Assessment of Concrete Susceptibility to Fire Spalling: A Report on ...

Available online at www.sciencedirect.com

ScienceDirect Procedia Engineering 108 (2015) 285 – 292

7th Scientific-Technical Conference Material Problems in Civil Engineering (MATBUD’2015)

Assessment of concrete susceptibility to fire spalling: A report on the state-of-the-art in testing procedures Katarzyna Krzemieńa,*, Izabela Hagera a

Cracow University of Technology, Warszawska 24, 31-155 Kraków, Poland

Abstract

The assessment of concrete behavior at high temperature is done by a variety of tests carried out on specimens of different sizes. Small-scale tests examine concrete's behavior when exposed to elevated temperature, while full-scale fire tests are carried out on full-sized concrete elements in which the boundary conditions, external load and conditioning correspond to design assumptions. Complementary to these is the medium-scale test carried out on a portion of a slab's surface area which has been exposed to fire, ca. 1m2. Such medium-scale tests are often used as a cost-effective solution to verify the behavior of a specific concrete mix in fire conditions. This paper reviews the existing furnaces, testing procedures and laboratory setups used to assess a material's tendency to spall. Its objective is to emphasize the need to unify spalling risk assessment procedures by establishing recommended guidelines for testing. © 2015 The Authors. Published by Elsevier Ltd. © 2015 Theand Authors. Publishedunder by Elsevier Ltd. This of is an open access article under license Selection peer-review responsibility organizing committee of the theCC 7thBY-NC-ND Scientific-Technical Conference Material (http://creativecommons.org/licenses/by-nc-nd/4.0/). Civilresponsibility Engineering.of organizing committee of the 7th Scientific-Technical Conference Material Problems in Civil Engineering Problems inunder Peer-review Keywords: fire concrete spalling; concrete structures; assessment; testing setups

1. Introduction Laboratory tests of concrete which aim to investigate its susceptibility to fire spalling employ various testing procedures carried out on specimens of different sizes and shapes. Due to the lack of standardized testing guidance, there is a wide range of approaches to concrete fire spalling assessment. During such tests, different measurements

* Corresponding author. Tel.: +48-12-628-2371. E-mail address: [email protected]

1877-7058 © 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license

(http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of organizing committee of the 7th Scientific-Technical Conference Material Problems in Civil Engineering

