Study on the compression performance of steel reactive powder concrete columns

Abstract

In this study, the mechanical properties and failure characteristics of steel reactive powder concrete columns with different strength grades were investigated through compression testing. Six steel reactive powder concrete columns were tested; three columns underwent axial compression testing and three columns underwent eccentric compression testing. The results of the axial compression testing showed that steel and reactive powder concrete could work cooperatively at the initial stage, and the final column failure mode was primarily splitting failure at the end of the column, with the formation of a main crack in the longitudinal direction extending to the middle of the column. The results of the eccentric compression testing showed that the eccentrically loaded steel reactive powder concrete columns had comparatively strong deformability. The columns presented ductile failure mode under the eccentric load with 0.2 eccentricity. The final failure of the column involved a sudden increase in the horizontal crack width on the tension side, the steel flange on the tension side reached the yield state, the reactive powder concrete in the middle of the compressive side was crushed, and the reactive powder concrete surface layer burst open and partially spalled off. According to the test results and with reference to the relevant standards, equations for calculating the approximate ultimate bearing capacities of axially and eccentrically compressed reactive powder concrete columns were proposed.

Keywords

axial compression eccentric compression experiment steel reactive powder concrete column ultimate bearing capacity

Introduction

Reactive powder concrete (RPC), a new cement-based composite with high strength, high toughness, and good reliability (Richard and Cheyrezy, 1995), has already been applied in areas such as bridge (Adeline et al., 1998), road (Dallaire et al., 1998), structure (Cavili and Rebentrost, 2006), ballastless track (Yang et al., 2015), and marine engineering (Wang et al., 2013). Previous researchers have successively performed the studies on the material mixes and properties, component performance, and so on, of RPC.

With regard to the studies on the material mixes and properties of RPC, Pierre Richard and Marcel Cheyrezy (1995) began research and development of RPC constituent components in 1995. Dugat et al. (1996) carried out experimental studies on the development of RPC 200 and RPC 800, and the effect of steel fiber content on the ductility and fracture properties of RPC in 1996. Hiremath and Yaragal (2017) investigated the early strength of RPC under four curing regimes including ambient air, hot air, hot water bath, and accelerated curing, and the results showed that the hot water bath curing resulted in higher strength, while the combined curing regimes gave excellent strength. Aydin et al. (2010) and Aydin and Baradan (2013) investigated the effect of the aggregate type and the composite material on the mechanical properties of RPC through experimental studies. Zheng et al. (2012a, 2012b, 2013) carried out experiments on compressive and tensile properties, compressive constitutive relationship of RPC reinforced with steel fiber. Canbaz (2014) studied the mechanical properties at different temperatures, such as 20°C, 100°C, 400°C, 700°C, and 1000°C, in addition, used ultrasonic test to investigate the compressive strength and the microstructure.

As for the RPC component performance, Ju et al. (2014) studied the effect of various axial compression ratios on the bearing capacity, ductility, and the energy dissipation capacity of RPC beam-column joint through finite element analysis. Gao et al. (2006) investigated the static properties of plain RPC beams on the basis of experimental study of four beams. Li et al. (2010) carried out the test of five two-span reinforced RPC beams to study the bending capacity and the moment modification coefficient. Fu et al. (2012) analyzed the crack characteristics of unbonded partially prestressed RPC beams on the basis of experimental research of different unbonded prestressed tendons. Shi et al. (2016) investigated the bearing features of eccentrically compressed RPC columns using compression test of 22 RPC columns, also set up a simple analytical method to calculated the ultimate bearing capacity. In addition, the bond–slip behavior between the reinforcement including steel and steel bar and concrete has been investigated through experimental and numerical studies (Chen et al., 2015; Feng et al., 2019a; Majdi et al., 2014).

In recent years, with the continuous development of high-rise and super high-rise buildings, while increasing the bearing capacity of vertical components, the cross-sectional component dimensions must expand as little as possible, becoming a research topic of interest for scholars in the industry. Therefore, steel tube concrete columns and steel concrete columns have emerged. The material mixes and properties, the plain and reinforced RPC beam and column performances, and RPC filled steel tube column performances have been mainly reported in the previous studies, however, relatively few studies exist concerning RPC steel columns.

In this article, static axial compression and eccentric compression tests were performed on RPC steel concrete columns of different strength grades with the aim of studying their axial and eccentric compression mechanical properties and failure modes. Equations for calculating the ultimate bearing capacities of the axially and eccentrically compressed columns were established in combination with existing research results.

