Experimental and theoretical studies on postfire behavior of functionally graded ultra-high performance concrete

Abstract

In order to improve structural fire-resistant behaviors, this paper designed a two-layer functionally graded ultra-high performance concrete (FGUHPC) structure composed of a UHPC layer and a lightweight aggregate concrete (LWAC) layer. UHPC layers are adopted to provide structural bearing capacity and protected by LWAC layers from elevated temperature. Splitting tensile tests and three-point flexural tests were conducted under ambient and elevated temperatures to evaluate interfacial bond performance and flexural bearing capacity, where two interfacial treatments were adopted and compared. The experimental results revealed that FGUHPC members exhibited good integrity during heating, no explosive spalling occurred and the maximal temperature at interfacial regions was 266°C. The interfaces showed desirable bond performance under ambient temperature while the splitting tensile strength was decreased by around 85% in the case of high temperature. Flexural test results indicated that the structural stiffness would be reduced by around 42% under elevated temperature, as a result, the maximal deflection was increased from 2.5 mm to 3.7 mm. SWM could significantly improve interfacial bond performance and prevent debonding failure of specimens at the postfire state, leading to higher structural bearing capacities. The bearing capacities of specimens with and without interfacial treatments were 42.7 kN and 38.4 kN respectively under ambient temperature, which remained about 88% after elevated temperature.

Keywords

functionally graded ultra-high performance concrete postfire performance interfacial treating methods interfacial bond behavior flexural bearing capacity

Introduction

Ultra-high performance concrete (UHPC) has been proven as a class of cementitious material capable of addressing a variety of needs in structural engineering (Graybeal et al., 2020). Compared with normal concrete (NC), it provides better strength, durability, energy absorption and fatigue behavior (Yoo and Yoon, 2016), these advantages can lead to economical construction by saving material and installation costs (Abbas et al., 2016). However, UHPC is usually deemed as an undesirable fire-resistant material due to the occurrence of explosive spalling subjected to elevated temperature, which hinders its applications under fire risk (Liang et al., 2018). Two theories were commonly accepted as the explanations for the phenomenon, named thermal stresses theory and pore pressure theory respectively. Thermal stresses theory is related to the thermomechanical process, a steep thermal gradient would be developed due to the rapid increase of temperature, as a result, the high thermal stress between heated surfaces and moisture clogs is generated, leading to the spalling of UHPC (Li and Zhang, 2021). Pore pressure theory involves the low permeability and the absence of capillary pores of UHPC. High pore pressure would be induced in the case of vaporization in moisture clogs and the thermal dilation of vapor, therefore, the explosive spalling would occur as the pore pressure exceeds the tensile strength of concrete (Kodur and Banerji, 2021; Peng et al., 2017b). It shall be noted that the fire-induced spalling is a complex phenomenon, and could be influenced by many factors, such as heating regimes, moisture contents, aggregate sizes and types, and so on (Banerji et al., 2020; Zhu et al., 2021).

Functionally graded concrete (FGC) is a cementitious composite whose material composition is spatially varied by two or several mixes (Torelli et al., 2020), as Figure 1 shows. Layered graded concrete refers to material with a step-wise spatial in composition, while continuously graded concrete is defined as concrete with a continuously variable composition. Layered graded concrete is more common and easy to prepare, which has attracted much attention by researchers. In comparison, most studies of continuously graded concrete were based on finite element analyses (Herrmann and Sobek, 2017; Strieder et al., 2018), it is because the inherent limitations in production technologies shall be overcome to translate the concepts into practice, and recent improvements of robotics promote the development of manufacturing techniques for continuously graded concrete members. By reasonably adopting multiple materials in different structural regions, the overall or local performance would be effectively improved with lower cement consumption, self-weight and long-term costs. For instance, Bajaj et al. (2013) developed a two-layered functionally graded concrete beam (FCB) composed of NC and fly ash concrete layers, where the structural compressive and flexural strength were 12.86% and 3.56% higher than those of conventional NC members respectively. Maalej et al. (2003) aimed at improving the structural corrosion durability, to this end, a two-layered FCB was designed, where the ductile fiber reinforced cementitious composite material was employed as the bottom layer. In this way, the proposed FGB had a noticeably higher durability than normal beams, where the time to achieve the same level of steel corrosion was enhanced by about 70%. Liu et al. (2018) assessed the potential of FGB that using recycled aggregates in place of a part of natural aggregates for environmental purposes. The results revealed that the maximal cost and CO₂ emission were reduced by 33.6% and 37.7% respectively compared with the traditional homogeneous material members. It is noticed that most of the studies aim at improving structural performances under ambient temperature, by contrast, there are few researches related to elevated temperature were reviewed. It is attributed to the incompatible deformation at interfacial regions between two materials subjected to high temperature, which might lead to delamination or debonding of interfaces (Shen et al., 2008). Zhang et al. (2022) established finite element models of a three-layer FCB, where the bond performance and flexural properties were discussed. According to the numerical results, the authors believed that the delamination failure could be avoided when the interfacial cohesive stiffness was higher than three. Du et al. (2023) designed a two-layered FGC member composed of UHPC layers and LWAC layers, debonding tests were conducted to evaluate the interfacial bond performance under different temperatures (20°C–1200°C), besides, the effect of cast-in bolts on bond strength were investigated. The results revealed that the bond strength was dramatically decreased with temperature, after 1200°C heating, the residual bond strength was reduced by around 95% compared with that under ambient temperature. The maximal improvement of bond strength was 56.5% when employing cast-in bolts, however, the effect became insignificant under high temperature due to the severe deterioration of LWAC, in this case, all specimens presented brittle failure. Nematzadeh and Mousavi (2021) studied the flexural behavior of FCB containing recycled tire crumb rubber under high temperature, where the effects of layers, fiber volume fractions, volume percentages of the rubber and heating regimes were discussed. The results indicated that the flexural capacity and stiffness of FGB were slightly increased under 300°C and markedly decreased by a range of 38% to 60% under 600°C.