doi:10.1016/j.proeng.2015.06.149

286

Katarzyna Krzemień and Izabela Hager / Procedia Engineering 108 (2015) 285 – 292

are carried out in order to better describe the processes taking place in concrete during heating. These parameters are temperature and vapor pore pressure development, thermal strains or specimen deflection. Numerous experimental tests results can be found in the literature, all attempting to investigate the fire spalling phenomenon by indicating different parameters that may enhance spalling risk, such as the mix composition, the heating scenario, the initial water content and the geometry of specimen or mechanical boundary conditions. Unfortunately, the influence of specific spalling parameters is difficult to assess and compare with others due to differences in the testing procedures used to obtain these results. Generally, we can distinguish three main categories of testing methods for spalling behavior investigations: small-scale, medium-scale and full-scale tests. The subsequent paragraphs describe the diversity of furnaces, their components, testing instrumentations and procedures. 2. Experimental methods for the assessment of concrete susceptibility to fire spalling 2.1. Small-scale concrete spalling tests Small-scale tests examine the material's behavior when exposed to elevated temperature. These tests are carried out on small concrete specimens: prisms, cubes or cylinders with the volume not exceeding ca. 4000 cm 3. High temperature conditions are mostly provided by electrical heating coils. In these tests only the spalling occurrence is detected and the number of specimens which spall is recorded. In more elaborate techniques, the specimens are mechanically loaded during heating. In most cases, the furnace is placed in a loading ram along with a concrete specimen which is stressed during heating, ex. Connolly [1], Phan [2], Hager and Pimienta [3], Phan [4], Mindeguia [5], Huismann et al. [6]. Following the recommendation of RILEM TC 200-HTC [7], two test methods can be distinguished in which the material's behavior is studied in stressed, and unstressed conditions. The stressed test method corresponds to conditions in which the specimen is uniaxially loaded during heating. A compressive load of 10–50 % of ultimate stress is applied to the specimen at room temperature, which remains constant during heating to the target temperature level (T). All spalling events are recorded as well as the temperature at its occurrence. Although these methods were mainly used to determine the material properties at the hot stage, the concrete's susceptibility to spalling can be also assessed using these procedures. The testing set developed by Connolly [1] employs cylindrical specimens of Ø 150 mm, H 100 mm. The specimen is mounted in steel ring and is loaded peripherally by hydraulic arms in both vertical and horizontal directions. The loading arms are also designed to restrain thermal expansion. Both the load and thermal expansion are recorded within the event by a load cell. Structural stability is obtained by the stiff restraint frame, which provides support for hydraulic jacks. Heating is delivered by electrical radiative heating elements by Kanthal Electrical Ltd., producing a heat flux level of 150 kW/m2. The temperature was measured with the use of thermocouples cast at different depths of the specimen, which was also equipped with a ceramic pipe connected to pore pressure transducers capable of reading the pressure within concrete up to 10 N/mm2. The assessment of spalling, if such occurred, was determined by counting the number of steel grid squares (10 x 10 mm) placed over the concrete surface which were damaged more than 50%. By this procedure the authors obtained the extent of spalling expressed in a percentage of the total surface area. The PTM test developed by Kalifa [8] is so far one of the most referenced and cited test methods. The examination of specimens – recording prisms done with additional pressure (P), temperature (T) and mass loss (M). These specimens can be instrumented with six gauges made of a sintered metal round plate, which are placed at casting. The last one is brazed to a thin metal tube (inner diameter 1.6 mm), which comes out of the rear face of the specimen. At the time of testing, a tight connector is placed at the free end of the tube. Firstly, it connects the gauge to a piezoelectric pressure transducer by means of a flexible tube filled with silicone oil. Secondly, a thermocouple is inserted in the tube through the connector down to the metal plate. The free volume of the gauge is around 250 mm3. The specimen, a prism (300 mm x 300 mm x 120 mm), is positioned horizontally. Its upper

Katarzyna Krzemień and Izabela Hager / Procedia Engineering 108 (2015) 285 – 292

surface is directly exposed to elevated temperature, such as a heating rate of 100 °C / hour, from a radiant electric heater (quasi-unidirectional thermal load), whereas the other surfaces are thermally insulated, Fig. 1a.

a)

b) Fig. 1. The experimental setups a) PTM test, after Kalifa [8], b) testing restrained concrete, after Hertz and Sørensen [9].

Similarly, Phan [4] proposes measuring pore pressure development inside a concrete specimen carried out on concrete blocks (100 x 200 x 200 mm3) which are molded with pressure gauges and thermocouples at different depths (13 mm, 25 mm, 50 mm and 75 mm from the heated surface). The blocks are insulated on all sides, except for one which is subjected to heating and placed inside the same electric furnace. One-sided heating is believed to reflect one-dimensional heat flow. The specimen is then exposed to a heating rate about 5 °C/min or 25 °C/min. In Kalifa's [8] and Phan’s [4] tests the heating rate was designed in a way that the spalling does not occur during the test, the test main goal being to measure internal pore pressure development along with temperature at different depths. Another approach to concrete spalling assessment is to mold a specimen in a steel mantle or steel ring and perform the restrained conditions for concrete surface, ex. Hertz and Sørensen [9], Tanibe et al. [10]. Detailed testing setups and procedures are provided below. The spalling investigations performed by Hertz and Sørensen [9] consist in tests carried out on cylindrical specimens Ø 150 mm, H 300 mm that are placed in a steel mantle, Fig. 1b. The two parts of a steel mantle (upper and bottom) are connected with the use of 12 bolts Ø 36 mm. The thin space between mantle and specimen is filled with a neoprene in order to reduce irregularities of the concrete surface. One plane side of the specimen is exposed to a temperature of 1000 °C from an electrical oven by a 100 mm hole. The steel mantle is believed to counteract the thermal stresses that occur in concrete specimens. By the use of a steel mantle, the concrete cylinder is restrained in its peripheral, while the cylindrical base of the specimen is exposed to fire. Thus, the conditions taking place in the cylindrical specimen reflect that of a concrete wall exposed to fire. Susceptibility to spalling is assessed by measuring the area of spalling. In addition, the specimen's acoustic emission is recorded in order to determine the time of spalling event. In Tanibe et al. [10] the experimental setup was designed to monitor the internal temperature, vapor pressure and restrained stress of the concrete specimen. A circular specimen of 284 mm in diameter and 100 mm in height was molded in two steel rings of 300 mm in external diameter, 50 mm in height and 8 mm in thickness. The modulus of elasticity Es and yield strength of the steel ring fy are known (Es = 210 GPa, fy = 295 MPa). The steel rings provide restraint for the concrete and enable its thermal strains to be measured by exposing the concrete surface to fire at a heating curve of RABT 30. The thermal strains are then monitored by recording the strains of the steel ring, two strain gauges being attached to its external surface (the thermal strain of concrete equals the measured stress reduced by the thermal strain of steel at a particular temperature). In order to measure the temperature profile inside a concrete specimen, six thermocouples are molded at different depths from the surface exposed to the fire. Additionally, two thermocouples are placed in the outer surface of steel ring, where the thermal strains are examined. Vapor pressure measurement is provided by the pore pressure setup, which consists of a steel pipes molded into the concrete specimen at depth of 10 mm and 20 mm and connected to pressure transducer located outside the furnace. The aim of this experimental test is to measure the thermal strains of concrete and

