Revision of concrete performance at elevated temperatures: A critical review and initial results

ISSN 1859-1531 - TẠP CHÍ KHOA HỌC VÀ CÔNG NGHỆ ĐẠI HỌC ĐÀ NẴNG, SỐ 5(126).2018, Quyển 2 33 REVISION OF CONCRETE PERFORMANCE AT ELEVATED TEMPERATURES: A CRITICAL REVIEW AND INITIAL RESULTS ĐÁNH GIÁ LẠI SỰ LÀM VIỆC CỦA BÊ TÔNG Ở ĐIỀU KIỆN NHIỆT ĐỘ CAO: PHÂN TÍCH CHUYÊN SÂU VÀ CÁC KẾT QUẢ BAN ĐẦU Quang Xuan Le1,2, Dinh Ba Le1,2, Son Duy Tran1, Vinh The Ngoc Dao1* 1School of Civil Engineering, The University of Queensland, Australia 2University of Science and Technology - The University of Danang; quang.le@uq.edu.au; v.dao@uq.edu.au; badinh.le@uq.net.au; duy.tran@uq.net.au Abstract - The performance of concrete at elevated temperatures depends on its mix constituents and proportions as well as their complex physio-chemical transformations during fire exposures. In this paper, some major influencing factors and spalling of concrete at elevated temperatures will be first reviewed. Limitations of current test setups for fire test, with a focus on the questionable reliability of thermal boundary condition and deformation capturing, will then be discussed to highlight the need for improved setups. On that basis, details of a new test setup will be presented. It will be shown that the use of radiant burner system and DIC technique with band-pass filter and blue illumination effectively addresses the identified limitations of existing systems: While the radiant burner system allows us to generate known and consistent thermal boundary conditions, the adoption of DIC technique with band-pass filter and blue illumination enables it to reliably capture deformation of concrete surfaces at elevated temperatures in a full-field, non-contact manner. With this new test setup established, a comprehensive set of reliable data can now be further collected, forming a solid basis for revising concrete performance in fire, taking into account the effects of temperature and temperature gradients. Tóm tắt - Sự làm việc của bê tông ở điều kiện nhiệt độ cao phụ thuộc vào thành phần, hàm lượng cốt liệu cấu thành và các chuyển hóa phức tạp xảy ra khi bê tông tiếp xúc với lửa. Sự bong tróc bề mặt và yếu tố chính ảnh hưởng đến sự làm việc của bê tông sẽ được phân tích trong bài báo này. Sau đó, bài báo sẽ nêu bật những hạn chế của phương pháp thí nghiệm bê tông ở nhiệt độ cao hiện tại bao gồm: độ tin cậy của điều kiện biên nhiệt độ, và đo đạc biến dạng của bê tông ở nhiệt độ cao. Từ đó, phương pháp thí nghiệm sử dụng hệ thống tấm bức xạ nhiệt để tạo ra điều kiện biên nhiệt độ có đồ đồng nhất cao và sử dụng hệ thống camera để đo biến dạng của bê tông một cách không tiếp xúc sẽ được trình bày. Phương pháp thí nghiệm mới sẽ thu thập các số liệu đáng tin cậy nhằm đánh giá lại sự làm việc của bê tông ở nhiệt độ cao có kể đến ảnh hưởng của cả nhiệt độ và gradient nhiệt độ. Key words - concrete; elevated temperatures; incident heat flux; thermal boundary condition; concrete deformation Từ khóa - bê tông; nhiệt độ cao; thông lượng nhiệt; điều kiện biên nhiệt độ; biến dạng của bê tông. 1. Introduction The outbreak of a fire in buildings and civil engineering structures can lead to severe structural damage, significant loss of content and possible loss of life. Adequate design for fire is, therefore, an important and essential requirement in the design process. Being the most commonly used construction material, concrete has favorable inherent characteristics with respect to fire: (i) It has low thermal conductivity and high heat capacity; (ii) It is non-combustible; and, (iii) It does not emit smoke or toxic gases. High levels of fire resistance for traditional concrete structures can thus generally be achieved by adopting certain member dimensions and cover for reinforcement [1]. Nevertheless, concrete also exhibits some less attractive aspects when exposed to elevated temperatures, including degradation of material properties and spalling. Consequently, both the load-carrying and separating/ insulating functions of concrete structures could be compromised. This paper will first provide an overview of the performance of concrete at elevated temperatures, with a focus on major influencing factors and spalling. The paper will then highlight the need to better characterize concrete performance in fire. The limitations of conventional furnace tests, regarding both thermal boundary conditions and deformation capturing at elevated temperatures, will be discussed. On that basis, an improved test setup that effectively addresses the identified limitations will be presented, together with evidence giving confidence to the reliability of such a new setup. 