Figure 1.

Layered and continuous functionally graded concrete.

Based on the FGC theory, a novel functionally graded ultra-high performance concrete (FGUHPC) was designed in this study. The structure is divided into two layers, including a UHPC layer and a lightweight aggregate concrete (LWAC) layer. UHPC layers are used to provide structural overall bearing capacities, stiffness, ductility and durability. LWAC, which is a desirable thermal insulation material with low thermal conductivity and good spalling resistance, is adopted as the protection layer from high temperature (Felicetti et al., 2013; Hassanpour et al., 2012), as Figure 2 diagrams. By this means, both the advantages of UHPC in mechanism and LWAC in thermology would be exploited. The structural self-weight and cross-sectional dimensions can be significantly decreased owing to the reduction of cement and reinforcement consumption (Nes and Overli, 2016), furthermore, extra fire protection techniques, such as insulation boards or coatings, could be removed.

Figure 2.

Configuration of FGUHPC.

This study aimed at evaluating the postfire performance of FGUHPC members in two aspects, including interfacial bond performance and flexural bearing capacity. To this end, splitting tensile tests and three-point flexural tests were conducted under both ambient and elevated temperatures, where two interfacial treating methods were adopted and compared. The interfacial temperature and spalling conditions were recorded and observed during the tests. The failure modes, loads, deflections, concrete strain and crack widths were measured and analyzed. Finally, the residual bearing capacities of FGUHPC members after fire were discussed and compared between the specification and experimental results.

Materials and experimental program

Mixture design and material properties

Two mixtures were adopted to prepare UHPC and LWAC respectively, where the proportions were listed in Table 1. For UHPC, 52.5# Portland cement, active admixtures (silica fume) and fine sands were employed, the steel fiber content was 2% by volume, where the length, diameter and strength were 13 mm, 0.2 mm and 2900 MPa respectively. For LWAC, ceramsites and fly ash were adopted as aggregates and active admixtures respectively. Polypropylene (PP) fibers are regarded as an effective method to improve spalling resistance of concrete, it is since that PP fibers would melt at 170°C, providing connected pores for moisture to escape, as a result, the pore pressure would be alleviated and the explosive spalling could be avoided to a great extent (Banerji et al., 2020; Du et al., 2020). In the tests, 3 kg/m³ PP fibers were added for two mixes according to Eurocode 2 (2001). The length, diameter and strength of PP fibers were 15 mm, 0.02 mm and 600 MPa respectively. The densities of UHPC and LWAC were around 2400 kg/m³ and 1500 kg/m³ respectively.

Table 1.

Mix proportions of UHPC and LWAC.

Ingredient (kg/m³)	UHPC	LWAC
Cement	852	435
Active admixture	61	115
Sand	1156	—
Ceramsite	—	770
Water	184	203
Steel fiber	155	—
PP fiber	3	3
Superplasticizer	13	11

The dry premix was firstly mixed for 3 min, and the water with superplasticizers was then added and mixed for 5 min. Afterwards, the fibers were added and mixed until the mixtures were homogeneously mixed. Slumps were tested before casting according to GB 50204 (2015), where those of UHPC and LWAC were 63±3 cm and 47±3 cm respectively.