287

288

Katarzyna Krzemień and Izabela Hager / Procedia Engineering 108 (2015) 285 – 292

assess any spalling that might occur. Spalling assessment is carried out by measuring the depth of spalling in the fire-exposed area and monitoring the moment of spalling at the time of testing. Table 1. Summary of reviewed small-scale experimental setups. Reference

Specimens type and dimensions [mm]

Set

Heating method

Heating curve

External load

additional measurements

Connolly [1]

cylindrical Ø 150 H 100

vertical

electrical

linear

vertical horizontal

T, P, thermal stresses, area of spalling

Kalifa [8]

prism 300 x 300 x 120

horizontal

electrical

100 °C/hour

-

T, P, mass loss

Hertz and Sørensen [9]

cylindrical Ø 150 H 300

horizontally

electrical

linear max 1000 °C

restrained

T, AE, thermal stresses area of spalling

Phan [4]

prism 100 x 200 x 200

vertical

electrical

5 °C/min 25 °C/min

-

T, P

Tanibe et al. [10]

cylindrical Ø 284 H 100

vertical

electrical

RABT 30

restrained

T, P, restrained stress thermal strains, depth of spalling

T – temperature

P – vapor pore pressure

AE – acoustic emission

2.2. Medium-scale furnaces Medium-scale tests are performed as a screening tests in order to verify the specific concrete mixture's behavior in fire conditions and select the one which does not tend to spall. These types of tests are cost effective when compared to the full scale tests performed on structural elements. The results of a medium scale test allow one to preselect the concrete mixes that are not susceptible to spall. This approach is intended to limit the number of specimens employed in full scale testing and thus to increase its cost-effectiveness. Concrete spalling due to fire is mainly evaluated in two ways. The first consists of tests carried out during high temperature exposure where the following measurements are made during the spalling event, i.e. its beginning, duration and frequency. Acoustic emission methods, such as Huismann et al. [11], allow the spalling to be characterized by counting the number of acoustic events, whereas the method reported by Carré et al. [12] employs a digital camera to register the volume and time of spalling. Moreover, there are other experimental techniques which aim to explain the resulting physical phenomena, such as internal vapor pore pressure or moisture profiles evaluation. These can be conducted during the heating of the specimen. The second approach to spalling effect characterization relies on measurements taken after the test's completion, i.e. the size, shape and amount (mass, volume) of the spalled concrete material. The extent of the spalling and its quantification may be presented as a spalling map, a diagram of maximum and average spalling depth or even spalling topography. Such medium-scale experiments are mostly carried out on a concrete slab whose average size, 1000 mm in length x 800 mm in width x 200 mm in thickness, does not exceed a volume of 1.5 m3. Depending on the testing set, specimens are placed horizontally on the top of furnace chamber Carré et al. [12], Heel and Kusterle [13], Jansson and Boström [14] or mounted vertically in the furnace wall Huismann et al. [11]. During the test, temperature development is measured with the use of thermocouples molded inside the concrete specimen, as well as being placed near the fire exposed surface. The high temperature conditions are provided mostly by propane or oil burners. The controlled gas or oil flows provide different fire scenarios, i.e. ISO 834-1, HC, RABT and RWS fire curve. In most cases, the furnace along with the specimen is placed in a loading ram and the concrete specimen is stressed during heating. Both unloaded tests by Mindeguia [5], Huismann et al. [11], Iglesias [15], Hager et al. [16], and loaded tests by Carré et al. [12], Heel and Kusterle [13] and Boström and Jansson [17] can be found in the literature. Compression can be applied in different ways. The first is to use post tensioning bars molded into specimen or mounted on the external perimeter of the slab. Another solution employs flat jacks or a prestressing clamping frame aimed at loading the slab in a uniaxial or biaxial compression. Example of testing sets and procedures are described below, distinguishing between unloaded and loaded tests.