Experimental study

Component design

In the Code for Design of Composite Structures JGJ 138:2016 (2016), the concrete strength grade for steel reinforced concrete should be less than C80, taking into account of the disadvantage of shrinkage and brittleness for high and ultra-high strength concrete. To connect with and extend the current specifications for composite structure, the RPC 120, 150, and 180, the strength of which are a little higher than that of C80 concrete, are investigated in the present article. To study the mechanical properties of RPC steel concrete with axially and eccentrically compressed columns of different strength grades, RPC dry mix was used in this study to separately prepare RPC steel columns with strength grades of 120 MPa, 150 MPa, and 180 MPa. The RPC concrete material used in this article was prepared as follows: P.O. 42.5 ordinary Portland cement, silica fume, quartz sand, quartz powder, water reducer, and steel fibers. And the mix proportions have been shown in Table 1.

Table 1.

Mix proportions of RPC.

Material	Cement	Silica fume	Quartz sand	Quartz powder	Water reducer	Steel fibers (%)	Water-binder ratio
RPC 120	1.00	0.25	1.10	0.30	0.025	2	0.20
RPC 150	1.00	0.25	1.10	0.30	0.025	2	0.18
RPC 180	1.00	0.25	1.10	0.30	0.025	2	0.16

RPC: reactive powder concrete.

The RPC steel columns and the samples for mechanical properties were casted under the same curing condition. For hot water curing, the humidity of 68% at the room temperature was adopted on the first day, then, they were curing under hot water with temperature of 70°C ∼80°C for 48 h. Finally, the nature curing was used with the temperature higher than 10°C and with surface moistening.

The design cross-sectional dimension of the axially compressed column (ACC) was 200×160 mm, and its design length was 1200 mm. The steel used was Q235 I-shaped steel with h×t×b×d = 140×9.1×80×5.5 mm. There was a shear-resistant pin with a diameter of 13 mm welded on the steel web every 200 mm, and 6-mm-diameter stirrups were welded onto the periphery of the steel at an interval of 200 mm. A schematic diagram of the specimen cross section is shown in Figure 1(a).

Figure 1.

Schematic diagram of the specimen cross sections (unit: mm): (a) axially compressed specimen; (b) eccentrically compressed specimen.

The cross-sectional design dimensions of the eccentrically compressed column (ECC) were 200×160 mm, with a design length of 1200 mm. The steel used was Q235 I-shaped steel with h×t×b×d = 140×9.1×80×5.5 mm. To meet the eccentric loading needs, brackets with cross-sectional design dimensions of 160×300 mm were located on both ends of the specimen, and 8-mm-thick steel plates were used to reliably weld the brackets to the steel. The column body was configured with 8-mm-diameter stirrups at an interval of 150 mm; within a range of approximately 250 mm of the two ends of the columns, the stirrups were reinforced, and the interval of the reinforced stirrups was 50 mm. A schematic diagram of the specimen is shown in Figure 1(b).

Material properties

To investigate the mechanical properties of RPC materials, the mechanical tests of three cubic samples with the dimensions of 100 mm×100 mm×100 mm have been performed for cubic compressive strength or splitting tensile strength, while the mechanical tests of three prism samples with the dimensions of 100 mm×100 mm×300 mm have been conducted for axial compressive strength or elastic modulus. The mean value for each mechanical property has been displayed in Table 2. According to the RPC constitutive relationship obtained from the study results (Lu et al., 2014), the RPC axial compressive strength $f_{c}$ and the cubic compressive strength $f_{cu}$ basically conform to a linear relationship

f_{c} = 0.845 f_{cu}

(1)

Table 2.

Basic mechanical property indices of RPC.

Strength grade	$f_{c} (MPa)$	$f_{cu} (MPa)$	$f_{t} (MPa)$	$E (GPa)$
RPC 120	103.3	122.2	9.2	41.6
RPC 150	129.6	153.4	11.4	44.3
RPC 180	145.0	171.6	13.7	45.8

Note: $f_{c}$ —mean value of RPC axial compressive strength; $f_{cu}$ —mean value of RPC cubic compressive strength; $f_{t}$ —mean value of RPC splitting tensile strength; E—mean value of RPC elastic modulus. RPC: reactive powder concrete.

And the relationship between the elastic modulus E and the compressive strength of the prism is given by

E = 2055 \sqrt{f_{c}} + 18897

(2)

For ordinary and RPC concrete, the randomness of concrete has been studied by nano- and micro- indentation (Liu et al., 2018), and the probabilistic behavior was also analyzed for reinforced concrete structure (Feng et al., 2019b). In the present article, the focus is on the first-order characteristics of the mechanical properties.

Boundary conditions, loading scheme, and measuring point arrangement

Boundary conditions

This test was carried out on a 20,000 kN electrohydraulic servo-coordinated loading system. The loading test was performed using a monotonous and continuous load-displacement hybrid control loading method. The upper and lower ends of the specimen were articulated and constrained by the fabricated ball hinge steel support. Before testing, a laser instrument was used for centering, to ensure the verticality of the components and the direction of the applied load. The specimen loading is shown in Figure 2.