The mechanical and thermal properties of two materials were measured according to GB/T 50081 (2019) and GB/T 10294 (2008). The compressive strength, tensile strength, elastic modulus and thermal conductivity were tested by 100 mm×100 mm×100 mm cubes, 100 mm×100 mm×500 mm dog bones, 100 mm×100 mm×300 mm prisms and 250 mm×250 mm×50 mm plates respectively. Electro servo-hydraulic pressure testing systems were adopted for measuring mechanical properties, where the load rates for testing compressive strength, tensile strength and elastic modulus elastic were 1.0 MPa/s, 0.2 mm/min and 1.2 MPa/s respectively. The thermal conductivity was tested by a conductivity tester. Each group was tested by three samples, the results are presented and summarized in Figure 3 and Table 2 respectively.

Figure 3.

Tests on mechanical and thermal properties.

Table 2.

Mechanical and thermal properties.

Material	UHPC	LWAC
Compressive strength (MPa)	144.1	25.5
Tensile strength (MPa)	9.5	1.9
Elastic modulus (GPa)	46.4	17.5
Thermal conductivity (W/m·K)	1.89	0.53

Specimen design and preparation

Splitting tensile tests and three-point flexural tests were conducted by 100 mm × 100 mm × 100 mm cubes and 150 mm × 150 mm × 550 mm prisms respectively. Both the thickness of UHPC layers and LWAC layers were 50 mm respectively for cubes while those of prisms were 100 mm and 50 mm in height respectively, as Figure 4 presents. The steel wire mesh (SWM) method was adopted to enhance the roughness of interfaces (Lu et al., 2022). SWM was pressed into the surfaces of concrete after casting and removed after hardening, the textures were shown at surfaces and the fibers were observed. Meanwhile, the specimens without interfacial treatments were also prepared as a contrast.

Figure 4.

Dimensions of specimens.

The LWAC layers were firstly cast, and the SWM treatment were subsequently conducted on the surfaces, after curing for 2 days, UHPC was cast against the hardened LWAC layers. All specimens were cured by 90°C steam for 2 days, afterwards, they were demolded and placed in a curing room until the follow-up process. Taking the cubes as an example, the casting procedure is exhibited in Figure 5.

Figure 5.

Casting procedure of FGUHPC members.

Experimental program

A 350 mm × 350 mm × 600 mm furnace was used for heating, and a steel cage was employed to prevent the furnace from damage caused by the explosive spalling of concrete, as Figure 6 shows. A severe heating regime was adopted in the tests, where the heating rate was 100°C/min to 1200°C and then kept at this temperature till 120 min. Bottom LWAC surfaces were regarded as the exposed surfaces, while the other surfaces were coated with fire retardant coatings in order to simulate one-side heating for both cube and prism specimens. For prism specimens, type K thermocouples were mounted at the depth of 50 mm (interfaces) from the exposed surface to determine the interfacial temperature. An electro servo-hydraulic pressure testing system with a load capacity of 300 kN. For the splitting tensile tests, the load was applied by 0.05 MPa/s and the strength $f_{t s}$ was calculated by

f_{t s} = 2 F / π A = 0.637 F / A

(1)

where

F

and

A

were maximal applied load and splitting area respectively. For three-point flexural tests, the applied load was 0.2 mm/min according to European Standards (2005) and the clear span was 500 mm. The loads and mid-span deflections were recorded by the system and a linear variable displacement transducer (LVDT) respectively (Gondokusumo et al., 2023). The strain gauges were mounted at 0 mm (bottom), 50 mm (interfaces), 100 mm and 150 mm (top) along with the height. Besides, the development and ultimate widths of cracks for both UHPC and LAWC layers were recorded during the tests (Wang et al., 2022). The typical testing diagrams are presented in Figure 6.

Figure 6.

Testing diagrams and measuring point arrangements.

Each specimen is named in the form of “experimental type”- “interfacial treatment”- “temperature condition”. For instance, S-SWM-AT represents the splitting tensile test on the specimen treated by SWM under ambient temperature. The details of specimens are summarized in Table 3.

Table 3.

Summaries of experimental specimens.