2. Performance of concrete at elevated temperatures 2.1. Factors influencing the fire performance of concrete Key factors affecting the performance of concrete at elevated temperatures can be divided into material and environmental categories, as summarised in Figure 1. Figure 1. Factors influencing fire performance of concrete. 2.1.1. Material factors a. Concrete ingredients Concrete is a complex composite made from cementitious materials, water, aggregates, and admixtures. These 34 Quang Xuan Le, Dinh Ba Le, Son Duy Tran, Vinh The Ngoc Dao component ingredients interact to form hardened concrete that typically consists of three phases: aggregates, hydrated cement paste and the interfacial transition zone between those two phases. The behavior of concrete during both heating and cooling process is therefore dependent not only on the physio- chemical changes of each individual phase but also on the complex interaction between these changes [2-4]. b. Concrete strength The initial strength of concrete plays an important role in concrete structure performance at elevated temperatures. The change in mechanical properties of concrete at elevated temperatures may differ significantly between normal strength concrete (NSC) and high strength concrete (HSC): The loss of strength due to exposure to elevated temperatures of HSC is typically significantly higher than that of NSC [5-7]. c. Moisture content Moisture content has been recognized as one of the most important factors that influence the behavior of concrete at elevated temperatures [3, 6, 8-10]. If all else is comparable, test specimens with higher moisture content tend to have an increased propensity for spalling mainly due to the higher pore pressures [3, 6, 10]. 2.1.2. Environmental factors a. Types of test Many studies [5-7, 9-12] have demonstrated that strength of concrete at elevated temperature is affected by the type of test. Three types of tests are typically used; stressed test, unstressed test, and unstressed residual test: - In the unstressed test, the unloaded specimen is heated to the target temperature and then loaded to failure while hot. - By contrast, in the stressed test, the concrete specimen is heated to the target temperature while under loading. The loading is subsequently increased until specimen failure. - In the unstressed residual test, the concrete specimen is heated to the target temperature, then allowed to slowly cool to ambient temperature. The concrete specimens are then loaded until failure. b. Target temperatures The performance of concrete at a given target temperature is influenced by the combined and accumulated effects of physio-chemical changes occurring in the concrete. Table 1 summarizes the key physio- chemical changes of concrete at elevated temperatures. Table 1. Major physio-chemical changes of concrete components at elevated temperatures [9] Temperature (°C) Major physio-chemical changes of concrete components 20-80 Increase in hydration, slow capillary water loss and reduction in cohesive forces as water expands 100 Marked increase (of up to hundred-fold) in water permeability 80-200 Increase in the rate of loss of capillary water and then physically-bound water 80-850 Loss of chemically-bound water 150 Peak for the first stage of decomposition of calcium silicate hydrate (CSH) >300 Marked increase in porosity and micro-cracking 350 Break-up of some river gravel aggregates 374 Critical point of water when no capillary water is possible 400-600 Dissociation of Ca(OH)2 into CaO and water 573 α-β transformation of quartz in aggregates and sands 550-600+ Marked increase in ‘basic’ creep c. Heating rate – Temperature gradient Different heating rates undoubtedly lead to different temperature and moisture gradients within concrete test specimens, thereby causing different moisture losses, thermal stresses and pore pressures therein. Depending on the actual combined effects of these for specific cases, the mechanical properties of concrete specimens subject to faster heating rates can be higher [13, 14] or lower [15] than