Figure 2.

Test loading device: (a) axially compressed specimen; (b) eccentrically compressed specimen.

Loading scheme

In the preloading stage, in accordance with the common characteristics of the axially and eccentrically compressed components, the preloading scheme was used: load grading loading control with a preaxial force of 200 kN for each grade, the preaxial force was loaded to 800 kN, the contact situation between the specimen and the loading device was inspected, and the loading device and testing instrument working conditions were inspected.

In the formal loading stage, in full consideration of the differences in the mechanical characteristics of the axially and eccentrically compressed components, different loading schemes were used:

For the axially compressed components, before 80% of the estimated ultimate load of the specimen reached, loading control was used for the loading method, load grading was applied with 200 kN at each grade; after reaching 80% of the estimated ultimate load, displacement control was used until the specimen failed.

For the eccentrically compressed components, before the specimen cracked, load grading was applied with 100 kN at each grade; after the specimen cracked, the load grade was reduced to 50 kN per grade.

Measuring point arrangement

To obtain the strain on the RPC specimen surface and the steel surface in the test, strain measuring points were arranged on the cross section in the middle part of the column body, as shown in Figure 3. In order to obtain the lateral deflection of the eccentrically compressed components, dial indicators were arranged on one side of the column body, as shown in Figure 3(b). R represents an RPC strain gauge and S represents a steel strain gauge. The strain was automatically collected by the DH3816 N static strain test system.

Load and deformation: the load and the specimen deformation along the load direction were directly read during the testing process through the electrohydraulic servo press control system.

Strain: the longitudinal strain of the outer RPC layer and the longitudinal strain of the internal steel were measured through the strain gauges affixed to the RPC specimen surface and the steel surface. For the axially compressed specimen, the mean value of the steel strain S1-S5 was taken as S-ACC, and the mean value of the RPC strain R1-R4 was taken as R-ACC. For the eccentrically compressed specimen, the mean value of strain S1-S3 for the compressed steel flange was taken as Sc-ECC, the mean value of strain S4-S6 for the flange subjected to tension was St-ECC, the strain S7 in the middle part of the web was Sw-ECC, the mean value of strain R0-1∼R0-3 for the tension side of the RPC was Rt-ECC, and the mean value of strain R6-1∼R6-3 for the compression side of the RPC was Rc-ECC.

Lateral deformation: the specimen’s lateral deflection deformation value for various grades of loading was measured through the dial indicators installed at the center and 1/4 of the specimen height on the tension side of the eccentrically compressed specimen.

Crack: a whitening treatment was applied to the component surface, the width of the crack on the specimen surface was observed using an instrument for the width of the crack, and the position, development trend, and distribution situation for the crack appearance was described by a water-based pen, and the maximum width of the crack and the corresponding load value in the crack development process were recorded.

Figure 3.

Arrangement of strain gauges and dial indicators: (a) strain gauges on the axially compressed specimen; (b) strain gauges and dial indicators on the eccentrically compressed specimen.

Test results and analysis

Axial compression

Specimen failure process

The test results indicated that specimens ACC120, ACC150, and ACC180 were in an elastic state at the initial stage of test loading, and there were no cracks produced. As the loading increased, small cracks appeared at the ends of the three specimens, after which the cracks gradually developed and slowly extended toward the column body. When the peak load was reached, main cracks formed on the surfaces of the three sets of specimens. After the peak load was reached, as loading continued, a vertical main crack that extended toward the column body formed on the three sets of specimens, the reactive powder concrete at the ends of the specimens was crushed, and the failure occurred.

Using ACC150 as an example, in the initial stage of test loading, the loading was increased to 2900 kN by a load of 200 kN per grade, the relative displacement was 7.19 mm, and no crack appeared on the specimen. Subsequently, displacement loading control was used to load to 3100 kN, the relative displacement was 7.47 mm, and small cracks appeared at the ends of the specimens; see Figure 4(a). As the displacement loading continued to increase, the cracks at the ends of the specimens continued to develop, accompanied by the generation of new cracks. When the loading reached 3624 kN, the relative displacement was 8.80 mm, the bearing capacity reached its peak, the cracks at the ends of the specimens developed obviously and some of the cracks connected with each other, and the development of cracks on the column body was slow; see Figure 4(b). In continued loading after the peak load, the bearing capacity of the specimens began to decrease; when the displacement reached 10.40 mm, the load fell to 3347 kN, the cracks at the ends of the specimens connected and formed a vertical main crack that extended toward the middle part of the column body, the reactive powder concrete at the feet of the columns was crushed, and the steel fibers at the cracks continued to snap; see Figure 4(c). When the loading reached 16.98 mm, loading stopped when the bearing capacity decreased to 1310 kN, and the specimens failed severely; see Figure 4(d).

