 200 F (left) and split tensile test of heated specimen (right). 73
J. Novak, A. Kohoutkova
Fire Safety Journal 95 (2018) 66–76
Fig. 10. Splitting tensile strength-displacement diagram of the concrete composites.
the porosity of HFRC at elevated temperature is higher than the porosity of SFRC and PC due to the presence of the polypropylene ﬁbres. The polypropylene ﬁbres melt shortly with increasing temperature and consequently the material porosity increases. As a consequence, the failure mode of HFRC specimens at 200 C was typical because most of steam was released during the 2 h heat treatment. Whereas the SFRC and PC specimens exhibited a explosive failure mode with the visible steam
temperature. Considering such statement, the detailed analysis of such results can be carried out. While the specimens of the investigated concrete composites were being tested, there had been already pore pressure inside them resulted from the heat treatment. The level of pore pressure mainly depends on temperature gradient, time of the heat treatment and material porosity. While the ﬁrst two parameters are identical for all concrete composites, 74
J. Novak, A. Kohoutkova
Fire Safety Journal 95 (2018) 66–76
200 C, 400 C and 600 C. Based on the obtained data the following conclusions are made: ▪ The long-term aging and drying of FRC specimens, regardless of the type of ﬁbre used, in normal conditions do not signiﬁcantly affect the weight of specimens as well as the water content. ▪ The presence of polypropylene ﬁbres in a concrete composite lowers the risk of explosive concrete spalling as the ﬁbres melt with increasing temperature due to the low melting point. Consequently the ﬁbres increase the concrete composite permeability and decrease pore pressure. Considering the heating rate 1000 C/h, HFRC, SFRC and PC exhibited no concrete spalling, low-level concrete spalling (in millimetres) and extensive concrete spalling, respectively. ▪ With increasing temperature the volumetric mass density of concrete gradually decreases approximately down to 2375 kg/ m3, 2325 kg/m3, 2275 kg/m3 and 2225 kg/m3 corresponding to the temperatures 20 C, 200 C, 400 C and 600 C, respectively. ▪ The presence of polypropylene ﬁbres in a concrete composite affects the peak compressive strength of the material at elevated temperature. In comparison with PC and SFRC, the peak compressive strength of HFRC decreased more rapidly with increasing temperature due to the higher porosity of specimens resulted from melted polymer ﬁbres. This phenomenon is particularly noticeable at SFRC whose compressive strength at 200 C, 400 C and 600 C corresponds to 99%, 99% and 55%, respectively, of its initial strength at ambient temperature and HFRC whose compressive strength at the same temperature levels represents 98%, 88% and 45%, respectively, of its initial strength. ▪ When conducting splitting tensile test at 200 C, the SFRC and PC specimens exhibited explosive failure with visible steam release. The main cause of such behaviour is pore pressure which together with applied load tend to tear the test specimens. As a consequence, the level of pore pressure inside concrete composites affects the material capability to withstand tensile stress. The presence of polypropylene ﬁbres in HFRC increased the porosity of the material, subsequently decreased pore pressure in the specimens and consequently increased the capability to withstand applied load at 200 C. While the peak splitting tensile strength of SFRC at 200 C, 400 C and 600 C correspond to 85%, 103% a 58% of its initial strength at ambient temperature, the peak strength of HFRC represents 97%, 88% and 41%, respectively, of its initial strength. ▪ The comparison between tensile strength recommended for plain concrete with silica aggregates stated in CSN EN 1992-1-2  and obtained results from the experimental investigation shows a signiﬁcant difference. While the standard recommends to use 80%, 40% and 0% of the initial tensile strength of concrete for the ﬁre design of structures exposed to 200 C, 400 C and 600 C, respectively, the peak strength of PC obtained from the experimental investigation corresponds to 57%, 81% and 52% of its initial strength. Higher strength of PC at 400 C than at 200 C is caused by the explosive failure at 200 C described in a detail above.
Fig. 11. Observed splitting tensile strength of concrete composites at ambient and elevated temperature.
