Experimental investigation on flexural behavior of prefabricated sandwich insulation wall panels with textile reinforced engineered cementitious composites as the wythes after freeze-thaw cycles

Abstract

This paper proposes a novel prefabricated sandwich insulation wall panels (PSIWP) with textile reinforced engineered cementitious composites (TRE) as inner and outer wythes. A parametric analysis conducted using an experimental approach focuses on the influence of the freezing-thawing conditions, the freezing-thawing cycle times, the thickness of the TRE wythes, the type of the thermal insulation layer, the textile treatment methods and the thickness of the thermal insulation layer. The derivation of calculation formulas of fully composite and non-composite TRE-PSIWP, as well as the analysis of the composite degree of TRE-PSIWP were carried out based on the concrete bending capacity calculation formula. It was found that the flexural bearing capacity and the cooperative working performance of TRE-PSIWP weakened after the freeze-thaw cycles. On the contrary, the ductility of TRE-PSIWP increased after the freeze-thaw cycles. Increasing the thickness of thermal insulation layer and TRE wythes were contributed to improve the flexural bearing capacity of TRE-PSIWP. Treating the textiles with impregnation and sand burning was also contributed to improve the flexural bearing capacity of TRE-PSIWP. For different thermal insulation materials, the cooperative working performance of the TRE wythes and the expanded polystyrene board (EPS) was the best.

Keywords

textile reinforced engineered cementitious composites sandwich insulation wall panels flexural behavior freeze-thaw environment composite degree

Introduction

The rise of prefabricated buildings has greatly promoted the development of prefabricated components and structures in recent years, especially prefabricated sandwich insulation wall panels (PSIWP). PSIWPs are consisted of inner and outer wythes, a middle insulation layer and a series of connectors connecting the three layers and transferring the in-plane shear force between two wythes. The shear force originates from the self-weight, the wind load, the seismic load and so on. PSIWPs originated in the cold regions of Europe and America, with a history of more than half a century (Salmon et al., 1994). The composite degree of PSIWPs, namely fully composite, partially composite and non-composite, is determined by the shear transfer capacity of the connectors between the inner and outer wythes (Benayoune et al., 2008). PSIWPs, usually used as enclosures, not only can reduce the energy consumption of the external wall (O’Hegarty and Kinnane, 2020) but also may sustain wind load, seismic load and self-weight. Since wind load and seismic load are the important factors affecting the safety use of PSIWPs, and the PSIWPs usually bend under the action of wind load, many scholars have studied the flexural behavior of PSIWPs.

Carbonari et al. (2012) studied the flexural properties of lightweight PSIWPs consisted of concrete and expanded polystyrene board (EPS) by tests, and found that reducing the angle between the connectors and the insulation panel was helpful to improve the overall bending resistance of the PSIWP. Kim and You (2015) compared the bending performance of PSIWPs using EPS and extruded polystyrene board (XPS) as middle insulation layer, indicating that PSIWPs with EPS were of higher composite degree. Amran et al. (2016) carried out research on the bending capacity of the precast foamed concrete sandwich panel (PFCSP), concluding that the distribution of cracks in PFCSP was similar to that in solid reinforced concrete sandwich panel and precast concrete sandwich panel, as well as the flexural behavior of PFCSP was consistent with that of fully composite wall panel before cracking, but it was the same as that of partial composite wall panel after cracking. Frankl et al. (2011) studied the bending capacity of the prestressed concrete sandwich insulation wall panel by tests, showing that the composite degree of the sandwich panels was close to 100%, especially that of the sandwich panels with carbon-fiber-reinforced polymer (CFRP) grid as connectors and EPS as insulation layer. Xie et al. (2021) investigated the impact resistance of paulownia wood core sandwich panels through the tests, and concluded that increasing face sheet thickness and changing the fiber lay-up sequence could effectively improve the impact damage resistance. Meanwhile, he also investigated and analyzed the influence of different panel and core thicknesses on the bending performance of a special composite sandwich panels (Xie et al., 2022). Chen et al. (2022) studied the flexural properties of precast ceramsite-concrete-insulated sandwich panel (PCCISP), and pointed out that increasing the thickness of the inner concrete wythe had a positive impact on the ultimate bearing capacity of the PCCISP, but reducing the thickness of insulation had no significant influence. Lal and Kumar (2019; Lal and Markad, 2021a) performed the nonlinear progressive failure analysis of piezolaminated composite shell panel under mechanical out-of-plane loading in the hygro-thermal environment, and presented the second-order statistics of hygro-thermo-electrically-induced progressive failure for laminated composite material plate (LCMP). Garhwal et al. (2022) assessed the flexural properties of EPS sandwich concrete panels under different corrosion levels, and the test results showed that the ultimate flexural bearing capacity and ductility of the structure reduced sharply with the increase of corrosion degrees. Daniel Ronald Joseph et al. (2023) studied the failure mode of concrete sandwich panel under axial compression and summarized its general failure mechanism, concluding that the ultimate failure of the panels occurred due to buckling of both wythes.

On the basis of experiment and numerical simulation, many scholars have put forward corresponding calculation theories. In the 1950s, Reissner sandwich theory and Hoff sandwich theory were proposed successively. The Reissner sandwich theory ignores the flexural rigidity of the partition and treats it as a membrane for analysis. However, the Hoff sandwich theory not only considers the flexural stiffness of the partition, but also the transverse shear deformation of the sandwich layer. In the 1990 s, Salmon et al. (1997, Salmon and Einea, 1995) conducted an experimental study on the concrete sandwich panels and established the theoretical system of the sandwich panel. Hassan and Rizkalla (2010) established the evaluation method for composite degree of PSIWPs under different loads. In modern, Markad and Lal (2022) investigated the thermomechanical properties of modified shape memory polymer carbon fiber-reinforced hybrid nanocomposites (SMPHC) through DMA from 30°C to 150°C. Meanwhile, they also performed the post-buckling, static, and dynamic analyses under uniform and non-uniform in-plane axial compressive periodic loading conditions for the first time through evaluating the dynamic instability region (Lal and Markad, 2021b). Choi et al. (2016) and Huang and Dai (2020) proposed two main methods to calculate the composite degree of sandwich panel at different stress stages, that is, the composite degree of stiffness before cracking and the composite degree of the bearing capacity after cracking. Based on the experimental investigation, Liu et al. (2022) proposed the formulas for calculating the cracking load and ultimate load of the sandwich panel composed of reinforcement, concrete and phenolic foam board. Based on the parallel axis theorem and sandwich beam theory, Balıkoğlu and Demircioğlu (2022) derived the formulas for calculating the equivalent and effective bending stiffness of the composite sandwich beams with grid-scored foam.

The inner and outer wythes of PSIWPs adopted in the above studies are mostly novel materials or ordinary concrete, the matrix of which belongs to brittle materials with very low ultimate tensile strain when cracking, that is 0.01% ∼ 0.02%. However, engineered cementitious composite (ECC) is a novel cementitious composite, and has excellent tensile properties. Its tensile strain is 3%∼7%, which is 200∼500 times that of ordinary concrete (Zhang and Li, 2002). Meanwhile, ECC also has the characteristics of multiple cracking and strain hardening (Hou et al., 2020), as well as ultra-high ductility, toughness and strong durability, which is suitable for application in environments with high requirements such as freezing and thawing (Li and Xu, 2009). In engineering applications, due to the stress direction of ECC determined by irregularly distributed short fibers is not clear, ECC matrix is usually combined with the textiles with light weight, corrosion resistance and clear stress direction to form textile reinforced ECC composites (TRE). According to previous studies (Li et al., 2019; Li and Xu, 2011), TRE is of better cracking control ability and tensile properties than TRC, and the number of layers and treatment methods of textiles can impact the bending properties of TRE thin plank.