Specimen	Experimental type	Dimensions (mm)	Interfacial treatment	Temperature (°C)
S-SWM-AT	Splitting tensile tests	100 mm × 100 mm × 100 mm cubes	SWM	20
S-0-AT			No treatment	20
S-SWM-HT			SWM	1200
S-0-HT			No treatment	1200
F-SWM-AT	Three-point flexural test	150 mm × 150 mm × 550 mm prisms	SWM	20
F-0-AT			No treatment	20
F-SWM-HT			SWM	1200
F-0-HT			No treatment	1200

Note: AT: ambient temperature; HT: high temperature.

Experimental results

Overall performance

Temperature-time curves of the furnace and interface were recorded and depicted in Figure 7, in which the maximal temperature at the interface was 266°C. No noise was heard during the heating for all specimens, and no explosive spalling was observed at exposed surfaces. The bottom surfaces of LWAC turned yellow, where many cracks distributed. UHPC layers kept in good condition, no visible cracks and spalling were observed since they were at a relatively low temperature with the protection of LWAC layers.

Figure 7.

Temperature distributions and experimental phenomena.

For the cube specimens, obvious cracks along with interfaces were observed for the specimen without interfacial treatments (S-0-HT), and a part of specimens were broken apart directly, which was deemed as failure under high temperature. It is because of the discrepancy of expansion coefficients between two materials under high temperature, which would result in the interlaminar shear and incompatible deformation between two layers (Gao et al., 2019). In comparison, the specimens treated by SWM (S-SWM-HT) showed significantly better bond performance and no debonding occurred in spite of the small cracks at interfacial regions. For the prism samples, owing to larger contact areas, both F-SWM-HT and F-0-HT remained desirable bond behavior although visible cracks were observed at some of interfacial regions, as Figure 7 presents.

Results of splitting tensile tests

The splitting tensile test results were summarized in Table 4, where each group was tested by three samples. Under room temperature, the splitting tensile strength of S-SWM-AT and S-0-AT was 5.95 MPa and 4.62 MPa respectively with low variable coefficients. Meanwhile, the splitting tensile strength of homogeneous UHPC and LWAC was also tested as a contrast, which was 16.6 MPa and 2.93 MPa respectively. It was observed that the splitting tensile strength of FGUHPC samples was between that of homogeneous LWAC and homogeneous UHPC. For all specimens, the cracks occurred and developed at LWAC layers instead of interfacial regions, indicating that the interfaces had higher tensile strength than LWAC layers, and there was no bond issue at interfacial regions under ambient temperature. The specimens were damaged in the form of ductile failure, where fibers could be observed at the failure surfaces, as Figure 8 exhibits. Besides, SWM treatments could effectively improve the bond strength of members, the splitting tensile strength of S-SWM-AT was 5.95 MPa, 28.8% higher than that of S-0-AT.

Table 4.

Splitting tensile test results.

Specimen	Splitting tensile strength (MPa)	Mean (MPa)	COV (%)
S-SWM-AT	5.56	5.95	6.32
	6.31
	5.98
S-0-AT	4.19	4.62	11.40
	4.47
	5.21
S-SWM-HT	0.77	0.91	23.85
	1.16
	0.80
S-0-HT	0.41	0.37	15.29
	0.33
	—

Note: COV: coefficient of variation.

Figure 8.

Results of splitting tensile tests.

In the case of high temperature, the splitting tensile strength was dramatically dropped, where the mean strength of S-SWM-HT and S-0-HT was 0.91 MPa and 0.37 MPa, approximately 15.3% and 8.0% of that under room temperature respectively. With the load increased, brittle fracture along with the interfaces suddenly occurred, leading to the failure of specimens. There was no crack developing process, and interfaces had become the weak zone at this time, as Figure 8 diagrams. Furthermore, high temperature would result in large scatters of the splitting tensile strength, where the variable coefficients of S-SWM-HT and S-0-HT were 23.85% and 15.29% respectively.

Results of three-point flexural tests

The load-deflection and load-strain curves are illustrated in Figures 9 and 10 respectively, where the details are summarized in Table 5. It is noted that the strain of LWAC bottom (0 mm) for specimens under high temperature was not recorded since the surface had been severely deteriorated at this moment. F-SWM-AT and F-0-AT exhibited similar overall performance under ambient temperature. At the initial state, two specimens presented similar stiffness, cracks were observed at the bottom of LWAC layers at a very early time, where the crack loads were 7.1 kN and 6.2 kN for F-SWM-AT and F-0-AT respectively. It is because LWAC had very low elastic modulus and tensile strength. With the load continued to increase, cracks developed upwards along with the mid-span, and the sound of steel fibers being pulled out was heard, leading to the ductile failure of specimens. F-SWM-AT showed slightly higher ultimate load, where the ultimate load and deflection were 42.7 kN and 2.6 mm, in comparison, those of F-0-AT were 38.4 kN and 2.4 mm respectively. Besides, the specimen treated by SWM exhibited higher stiffness, which was 9.4% higher than the sample without interfacial treatments.