Figure 4.

Test failure mode of axially compressed specimen ACC150, with the loading of (a) 3100 kN, (b) 3624 kN, (c) 3347 kN and (d) 1310 kN.

Load-displacement curve

The crack development and failure features of the three axially compressed specimens in the test loading process were basically the same, and the load-displacement curves are shown in Figure 5. Roughly five stages occurred in the whole processes: (1) elastic stage: at the initial stage of loading, due to the small test load applied, the RPC and steel worked cooperatively through the interface cohesive force, and the specimens were in an elastic state; (2) anti-slip phase: as the test load increased, the shear-resistant pin did not participate in the work, the interface cohesive force between the RPC and steel decreased, which increased the relative slippage between the two, and the axial deformation increased relative to the previous stage; (3) cooperative work stage: as the slippage continued to increase, the dowel action of the shear-resistant pin between the RPC and the steel was unleashed and, under the joint action of the shear-resistant pin and the stirrups, the renewed cooperative work between the RPC and the steel decreased the relative slippage, and the specimens maintained their elastic states; (4) plastic stage: as the test load continued to increase, the dowel action of the shear-resistant pin gradually decreased, the inside of the specimens gradually failed, accompanied by a crisp and bright sound that was more concentrated during a certain time segment, the relative slippage between the RPC and the steel gradually increased, the bearing capacity was gradually enhanced, and an inelastic state presented in the specimens; (5) failure stage: as the loading continued to proceed after the peak load, the crack gradually developed until the bearing capacity experienced a sudden drop, and the failure occurred.

Figure 5.

Load-displacement curves of the axially compressed specimens.

According to the load-displacement curves shown in Figure 5, the bearing capacity of the specimens and the stiffness at each stage increased as the RPC strength increased, but when the strength exceeded a certain value, the magnitude of increase in the bearing capacity obviously decreased. It should be noted that the stiffness of specimens ACC150 and ACC180 was comparatively close at all stages.

Load-strain curve

The RPC and steel load-strain curves of the three specimens are shown in Figure 6. At the initial stage of loading, the RPC strain and the steel strain were basically the same, and they both increased linearly. Subsequently, as the load grade continued to increase, the difference between the RPC strain and the steel strain gradually increased, the internal forces of the RPC and the steel were redistributed, the strain curves changed from linear to nonlinear, and the steel gradually started to yield. In the loading process at various grades for the same specimen, the RPC strain was notably smaller than the steel strain. In comparing the three sets of different specimens, for the same load situation, the steel and RPC strains of the three sets of specimens did not change greatly in the initial loading stage. As the loading grade increased, the variable quantity in the RPC strain of specimen ACC150 was less than that of ACC120, but the variable quantity in the steel strain was greater; the variable quantity in the RPC strain of specimen ACC180 was greater than that of ACC150, but the variable quantity in the steel strain was less. As the RPC strength grade of the specimen increased, changes were generated in the working mechanism of the RPC and steel in the specimen, which led to a redistribution of internal forces. When the RPC strength was relatively low, the interaction between the steel pin and the RPC caused the inside of the RPC to prematurely generate cracks or even crush locally, and the cooperation between the RPC and the steel was reduced, leading to an increase in the internal steel bearing load and even yielding. After the RPC strength increased to a certain degree, the internal fracture development of the RPC slowed, and an improved cooperation between the RPC and the steel caused the bearable load of the RPC to increase.

Figure 6.

Load-strain curves of the axially compressed specimens.

Eccentric compression

Specimen failure process

The test results showed that specimens ECC120, ECC150, and ECC180 were in an elastic state at the initial stage of test loading. The specimens were slightly bent, but no associated cracks were generated. As the loading grade increased, the specimens increased their degree of bending, accompanied by the generation of cracks, and the increasing speed of lateral deflection accelerated. Subsequently, new cracks were generated continuously in the tension zone, the RPC concrete in the compression zone peeled off and bulged up, and the increasing speed of lateral deflection slowed. Near the peak load, the lateral deflection and crack width of the specimens increased suddenly. At the peak load, the RPC concrete in the compression zone was crushed and peeled off, and the specimens failed.