Katarzyna Krzemień and Izabela Hager / Procedia Engineering 108 (2015) 285 – 292

2.2.1. Unloaded tests A specimen placed horizontally on the top of the furnace is subjected to a dead load, without additional external loading. The temperature in the slab is measured by thermocouples molded during casting and in the furnace with a flat thermocouple. This test does not provide external load and the concrete specimen is molded without reinforcement. Such a solution was employed in VersuchsStollen Hagerbach by Iglesias [15], in Cracow University of Technology by Hager, et al. [16] and in Politecnico di Milano. The Dragon furnace developed and used in Cracow University of Technology consists of a steel shell and an internal fireproof lining. The specimen, a concrete slab (1200 mm x 1000 mm x thickness 300 mm), was placed horizontally on the top of furnace and exposed to a fire action surface of ca. 750 mm x 950 mm. The furnace has two ventilation pipes 120 mm in diameter with draft regulators. The fire conditions are provided by gas burners (140 kW) fueled with propane-butane. Specimens are subjected to a time – temperature scenario, which is presently an ISO 834-1 curve. The BAM furnace developed by Huismann et al. [11] used for vertical investigation of a concrete slab 1000 mm x 1000 mm x 1000 mm runs with two oil burners that provide a fire scenario on the hydrocarbon (HC) fire curve. The fire exposed surface of the vertically placed concrete specimen is 500 mm x 500 mm. The furnace temperature is controlled by insulated thermocouples. During fire tests the temperature is measured inside the furnace as well as inside the specimen, an AE analysis is also being performed. Sensors and amplifiers are connected at five points on the unexposed surface. Additionally, the pore pressure is measured by pressure gauges which have been cast into the concrete. These are connected to the pressure sensors and the thermocouples. The entire setup does not provide a mechanical load. The spalling phenomenon, if it occurs, is described at the time of spalling and is presented in three ways – by means of Acoustic Emission within the test, by a diagram of the maximum and average spalling depth and by spalling topography performed after completing the test. 2.2.2. Loaded tests The experimental setup proposed by Heel and Kusterle [13] provides tests on flat slabs with the dimensions of 1400 mm x 1800 mm x 500 (or 300) mm under an RWS fire scenario and loading conditions. The loading is provided by selecting either of two available systems: unbounded prestressing tendons or a prestressing clamping frame. The two-stage oil burner placed in the central part of the furnace housing's shorter side is controlled on the basis of digitally recorded temperatures. The fire exposure area is 800 mm x 1200 mm. Externally applied compressive force is performed as follows: the main loading force of 1.16 MPa or 9 MPa in a transverse direction and an accompanying force of 0.5 MPa in a longitudinal direction in order to reduce slab deformation. The tensile stress in the tendons induces compressive stress in the specimens, and is controlled by a readout of the prestressing force indicated on the prestressing jacks. Additionally, a manometer is used to measure increase of pressure in tendon ducts caused by thermal expansion of anti-corrosion grease protecting tendons. The mix composition, reinforcement and curing conditions are consistent with the design of a real structure. During the test, temperature is measured by thermocouples placed centrally at 10 mm and 40 mm from the fire exposed surface. For spalling assessment, temperature and pore pressure development is recorded at time of testing. When a spalling event takes place, the values of temperature, pore pressure and loading force are recorded. Additionally, after test completion, the maximum and average spalling depth is measured. The setup developed by Jansson and Boström [14] allows the testing of concrete slab specimens molded without reinforcement with dimensions of 600 mm x 500 mm x 200 mm. The slab is placed centrally on the top of furnace and is subjected to a hydrocarbon (HC) fire curve or linear heating with the rate of 10 °C/minute. Three aluminum pipes are placed in each specimen into which post-stress bars can be placed after the casting (Fig. 2a). Posttensioning bars are used to apply the external compressive load. For the purpose of measuring temperature, thermocouples are centrally placed in each specimen at a depth of 10 mm and 40 mm from the fire exposed surface. The loading system developed by Jansson and Boström [14] is performed with a post tensioning system made of four Dywidag bars, Ø 28 mm, placed horizontally inside the concrete slab. The load level of the post-