Figure 9.

Load-deflection curves.

Figure 10.

Load-strain curves.

Table 5.

Summaries of testing results.

Specimen	Initial crack state		Ultimate stage
Specimen	$P_{c r} (k N)$	$∆_{c r} (m m)$	$P_{u} (k N)$	$∆_{u} (m m)$	$δ_{L} (m m)$	$δ_{U} (m m)$	Failure mode
F-SWM-AT	7.1	0.4	42.7	2.6	0.9	0.8	Flexure
F-0-AT	6.2	0.5	38.4	2.4	0.9	0.9	Flexure
F-SWM-HT	—	—	37.6	3.7	3.8	1.1	Flexure
F-0-HT	—	-—	33.9	3.8	4.5	1.0	Debonding

Note: $P_{c r}$ and $∆_{c r}$ are load and deflection at the initial crack state; $P_{u}$ and $∆_{u}$ are load and deflection at the ultimate state; $δ_{L}$ and $δ_{U}$ are crack widths of LWAC and UHPC layers.

Stress redistributions were observed for both specimens during the tests, as Figures 9 and 10 diagram. They occurred at the load of about 35 kN, indicating the total fracture of LWAC layers, and all loads were bore by UHPC layers at this time. As a result, the strain of UHPC rapidly increase, and the load continued to rise after a sudden drop. Major cracks along the mid-span location were observed, the cracks crossed interfaces continuously while no debonding or delamination between layers was detected, as Figure 11 shows.

Figure 11.

Phenomena of three-point flexural tests.

In the case of high temperature, LWAC layers had been severely damaged during heating, where a large number of visible cracks were observed. In this case, LWAC might no longer provide bearing capacities and no stress redistribution was observed from experimental results. Compared with ambient temperature, the stiffness of specimens subjected to elevated temperature was decreased by 41.7% to 43.4%, leading to greater deflections under the same load. F-SWM-HT remained good integrity during the test and showed flexural failure, however, for F-0-HT, interfacial debonding occurred due to the weakness of interfacial bond strength, leading to the occurrence of debonding failure, as Figure 11 diagrams. The ultimate loads of F-SWM-HT and F-0-HT were 37.6 kN and 33.9 kN respectively, 88.1% and 88.3% of those under ambient temperature. Besides, it was observed that the cracks were not continuous at interfaces. The maximal crack widths of LWAC layers of F-SWM-HT and F-0-HT were 3.8 mm and 4.5 mm respectively, significantly larger than those under ambient temperature while the discrepancies of crack widths of UHPC layers among specimens were not evident, as Table 5 lists.

Discussions

Mechanism of interfacial bond performance

This section discusses the mechanism of interfacial bond performance of FGUHPC members and the reason for the dramatic drop after high temperature. The interfacial bond strength between layers is mainly provided by two parts, one is the mechanical force, including bite force between layers and the bridging effect of fibers crossing interfaces, the other is the chemical bond force (van der Waals force) among molecules in the transition interfacial layer between two materials (Gao et al., 2019). UHPC contains a large number of active admixtures whose main components are SiO₂ and Al2O₃, they can absorb Ca(OH)₂ to produce C-S-H gel, promote the hydration of C₃S and C₂S and reduce the water contents in pastes, as a result, the concentration of pastes is increased and the chemical bond force between layers will be improved. Meanwhile, the C-S-H gel can penetrate into gelled pores, which is beneficial to enhance the interfacial roughness (Graybeal et al., 2017). Besides, fibers crossing interfaces could contribute to the bond strength through the bridging effect. All these factors provide FGUHPC members high interfacial bond strength under room temperature, and specimens with SWM treatments presented slightly higher bond strength owing to higher mechanical bite force caused by rougher interfaces (Semendary and Svecova, 2021). It is concluded that there is no interfacial bond issue for FGUHPC elements under room temperature even if the interfaces were not treated on purpose.