Research has been conducted to evaluate the flexural properties of PSIWPs with the novel material as the inner and outer wythes, but basically focused on the conventional environment. In cold regions, the free water inside the building will freeze at low temperatures and melt when the temperature rises, seriously affecting the bearing capacity of the building. In coastal areas, PSIWPs not only crack due to volume expansion caused by freeze-thaw cycles, but are also corroded by the large amount of Na⁺ and O₄²⁻ ions in the matrix. Therefore, it is important to study the degradation mechanism of bearing capacity of PSIWPs under freezing-thawing cycles. TRE components with excellent mechanical properties have received extensive attention. Some studies have shown that increasing the number of freezing-thawing cycles can reduce the tensile and flexural performance of TRE thin plank (Yin et al., 2021), as well as the flexural performance of PSIWPs vary with the type and thickness of thermal insulation materials. However, there are limited studies on the flexural performance of PSIWPs with TRE as the inner and outer wythes, especially in the freezing-thawing environment in which PSIWPs repeatedly undergo freezing and thawing.

Therefore, the influences of the freezing-thawing conditions, the freezing-thawing cycle times, the thickness of the TRE wythes, the type of the thermal insulation layer, the textile treatment methods and the thickness of the thermal insulation layer on the flexural properties of TRE prefabricated sandwich insulation wall panels (TRE-PSIWP) were investigated using an experimental approach. The analysis of the failure modes, load-midspan deflection relation curve, flexural bearing capacity, bending stiffness and ductility of TRE-PSIWP, as well as the calculation of the composite degree of TRE-PSIWP were carried out based on the experimental investigation, providing a theoretical basis for the design of TRE-PSIWP in the future.

Test specimens and scheme design

Raw material

The thermal insulation materials adopted in this test are XPS, EPS and rock wool board, which were produced by Jiangsu Huatai Energy Saving and Thermal Insulation Technology Ltd, and their characteristics are shown in Table 1. The textile used in this test is a carbon-glass fiber hybrid mesh, which were produced by Toray (Japan) Ltd. The type of carbon fiber and its properties and geometric parameters are shown in Table 2. It should be noted that the carbon-glass fiber hybrid mesh consists of the latitudinal carbon fiber bundle arranged along the force direction and the radial alkali-resistant glass fiber bundle playing a fixed role. The connectors are typically made of concrete, steel, and fiber-reinforced polymer (FRP). Due to the higher thermal conductivity of steel and concrete, the thermal bridge effect occurs in the PSIWPs, which compromises the overall thermal efficiency of the PSIWPs. In contrast, FRP has lower thermal conductivity and high strength, so it is more suitable for making connectors. The main types of FRP are Carbon Fiber-reinforced Polymer (CFRP), Glass Fiber-reinforced Polymer (GFRP) and Basalt Fiber-reinforced Polymer (BFRP), of which BFRP is cheaper and more environmentally friendly. Therefore, BFRP bars with a diameter of 6 mm were used as connectors in this test, which were produced by Jiangsu Greenwood Valley New Material Technology Development Ltd, and its ultimate tensile strength, elastic modulus and elongation were 1281 MPa, 60.13 GPa, 2.14%, respectively.

Table 1.

Properties and geometric parameters of EPS and XPS.

Thermal insulation type	Thermal conductivity (W/(m·K))	Density (kg/m³)	Tensile strength (MPa)	Compressive strength (MPa)
EPS	0.039	16.1	0.17	0.16
XPS	0.030	27.2	0.28	0.19
Rock wool	0.038	121.8	0.013	0.073

Table 2.

Properties and geometric parameters of carbon fiber.

Fiber type	Carbon (T700S)	Impregnated carbon fiber	Carbon fiber with impregnation and sand burning
Number of filaments	12,000	12,000	12,000
Tensile strength (MPa)	4660	4100	4100
Modulus of elasticity (GPa)	231	180	180
Yarn tex (g/cm³)	801	—	—
Density (g/cm³)	1.78	—	—
Theory area (mm²)	—	0.45	0.45

The ECC matrix is mainly made of cement, fly ash, quartz sand, water, water reducer and polyvinyl alcohol (PVA) fiber, and their contents are 379 kg/m³, 885 kg/m³, 455 kg/m³, 379 kg/m³, 17.4 kg/m³, 26 kg/m³ (equivalent to 2% volume content), respectively. The suppliers of the cement, fly ash, quartz sand, water reducer and PVA fiber were Xuzhou Zhonglian Cement Ltd, Xuzhou Foundry Concrete Ltd, Inner Mongolia Baotou Fumiao Environmental Protection Technology Ltd, Sika (China) Ltd, Kuraray (Japan) Ltd, respectively. The mechanical properties of the ECC matrix were investigated. The cubic compressive strength, ultimate tensile strength and tensile strain of ECC were 52 MPa, 7.41 MPa, and 3.12%, respectively. The flexural properties of TRE panels with the size of 400 mm × 100 mm × 15 mm were investigated to clearly understand the force mechanism of TRE-PSIWP, and three specimens were used to avoid accidental errors. The four-point bending test was carried out with a self-made test loading device at a loading rate of 0.5 mm/min. The mid-span load and deflection of the specimen were measured by a 2T load sensor and a displacement meter with a range of 50 mm, as shown in Figures 1 and 2. The test results of TRE panels were showed in the Table 3, and TRE-1, TRE-2, TRE-3 are the number of three specimens respectively.

Figure 1.

Device used for loading.

Figure 2.

Schematic of four-point bending test.

Table 3.

Four-point bending test results of TRE panels.

Specimens id	Initial cracking load (kN)	Midspan deflection of initial crack (mm)	Peak load (kN)	Midspan deflection corresponding to peak load (mm)
TRE-1	0.55	0.78	1.60	11.01
TRE-2	0.57	0.75	1.93	11.35
TRE-3	0.51	0.70	1.66	11.62
Average value	0.54	0.74	1.73	11.32
Standard deviation	0.03	0.03	0.14	0.25