release while being tested. The main cause of such failure is that the low material porosity enabled only little steam to be released within 2 h heat treatment. Since the SFRC and PC specimens were exposed to high pore pressure right at the time of testing, they failed at a lower level of applied load. From my point of view, such phenomenon has not been investigated yet as the most experimental tests are conducted on specimens cooled down after ﬁre exposure and thus with no pore pressure inside them. The reason why such failure mode was not such distinctive during the splitting tensile test of PC and SFRC specimens at 400 C and 600 C is associated with the time of the heat treatment. To heat the specimens up to 400 C and 600 C lasted twice and three times longer, respectively. Such time periods enabled most of steam to be gradually released through micro pores in the composites and consequently to reduce the pore pressure right at the time of testing. No explosive failure mode was also observed while the SFRC and PC specimens being tested at 400 C and 600 C which it only proves the statement. The pore pressure inside the SFRC and PC specimens gradually fell down with increasing temperature to the value the HFRC specimens had. Therefore, the peak splitting tensile strength of the investigated concrete composites at 600 C is almost identical. While the peak splitting tensile strength of SFRC at 200 C, 400 C and 600 C correspond to 85%, 103% a 58% of its initial strength at ambient temperature, the peak strength of HFRC represents 97%, 88% and 41%, respectively, of its initial strength. The increase in strength of SFRC up to 103% at 400 C might result from uncertainties and variance in the production and testing of the specimens. Nevertheless, increasing strength in a range from ambient temperature to 400 C is rather usual for reactive powder concrete composites [15,18,26] but was also observed within the experimental investigation of concrete composites composed of conventional concrete components . The comparison between ﬁgures stated in CSN EN 1992-1-2  for plain concrete with silica aggregates and obtained results from experimental tests shows a signiﬁcant difference. While the standard recommends to use 80%, 40% and 0% of the initial strength for the ﬁre design of concrete structures exposed to 200 C, 400 C and 600 C, respectively, the peak strength of PC obtained from the tests correspond to 57%, 81% and 52% of its initial strength.
Based on the obtained ﬁndings, HFRC with addition of steel and polypropylene ﬁbres is suggested to use for structures with the high risk of ﬁre loading. Despite the HFRC exhibited lower compressive strength at elevated temperature in comparison with the other concrete composites, owing to the presence of polypropylene ﬁbres no explosive failure occurred during the splitting tensile test. Moreover, polypropylene ﬁbres reduce the risk of concrete spalling which is usually the main cause of concrete structure collapse. The experimental investigation also shows that a tensile failure mode of hot specimens and specimens cooled down after ﬁre exposure might be different. Further within the ongoing project, the obtained results will be validated on real-size concrete elements.
3. Conclusion The paper presents the case study of the behaviour of the concrete composites under ambient and elevated temperature with the aim to determine the mechanical properties of the materials. The experimental tests were conducted on three investigated concrete composites with a various type and content of ﬁbrous reinforcement - plain concrete (PC), steel ﬁbre reinforced concrete (SFRC) and hybrid ﬁbre reinforced concrete (HFRC) with addition of steel and polypropylene ﬁbres. Within the scope of work the heat transport test, compressive and splitting tensile tests were conducted at ambient temperature and elevated temperatures 75
J. Novak, A. Kohoutkova
Fire Safety Journal 95 (2018) 66–76
This work was supported by the grant Models of steel and ﬁber composite columns exposed to ﬁre, no. GACR 15-19073S, from the Grant Agency of the Czech Republic (GACR).