Using ECC120 as an example, in the initial stage of test loading, loading was carried out using a load increment of 100 kN per grade, and nothing abnormal appeared in the specimen. When the loading reached 600 kN, the lateral deflection was 4.59 mm, vertical cracks appeared at the end of the specimen near the support, a horizontal crack appeared in the tension zone in the center of the specimen, and the width of the crack was approximately 0.10 mm; see Figure 7(a). Afterward, loading was carried out using a load increment of 50 kN per grade, and when the loading reached 1030 kN, the lateral deflection was 9.34 mm, vertical cracks developed to the root of the brackets, a second horizontal crack appeared at 1/4 in height in the lower part of the tension side of the specimen, and the RPC concrete in the middle part of the compression side peeled off and bulged up; see Figure 7(b). When the loading reached 1340 kN, the lateral deflection was 11.80 mm, and the width of the horizontal crack in the middle part of the tension side suddenly increased to 3 mm; see Figure 7(c). When loading reached 1450 kN, the lateral deflection was 12.50 mm, the loading capacity reached its peak, and the specimen deformation and failure were severe; see Figure 7(d).

Figure 7.

Test failure mode of eccentrically compressed specimen ECC120, with the loading of (a) 600 kN, (b) 1030 kN, (c) 1340 kN and (d) 1450 kN.

Load-displacement curve

The crack development and failure features in the test loading process of the three eccentrically compressed specimens were basically the same, and the load-displacement curves are shown in Figure 8. Roughly four stages were observed in the whole processes: (1) elastic stage: at the initial stage of loading, due to the comparatively small load applied, the specimens were slightly bent, the RPC and steel deformations were coordinated with each other, the lateral deflection was comparatively small, and the specimens were in an elastic state. (2) cracking stage: when the loading reached approximately 40% of the peak load, vertical cracks and horizontal cracks appeared at the ends of the specimens and the center of the tension side, respectively, the stiffness of the specimens started to degenerate, the lateral deflection grew comparatively quickly, there was a clearer slippage segment in the curves, and the internal forces of the steel and the RPC started to redistribute. (3) Crack development stage: as the loading continued, the vertical and horizontal cracks continued to develop, the specimen stiffness gradually degenerated and presented an inelastic state; when the loading was to approximately 70% of the peak load, a second horizontal crack appeared in the center of the specimens on the tension side, the RPC concrete on the compressive side peeled off and bulged up, and the lateral deflection continued to increase. (4) Failure stage: when the loading reached approximately 90% of the peak load, the horizontal crack increased suddenly; when the peak load was reached, the RPC concrete in the center of the specimens on the compression side was crushed and peeled off, the bearing capacity suddenly dropped and entered a descending stage, the displacement gradually increased, the load gradually decreased, and the specimens failed and exhibited stronger ductility. After the tests ended, a gradual chiseling of RPC observations showed that the lateral RPC of the steel flange on the compression side was crushed, and the RPC inside the flange was relatively intact, indicating that the internal RPC constraint by the steel was better.

Figure 8.

Load-displacement curves of the eccentrically compressed specimens.

The load-displacement curves in Figure 8 show that, in the initial stage of loading, the three sets of specimen curves basically coincided. After the specimens cracked, there was a clearer slippage segment in the curves, the stiffness of the specimens gradually degenerated, and the internal forces between the RPC and the steel were redistributed. As the loading grade increased, the degeneration in the stiffness of specimen ECC120 was more notable than that of ECC150 and ECC180, and the stiffness degeneration of specimen ECC150 and specimen ECC180 was basically the same. When the bearing capacity reached its peak, the higher the RPC strength, the larger the bearing capacity. After the three sets of specimens reached the peak load, the residual strength value was comparatively high, at approximately 75%–90% of the peak load.

Load-strain curve

Figure 9(a) compares the steel and RPC load-strain curves for the three specimens. The compressive, tensile steel flange and steel web strains are Sc, St, and Sw, respectively, and the RPC strains are R1, R3, and R5. Figure 9(b) shows the RPC load-strain curves for the three specimens. In the initial loading stage, the RPC strain and the steel strain of the three specimens were basically the same, all increasing linearly. As the loading continued to proceed, the specimens cracked, the internal forces of the RPC and the steel on the tension side started to redistribute, the RPC strain and the steel strain on the tension side of specimen ECC120 were clearly larger than those of ECC150 and ECC180, which were basically equal; in addition, the RPC strain on the compression side of the three sets of specimens basically remained consistent. Due to the relatively low RPC strength, for ECC120, the RPC on the tension side of the specimen cracked open prematurely, the cooperative effect between the RPC and the steel was reduced, the internal forces started to redistribute, the tensile steel flange yielded due to premature bearing of larger loads, the cracks and the neutral axis continued to extend from the tension side toward the compression side, leading to a decrease in the cross-sectional area of the compression zone, the RPC concrete in the compression zone was crushed, and the specimen failed. In addition, after the RPC strength was increased to a certain degree, the specimen cracked relatively late; there is good synergy between the RPC and the steel, therefore the RPC and steel strains of specimens ECC150 and ECC180 basically remained consistent.

Figure 9.

Load-strain curves of the eccentrically compressed specimens: (a) the steel and RPC load-strain curves, and (b) RPC load-strain curves.