Design and preparation of specimens

All specimens are composed of the inner and outer TRE wythes with the same thickness, a thermal insulation panel, and a series of BFRP bars connecting three layers into a whole. The materials of thermal insulation layer are XPS, EPS and rock wool. Based on the experimental results, Carbonari et al. (2012) suggested that a certain angle should be adopted when inserting the connector to increase its contribution to flexural stiffness. However, the insertion angle of the connector is too small to facilitate construction operation. Therefore, considering the connector’s contribution to flexural capacity and ductility as well as the difficulty of processing, the insert angle of BFRP bars in this test was set as 60°. Two rows of BFRP bars were arranged in each specimen to ensure the connection of the inner and outer TRE wythes. In principle, the spacing of BFRP bars along the length and width direction is the same. But there was a certain angle of the inserted connector, so the spacing along the length direction needed to be greater than that along the width direction to avoid the mutual influence caused by the small end spacing between the BFRP bars. Therefore, the spacing of BFRP bars along the length direction was set at 150 mm, and the spacing along the width direction was set at 100 mm. Figures 3 and 4 show the detailed dimensions of the specimen and the spacing of the BFRP bars. According to the test results, Yin et al. (2021) pointed out that the flexural capacity of TRE increased by 123.3% when the number of textiles increased from one to two. However, according to the existing studies (Daskiran et al., 2022; Zargaran et al., 2023), the degree of improvement in the flexural bearing capacity of TRC decreases with the increase of the number of textiles, while the impact strength of TRC is almost unaffected. In the meantime, according to the research results of Contamine and Si Larbi (2016), it is found that the efficiency ratio of weft yarn for stiffness is the largest when the two-layer textiles is set. Therefore, each TRE wythe was equipped with two layers of textiles, and its thickness (15, 20 and 30 mm) was one of the parameters in this study. In engineering practice, the thickness of insulation panels is usually 50 mm to 100 mm, so the insulation board thickness of the specimen is selected as 50 mm, 70 mm and 100 mm. The size of each specimen is 1200 mm × 300 mm, and the total thickness is determined by the thickness of insulation panel and TRE wythes. The prepared specimen is shown in Figure 5.

Figure 3.

Design diagram of specimens (unit: mm).

Figure 4.

Schematic diagram of connector layout (unit: mm).

Figure 5.

Specimen after demoulding.

In order to get closer to the real freeze-thaw cycles, the weather resistant adhesive was smeared on the side of the insulation panel to prevent the water and Na₂SO₄ solution from corroding the insulation board from the side. The freeze-thaw cycle was carried out in the temperature cycle rain box. The operating procedures of freeze-thaw cycles are as follows: The specimens were placed at (20 ± 2)°C for 8 h and sprayed with water at a rate of (1.0 ± 0.1) L/(m²·min), and then the temperature of the box was reduced to −20°C and maintained at (−20 ± 2)°C for 16 h, which was a cycle. The cracks and shedding on the specimen surface were observed and recorded after each three cycles.

Test group

The component information of each group of specimens is shown in Table 4, which gives 13 individual specimens in total. Among them, a group of specimens without freeze-thaw cycles were considered for comparison. All specimens were named according to a certain parameter sequence, which was successively the freeze-thaw environment, the number of freeze-thaw cycles, the type of thermal insulation layer, the thickness of thermal insulation layer, the textile treatment methods and the thickness of TRE wythes. For example, W30-X70-S30 refers to that the specimen’s freeze-thaw environment was water solution, the number of freeze-thaw cycles was 30, the insulation material was XPS board, the thickness of the insulation panel was 70 mm, the textile treatment methods was impregnation, and the thickness of TRE wythes was 30 mm. It should be noted that the insertion angle (the angle between the BFRP bars and the thermal insulation layer) of the connectors was 60° in all specimens, as well as the anchorage depth of BFRP bars of the specimens named W30-X70-S15*, W30-X70-S20* and W30-X70-S30* was 15 mm, and that of other specimens was 20 mm.

Table 4.

The component information of specimen in freeze-thaw environment.

Specimen id	Freeze-thaw environment	Number of freeze-thaw cycles	Thermal insulation layer type	Thermal insulation layer thickness (mm)	Textile treatment methods	TRE wythes thickness (mm)
X70-S30	—	—	XPS	70	Impregnation	30
W30-X70-S15*	Water	30	XPS	70	Impregnation	15
W30-X70-S20*	Water	30	XPS	70	Impregnation	20
W30-X70-S30*	Water	30	XPS	70	Impregnation	30
W10-X70-S30	Water	10	XPS	70	Impregnation	30
W20-X70-S30	Water	20	XPS	70	Impregnation	30
W30-X70-S30	Water	30	XPS	70	Impregnation	30
L30-X70-S30	5% Na₂SO₄ solution	30	XPS	70	Impregnation	30
W30-E70-S30	Water	30	EPS	70	Impregnation	30
W30-R70-S30	Water	30	Rock wool	70	Impregnation	30
W30-X70-N30	Water	30	XPS	70	Unprocessed	30
W30-X70-I30	Water	30	XPS	70	Impregnation and sand burning	30
W30-X50-S30	Water	30	XPS	50	Impregnation	30
W30-X100-S30	Water	30	XPS	100	Impregnation	30

Test loading scheme

In this paper, the flexural properties of TRE-PSIWP were studied by four-point bending test. The specimen was placed on two fixed bearings. A steel plate with a size of 400 mm × 80 mm × 10 mm and a steel roller were arranged between the specimen and the bearing. The four-point bending test diagram was shown in Figure 6. A vertical load, provided by a 10-ton hydraulic jack, was applied by the distribution beam. The loading points were arranged symmetrically on the upper surface of the specimen 150 mm away from the central axis. The load was applied with force control at a loading per stage of 1 kN for 10 min. A 15-ton load cell was placed above the distribution beam to measure the vertical load. Five displacement gauges were used to measure the vertical displacement of the midspan and bearings, as well as the horizontal displacement of the TRE panels. The measuring range of the displacement gauge at the midspan was 50 mm, and the rest were 30 mm.

Figure 6.

Test loading and measurement device. (a) Sketch of the test loading device (unit: mm). (b) Photograth of the test device.

Test results

Damage and failure modes

Figure 7 shows the surface characteristics of TRE-PSIWPs after freeze-thaw cycles. After freeze-thaw cycles, all specimens remained intact without cracking and shedding, as well as salt weathering appeared on the surface of the wythes immersed in 5% Na₂SO₄ solution.

Figure 7.

Surface shape characteristics of the PSIWPs after freeze-thaw cycles. (a) Specimen surface after freeze-thaw cycles. (b) Salt weathering.

Table 5 shows the test results of each specimen after freeze-thaw cycles, including failure mode, cracking load, peak load and corresponding midspan deflection. A flexural crack appeared near the loading points of the bottom wythe when the load reached about 50% of the peak load varying with the change of parameters. This crack formed may be because this location being subjected to a large bending moment and shear force at the same time. Many new flexural cracks mostly appeared between the two loading points with the increase of load, and then the bond between the wythes and the thermal insulation panel gradually failed. However, the debonding modes varied with the change of the parameters. Among them, the debonding mode of the specimens with freeze-thaw cycles of 30 times in water, 30 mm or 15 mm TRE wythes, impregnated textile, or XPS or rock wool as thermal insulation panel was the end debonding between the bottom wythe and the thermal insulation panel, and that of the rest was the end debonding between the upper wythe and the thermal insulation panel Among the specimens with debonding between the upper wythe and the thermal insulation layer, except with 20 mm TRE wythes and freeze-thaw cycles in 5% Na₂SO₄ solution, the connector inside the specimen was broken, that was, the connector was pulled out or pulled off. The failure mode of the specimens named X70-S30, W30-X70-S30 and W30-X100-S30 was thermal insulation panel fracture. Among the residual four specimens, the failure mode of the specimen with rock wool board was the shear failure of rock wool board, and that of the specimen with 50 mm insulation board was splitting failure of connectors. Penetrating cracks extending from the side to the top of the upper TRE wythes appeared near the bearings in the specimens with 15 mm or 20 mm TRE wythes.

Table 5.

