Experimental Study on the Static Behavior and Recovery Properties of CFRP/SMA Composites

Abstract

Strengthening reinforced concrete elements with externally bonded prestressed fiber reinforced polymer (FRP) sheets has become a popular reinforcement technology in recent years. However, in practical engineering applications, due to the limitations of construction operation space and the need for specialized design of tensioning and anchoring devices, it is very cumbersome to apply prestressing force to FRP sheets. Therefore, using the recovery effect of shape memory alloys (SMA) to introduce prestressing into FRP sheets can innovate a new approach by combining FRP sheets and SMA wires. In order to study the basic mechanical properties of FRP/SMA composites, carbon fiber reinforced polymer and shape memory alloys were used to make the composite specimens, and uniaxial tensile tests were carried out on them. The mechanical properties such as the stress-strain curve, failure mode, ultimate tensile strength and fracture strain were obtained. The test results show that CFRP sheet exhibits obvious linear elastic behavior in tensile tests. The stress-strain curve of SMA wire can be divided into four stages: the linear elastic stage, yield stage, strengthening stage and failure stage. The fracture strain at failure can reach 7%, which indicates excellent deformation properties. The loading and unloading cycles have little effect on the mechanical properties of SMA wire. With the increase in the loading rate, the ‘stress plateau’ section of the phase transformation section of the SMA wire hysteresis curve gradually transits to an oblique upward curve. Increasing the pre-strain value within a certain range can improve the resilience of SMA wires. SMA wires with a pre-strain value of 8% can provide a maximum resilience of 514 MPa after heating to the austenitic state. A prediction model for the number of temperature cycles and maximum recovery force of SMA was proposed and validated. According to this model, the SMA wires can still provide stable resilience after 30 cycles. Increasing the amount of wire (volume ratio) can improve the maximum fracture strain and ultimate tensile strength of CFRP/SMA composite specimens, and the more wire is added, the greater the residual strength after fracture. The diameter of the fiber can significantly reduce the maximum fracture strain and ultimate tensile strength of the FRP/SMA composite specimen.

1. Introduction

In reinforced concrete structures, rebar corrosion is an important factor affecting the load-bearing capacity and seismic resistance of structures [1,2]. To address this issue, fiber reinforced polymer (FRP), one of the new materials used in civil engineering structures in the 20th century, has become an important choice for improving concrete structures. The advantages of FRP in corrosion resistance, light weight and high strength [3,4,5,6] make it mainly used in two aspects in structural optimization: (i) Used to replace steel materials in structures [7] and make FRP-concrete composite with light weight and good durability [8,9]. (ii) Applied to the repair, modification and reinforcement of civil engineering structures [10,11]. With the passage of time and socio-economic development, more and more early built old reinforced concrete bridges have quality problems or need to be updated. It is obvious that dismantling and rebuilding reinforced concrete structures require a large amount of resources, making it difficult to meet the goals of green and sustainable development. In contrast, using bridge reinforcement technology can prolong the service life of bridge structures and achieve sustainability with less resource consumption. Therefore, new green materials such as FRP have become one of the excellent choices for structural reinforcement of existing bridges. In the repair and reinforcement of structures using non-prestressed FRP, a common method is to bond FRP sheets on the outside of the structure [12,13], but this reinforcement method cannot fully exploit the lightweight and high-strength properties of FRP materials. Some researchers have shown that introducing prestress into FRP sheets is an effective way to utilize their performance [14,15]. However, the traditional mechanical prestressing method has the disadvantages of complex construction, small construction space and large prestress loss, which limits the promotion and application of prestressed FRP sheets or plates [16,17,18,19,20,21]; in addition, the use of a large number of anchoring and stretching tools has resulted in the consumption of human and material resources and low work efficiency, and is not conducive to sustainable development. Therefore, there is a need to develop an effective method to introduce the prestress into the FRP sheets/plates.

Shape memory alloy (SMA) is a new outstanding material with super-plasticity and shape memory effect (SME) [22,23], the SME of which makes it recoverable and produces resilience. SMA has now entered the field of civil engineering applications [22,24,25,26,27,28,29], and is applied to engineering structure seismic resistance with its macroscopic super-elasticity to achieve energy dissipation [30,31].

In the early days, FRP/SMA composites were proposed for use in the field of mechanical and aviation engineering [32,33]. In civil engineering research, Dawood et al. developed the combination of SMA and CFRP to form composite materials, using SMA wires to introduce prestress into the steel plate to reduce the stress amplitude of steel plate and improve the fatigue performance of steel plate [34,35]. Andrawes et al. [36,37,38,39] applied SMA to achieve seismic reinforcement of reinforced concrete (RC) columns, and manufactured glass fiber reinforced polymer (GFRP)-SMA composites for seismic reinforcement. Wang et al. [15] studied the mechanical behavior and recovery properties of CFRP/SMA composites in two types; a stress-strain relationship model and a model for predicting the mechanical properties and recovery properties of CFRP/SMA composites were proposed as well.