those subject to slower heating rates. Despite such significant effects of heating rates, currently available constitutive models for concrete have largely been derived from standardized tests where temperature gradients within the concrete test specimens is intentionally minimized [16, 17], with the aim of separating, as far as possible, the “material” effects from the “structural” effects [18]. Limitations of these constitutive models include: - Mass transfer processes affected by heat are different from typical real fire situations (with higher temperature gradients) because the very slow heating rates not only allow dissipation of heat through the specimen but also slow dissipation of water vapor with minimal pore pressure increase. - Components of the model linked with temperature gradients have not been explicitly addressed, nor have the couplings between different processes linked to temperature gradients (including moisture transport, vapor pressure and thermal gradient induced stresses). It is thus questionable whether such constitutive models developed based on tests under minimized temperature gradients are representative of concrete in structures with substantial temperature gradients, as typically the case for those in fire. 2.2. Spalling of concrete in fire Spalling is the violent or non-violent breaking off of layers or pieces of concrete from the surface when it is exposed to high and rapidly rising temperature as experienced in fire [19]. Notable features of concrete spalling upon heating are highlighted as follows: • There seems a general consensus that spalling is due ISSN 1859-1531 - TẠP CHÍ KHOA HỌC VÀ CÔNG NGHỆ ĐẠI HỌC ĐÀ NẴNG, SỐ 5(126).2018, Quyển 2 35 to a combination of thermal stresses, pore pressures, and load-induced stresses (Figure 2), although the relative importance of these contributing mechanisms is still debatable [20-22]. • Compared to normal-strength concrete, high-strength concrete is subjected to (i) higher thermal stresses because of its greater rigidity and higher coefficient of expansion; and (ii) higher pore pressures due to its lower degree of porosity and permeability that does not readily allow moisture to escape from the heated concrete. Accordingly, despite having higher tensile strength, high-strength concrete appears more susceptible to spalling than normal-strength concrete. • Pore pressure is influenced by concrete permeability, moisture condition, and heating rate. o Polypropylene fibers appear very effective in spalling mitigation due to their melting at about 160oC providing channels for moisture to escape, especially when the pore pressure mechanism governs. o The higher the water saturation level at fire exposure, the higher the risk of spalling: It seems thus desirable to dry the concrete after sufficient initial curing to achieve the required strength. Figure 2. Combination of compressive, thermal stress and pore pressure [21, 25] • Spalling results in loss of material, reduction in section size and early exposure of the reinforcing steel to excessive temperatures. The reduced performance of a concrete structure due to spalling can render fire design calculations inaccurate and result in significantly reduced levels of fire safety. Unfortunately, our current understanding of spalling remains inadequate mainly due to the highly complex nature of the phenomenon [16, 23, 24] and the lack of appropriate test methods to reliably quantify and separate the effects of influencing factors, as discussed further in the next sections. 3. Limitations of conventional fire testing of concrete material and structures 3.1. Establishing reliable thermal boundary condition The increase in temperature in a test specimen is a function of the incident heat flux on the specimen’s surface. In conventional furnace tests, this incident heat flux on the test specimen’s surface results from the combination of the convective and radiative heat flux from the gas phase, the radiative heat flux from the furnace walls and possibly also from the other test specimens (Figure 3). Figure 3. Schematic of key heat transfer processes in the conventional furnace [26] Through consideration of energy balance at the surface of specimen, the absorbed heat flux at the specimen surface (𝑞𝑠 ") can be approximated as: '' 4 4 , . . . ( ) . .s g s g g c g s s sq F T h T T T   = + − − (1) where, Fg,s