Test results of specimens under freeze-thaw environment.

Specimen id	Initial cracking load (kN)	Midspan deflection of initial crack (mm)	Peak load (kN)	Midspan deflection corresponding to peak load (mm)	Ultimate midspan deflection (mm)	Ductility index	Failure modes
X70-S30	12	2.63	22	13.38	19.15	5.09	End debonding between the bottom wythe and the thermal insulation panel, thermal insulation panel fracture.
W30-X70-S15*	4.5	2.18	10.75	13.12	15.4	6.02	Local debonding between the bottom wythe and the thermal insulation panel, penetrating cracks at the upper wythe.
W30-X70-S20*	5	1.47	11	9.83	17.13	6.69	End debonding between the upper wythe and the thermal insulation panel, penetrating cracks at the upper wythe.
W30-X70-S30*	6	1.43	16	15.91	16.36	11.12	End debonding between the bottom wythe and the thermal insulation panel.
W10-X70-S30	8	1.38	15.5	9.55	18.24	6.92	End debonding between the upper wythe and the thermal insulation panel, connector pulled-out.
W20-X70-S30	8	1.56	12.5	9.15	27.4	5.87	End debonding between the upper wythe and the thermal insulation panel, connector pulled-out.
W30-X70-S30	7	1.38	13	14.9	17.72	10.80	End debonding between the bottom wythe and the thermal insulation panel, thermal insulation panel fracture
L30-X70-S30	6	1.52	13.5	20.29	21.14	13.35	End debonding between the upper wythe and the thermal insulation panel, upper wythe local punching failure.
W30-E70-S30	9	2.68	16.5	17.24	23.04	6.43	End debonding between the upper wythe and the thermal insulation panel, connector pulled-off.
W30-R70-S30	7	3.3	11	19.13	20.73	5.80	End debonding between the bottom wythe and the thermal insulation panel, interface shear failure of rock wool panel.
W30-X70-N30	9	2.26	16.5	19.81	21.26	8.76	End debonding between the upper wythe and the thermal insulation panel, connector pulled-out.
W30-X70-I30	9	1.65	20	6.6	17.82	4	End debonding between the upper wythe and the thermal insulation panel, connector pulled-out.
W30-X50-S30	4	1.57	11.5	17.74	20.94	11.30	End debonding between the bottom wythe and the thermal insulation panel, connector splitting failure.
W30-X100-S30	7	1.49	21	12.85	15.44	8.62	End debonding between the bottom wythe and the thermal insulation panel, thermal insulation panel local fracture.

Figure 8 shows the main failure modes of the specimens. Combined with Table 5 and Figure 8, it can be seen that the freeze-thaw environment, the number of freeze-thaw cycles, the thickness of TRE wythes, the type and thickness of thermal insulation layer and the textile treatment methods all have an impact on the failure mode. Changing the freeze-thaw environment would lead to the failure mode from the insulation panel fracture to the punching failure of the upper wythe, and the failure mode changed to the pull out of connectors with the decrease of the number of freeze-thaw cycles. However, the failure mode of the specimen undergoing 30 times freeze-thaw cycles in water was consistent with that of the specimen without freeze-thaw cycles. This may be because the damage to the wythes was more serious at the early stage of the freeze-thaw cycles, while the damage to the insulation material was increased at the later stage of the freeze-thaw cycles. Therefore, when the number of freeze-thaw cycles increased to 30, the failure mode changed to the insulation panel fracture. Due to the low stiffness of the wythes with a small thickness, a penetrating crack would appear on the top surface of the upper wythe. However, when its thickness increased to 30 mm, the upper wythe was sufficient to bear the bending load and avoid penetrating cracks. The insulation panel was damaged in all specimens except those with EPS insulation panel. Among them, XPS suffered fracture failure and rockwool board suffered shear failure, indicating that the failure mode of PSIWP under bending load was closely related to the stiffness of insulation material. The pulling out of the connectors occurred in the specimens whose textiles was sanded, and whose textiles was not treated, while insulation panel suffered fracture failure in the specimen whose textiles was impregnated with epoxy resin. The failure mode of the specimen with the 50 mm insulation panel was the splitting failure of BFRP bars. This may be because the stiffness of the insulation panel decreased with the decrease of its thickness. Therefore, the splitting failure of BFRP bars was easy to occur under the interaction of bending and shearing, while the insulation panel fracture occurred in the specimen with higher insulation panel thickness.

Figure 8.

Main failure modes of PSIWPs. (a) End debonding between the upper wythe and the thermal insulation panel. (b) End debonding between the bottom wythe and the thermal insulation panel. (c) Interface shear failure of rock wool panel. (d) Connector splitting failure. (e) Thermal insulation panel fracture. (f) Penetrating cracks at the upper TRE wythe. (g) Crack distribution of the bottom TRE wythe.

Bearing capacity and load-midspan deflection relation curve

This section analyzed the effect of research parameters on the flexural properties of the PSIWPs under freeze-thaw environment from the aspects of bearing capacity and load-midspan deflection relation curve.

Number of freeze-thaw cycles

Figure 9(a) shows the load-midspan deflection curves under different number of freeze-thaw cycles. Table 5 shows the peak load and midspan deflection corresponding to peak load of the specimen. The initial cracking load and peak load were reduced after freeze-thaw cycles, and those of the specimen undergoing 30 times freeze-thaw cycles suffered the most significant reduction, which decreased by 41.67% and 40.91%, respectively. The initial cracking load and peak load showed a decreasing trend with the increase of the number of freeze-thaw cycles, but the midspan deflection corresponding to peak load increased. The initial cracking load, peak load and midspan deflection corresponding to peak load of the specimens undergoing 30 times freeze-thaw cycles were 12.5% and 16.13% lower, as well as 56.02% higher than those of the specimens undergoing 10 times freeze-thaw cycles, respectively. This decrease in the initial cracking load associated with the tensile strength of the ECC was because increasing the number of freeze-thaw cycles aggravated the frost heaving damage of the ECC matrix. The reason for this reduction might be that the aggravation of the frost heaving damage of the ECC matrix reduced the interface bonding capacity between the wythes and the insulation panel and the flexural properties of the insulation panel itself, as well as expanded the crack width of the bottom wythe, resulting in the decrease of the peak load and the increase of the midspan deflection corresponding to the peak load. As shown in Figure 9(a), different from the expectation, with the increase of the number of freeze-thaw cycles, the initial stiffness was similar, but the ultimate bearing capacity decreased, indicating that the freeze-thaw cycles had little effect on the initial flexural stiffness of the specimens. The possible reason for this result was that the structure size and cross-section shape were not impact by the freeze-thaw cycles before the cracking of the specimen, but the internal micro-cracks caused by the freeze-thaw cycles developed rapidly after the cracking, reducing the effective section area. Therefore, the flexural stiffness after cracking decreased with the increase of the number of freeze-thaw cycles, and the flexural capacity also decreased.

Figure 9.

Load-midspan deflection curve of the specimens after freeze-thaw cycles. (a) Number of freeze-thaw cycles. (b) Freeze-thaw environment. (c) Type of thermal insulation layer. (d) Textile treatment methods. (e) TRE wythes thickness. (f) Thermal insulation layer thickness.