289

290

Katarzyna Krzemień and Izabela Hager / Procedia Engineering 108 (2015) 285 – 292

stressing force is 10 % of the cube compressive strength tested prior to the fire test. The resulting compressive force is transferred to the concrete specimen by the steel plates fixed to its loaded side. The tensile stress in the bars, which induces compressive stress in the concrete specimens, is controlled with load cells connected to the MGC Plus system and computer for data acquisition. Load can also be applied by externally mounted post-stressed bars (Fig. 2b), which allows the influence of the steel pipes' thermal expansion to be limited. In such cases, the concrete slab is molded without reinforcement and aluminum pipes. During the test, temperature development and internal pore pressure are recorded. When spalling occurs, the corresponding values of T and P are registered, further spalling quantification being performed afterwards. To assess tendency to spall, average and maximum spalling depth is measured and a 3D spalling profile is prepared.

Fig. 2. Testing sets in medium scale: a) with the inner post-tensioning bars, b) right with the outer post-tensioning bars, after Jansson and Boström [14], c) with 4 flat hydraulic jacks, after Carré et al. [12]. Table 2. Summary of reviewed medium-scale experimental loaded sets. Reference Heel, Kusterle [13] (A) Heel, Kusterle [13] (B) Boström, Jansson, [17] Carré, et al. [12]

Fire exposed area [mm x mm]

Fire curve

Reinforce ment

Location of load

Type of load

Fuel

Stress state

Additional measurements

1400 x 1800

RWS

yes

internally

post stressing bars

oil

compression

T, P, spalling depth

1400 x 1800

RWS

yes

externally

oil

biaxial compression

T, P, spalling depth

600 x 500

ISO 834-1 10 °C/min

no

internally

gaspropane

compression

580 x 680

ISO 834-1

no

externally

gaspropane

uniaxial and biaxial compression

T – temperature

P – vapor pore pressure

prestressing clamping frame post tensioning bars flat jacks

T, P, spalling depth and profile T, digital camera, acoustic events

AE – acoustic emission

In the CSTB Laboratory, tests are carried out according to an ISO 834-1 time-temperature curve, Carré et al. [12]. The external part of the furnace is made of steel whereas the internal lining is made of high temperature resistant bricks. The fire source consists of 8 gas burners using propane. This setup tests concrete slab specimens (680 mm x 580 mm x thickness (ex. 150 mm)) placed horizontally on the furnace and subjected to fire from beneath, the fire exposed surface being 600 mm by 420 mm. The examined slab is loaded horizontally with flat jacks (type Freyssinet), which enable uniaxial or biaxial stress state application during the test (2 or 4 flat jacks respectively, Fig. 2c). Thermocouples inside the furnace, located 10 cm from the exposed part of the slab, are hand controlled and monitored during the test, while pressure gauges are employed at various depths - 10 mm, 20 mm