is view factor of the compartment to specimen surface; εg: thermal emissivity of gases; Tg, Ts: temperature of the gas and of specimen surface, respectively (K); σ: Stefan-Boltzmann constant (5.67x10-8 W/m2K4); hc: average convective heat transfer coefficient by both radiation and convection modes of gas (W/m2K). The evolution of 𝑞𝑠 " in time and space is thus highly complex, and accordingly very difficult to be controlled accurately and consistently. This inconsistent thermal loading imposed on test specimens has serious implications for the case of concrete, where proper characterization of thermal boundary conditions is required [27, 28]. Poor definition of 𝑞𝑠 " makes it challenging to achieve reliable control of the temperature evolution as well as temperature gradients in test specimens in conventional furnace tests. This has contributed to the significant variation in test results regarding both strength deterioration and spalling of concrete upon heating [16, 18]. 3.2. Deformation capturing of concrete at elevated temperatures Reliable deformation measurement is required for proper quantification of fire performance of concrete structures. Predictive capability of models for many critical properties, including Young’s moduli, stress-strain relationships and load-induced thermal strains, is first and foremost dependent on such reliable deformation capturing. 36 Quang Xuan Le, Dinh Ba Le, Son Duy Tran, Vinh The Ngoc Dao However, conventional fire testing generally does not allow reliable direct measurement of deformation and strain of high-temperature concrete. It has been shown that measurement of structural displacements and strains has been mostly performed outside the heated zone in previous fire tests of concrete [29]. As a result, the validity of models for many critical properties, including Young’s moduli, stress-strain relationships and load-induced thermal strains, remains questionable. A major research program is thus ongoing at the University of Queensland to address the above-discussed limitations in testing of concrete at elevated temperatures, thereby allowing collecting reliable data for revision of concrete performance in fire taking on account of the effects of temperature and temperature gradients. 4. New test setup with improved thermal boundary conditions and reliable deformation capturing of concrete at elevated temperatures 4.1. Reliable thermal boundary condition A novel test method, by using high-performance radiant burners as shown in Figure 4, has been proposed to establish known and consistent thermal boundary condition on test specimens [30, 31]. The radiant burners can be arranged to create a known incident heat flux on different types and sizes of sample surfaces. Besides, the intensity of incident heat flux can be adjusted by changing the distance between the radiant burners to sample surfaces. This test setup also enables the simultaneous application of thermal loading and other types of loadings, allowing us to test specimens under various stress- temperature paths. Figure 4. Radiant burners array Figure 5. Thermal boundary condition on the cylindrical specimen By adopting this novel testing method, the setup of using four radiant burners has been successfully used to generate well-defined, homogeneous and reproducible heating regimes on concrete cylinder specimens (Figure 5), enabling us to re-examine the effects of temperature and temperature gradient on mechanical properties and spalling of concrete [32, 33]. The good repeatability, consistency, and uniformity of the thermal boundary conditions on test cylinder specimens have been clearly demonstrated [32, 33]. 4.2. Deformation capturing of concrete at elevated temperatures A review of major relevant existing methods has shown that the Digital Image Correlation (DIC) technique is one of the most promising techniques for deformation capturing of concrete surfaces at elevated temperatures [34]. The question then is how the remaining technical challenges can be adequately addressed to enable such reliable deformation capturing using DIC technique. The new setup with radiant burner system allows test specimens to be heated with virtually no smoke, flame or soot particles produced, which greatly facilitates the application of optical measuring techniques. In addition, the available open space enables the adoption of