In the case of elevated temperature, fibers have been melted, and Ca(OH)₂ is decomposed into CaO and H₂O, water in concrete has been evaporated, which increases the pore-water pressure in the matrix (Way and Wille, 2016). The transition layers would be damaged, resulting in the reduction or loss of the bond strength. Besides, the different thermal expansions between two materials may lead to an incompatible deformation and interlaminar shear at interfacial regions during heating, further reducing the bond strength. As a result, mechanical bite force becomes the main factor for the interfacial bond strength. In the test, the splitting tensile strength of specimens after heating was reduced by 85% and 92% respectively. Compared with specimens without interfacial treatments, SWM could slightly improve the bond strength by increasing the roughness of interfaces (Farzad et al., 2019).

Residual bearing capacities of FGUHPC members

The mechanical properties of UHPC show a tendency that first increase and then decline with temperature (Abid et al., 2017; Peng et al., 2017a). Based on the previous studies, the residual compressive strength, tensile strength and elastic modulus of UHPC under different temperatures had been tested, where the relative results are depicted in Figure 12. They increase with temperature before 200°C and decrease followed by. In the case of 400°C, the compressive strength of UHPC is almost the same with that under room temperature while the tensile strength and elastic modulus remain about 67% and 60% respectively. During the tests, the maximal temperature of UHPC layers was 266°C. Considering the excess temperature transferred from LWAC layers, the temperature of UHPC might continue to rise slightly after heating (Liu et al., 2023). In this case, UHPC would exhibit comparable compressive strength and lower tensile strength and elastic modulus, leading to a slight reduction of bearing capacity whereas a significant drop of stiffness compared with ambient temperature.

Figure 12.

Relative mechanical properties of UHPC after high temperature.

Assuming that the cross-sections of FGUHPC elements are satisfied with the plane section theory, hence the flexural bearing capacities under ambient temperature can be predicted by equaling LWAC layers to UHPC layers under the same flexural stiffness (Herrmann and Sobek, 2017), where the reduction factor $n$ is defined as:

μ = E_{L} / E_{U}

(2)

where

E_{L}

and

E_{U}

are elastic modulus of LWAC and UHPC respectively. In this case, the cross-section is expressed by a T-shape section with homogeneous UHPC, where the LWAC was equivalent to a UHPC cross-section with the same height and smaller width, as Figure 13 presents.

Figure 13.

Strain and stress profiles of FGUHPC elements.

The nominal flexural bearing capacity is calculated by T/CCPA 35-2022 (2022), namely the technical specification for UHPC structures:

M_{n} = 0.92 f_{c} b x (h - \frac{x}{2}) - 0.45 f_{t} b (h - x) [0.45 (h - x)] - 0.25 f_{t} (b_{f} - b) h_{f}^{2}

(3)

where

f_{c}

and

f_{t}

are nominal compressive and tensile strength of UHPC,

b

and

h

are width and height of the cross-section,

b_{f}

and

h_{f}

are width and height of the flange respectively where

b_{f}

is equal to

μ b

x

is the height of the compressive zone, which is determined by:

0.92 f_{c} b x = 0.45 f_{t} b (h - x) + 0.5 f_{t} (b_{f} - b) h_{f}

(4)

It shall be noted that the contribution of longitudinal reinforcing bars is ignored in equation (3) and equation (4) since there are no longitudinal reinforcing bars in the tests.

Table 6 compares the flexural bearing capacities of FGUHPC between experimental results and the specification, which show a good agreement under ambient temperature. Nevertheless, the flexural bearing capacity calculated by specification is significantly lower than experimental results in the case of high temperature. Defining the residual capacity

η = M_{n_H T} / M_{n_A T}

η

is 0.72 for the specification, lower that of experimental results (0.89 and 0.88 for group F-SWM- and F-0- respectively), it might be because the high temperature within a range improved mechanical properties of UHPC, leading to higher flexural bearing capacities of FGUHPC members subjected to elevated temperature compared with the specification (Kodur, 2018).

Table 6.

Flexural strength of experimental results and specification.

Specimen	Flexural bearing capacity $(k N \cdot m)$		$M_{n_H T} / M_{n_A T}$
Specimen	$M_{n_A T}$	$M_{n_H T}$	$M_{n_H T} / M_{n_A T}$
F-SWM-	5.3	4.7	0.89
F-0-	4.8	4.2	0.88
Specification	5.0	3.6	0.72

Note: $M_{n_A T}$ and $M_{n_H T}$ are flexural bearing capacities of FGUHPC members under ambient and high temperatures respectively.