Freeze-thaw environment

Figure 9(b) shows the load-midspan deflection curves under different freeze-thaw environment. Compared with those of the specimens undergoing the freeze-thaw cycles in water, the flexural stiffness of the specimen undergoing freeze-thaw cycles in 5% Na₂SO₄ solution decreased significantly before cracking, which might be due to the sulfate ions into the ECC matrix reacted with the cement hydration products to generate expansion substances such as ettringite, resulting in the formation of early microcracks and the aggravation of freeze-thaw damage. Meanwhile, the initial cracking load and midspan deflection corresponding to the initial cracking load of the specimens undergoing freeze-thaw cycles in 5% Na₂SO₄ solution were 14.29% lower and 10.15% higher than those undergoing the freeze-thaw cycles in water, respectively. However, the peak load and midspan deflection corresponding to the peak load of the former were 3.85% and 19.30% higher than that of the latter respectively, which might be due to the early microcracks in the specimens could self-healing at the late stage of the freeze-thaw cycles (Şahmaran and Li, 2007), contributing to the improvement of the ultimate flexural bearing capacity after cracking. Furthermore, the fluctuation of the load-midspan deflection curves under different freeze-thaw environment appeared after cracking, and that undergoing freeze-thaw cycles in 5% Na₂SO₄ solution was more obvious.

Type of thermal insulation layer

Figure 9(c) shows the load-midspan deflection curves under different type of thermal insulation layer. The initial flexural stiffness of the specimens with XPS insulation material was the highest, followed by EPS and rock wool insulation material, indicating that the initial flexural stiffness was related to the performance of the thermal insulation panel. Due to the good bonding between the wythes and the thermal insulation panel before cracking, the specimens possessed a high degree of composite and the ability of working together of the TRE wythes and the thermal insulation panel. As shown in Table 1, the tensile and compressive strength of XPS was the highest, so the specimen with XPS could be of higher flexural stiffness compared with the specimen with EPS and rock wool to resist bending load. However, the peak load and midspan deflection corresponding to peak load of the specimen with XPS after cracking decreased by 26.92% and 30.02% compared with that of the specimens with EPS, respectively. The reason for this reduction was that the interfacial bonding performance between the wythe and the EPS after freeze-thaw cycles was better than that between the wythe and the XPS, leading to the decrease of the composite degree of the specimens with XPS after cracking was larger than that of the specimens with EPS. In addition, because the compressive strength of rock wool was the worst and less than 50% of the XPS and EPS, the peak load of the specimens with rock wool was the lowest.

Textile treatment methods

Figure 9(d) shows the load-midspan deflection curves under different textile treatment methods. The load-midspan deflection curves of these specimens before cracking were basically coincident, indicating that changing the treatment methods of textiles had no effect on the initial flexural stiffness. The cracking load and peak load of the specimen with textiles treated by impregnating and sand burning were higher than those of others, but the midspan deflection corresponding to peak load was lower than that of others. This increase of bearing capacity was due to the bonding performance between the textiles treated with impregnation and sand burning and the ECC matrix was best, which contributed to improve the overall working ability and mechanical performance of the TRE wythes. However, the flexural capacity of the specimens with impregnated textiles were the worst, which might be due to the degradation of the performance of the epoxy resin adhesive caused by the freeze-thaw cycles, resulting in the decrease of the interfacial bonding between the textiles and the ECC matrix. At the same time, the force of loose carbon yarn caused by impregnation was uneven, which made it was difficult to give full play to the tensile strength of textiles.

Thickness of TRE wythes

Figure 9(e) shows the load-midspan deflection curves under different thickness of TRE wythes. The load-midspan deflection curves of the specimens with 20 mm or 30 mm TRE wythes basically coincided before cracking, while the curve slope of the specimens with 15 mm TRE wythes was lower than that of others. This suggested that increasing the thickness of the TRE wythes was helpful to improve the initial flexural stiffness, but the improvement degree of the initial flexural stiffness was reduced with the increase of the thickness of TRE wythes. Increasing the thickness of TRE wythes could help to improve the cracking load and the peak load, but would slightly reduce the midspan deflection corresponding to peak load. Compared with those of others, the cracking load of the specimens with 30 mm TRE wythes increased by 33.33% and 20% respectively, and the peak load increased by 48.83% and 45.45% respectively. Owing to the thinner TRE wythes, the bearing capacity and ability to resist cracking of the TRE wythes was smaller. Therefore, the bonding interface between the TRE wythes and the thermal insulation panel delaminated early after freeze-thaw cycles, resulting in the decrease of the ultimate bearing capacity.

Thickness of thermal insulation layer

Figure 9(f) shows the load-midspan deflection curves under different thickness of thermal insulation layer. The load-midspan deflection curves of these specimens were almost coincident at the beginning of loading, indicating that they possessed the same initial flexural stiffness. And then, the specimens with the 50 mm thermal insulation layer cracked first, and the flexural stiffness decreased rapidly. The initial cracking load of the specimens with 70 mm and 100 mm thermal insulation layer was similar, but the midspan deflection corresponding to the cracking load of the former was lower. Increasing the thickness of thermal insulation panel could help to improve the peak load, but would reduce the midspan deflection corresponding to peak load. Compared with those of others, the peak load of the specimens with 100 mm thermal insulation layer increased by 82.61% and 61.54%, respectively, and the midspan deflection corresponding to the peak load decreased by 26.27% and 12.87%, respectively. Therefore, increasing the thickness of the thermal insulation panel contributed to improve the flexural bearing capacity of the specimens after freeze-thaw cycles in water, but would reduce the deformation capacity.

Ductility analysis

Ductility, a very important index to evaluate the behavior of PSIWPs, directly affects the failure mode and deformation capacity of the specimens. Owing to PSIWPs with different research parameters are of different ductility under the bending load, the ductility of each specimens needs to be calculated and analyzed. In general, ductility is expressed by the ratio of midspan deflection corresponding to the peak load to midspan deflection corresponding to the yield load (Tomlinson and Fam, 2015). However, the specimens studied in this paper did not have yield stage, so the midspan deflection corresponding to the initial cracking load was applied to replace the midspan deflection corresponding to the yield load. The calculation formula is as follows.

μ = \frac{∆_{u}}{∆_{cr}}

(1)

where, μ is the ductility index of the specimens; Δ_u and Δ_cr are the midspan deflection corresponding to the peak load and the midspan deflection corresponding to the initial cracking load of the specimens, respectively.

The ductility index of each specimen after freeze-thaw cycles is shown in Table 5. The specimens still maintained superior ductility after freeze-thaw cycles, and the ductility had an increasing trend with the increase of the number of freeze-thaw cycles. The ductility of the specimens with 15 mm or 20 mm TRE wythes was similar, but that of the specimens with 30 mm TRE wythes was increased by 45.9% and 39.8% respectively compared with others, indicating that increasing the thickness of TRE wythes could effectively improve the ductility of the specimens. The textile treatment methods had a great impact on the ductility, and the ductility of the specimens with textile treated by impregnating and sand burning was the lowest. This change occurred because the bonding interface between the textile treated by impregnating and sand burning and the ECC matrix was still intact after freeze-thaw cycles, and the specimens possessed a high degree of composite, which could better sustain the bending load. Owing to the deformation capacity of the insulation material was weaker than that of the TRE wythes, the stiffness increased with the increase of the thickness of thermal insulation panel, resulting in the ductility decreased. The ductility index of the specimens after freeze-thaw cycles in the 5% Na₂SO₄ solution was higher than that in the water, indicating that the corrosion of sulfate on the specimens and the frost heaving damage led to the decrease of the stiffness, inducing a large bending deformation in the later loading stage.