non- contact measuring equipment outside the hot environment to capture the deformation of hot areas on sample surfaces. Figure 6. Test setup for deformation capturing of concrete at elevated temperatures However, a key factor affecting the performance of DIC when using radiant burner system is thermal radiation. To overcome this issue, a Midopt BP470 filter is employed to minimise the effect of thermal radiation, together with an EFFI-Lase-Power-CM-C02-465 blue light to enhance the lighting condition (Figure 6). Such use of band-pass filter and blue illumination has proved to be very effective in stabilizing the lighting condition during the whole testing time. Initial results demonstrate a good agreement between DIC measurement and data collected from the loading machine, at both ambient and elevated temperatures. Figure 7 and 8 show a good agreement between load-displacement curves, deformations captured Camera with blue filter Camera with blue filter Blue LED light Concrete cylinder Actuator Radiant panels ISSN 1859-1531 - TẠP CHÍ KHOA HỌC VÀ CÔNG NGHỆ ĐẠI HỌC ĐÀ NẴNG, SỐ 5(126).2018, Quyển 2 37 by DIC and actuator of three heated concrete cylinders, giving confidence to the reliability of this measuring technique. Additional tests will be conducted for further verification. Figure 7. Good agreement between load-displacement curves captured by DIC and actuator Figure 8. Good agreement between deformations captured by DIC and actuator 5. Summary and conclusion When subjected to fire condition, concrete undergoes various complex physio-chemical changes. Some major influencing factors and spalling of concrete at elevated temperatures are reviewed in this paper. Limitations of current test setups for fire testing are then discussed, highlighting the need for new setups with improved thermal boundary condition and deformation capturing of concrete at elevated temperatures. On that basis, a new test setup that effectively addresses the identified limitations is briefed: The use of radiant burner system and DIC technique with band-pass filter and blue illumination has allowed us to (i) create known and consistent thermal boundary conditions and (ii) reliably capture deformation of concrete surfaces at elevated temperatures in a full-field, non-contact manner. Further work is ongoing to utilize this new test setup to generate a comprehensive set of reliable data, allowing us to revise concrete performance in fire taking on account of the effects of temperature and temperature gradients. REFERENCES [1] Australian Standard, AS 3600-2009. Concrete structures. 2009, SAI Global Limited: Standards Australia Limited. [2] Savva, A., P. Manita, and K.K. Sideris, Influence of elevated temperatures on the mechanical properties of blended cement concretes prepared with limestone and siliceous aggregates. Cement and Concrete Composites, 2005. 27(2): p. 239-248. [3] Hertz, K.D., Concrete strength for fire safety design. Magazine of Concrete Research, 2005. 57(8): p. 445-453. [4] Sakr, K. and E. El-Hakim, Effect of high temperature or fire on heavy weight concrete properties. Cement and Concrete Research, 2005. 35(3): p. 590-596. [5] Knaack, A.M., Y.C. Kurama, and D.J. Kirkner, Compressive strength relationships for concrete under elevated temperatures. ACI Materials Journal, 2010. 107(2): p. 164-175. [6] Kodur, V.K.R. and L.T. Phan, Critical factors governing the fire performance of high strength concrete systems. Fire Safety Journal, 2007. 42(6-7): p. 482-488. [7] Phan, L.T., Fire performance of high-strength concrete: A report of the state-of-the-art. The National Institute of Standards and Technology (NIST), 1996: p. 118. [8] Castillo, C. and A.J. Durrani, Effect of transient high temperature on high-strength concrete. Materials Journal, 1990. 87(1): p. 47-53. [9] Khoury, G.A., Effect of fire on concrete and concrete structures. Progress In Structural Engineering And Materials, 2000. 2(4): p. 429-447. [10] Kodur, V.K.R. and R. McGrath, Fire endurance of high strength concrete columns. Fire Technology, 2003. 39: p. 73-87. [11] Schneider, U., et al., Recommendation of RILEM TC 200-HTC: Mechanical concrete properties at high temperatures-modelling and applications: Part 2: stress-strain relation. Materials and Structures, 2007. 