References   Meacham, Brian, et al. “Fire and Collapse, Faculty of Architecture Building, Delft University of Technology: Data Collection and Preliminary Analyses.” http://www. dr.dk/NR/rdonlyres/FECDDBA8-7428-460D-846F-0C3D424D3A81/2748348/ Delft_brandrapport.pdf (accessed March 30, 2016).  I. Fletcher, S. Welch, J.A. Capote, D. Alvear, M. Lazaro, Model-based analysis of a concrete building subjected to ﬁre, in: Advanced Research Workshop on Fire Computer Modelling, Santander, 2007.  On Stop Shop in Structural Fire Engineering. http://www.mace.manchester.ac.uk/ project/research/structures/strucﬁre/CaseStudy/HistoricFires/BuildingFires/ default.htm (accessed March 31, 2016).  IrJ.F. Deno€ el, Fire Safety and Concrete Structures, FEBELCEM, 2007.  CSN EN 1991-1-2, Eurocode 1: Actions on Structures - Part 1-2: General Actions Actions on Structures Exposed to Fire, Cesky normalizacni institut, November 2006.  Qianmin Ma, Rongxin Guo, Zhiman Zhao, Zhiwei Lin, Kecheng He, Mechanical properties of concrete at high temperature—a review, Constr. Build. Mater. 93 (14 June 2015) 371–383.  C.S. Poon, Z.H. Shui, L. Lam, Compressive behavior of ﬁber reinforced highperformance concrete subjected to elevated temperatures, Cem. Concr. Res. 34 (13 February 2004) 2215–2222.  Jaroslava Drchalov a, Eva M nahoncakova, Roman Vejmelka, Jirí Kolísko,
Patrik Bayer, Robert Cerný, Hydric and mechanical properties of carbon ﬁber reinforced cement composites subjected to thermal load, Constr. Build. Mater. 18 (6 April 2004) 567–578.  A. Noumowe, Mechanical properties and microstructure of high strength concrete containing polypropylene ﬁbres exposed to temperatures up to 200 C, Cem. Concr. Res. 35 (15 March 2005) 2192–2198.  A. Lau, M. Anson, Effect of high temperatures on high performance steel ﬁbre reinforced concrete, Cem. Concr. Res. 36 (23 March 2006) 1698–1707.  Gai-Fei Peng, Wen-Wu Yang, Jie Zhao, Ye-Feng Liu, Song-Hua Bian, Li-Hong Zhao, Explosive spalling and residual mechanical properties of ﬁber-toughened highperformance concrete subjected to high temperatures, Cem. Concr. Res. 36 (14 December 2006) 723–727.  Harun Tanyildizi, Statistical analysis for mechanical properties of polypropylene ﬁber reinforced lightweight concrete containing silica fume exposed to high temperature, Mater. Des. 30 (20 January 2009) 3252–3258.  K.K. Sideris, P. Manita, E. Chaniotakis, Performance of thermally damaged ﬁbre reinforced concretes, Constr. Build. Mater. 23 (27 September 2009) 1232–1239.  Piti Sukontasukkul, Worachet Pomchiengpin, Smith Songpiriyakij, Post-crack (or post-peak) ﬂexural response and toughness of ﬁber reinforced concrete after exposure to high temperature, Constr. Build. Mater. 24 (24 April 2010) 1967–1974.  Yuh-Shiou Tai, Huang-Hsing Pan, Ying-Nien Kung, Mechanical properties of steel ﬁber reinforced reactive powder concrete following exposure to high temperature reaching 800 C, Nucl. Eng. Des. 241 (1 April 2011) 2416–2424.  Yining Ding, Cecília Azevedo, J.B. Aguiar, Said Jalali, Study on residual behaviour and ﬂexural toughness of ﬁbre cocktail reinforced self compacting high performance concrete after exposure to high temperature, Constr. Build. Mater. 26 (2 July 2011) 21–31.  Ahmet Çavdar, A study on the effects of high temperature on mechanical properties of ﬁber reinforced cementitious composites, Compos. Part B 43 (19 October 2012) 2452–2463.  Wenzhong Zheng, Haiyan Li, Ying Wang, Compressive behaviour of hybrid ﬁberreinforced reactive powder concrete after high temperature, Mater. Des. 41 (21 May 2012) 403–409.  Wenzhong Zheng, Haiyan Li, Ying Wang, Compressive stress–strain relationship of steel ﬁber-reinforced reactive powder concrete after exposure to elevated temperatures, Constr. Build. Mater. 35 (15 June 2012) 931–940.  K.K. Sideris, P. Manita, Residual mechanical characteristics and spalling resistance of ﬁber reinforced self-compacting concretes exposed to elevated temperatures, Constr. Build. Mater. 41 (11 January 2013) 296–302.  G.M. Chen, Y.H. He, H. Yang, J.F. Chen, Y.C. Guo, Compressive behavior of steel ﬁber reinforced recycled aggregate concrete after exposure to elevated temperatures, Constr. Build. Mater. 71 (7 September 2014) 1–15.  Arabi N.S. Al-Qadi, Sleiman M. Al-Zaidyeen, Effect of ﬁbre content and specimen shape on residual strength of polypropylene ﬁbre self-compacting concrete exposed
   