Katarzyna Krzemień and Izabela Hager / Procedia Engineering 108 (2015) 285 – 292

and 30 mm from the exposed surface. During the fire test, spalling events are observed through a small hole and recorded using a digital camera. In such tests, acoustic events are recorded using a microphone located near the specimen. In order to secure a stress state provided by flat jacks, a reference test without thermal load is made using strain gauges placed on the specimen's surface. 3. Full scale fire tests Full scale test setups allow tests to be performed on concrete elements (beams, columns walls, tunnel linings) by applying external load and choosing various fire scenarios. Full scale fire tests which permit real structural members to be tested are carried out on the fully sized concrete elements in which the boundary conditions, external load and conditioning correspond to design assumptions. Full scale tests, being the most representative, directly present the qualitative and quantitative behavior of structural concrete element subjected to fire. Since, the description of testing procedure of a particular setup strongly depends on the testing configuration, this paper describes two testing setups as examples of full scale tests. The furnace described by Richter [18] is used for examination of tunnel segments. The internal dimensions of the furnace chamber were: 2600 mm in width, 3900 mm in length and 1300 mm in height. The specimen is placed on the top of furnace and subjected to any fire curve. The specimen is heated with seven oil burners, two on each wall except for one longitudinal wall, where only one is installed. The boundary conditions as well as the load can be adjusted to needs - restrictions of elongations and additional vertical (6 hydraulic jacks) or horizontal (2 hydraulic jacks) loads can be applied. During the tests temperature development can be measured 100 mm from the fire-exposed surface at three points. The vertical deflections of specimen and horizontal displacement at the support can be recorded continuously. Spalling and cracking behavior as well as the emergence of moisture on the unexposed surface of the tested element can be checked visually during the heating. If spalling occurs, the temperature accompanying this event is noted. After the test, spalling is quantified visually by observations of micro-cracks on the fire-exposed surface and detailed photographic evidence of particular specimen. The furnace used in SP Technical Research institute of Sweden by Boström and Jansson [17] is a large horizontal furnace with a clear opening of 5000 mm x 3000 mm. It is used for the examination of different structural members, beams, columns, slabs, etc. Oil is used for fire development with examinations done according to ISO 834-1 and HC fire curves. Specimens, such as slabs, are located horizontally and supported on the walls from two sides. During the test for slab elements, flat thermometers are placed 10 mm below the fire exposed concrete surface in order to measure temperature development. Additionally, in the cited work the measurements of internal pore pressure was performed along with temperature measurements at different element depths. The specimens were loaded with compressive load by post tensioning threaded bars (Dywidag 36 mm) in aluminum pipes, with load cells mounted in order to ensure that a correct load level was applied, thus enabling continuous measurement during the fire tests. Such full scale tests are the most representative, directly showing the qualitative and quantitative behavior of structural concrete elements subjected to fire. The disadvantage is that tests performed using large furnaces are both costly and are unsuitable for testing a large number of different concrete mixes. To overcome this limitation, and thanks to the dimensions of a large furnace (Mindeguia [5], Boström and Jansson [17], Taillefer, et al. [19]), many small-sized slabs can be tested at the same time, thereby making it possible to examine different concrete mixes, or various specimen sizes during a single test. 4. Summary Most experimental tests used for concrete spalling attempt to explain its causes and define the factors influencing its occurrence. However, for the purpose of assessing concrete's tendency to spall, researchers have considered different approaches. Experimental tests carried out on small specimens with the use of an electrical heater mostly provides information regarding the material's response to fire, while the results of medium and large-

291

292

Katarzyna Krzemień and Izabela Hager / Procedia Engineering 108 (2015) 285 – 292

scale tests take into consideration structural effects, such as loading, restraint and boundary conditions. Currently, there are no guidelines for medium-scale tests and the procedures for large-scale fire tests do not provide a means to quantify spalling and assess the damage extent. Accordingly, medium scale tests seem to be a good way to screen specific concrete mixtures for their susceptibility to fire spalling. Unfortunately, there is a lack of specified testing procedures, which means that the results from different setups and test procedures are often inconclusive and easily compared. Therefore, guidelines for optimal test conditions, equipment and specimen instrumentation should be introduced. Furthermore, testing procedures (i.e. loading application) should be described in detail, providing criteria to evaluate spalling. Specifically, the methods employed to quantify the spalling effect need to be more unified, which would enable the results of different research to be compared. Any guidelines established for medium and full scale tests should arise from the experience of researchers in this field. Another issue requiring in-depth discussion is the procedure for determining a specimen's moisture content before testing. Many researchers have pointed this out as a problem particularly in need of advised unification, Jansson [20]. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]