40(9): p. 855-864. [12] Schneider, U., et al., Recommendation of RILEM TC 200-HTC: Mechanical concrete properties at high temperatures-modelling and applications: Part 1: General presentation. Materials And Structures, 2007. 40(9): p. 841-853. [13] Khoury, G.A., Compressive strength of concrete at high- temperatures - a reassessment. Magazine of Concrete Research, 1992. 44(161): p. 291-309. [14] Sarshar, R. and G.A. Khoury, Material and environmental-factors influencing the compressive strength of unsealed cement paste and concrete at high-temperatures. Magazine of Concrete Research, 1993. 45(162): p. 51-61. [15] Mohamedbhai, G.T.G., Effect of exposure time and rates of heating and cooling on residual strength of heated concrete. Magazine of Concrete Research, 1986. 38(136): p. 151-158. [16] Phan, L.T., Fire performance of high-strength concrete: A report of the state-of-the-art. 1996, The National Institute of Standards and Technology (NIST). p. 118. [17] Cheng, F.-P., V.K.R. Kodur, and T.-C. Wang, Stress-strain curves for high strength concrete at elevated temperatures. Journal of Materials in Civil Engineering, 2004. 16(1): p. 84-90. [18] Khoury, G.A., P.J.E. Sullivan, and B.N. Grainger, Radial temperature distributions within solid concrete cylinders under transient thermal states. Magazine of Concrete Research, 1984. 36(128): p. 146-156. [19] Fédération Internationale du Béton, Fire design of concrete structures: Materials, structures and modelling, in fib Bulletin 38. 2008. p. 77-90. [20] Bažant, Z.P., Analysis of pore pressure, thermal stress and fracture in rapidly heated concrete, in International Workshop on Fire Performance of High Strength Concrete, NIST SP 919, Gaithersburg, MD. 1997. p. 155-164. [21] Khoury, G.A., Passive fire protection of concrete structures. Proceedings of the Institution of Civil Engineers: Structures and Buildings, 2008. 161(3): p. 135-145. [22] Hertz, K.D., Limits of spalling of fire-exposed concrete. Fire Safety Journal, 2003. 38(2): p. 103-116. [23] Sanjayan, G. and L.J. Stocks, Spalling of high-strength silica fume concrete in fire. ACI Materials Journal, 1993. 90(2): p. 170-173. [24] Kalifa, P., F.-D. Menneteau, and D. Quenard, Spalling and pore pressure in HPC at high temperatures. Cement and Concrete 38 Quang Xuan Le, Dinh Ba Le, Son Duy Tran, Vinh The Ngoc Dao Research, 2000. 30(12): p. 1915-1927. [25] Zhukov, V., Explosive failure of concrete during a fire. Joint Fire Research Organisation, 1975. Translation No. DT 2124. [26] Yabuki Y., Harada K., and Terai T., Heat transfer in fire resistance test furnace with particular references to gas radiation. Proceedings of the 2nd Asia - Oceanian Conference on Fire Science and Technology, 1995: p. 262-273. [27] Torero, J., Assessing the performance of concrete structures in fires. Concrete in Australia - Special Issue on "Concrete Performance in Fire", 2014. 40(3): p. 44-49. [28] Bisby, L., J. Gales, and C. Maluk, A contemporary review of large- scale non-standard structural fire testing. Fire Science Reviews, 2013. 2(1): p. 27. [29] McAllister, T., et al., Measurement of temperature, displacement, and strain in structural components subject to fire effects: Concept and candidature approaches, in NIST Technical Note 1768, P.D. Gallagher, Editor. 2012, Department of Commerce - USA: National Institute of Standards and Technology. p. 73. [30] Maluk, C., et al., A heat-transfer rate inducing system (H-TRIS) test method. Fire Safety Journal, 2016. [31] Le, Q.X., et al., Effects of temperature and temperature gradient on concrete performance at elevated temperatures. Advances in Structural Engineering, 2017. Accepted on 03-11-2017. [32] Le, Q.X., A study of temperature gradient effects on mechanical properties of concrete at elevated temperatures, in School of Civil Engineering. 2016, MPhil Thesis, The University of Queensland: Brisbane, Australia. [33] Maluk, C., Development and application of a novel test method for studying the fire behaviour of CFRP prestressed concrete structural elements. 2014, PhD Thesis, University of Edinburgh, UK. [34] Le, B.-D., et al., Deformation capturing of concrete structures at elevated temperatures, in Proceedings of the 6th International Workshop on Performance, Protection & Strengthening of Structures under Extreme Loading, PROTECT2017, 11-12 December 2017. 2017: Guangzhou (Canton), China. (The Board of Editors received the paper on 25/11/2017, its review was completed on 14/12/2017)