Conclusions

This study aimed at improving structural fire-resistant performances and postfire behaviors. To this end, a two-layered functionally graded ultra-high performance concrete (FGUHPC) member was designed, and its interfacial bond and flexural performance subjected to ambient and elevated temperatures were evaluated by splitting tensile tests and three-point flexural tests respectively. The effects of two interfacial treating methods on interfacial bond performance were discussed. Based on the experimental results, the following conclusions can be drawn:

1. The designed FGUHPC elements exhibited desirable fire resistance, all specimens exhibited good integrity and no explosive spalling of UHPC occurred during 1200°C heating for 2 hours. The temperature at interfaces was 266°C, in this case, the mechanical properties of UHPC were not deteriorated owing to the thermal insulation of LWAC layers.

2. There was no interfacial bond issue for all specimens under room temperature. However, the splitting tensile strength was reduced by 85% to 92% under high temperature, SWM would efficiently improve the bond performance and prevent debonding failure of specimens. The stiffness and bearing capacity of the specimen treated by SWM was 9.4% and 11.2% higher than that without interfacial treatments.

3. At the postfire state, FGUHPC specimens showed lower stiffness, lower ultimate loads and larger deflections than those under ambient temperature. The reductions of structural stiffness and bearing capacity at ultimate state were around 42% and 12% respectively while the average deflection was increased from 2.5 mm to 3.7 mm.

4. The bearing capacities of FGUHPC members showed good agreement between experiments and the specification under room temperature. However, in the case of high temperature, the residual bearing capacities based on the experimental results were higher than those calculated by the specification. It might be because the mechanical properties of UHPC could be enhanced by elevated temperature within a limit.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by the National Natural Science Foundation of China (grant numbers: U1934205) and Academician special project of China Communications Construction Company (YSZX-01-2022-02-B).

ORCID iD

Jingquan Wang

References

Abbas

Nehdi

Saleem

(2016) Ultra-high performance concrete: mechanical performance, durability, sustainability and implementation challenges. International Journal of Concrete Structures and Materials 10(3): 271–295.

Abid

Hou

Zheng

, et al. (2017) High temperature and residual properties of reactive powder concrete – a review. Construction and Building Materials 147(519): 339–351.

Bajaj

Shrivastava

Dhoke

(2013) Experimental study of functionally graded beam with fly ash. Journal of The Institution of Engineers (India): Series A 94(4): 219–227.

Banerji

Kodur

Solhmirzaei

(2020) Experimental behavior of ultra high performance fiber reinforced concrete beams under fire conditions. Engineering Structures 208: 110316.

Huang

, et al. (2020) Experimental study on the spalling behaviour of ultra-high strength concrete in fire. Construction and Building Materials 258: 120334.

, et al. (2023) Evaluation of bond behaviors on functionally graded ultra-high performance concrete (FGUHPC) subjected to elevated temperature. Engineering Structures 274: 115112.

European Standards (2005) Test Method for Metallic Fibered Concrete - Measuring the Flexural Tensile Strength (Limit of Proportionality (LOP), Residual), Brussels: BSI.

Farzad

Shafieifar

Azizinamini

(2019) Experimental and numerical study on bond strength between conventional concrete and ultra high-performance concrete (UHPC). Engineering Structures 186: 297–305.

Felicetti

Gambarova

Bamonte

(2013) Thermal and mechanical properties of light-weight concrete exposed to high temperature. Fire and Materials 37(3): 200–216.

10.

Gao

Zhao

Qiao

, et al. (2019) Study on the bonding properties of engineered cementitious composites (ECC) and existing concrete exposed to high temperature. Construction and Building Materials 196: 330–344.

11.

GB 50204-2015 (2015) Code for Acceptance of Constructional Quality of Concrete Structures, Beijing: China Architecture and Building Press.

12.

GB/T 10294-2008 (2008) Thermal Insulation-Determination of Steady-State Thermal Resistance and Related Properties-Guarded Hot Plate Apparatus, Beijing: China Architecture and Building Press.

13.

GB/T 50081-2009 (2019) Standard for Test Method of Mechanical Properties on Ordinary Concrete, Beijing: China Architecture and Building Press.

14.

Gondokusumo

Venkateshwaran

, et al. (2023) Residual flexural tensile strength of normal-weight and lightweight steel fibre-reinforced concrete at elevated temperatures. Construction and Building Materials 367: 130221.

15.

Graybeal

De la Varga

Haber

(2017) Bond of field-cast grouts to precast concrete elements. Available at: www.fhwa.dot.gov/research

16.

Graybeal

Brühwiler

Kim

B-S

, et al. (2020) International perspective on UHPC in bridge engineering. Journal of Bridge Engineering 25(11): 04020094.

17.