Calculation and analysis of composite degree

The performance of PSIWPs under bending load is closely related to the composite degree of PSIWPs. Therefore, in order to further evaluate the flexural properties of TRE-PSIWPs, it is necessary to understand the influence of study parameters on the composite degree of TRE-PSIWPs. There are two main methods to determine the composite degree of PSIWPs. One is to determine the standard of fully composite and non-composite wall panel by test results, that is, the test results of PSIWPs without connectors are used as the standard of non-composite wall panels, and the test results of concrete solid panels are used as the standard of fully composite wall panels (Joseph et al., 2017). The other is to obtain the ultimate flexural bearing capacity of fully composite wall panels and non-composite wall panels through theoretical calculation (Amran et al., 2016; Lee and Pessiki, 2008; Mohamad et al., 2014; Pessiki and Mlynarczyk, 2003). In this section, the simplified calculation method was applied to determine the composite degree of PSIWPs, that is, the flexural bearing capacity of the equivalent model of the fully composite wall panels and the non-composite wall panels was calculated respectively, and then the composite degree of the specimens was obtained by the strength calculation theory. The stress distribution of fully composite wall panels and non-composite wall panels is shown in Figure 10. As shown in Figure 10(a), the inner and outer TRE wythes of the non-composite wall panels are subjected to load separately, and the connectors do not transmit longitudinal shear force. As shown in Figure 10(b), the inner and outer TRE wythes of the fully composite wall panels can form a whole to resist the bending load, and the connectors must transmit the required longitudinal shear force.

Figure 10.

Stress distribution of PSIWPs. (a) Stress distribution of non-composite wall panels. (b) Stress distribution of fully composite wall panels.

Flexural bearing capacity of non-composite PSIWPs

The upper and bottom TRE wythes were regarded as two fully independent panels when calculating the flexural bearing capacity of non-composite PSIWPs, and the stress distribution was also fully independent.

Owing to the excellent tensile capacity of ECC matrix, the tensile stress of ECC in TRE wythes should be taken into account when calculating the cross-sectional stress of the tensile zone. However, only the bearing capacity of the ECC between the bottom textile and the neutral axis was considered when calculating the tensile stress of ECC since the ECC with cracking no longer possessed the bearing capacity. The calculation formula of the flexural bearing capacity of non-composite wall panels are described as follows.

{\begin{cases} F_{e c} = 0.85 b h_{c} f_{e c} \\ F_{e t} = \frac{1}{2} b f_{e t} (\frac{2}{3} d - h_{c}) \end{cases}

(2)

F_{t} = k_{1} k_{2} A_{t} σ_{t u}

(3)

F_{e t} + F_{t} = F_{e c}

(4)

M_{u 1} = F_{t} (\frac{d}{2} - \frac{h_{c}}{2}) + F_{e t} (\frac{4 d}{9} - \frac{h_{c}}{6})

(5)

F = \frac{2 M_{u 1}}{l}

(6)

where F_ec and F_et are the pressure and the tension of ECC in a single TRE wythes, respectively; F_t is the tension of textiles in a single TRE wythes; F and M_u1 are the theoretical value of the flexural bearing capacity and the bending moment of a single TRE wythes, respectively; b is the width of the specimen; h_c is the height of compression zone; f_ec and f_et are the compressive strength and the tensile strength of ECC matrix; σ_tu is the tensile strength of carbon yarn; A_t is the total cross-sectional area of the tensile carbon textiles in a single TRE wythes; d is the height of a single TRE wythes; χ is the rate of distributing textiles, which is the ratio of the total cross-sectional area of the tensile carbon yarns to the cross-sectional area of the TRE wythes; l is the distance from the bearing to the adjacent loading point.

The tensile strength of the two-layer textiles needs to be reduced because the tensile stress of the textile near the compression zone does not reach its tensile strength under bending load. The reduction coefficient k₁ is related to the thickness of TRE wythes and the height of compression zone, and is presented as

k_{1} = \frac{3 d - 6 h_{c}}{4 d - 6 h_{c}}

(7)

Meanwhile, the effectiveness of the textiles gradually decreases with increasing reinforcement ratio (Hegger et al., 2006), so the tensile strength of the textiles needs to be reduced. The reduction coefficient k₂ is linear with the rate of distributing textiles, and is presented as

k_{2} = - 24.8 χ + 0.7

(8)

According to the above mechanical performance parameters of the ECC and the textiles, as well as the test results of TRE panels, it can be obtained that f_ec = 52 MPa, f_et = 7.41 MPa, b = 100 mm, d = 15 mm, A_t = 7.2 mm², and σ_tu = 4660 MPa. It can be obtained by combining formula (2)–(8) that the theoretical value of the height of compression zone, the flexural bearing capacity and the bending moment of a single TRE wythe are 3.32 mm, 1.726 kN and 86.336 kN·m, respectively. The theoretical value of the peak load was compared with the test results of TRE panels, and the error was 0.003%. Therefore, it is feasible to apply the above formulas to calculate the flexural bearing capacity of a single TRE wythe.

Since the TRE-PSIWPs contain two TRE wythes, the flexural bearing capacity of non-composite wall panels should be twice of F, and is presented as

F_{n u} = 2 F

(9)

where F_nu is the theoretical value of the flexural bearing capacity of non-composite wall panels.

Flexural bearing capacity of full composite PSIWPs

The upper and bottom TRE wythes and the middle thermal insulation layer are regarded as a whole when calculating the flexural bearing capacity of the fully composite PSIWPs, and there is no relative slip between the thermal insulation layer and the TRE wythes. The stress distribution of TRE-PSIWPs is similar with that of solid plate when calculating the bending moment, that is, the bottom TRE wythes is in full section tensile state, and the stress distribution of it almost completely reaches the tensile strength. Therefore, it is considered that the stress of all ECC in bottom TRE wythes reaches its tensile strength without considering the tensile force of ECC below the bottom textile and in the upper TRE wythes to facilitate calculation. The effect of reduction coefficient k₁ is no longer considered when calculating the tension of textiles. The calculation formula of the flexural bearing capacity of full composite wall panels are described as follows.

{\begin{cases} F_{e c} = 0.85 b h_{c} f_{e c} \\ F_{e t} = \frac{2}{3} b f_{e t} d \end{cases}

(10)

F_{t} = k_{2} A_{t} σ_{t u}

(11)

F_{e t} + F_{t} = F_{e c}

(12)

M_{c u} = F_{t} (h - \frac{d}{2} - \frac{h_{c}}{2}) + F_{e t} (h - \frac{2 d}{3} - \frac{h_{c}}{2})

(13)

F_{c u} = \frac{2 M_{c u}}{l}

(14)

where F_cu and M_cu are the theoretical value of the flexural bearing capacity and the bending moment of full composite wall panels, respectively; h is the section height of the specimens.

Composite degree of the specimens

The calculation formula of the composite degree of the specimens in the ultimate load is described as follows.

κ_{u} = \frac{F_{e} - F_{n u}}{F_{c u} - F_{n u}} \times 100 (%)

(15)

where κ_u is the composite degree of the specimens in the ultimate load; F_e is the experiment value of the flexural bearing capacity of the specimens.