[11] [12] [13] [14] [15] [16] [17] [18] [19] [20]

Connolly R. The Spalling of Concrete in Fires. PhD Thesis. The University of Aston in Birmingham; 1995. Phan L. and Carino N. Effects of Test Conditions and Mixture Proportions on Behavior of High-Strength Concrete Exposed to High Temperatures. ACI Mater J 2002;99(1):54-66. Hager I. and Pimienta P. Mechanical properties of HPC at high temperature. Workshop: Fire Design of Concrete Structures: What now? What next?; 2004, p. 95-100. Phan T. High-strength Concrete at High Temperature - An Overview. International Conference on Construction Materials. Vancouver; 2005. Mindeguia JC. Contribution Expérimentale A La Compréhension Des Risques D'instabilité Thermique Des Bétons. PhD Thesis. Universitè de Pau et des Pays de l'adour; 2009. Huismann S, Weise F, Meng B. and Schneider U. Transient strain of high strength concrete at elevated temperatures and the impact of polypropylene fibers. Mater Structures 2012;45:793-801. RILEM TC 200-HTC. Mechanical concrete properties at high temperatures – modelling and applications. Part 2 : Stress-strain relation. Mater Structures 2007;40:855–64. Kalifa P, Chene G. and Galle C. High-temperature behaviour of HPC with polypropylene fibers: From spalling to microstructure. Cem Concr Res 2001;31, 1487-1499. Hertz K. and Sørensen L.. Test method for spalling of fire exposed concrete. Fire Saf J 2005;40, 466-476. Tanibe T, Ozawa M, Lustoza RL, Kikuchi K, Morimoto H. Explosive spalling behaviour if restrained concrete in the event of fire. In: E. A. Koenders, E.A. and Dehn, F. (Ed.), Concrete Spalling due to Fire Exposure - Proceeding of the 2nd International RILEM Workshop. Delft, The Netherlands: RILEM Publications S.A.R.L.;2011, p. 319-26. Huismann S, Korzen M, Weise F. and Meng B. Concrete spalling due to fire exposure and the influence of polypropylene fibres on microcracking. RILEM Publications, Delft; 2011. Carré H, Pimienta P, La Borderie C, Pereira F, Mindeguia JC. Effect of compressive loading on the risk of spalling. MATEC Web of Conferences 2013;6(01007). Heel A. and Kusterle W. Die Brandbeständigkeit von Faser-, Stahl- und Spannbeton [Fire resistance of fiber-reinforced, reinforced, and prestressed concrete] (in German), Tech. Rep. 544, Bundesministerium für Verkehr, Innovation und Technologie, Vienna; 2004. Jansson R. and Boström L. Spalling of concrete exposed to fire, SP Technical Research Institute of Sweden, Borås; 2008. Iglesias E. and Wetzig V. Influence of porosity and specimen stiffness on the fire resistance of pp fibre mixed concrete. ITA-AITES World Tunnel Congress, Budapest; 2009. Hager I, Tracz T. and Krzemień K. The usefulness of selected non-destructive and destructive methods in the assessment of concrete after fire. Cem. Wapno Beton 2014;3:145-51. Boström L. and Jansson R. Self-Compacting Concrete Exposed to Fire, Borås: SP Technical Research Institute of Sweden; 2008. Richter E. Fire test on single-shell tunnel segments made of a new high-performance fireproof concrete. Workshop: Fire Design of Concrete Structures: What now? What next?; 2004, p. 261-270. Taillefer N, Pimienta P, Dhima D. Spalling of concrete: A synthesis of experimental tests on slabs. MATEC Web of Conferences 2013;6(01008). Jansson R. Fire Spalling of Concrete. Theoretical and Experimental Studies. PhD Thesis. Borås, KTH Architecture and the Build Environment; 2013.