Calculation of the ultimate bearing capacity

Axial compression

The test results show that there was a good cooperation between the RPC and the steel of the axially compressed columns, which satisfies the plane section assumption when calculating the ultimate bearing capacity. Therefore, the axial compression bearing capacity of the axially compressed column could be given as follows

N_{u} = 0.9 φ (f_{c} A_{c} + f_{ssy} A_{ss})

(3)

where $φ$ the axial compression stability coefficient related to the component slenderness ratio, the component slenderness ratio in this test was 7.5, and $φ$ was 1.0; $f_{c}$ is the mean value of the RPC axial compressive strength; $f_{ssy}$ is the yield strength of the steel; $A_{c}$ is the RPC cross-sectional area; and $A_{ss}$ is the area of the steel.

The axial compression bearing capacities in the three sets of tests were calculated in accordance with equation (3); see Table 3 for the calculation results. The calculation results show that the calculated values and the test values had a good fit.

Table 3.

Comparison of the bearing capacity of axially compressed columns between the test value and calculated value.

Specimen No.	$f_{c} (MPa)$	$N_{a} (kN)$	$N_{b} (kN)$	$N_{a} / N_{b}$
ACC-120	103.3	3227	3017	1.069
ACC-150	129.6	3934	3624	1.085
ACC-180	145.0	4348	3890	1.117

$N_{a}$ and $N_{b}$ are the calculated value and test value, respectively, of the axial compression bearing capacity of the tested columns. ACC: axially compressed column.

Eccentric compression

For the calculation of the ultimate bearing capacity of the RPC steel columns, there is still a lack of research data at home and abroad. The test results showed that prior to 0.60P_u, the cross-sectional strain distribution basically conformed with the plane section assumption; after the load reached 0.7P_u–0.8P_u, however, the plane section assumption was no longer valid. To facilitate calculation, the previous research (Zhao, 2005) was used as a reference for revising the plane section assumption when calculating the ultimate bearing capacity, and the ultimate compressive strain value taken for the RPC after the revision was $ε_{cu} = 0.003$ .

Using the calculation method for the ultimate bearing capacity of the steel-concrete compression column in the Code for Design of Composite Structures JGJ 138:2016 (2016) as a reference, the ultimate bearing capacity of the RPC steel column was calculated. Since a longitudinal steel bar was not arranged in the specimen, the terms in the equation associated with the longitudinal steel bar were discarded, and the simplified calculation diagram is shown in Figure 10.

Figure 10.

Simplified diagram for calculating the RPC steel eccentrically compressed column.

From the internal force balance conditions: $Σ N = 0$ , $Σ M = 0$ , it follows that

N = α_{1} f_{c} bx + f_{a}^{'} A_{af}' - σ_{a} A_{af} + N_{aw}

(4)

Ne = α_{1} f_{c} bx (h_{0} - x / 2) + f_{a}^{'} A_{af}' (h_{0} - a_{a}') + M_{aw}

(5)

In equation (5)

h_{0} = h - a_{a}

(6)

e = e_{i} + \frac{h}{2} - a_{a}

(7)

e_{i} = e_{0} + e_{a}

(8)

When $δ_{1} h_{0} < \frac{x}{β_{1}}$ and $δ_{2} h_{0} > \frac{x}{β_{1}}$

N_{aw} = [\frac{2 x}{β_{1} h_{0}} - (δ_{2} - δ_{1})] t_{w} h_{0} f_{a}

(9)

M_{aw} = [0.5 (δ_{1}^{2} + δ_{2}^{2}) + (δ_{2} + δ_{1}) + \frac{2 x}{β_{1} h_{0}} - {(\frac{x}{β_{1} h_{0}})}^{2}] t_{w} h_{0}^{2} f_{a}

(10)

When $δ_{1} h_{0} < \frac{x}{β_{1}}$ and $δ_{2} h_{0} < \frac{x}{β_{1}}$

N_{aw} = (δ_{2} - δ_{1}) t_{w} h_{0} f_{a}

(11)

M_{aw} = [0.5 (δ_{1}^{2} - δ_{2}^{2}) + (δ_{2} - δ_{1})] t_{w} h_{0}^{2} f_{a}

(12)

When $x < ξ_{b} h_{0}$ , and it is a large, eccentrically compressed component, then

σ_{a} = f_{a}

(13)

When $x \geq ξ_{b} h_{0}$ , and it is a small, eccentrically compressed component, then

σ_{a} = \frac{f_{a}}{ξ_{b} - β_{1}} (\frac{x}{h_{0}} - β_{1})

(14)