The theoretical value of the flexural bearing capacity of the specimens after freeze-thaw cycles needs to be reduced since the flexural bearing capacity of TRE wythes is weakened by freeze-thaw cycles. The relationship between the influence coefficient of freeze-thaw cycles and the cycle times of freeze-thaw is as follows (Yin et al., 2021).

λ = - 0.0048 N + 0.98

(16)

Table 6 shows the composite degree of the specimens in the ultimate load. The composite degree of the specimens decreased sharply after freeze-thaw cycles, indicating that the freeze-thaw cycles weakened the interfacial bonding between the insulation layer and the TRE wythes. The composite degree of the specimens reduces with the increase of the thickness of TRE wythes, suggesting that the TRE wythes with lower thickness had better cooperative work ability with the insulation layer. The composite degree of the specimens with textiles treated by impregnating and sand burning was higher than that of the specimens with impregnated textiles and untreated textiles. This situation occurred was due to the better bonding properties between the textiles treated by impregnating and sand burning and the ECC matrix, which could control the deformation of ECC to a certain extent, reducing the degree of the debonding between the insulation layer and the TRE wythes caused by the uncooperative deformation. The composite degree of the specimens increased with the increase of the thickness of thermal insulation layer. The reason for this increase was that the midspan deflection corresponding to the peak load reduced with the increase of the thickness of thermal insulation layer, which led to the reduction of the debonding degree between the thermal insulation panel and the TRE wythes. In addition, the specimens with rock wool as the insulation layer was of the lowest composite degree, indicating that the cooperative working ability of rock wool panel with TRE wythes was very poor. Therefore, it is not recommended to use the rock wool panels as the insulation layer of TRE-PSIWP.

Table 6.

The composite degree of the specimens in the ultimate load.

Specimen id	F_e (kN)	The theoretical value of the flexural bearing capacity (kN)		κ_u (%)
Specimen id	F_e (kN)	F _nu	F _cu	κ_u (%)
X70-S30	22	9.92	76.55	18.1
W30-X70-S15*	10.75	2.78	38.15	22.5
W30-X70-S20*	11	4.43	46.62	15.6
W30-X70-S30*	16	8.32	64.00	13.8
W10-X70-S30	15.5	9.23	71.35	10.1
W20-X70-S30	12.5	8.80	67.67	6.3
W30-X70-S30	13	8.32	64.00	8.4
L30-X70-S30	13.5	8.32	64.00	9.3
W30-E70-S30	16.5	8.32	64.00	14.7
W30-R70-S30	11	8.32	64.00	4.8
W30-X70-N30	16.5	8.92	68.95	12.6
W30-X70-I30	20	8.32	64.00	21.0
W30-X50-S30	11.5	8.32	56.27	6.6
W30-X100-S30	21	8.32	87.97	15.9

Conclusions

A precast sandwich insulation wall panel with TRE as the inner and outer wythes was proposed. Four-point bending tests were conducted to obtain the flexural properties (i.e., ductility, load-midspan deflection relation curve, and failure modes) of the TRE-PSIWPs. The effects of the freeze-thaw environment, the number of freeze-thaw cycles, the thickness of TRE wythes, the type and thickness of thermal insulation layer and the textile treatment methods were discussed. Furthermore, the derivation of the calculation formula for the flexural bearing capacity of fully composite and non-composite PSIWPs as well as the analysis of the composite degree of the specimens were conducted based on the calculation formula of flexural bearing capacity of concrete. However, this study does not involve the thermal performance of TRE-PSIWPs and the influence of cracks caused by high temperature on the flexural properties of TRE-PSIWPs. Because of these properties of TRE-PSIWPs as insulation member cannot be ignored, further research is needed on thermal performance of TRE-PSIWPs and flexural properties of TRE-PSIWPs after high temperature. In summary, the following conclusions were drawn from the study.

(1) For failure mode, in addition to the thickness of TRE wythes, other factors would cause the change of the failure mode. The failure of most specimens was due to the connectors were pulled out, while a few was due to the connector splitting, the TRE surface punched, the insulation panel fracture or the interface shear failure. The interface debonding between the TRE wythes and the insulation layer occurred in all specimens, as well as the location of debonding was associated with all the research variables except the thermal insulation layer thickness.

(2) For the bearing capacity and the load-midspan deflection relation, the freeze-thaw cycles weakened the flexural properties of TRE-PSIWPs, and the degree of declining increased with the increase of the number of freeze-thaw cycles. The microcracks generated by the freeze-thaw cycles in 5% Na₂SO₄ solution of the specimen can self-healing in the later stage, thereby improving the flexural bearing capacity of TRE-PSIWPs. The textile treatment methods had no significant impact on the initial bending stiffness of TRE-PSIWPs, but the textiles treated by impregnating and sand burning were contributed to enhance the flexural bearing capacity of TRE-PSIWPs. Increasing the thickness of the TRE wythes can help to improve the initial bending stiffness and flexural bearing capacity of TRE-PSIWPs, but increasing the thickness of the insulation layer can only improve the flexural bearing capacity of the TRE-PSIWPs. The flexural bearing capacity of the specimens with EPS as the thermal insulation layer was the highest, followed by XPS and rock wool.

(3) For ductility, the freeze-thaw cycles could improve the ductility of TRE-PSIWPs, and the enhancement amplitude increased with the increase of the number of freeze-thaw cycles. Increasing the thickness of TRE wythes was contributed to effectively improve the ductility of TRE-PSIWPs, while increasing the thermal insulation panel thickness did the opposite. The textiles treated by impregnating and sand burning had the best bonding performance to ECC matrix, which could constraint the deformation of ECC matrix, leading to the decrease of ductility of TRE-PSIWPs. Sulfate corrosion reduced the bending stiffness of TRE-PSIWPs and enhanced the ductility of TRE-PSIWPs.

(4) For composite degree, the freeze-thaw cycles caused the poor interfacial bonding between the TRE wythes and the thermal insulation panel, which seriously reduced their cooperative work ability. Increasing the thickness of the insulation layer was contributed to improve the composite degree of TRE-PSIWPs, but increasing the thickness of the TRE wythes could reduce the composite degree of TRE-PSIWPs. The textiles treated by impregnating and sand burning could better constrain the deformation of ECC matrix, which reduced the debonding degree between TRE wythes and thermal insulation layer, thereby improving their cooperative work ability. Among the three insulation materials, rock wool panels had the worst cooperative work ability with the TRE wythes, followed by XPS insulation panels and EPS insulation panels.

Footnotes

Acknowledgements

The experimental work described in this paper was conducted at the Jiangsu Key Laboratory of Environmental Impact and Structural Safety in Civil Engineering in the China University of Mining and Technology. Helps during the testing from staffs and students at laboratory are greatly acknowledged.

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: The authors gratefully acknowledge the financial support from the Xuzhou Key Research and Development Program (Industry Prospect and Common Key Technology) (KC18106), the Postgraduate Research & Practice Innovation Program of Jiangsu Province (KYCX23_2739) and the Graduate Innovation Program of China University of Mining and Technology (2023WLJCRCZL049).

ORCID iD

Shiping Yin

Data availability statement

All data, models, and code generated or used during the study appear in the submitted article.