Hassanpour

Shafigh

Mahmud

(2012) Lightweight aggregate concrete fiber reinforcement - a review. Construction and Building Materials 37: 452–461.

18.

Herrmann

Sobek

(2017) Functionally graded concrete: numerical design methods and experimental tests of mass-optimized structural components. Structural Concrete 18(1): 54–66.

19.

Kodur

VKR

(2018) Innovative strategies for enhancing fire performance of high-strength concrete structures. Advances in Structural Engineering 21(11): 1723–1732.

20.

Kodur

Banerji

(2021) Modeling the fire-induced spalling in concrete structures incorporating hydro-thermo-mechanical stresses. Cement and Concrete Composites 117: 103902.

21.

Zhang

(2021) Effect of lateral restraint and inclusion of polypropylene and steel fibers on spalling behavior, pore pressure, and thermal stress in ultra-high-performance concrete (UHPC) at elevated temperature. Construction and Building Materials 271: 121879.

22.

Liang

, et al. (2018) Development of ultra-high performance concrete with high fire resistance. Construction and Building Materials 179: 400–412.

23.

Liu

Yan

Galobardes

, et al. (2018) Assessing the potential of functionally graded concrete using fibre reinforced and recycled aggregate concrete. Construction and Building Materials 171: 793–801.

24.

Liu

Yue

, et al. (2023) Experimental investigation on fire endurance and residual strength of reinforced concrete beams strengthened with CFRP grids. Advances in Structural Engineering 26(5): 858–869.

25.

Pang

, et al. (2022) Bond strength between substrate and post-cast UHPC with innovative interface treatment. Cement and Concrete Composites 133: 104691.

26.

Maalej

Ahmed

SFU

Paramasivam

(2003) Corrosion durability and structural response of functionally-graded concrete beams. Journal of Advanced Concrete Technology 1(3): 307–316.

27.

Narayanan

(2001) EN1992 Eurocode 2: design of concrete structures. Proceedings of the Institution of Civil Engineers - Civil Engineering 144(6): 23–28.

28.

Nematzadeh

Mousavi

(2021) Post-fire flexural behavior of functionally graded fiber-reinforced concrete containing rubber. Computers and Concrete 27(5): 417–435.

29.

Nes

Overli

(2016) Structural behaviour of layered beams with fibre-reinforced LWAC and normal density concrete. Materials and Structures 49: 689–703.

30.

Peng

Yang

Shi

(2017a) Experimental study on residual mechanical properties of ultra-high performance concrete exposed to high temperature. China Civil Engineering Journal 50(4): 73–79.

31.

Peng

Niu

Cheng

(2017b) Research on fire resistance of ultra-high-performance concrete: a review. Materials Review 31: 17–23.

32.

Semendary

Svecova

(2021) Bond characterization and interfacial coefficients at precast UHPC and cast-in-place UHPC interface. Journal of Materials in Civil Engineering 33(7): 1–20.

33.

Shen

Hubler

Paulino

, et al. (2008) Functionally-graded fiber-reinforced cement composite: processing, microstructure, and properties. Cement and Concrete Composites 30(8): 663–673.

34.

Strieder

Hilber

Stierschneider

, et al. (2018) FE-study on the effect of gradient concrete on early constraint and crack risk. Applied Sciences 8(2): 246.

35.

T/CCPA 35-2022 (2022) Technical Specification for Ultra-high Performance Concrete Structures.

36.

Torelli

Fernández

Lees

(2020) Functionally graded concrete: design objectives, production techniques and analysis methods for layered and continuously graded elements. Construction and Building Materials 242: 118040.

37.

Wang

Guo

Wang

, et al. (2022) Flexural behaviour analysis of wet joint of precast bridge deck with UHPC functional gradient. Advances in Structural Engineering 26(3): 448–458.

38.

Way

Wille

(2016) Effect of heat-induced chemical degradation on the residual mechanical properties of ultrahigh-performance fiber-reinforced concrete. Journal of Materials in Civil Engineering 28(4): 1–10.

39.

Yoo

Yoon

(2016) A review on structural behavior, design, and application of ultra-high-performance fiber-reinforced Concrete. International Journal of Concrete Structures and Materials 10(2): 125–142.

40.

Zhang

Yang

Xie

, et al. (2022) Flexural behaviour and cost effectiveness of layered UHPC-NC composite beams. Engineering Structures 273: 115060.

41.

Zhu

Hussein

Kumar

, et al. (2021) A review: material and structural properties of UHPC at elevated temperatures or fire conditions. Cement and Concrete Composites 123: 104212.