The height relative to the limit compression zone is

ξ_{b} = \frac{β_{1}}{1 + \frac{f_{a}}{0.003 E_{a}}}

(15)

where e—distance between the action point of the axial force and the resultant force point of the steel tensile flange; $e_{0}$ —eccentricity distance of the axial force to the cross-section center of gravity; $e_{i}$ —initial eccentricity distance; $e_{a}$ —additional eccentricity distance, which took the greater of 20 mm and $h / 30$ ; $α_{1}$ —influence coefficient of the concrete compressive stress in the compression zone; $β_{1}$ —influence coefficient of the concrete stress pattern in the compression zone; $M_{aw}$ —moment of the axial resultant force borne by the steel web to the point of the resultant force for the steel flange on the tension side; $N_{aw}$ —axial resultant force borne by the steel web; $f_{c}$ —design value for the RPC axial compressive strength; $f_{a}$ , $f_{a}^{'}$ —design values of the steel tensile and compressive strengths; $A_{af}$ , $A_{af}'$ —cross-sectional areas of the steel tensile and compressive flanges; b—cross-sectional width; h—cross-sectional height; $h_{0}$ —effective cross-sectional height; $t_{w}$ —thickness of the steel web; $ξ_{b}$ —height relative to the limit compression zone; $E_{a}$ —steel elastic modulus; x—height of the RPC equivalent compression zone; $a_{a}$ —distance from the point of resultant force for the steel tensile flange to the cross-sectional tensile edge; $a_{a}'$ —distance from the point of resultant force for the steel compressive flange to the cross-sectional compressive edge; $δ_{1}$ —ratio of the distance from the upper end of the steel web to the upper edge of the cross section with $h_{0}$ ; $δ_{2}$ —ratio of the distance from the lower end of the steel web to the upper edge of the cross section with $h_{0}$ .

The calculation process was to first substitute equations (9) and (10) into equations (4) and (5), and then judge whether or not the calculated x values conform to the corresponding conditions. If they conform, then the corresponding calculation results were used; if they did not conform, then equations (11) and (12) were substituted into equations (4) and (5) to perform the calculation.

Using the literature (Cao et al., 2015) as a reference and taking $α_{1} = 0.88$ and $β_{1} = 0.70$ , the calculation results showed: when the eccentricity distance $e_{0} = 60 mm$ , the three steel RPC columns with large eccentricity presented ductile failure mode, which was consistent with the test results. See Table 4 for the specific calculation results.

Table 4.

Comparison of the bearing capacity of eccentrically compressed columns between the calculated value and test value.

Specimen No.	$P_{cd} (kN)$	$P_{u} (kN)$	$P_{cd} / P_{u}$ (%)	$\bar{P_{cd} / P_{u}}$ (%)
ECC-120	1126	1452	77.5	77.1
ECC-150	1375	1781	77.2
ECC-180	1618	2116	76.5

$P_{cd}$ and $P_{u}$ are the calculated value and test value, respectively, for the ultimate bearing capacity of the eccentrically compressed specimens. ECC: eccentrically compressed column.

Table 4 shows that the ultimate bearing capacity of the RPC steel eccentrically compressed column obtained from calculation is smaller than the experimentally measured ultimate bearing capacity. The mean deviation between the calculated value and the measured value is 22.9%, which is on the safe side. This is because the steel fibers in the RPC enabled the RPC to bear some tensile force, but the tensile stress borne by the RPC was not considered during the calculation.

Conclusion

Axial and eccentric compression performances of steel RPC columns have been comprehensively studied based on the experimental testing and the approximate ultimate bearing capacities analysis. The following conclusions could be drawn from the results and discussions obtained above.

When the axially compressed RPC steel column reached the ultimate bearing capacity, the body of the column did not have obvious cracks, and only a few small cracks appeared at the ends. As loading continued, the cracks at the ends gradually developed toward the center of the column body, forming a longitudinal splitting main crack near the steel flange. The main failure mode of the specimen was failure at the end.

Under the eccentric load with an eccentricity of 0.2, the eccentrically compressed RPC steel column presented ductile failure mode with large eccentric compression, and the behavior and the failure mode were similar to the large eccentric compression failure of ordinary steel-concrete columns. The difference was that the steel fibers in the RPC limited the excessively fast crack development, which made the residual strength of the RPC steel columns higher after large eccentric compression failure occurred. The ends of the specimens generated cracks first, then the main crack formed in the center of the specimens, and the final crack width was far larger than the crack widths of the ordinary steel-concrete components and the reinforced concrete components during failure.

The calculated value of the ultimate bearing capacity for the axially compressed columns based on the Code for Design of Composite Structures JBJ 138:2016 (2016) was larger than the test value. The ultimate bearing capacity of the eccentrically compressed RPC steel columns calculated based on the calculation method of the Code for Design of Composite Structures JBJ 138:2016 (2016) and in combination with the revised plane section assumption was on the safe side, and the calculated value was some 77.1% of the measured value. Therefore, further research is needed on the theoretical equation for the ultimate bearing capacity of eccentrically compressed RPC steel columns.