References

Amran

Rashid

Hejazi

, et al. (2016) Response of precast foamed concrete sandwich panels to flexural loading. Journal of Building Engineering 7: 143–158.

Balıkoğlu

Demircioğlu

(2022) Flexural behaviour of the composite sandwich beams with grid-scored foam: experimental and theoretical approach. Thin-Walled Structures 171: 108691.

Benayoune

Samad

Trikha

, et al. (2008) Flexural behaviour of pre-cast concrete sandwich composite panel–experimental and theoretical investigations. Construction and Building Materials 22(4): 580–592.

Carbonari

Cavalaro

SHP

Cansario

, et al. (2012) Flexural behaviour of light-weight sandwich panels composed by concrete and EPS. Construction and Building Materials 35: 792–799.

Chen

Kang

, et al. (2022) Bending performance of precast ceramsite-concrete-insulated sandwich panel with stainless steel shear connectors. Buildings 12(10): 1640.

Choi

Kim

You

(2016) Effect of cyclic loading on composite behavior of insulated concrete sandwich wall panels with GFRP shear connectors. Composites Part B: Engineering 96: 7–19.

Contamine

Si Larbi

(2016) Development of a textile reinforced concrete (TRC) to retrofit reinforced concrete structures. European Journal of Environmental and Civil Engineering 20(6): 626–642.

Daniel Ronald Joseph

Prabakar

Alagusundaramoorthy

(2023) Experimental and analytical investigations on the failure modes of concrete sandwich panels under axial compression. European Journal of Environmental and Civil Engineering 27(2): 733–762.

Daskiran

Gencoglu

(2022) Experimental investigation on impact strength of AR glass, basalt and PVA textile reinforced cementitious composites. European Journal of Environmental and Civil Engineering 26(6): 2037–2056.

10.

Frankl

Lucier

Hassan

, et al. (2011) Behavior of precast, prestressed concrete sandwich wall panels reinforced with CFRP shear grid. PCI Journal 56(2): 42–54.

11.

Garhwal

Sharma

Roy

ABD

(2022) Performance of expanded polystyrene (EPS) sandwiched concrete panels subjected to accelerated corrosion. Structures 43: 1057–1072.

12.

Hassan

Rizkalla

(2010) Analysis and design guidelines of precast, prestressed concrete, composite load-bearing sandwich wall panels reinforced with CFRP grid. PCI Journal 55(2): 147–162.

13.

Hegger

Will

Bruckermann

, et al. (2006) Load–bearing behaviour and simulation of textile reinforced concrete. Materials and Structures 39: 765–776.

14.

Hou

Chen

, et al. (2020) Seismic behavior of reinforced engineered cementitious composite members and reinforced concrete/engineered cementitious composite members: a review. Structural Concrete 21(1): 199–219.

15.

Huang

Dai

(2020) Flexural performance of precast geopolymer concrete sandwich panel enabled by FRP connector. Composite Structures 248: 112563.

16.

Joseph

JDR

Prabakar

Alagusundaramoorthy

(2017) Precast concrete sandwich one-way slabs under flexural loading. Engineering Structures 138: 447–457.

17.

Kim

You

(2015) Composite behavior of a novel insulated concrete sandwich wall panel reinforced with GFRP shear grids: effects of insulation types. Materials 8(3): 899–913.

18.

Lal

Kumar

(2019) Finite element-based stochastic nonlinear progressive failure response of piezo-laminated composite plate with elliptical cutouts. International Journal of Applied Mechanics 11(08): 1950076.

19.

Lal

Markad

(2021a) Probabilistic-based nonlinear progressive failure analysis of piezoelectric laminated composite shell panels in hygrothermal environment. Journal of Aerospace Engineering 34(6): 04021099.

20.

Lal

Markad

(2021b) Static and dynamic nonlinear stability analyses of hybrid sandwich composite beams under variable in-plane loads. Journal of Mechanical Science and Technology 35: 3895–3908.

21.

Lee

Pessiki

(2008) Experimental evaluation of precast, prestressed concrete, three-wythe sandwich wall panels. PCI Journal 53(2): 95–115.

22.

(2009) Experimental investigation and analysis on flexural performance of functionally graded composite beam crack-controlled by ultrahigh toughness cementitious composites. Science in China - Series E: Technological Sciences 52(6): 1648–1664.

23.

(2011) Experimental research on mechanical performance of hybrid fiber reinforced cementitious composites with polyvinyl alcohol short fiber and carbon textile. Journal of Composite Materials 45(1): 5–28.

24.

Xiong

Jiang

, et al. (2019) Tensile behavior of basalt textile grid reinforced engineering cementitious composite. Composites Part B: Engineering 156: 185–200.

25.

Liu

Xie

Liu

, et al. (2022) Research on bending performance of concrete sandwich laminated floor slabs with integrated thermal and sound insulation. Coatings 12(8): 1075.

26.

Markad

Lal

(2022) Synthesis of the multiphase shape memory hybrid composites hybridized with functionalized MWCNT to improve mechanical and interfacial properties. Polymer-Plastics Technology and Materials 61(6): 650–664.

27.

Mohamad

Khalil

Abdul Samad

, et al. (2014) Structural behavior of precast lightweight foam concrete sandwich panel with double shear truss connectors under flexural load. ISRN Civil Engineering 2014: 1–7.

28.

O’Hegarty

Kinnane

(2020) Review of precast concrete sandwich panels and their innovations. Construction and Building Materials 233: 117145.

29.

Pessik

Mlynarczyk

(2003) Experimental evaluation of the composite behavior of precast concrete sandwich wall panels. PCI Journal 48(2): 54–71.

30.

Şahmaran

(2007) De-icing salt scaling resistance of mechanically loaded engineered cementitious composites. Cement and Concrete Research 37(7): 1035–1046.

31.

Salmon

Einea

(1995) Partially composite sandwich panel deflections. Journal of Structural Engineering 121(4): 778–783.

32.

Salmon

Einea

Tadros

, et al. (1994) A new structurally and thermally efficient precast sandwich panel system. PCI Journal 39(4): 90–101.

33.

Salmon

Einea

Tadros

, et al. (1997) Full scale testing of precast concrete sandwich panels. Structural Journal 94(4): 354–362.

34.

Tomlinson

Fam

(2015) Flexural behavior of precast concrete sandwich wall panels with basalt FRP and steel reinforcement. PCI Journal 60: 51–71.

35.

Xie

Fang

, et al. (2021) Low-velocity impact damage detection and characterization in composite sandwich panels using infrared thermography. Composite Structures 269: 114008.

36.

Xie

Shen

Fang

, et al. (2022) Flexural property evaluation of web reinforced GFRP-PET foam sandwich panel: experimental study and numerical simulation. Composites Part B: Engineering 234: 109725.

37.

Shiping

Linli

Junling

(2021) Study on the tensile and flexural mechanical properties of TRE in freezing-thawing environments. Construction and Building Materials 268: 121150.

38.

Zargaran

Attari

Alizadeh

(2023) Flexural behavior of high-performance and non-high-performance textile reinforced concrete composites. European Journal of Environmental and Civil Engineering 27(2): 893–907.

39.

Zhang

(2002) Monotonic and fatigue performance in bending of fiber-reinforced engineered cementitious composite in overlay system. Cement and Concrete Research 32(3): 415–423.