to elevated temperatures, J. King Saud Univ. – Eng. Sci. 26 (10 December 2014) 33–39. Prakash S. Bhat, Vivian Chang, Mo Li, Effect of elevated temperature on strainhardening engineered cementitious composites, Constr. Build. Mater. 69 (15 August 2014) 370–380. Aminuddin Jameran, Izni S. Ibrahim, Siti Hamizah S. Yazan, Siti Nor A.A. Rahim, Mechanical properties of steel-polypropylene ﬁbre reinforced concrete under elevated temperature, in: Procedia Engineering (125) - the 5th International Conference of Euro Asia Civil Engineering Forum (EACEF-5), 2015, pp. 818–824. Jihwan Kim, Gyu-Phil Lee, Do Young Moon, Evaluation of mechanical properties of steel-ﬁbre-reinforced concrete exposed to high temperatures by double-punch test, Constr. Build. Mater. 79 (23 January 2015) 182–191. Faiz Uddin Ahmed Shaikh, Anwar Hosan, Mechanical properties of steel ﬁbre reinforced geopolymer concretes at elevated temperatures, Constr. Build. Mater. 114 (31 March 2016) 15–28. D. Choumanidis, E. Badogiannis, P. Nomikos, A. Soﬁanos, The effect of different ﬁbres on the ﬂexural behaviour of concrete exposed to normal and elevated temperatures, Constr. Build. Mater. 129 (3 November 2016) 266–277. Ruben Serrano, Alfonso Cobo, María Isabel Prieto, María de las Nieves Gonzalez, Analysis of ﬁre resistance of concrete with polypropylene or steel ﬁbers, Constr. Build. Mater. 122 (9 July 2016) 302–309. Wenzhong Zheng, Baifu Luo, Ying Wang, Stress–strain relationship of steel-ﬁbre reinforced reactive powder concrete at elevated temperatures, Mater. Struct. 48 (18 April 2015) 2299–2314. Farhad Aslani, Bijan Samali, High strength polypropylene ﬁbre reinforcement concrete at high temperature, Fire Technol. 50 (4 March 2014) 1229–1247. Farhad Aslani, Bijan Samali, Constitutive relationships for steel ﬁbre reinforced concrete at elevated temperatures, Fire Technol. 50 (4 September 2014) 1249–1268. Xiangjun Dong, Yining Ding, Tianfeng Wang, Spalling and mechanical properties of ﬁber reinforced high-performance concrete subjected to ﬁre, J. Wuhan Univ. Technology-Mater. Sci. Ed (October 2008) 743–749. G.M. Giaccio, R.L. Zerbino, Mechanical behaviour of thermally damaged highstrength steel ﬁbre reinforced concrete, Mater. Struct. 38 (April 2005) 335–342. Wasim Khaliq, Venkatesh Kodur, Thermal and mechanical properties of ﬁber reinforced high performance self-consolidating concrete at elevated temperatures, Cem. Concr. Res. 41 (24 June 2011) 1112–1122. Sriskandarajah Sanchayan, Stephen J. Foster, High temperature behaviour of hybrid steel–PVA ﬁbre reinforced reactive powder concrete, Mater. Struct. 46 (30 January 2016) 769–782. Jihwan Kim, Gyu-Phil Lee, Do Young Moon, Evaluation of mechanical properties of steel-ﬁbre-reinforced concrete exposed to high temperatures by double-punch test, Constr. Build. Mater. 79 (23 January 2015) 182–191. P. Bamonte, P.G. Gambarova, Thermal and mechanical properties at high temperature of a very high-strength durable concrete, J. Material Civ. Eng. (June 2010) 545–555. P. Bamonte, P.G. Gambarova, A study on the mechanical properties of selfcompacting concrete at high temperature and after cooling, Mater. Struct. 45 (16 February 2012) 1375–1387. Zhi Xing, Anne-Lise Beaucour, Ronan Hebert, Albert Noumowe, Beatrice Ledesert, Inﬂuence of the nature of aggregates on the behaviour of concrete subjected to elevated temperature, Cem. Concr. Res. 41 (11 January 2011) 392–402. CSN EN 197–1, Cement - Part 1: Composition, Speciﬁcations and Conformity Criteria for Common Cements, Cesky normalizacni institut, April 2012. Guncheol Lee, Dongyeop Han, Min-Cheol Han, Cheon-Goo Han, Ho-Jeong Son, Combining polypropylene and nylon ﬁbers to optimize ﬁber addition for spalling protection of high-strength concrete, Constr. Build. Mater. 34 (2012) 313–320. M.B. Dwaikat, V.K.R. Kodur, Hydrothermal model for predicting ﬁre-induced spalling in concrete structural systems, Fire Saf. J. 44 (2009) 425–434. M.B. Dwaikat, V.K.R. Kodur, Hydrothermal model for predicting ﬁre-induced spalling in concrete structural systems, Fire Saf. J. 44 (2009) 425–434. Robert Rowalski, Mechanical properties of concrete subjected to high temperature, Archit. Civ. Eng. Environ. (February 2010) 61–70. CSN EN 12390–12393, Testing Hardened Concrete Part 3: Compressive Strength of Test Specimens, Urad pro technicku normalizaci, metrologii a statni zkusebnictvi (UNMZ), 2009. A. Fina, F. Cuttica, G. Camino, Ignition of polypropylene/montmorillonite nanocomposites, Polym. Degrad. Stab. 97 (2012) 2619–2626. CSN EN 12390–12396, Zkouseni Ztvrdleho Betonu: Cast 6-Pevnost V Pricnem Tahu Zkusebnich Teles, Urad pro technicku normalizaci, metrologii a statní zkusebnictví (UNMZ), 2010. International Federation for Structural Concrete (ﬁb), Fib Model Code for Concrete Structures 2010, Wilhelm Ernst & Sohn, Verlag für Architektur und technische Wissenschaften GmbH & Co. KG, Berlin, Germany, 2013.