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 work was supported by the China Torch Program (Grant No. 2013GH561393).

References

Adeline

Lachemi

Blais

(1998) Design of behavior of the Sherbrook bridge. In: Proceedings of the international symposium on high-performance and reactive powder concretes, Sherbrooke, QC, Canada, 16–20 August 1998, pp. 89–98.

Aydin

Baradan

(2013) Engineering properties of reactive powder concrete without Portland cement. ACI Materials Journal 110(6): 619–627.

Aydin

Yazici

Yardimci

(2010) Effect of aggregate type on mechanical properties of reactive powder concrete. ACI Materials Journal 107: 441–449.

Canbaz

(2014) The effect of high temperature on reactive powder concrete. Construction & Building Materials 70: 508–513.

Cao

Chang

Wang

(2015) Experimental study on flexural capacity of normal section of high strength reinforced reactive powder concrete beam. Journal of Henan Polytechnic University (Natural Science) 34(1): 109–115. (in Chinese)

Cavili

Rebentrost

(2006) Ductal – A high-performance material for resistance to blasts and impacts. Australian Journal of Structural Engineering 7(1): 37–45.

Chen

Dai

Jin

(2015) Refining bond-slip constitutive relationship between checkered steel tube and concrete. Construction & Building Materials 21: 153–164.

Dallaire

Aïtcin

Lachemi

(1998) High-performance powder. Civil Engineering 68(1): 49–51.

Dugat

Roux

Bernier

(1996) Mechanical properties of reactive powder concrete. Materials and Structures 29(4): 233–240.

10.

Feng

Wang

(2019a) Progressive collapse performance analysis of precast reinforced concrete structures. The Structural Design of Tall and Special Buildings 28(5): e1588.

11.

Feng

Xie

Deng

, et al. (2019b) Probabilistic failure analysis of reinforced concrete beam-column sub-assemblage under column removal scenario. Engineering Failure Analysis 100: 381–392.

12.

Cao

, et al. (2012) Brief analysis on crack of unbonded partially prestressed RPC beam. Applied Mechanics and Materials 238: 181–184.

13.

Gao

Liu

Zhang

, et al. (2006) Static properties of plain reactive powder concrete beams. Key Engineering Materials 302–303: 521–527.

14.

Hiremath

Yaragal

(2017) Effect of different curing regimes and durations on early strength development of reactive powder concrete. Construction & Building Materials 154: 72–87.

15.

JGJ 138:2016 (2016) Code for design of composite structures. (in Chinese)

16.

Wang

(2014) Influence of axial compression ratio on seismic behavior of reactive powder concrete (RPC) beam-column joints. Applied Mechanics and Materials 597: 312–315.

17.

Zheng

(2010) Experimental study on mechanical properties of reactive powder concrete. Journal of Harbin Institute of Technology 17(6): 795–800.

18.

Liu

Ren

(2018) Indentation tests based multi-scale random media modeling of concrete. Construction & Building Materials 168: 209–220.

19.

Wang

(2014) Basic mechanical property indexes of reactive powder concrete. Journal of Harbin Institute of Technology 46(10): 1–9. (in Chinese)

20.

Majdi

Hsu

CTT

Sun

(2014) Local bond-slip behavior between cold-formed metal and concrete. Engineering Structures 69(9): 271–284.

21.

Richard

Cheyrezy

(1995) Composition of reactive powder concretes. Cement & Concretce Research 25(7): 1501–1511.

22.

Shi

Long

Cao

, et al. (2016) Mechanical property test and analytical method for reactive powder concrete columns under eccentric compression. KSCE Journal of Civil Engineering 21(2): 1–12.

23.

Wang

(2013) Durability of Green Reactive Powder Concrete. Berlin; Heidelberg: Springer.

24.

Yang

Wang

Peng

(2015) Application and outlook of RPC in ballastless track. Journal of Railway Science and Engineering 12(1): 53–58. (in Chinese)

25.

Zhao

(2005) Design Principle of Composite Structures. Beijing, China: China Higher Education Press. (in Chinese)

26.

Zheng

Wang

(2012a) Compressive behaviour of hybrid fiber-reinforced reactive powder concrete after high temperature. Materials & Design 41: 403–409.

27.

Zheng

Wang

(2012b) Compressive stress-strain relationship of steel fiber-reinforced reactive powder concrete after exposure to elevated temperatures. Construction and Building Materials 35: 931–940.

28.

Zheng

Luo

Wang

(2013) Compressive and tensile properties of reactive powder concrete with steel fibres at elevated temperatures. Construction and Building Materials 41: 844–851.