Carbon fiber reinforced polymer (CFRP) is a type of FRP material, and research has shown that CFRP has excellent tensile properties, fatigue resistance [40,41], corrosion resistance and creep resistance. Therefore, it can exhibit good mechanical properties and excellent long-term performance in engineering applications [42]. Despite the aforementioned advantages of CFRP, current reinforcement methods using CFRP cannot fully utilize the excellent properties of CFRP. Therefore, it is necessary to introduce prestress into CFRP in advance to enable it to participate in the work and give full play to its advantages. In view of the shortcomings of the traditional method of introducing prestress mentioned above, if the CFRP sheets and SMA wires are combined to form a CFRP/SMA composite, the resilience of SMA wires can be introduced into the CFRP sheets to form prestressed CFRP sheets, effectively avoiding the defects of mechanical tension leading to prestress. At the same time, no additional equipment is needed to introduce prestressing, simplifying the construction process and improving work efficiency. As a well-known high-performance green material, SMA not only conforms to low-carbon and sustainable requirements, but also achieves the goal of rapid repair and reinforcement of concrete bridge structures.

In the authors’ previous research [15], a CFRP/SMA composite strengthening system has been proposed and some mechanical behaviors of composite materials have been studied. In order to further study the mechanical properties of CFRP and SMA composites, this paper carried out uniaxial tensile tests of the basic mechanical properties of CFRP and SMA composites, studied the stress-strain curve, failure morphology and tensile strength and demonstrated the feasibility of the composite of the two materials from the perspective of basic mechanical properties.

2. Materials and Test Program

2.1. Mechanical Property Test of CFRP/SMA Single Material

2.1.1. Material Properties

The FRP material adapted in the test was CFS-I-300 high-strength carbon fiber reinforced polymer (CFRP) unidirectional cloth produced by Carbon Composite Co., Ltd. (Tianjin, China) CFSR-A/B epoxy resin was used to impregnate CFRP cloth to form CFRP sheet. Typical material property parameters for CFS-I-300 high-strength CFRP unidirectional cloth and CFSR-A/B epoxy resin are shown in

Table 1. Typical material performance parameters.

Material	Material Parameters	Performance Index
CFS-I-300 high-strength CFRP unidirectional cloth	Tensile strength (MPa)	3467
	Elastic modulus (GPa)	241
	Theoretical thickness (mm)	0.167
	Density (g/m3)	300
CFSR-A/B epoxy resin	Tensile strength (MPa)	26
	Tensile modulus of elasticity (GPa)	4.0
	Ultimate elongation rate (%)	0.7
	Glass transition temperature (°C)	85
	Weight ratio (recommended)	2:1
	Density after mixing (g/cm3)	1.05–1.25
	Positive tensile bonding strength with concrete (MPa)	≥2.5 (Cohesive failure of concrete)

. Epoxy resin is a two-component adhesive, with a ratio of 2:1 in weight between resin and polyamide curing agent. The specimen was in the form of rectangular sheet with the size of 250 mm × 25 mm × 0.667 mm, and the gauge length was 150 mm.

NiTi alloy wire produced by Xi’an Siwei Metal Materials Co., Ltd. (Xi’an, China) was selected as the SMA, with the mass fraction of Ni of 55.01%. Two kinds of SMA wires with diameter of 0.5 mm and 1 mm were used in the test. The length of the sample was 300 mm, and the gauge length was 200 mm. According to the differential scanning calorimetry (DSC) method, the austenitic starting temperature (A_s) of 0.5 mm SMA wire was 106.12 °C, and the austenitic finishing temperature (A_f) was 121.40 °C. The value of the martensite transformation starting temperature (M_s) was 55.34 °C, and the value of the martensite completion temperature (M_f) was 27.34 °C. The A_s of 1 mm SMA wire was 79.13 °C, A_f was 105.46 °C, M_s was 59.00 °C and M_f was 35.81 °C. Therefore, NiTi alloy wire was martensite at normal temperature (20 °C).

2.1.2. Loading Conditions and Regimes

The loading parameters of CFRP and SMA meet the requirements of the current Chinese standard [43]. A total of five CFRP sheet specimens were set up and tested on the DYD-50-5 series electronic universal material testing machine. Strain data were collected by pasting strain gauges on the specimen and connecting to the TST3827E dynamic and static signal testing and analysis system. All data were automatically collected, using an equal displacement loading method. According to Paragraph 8.3.2 of the standard [43], the standard displacement rate was set at 2 mm/min and the test piece was stretched until it was damaged. The SMA wire test device was a UTM4000 series electronic universal testing machine, which adopted the constant strain rate control. The strain of SMA wire was obtained through an extension gauge equipped with a universal testing machine, and the strain data were automatically collected through the PowerTest software equipped with the testing machine. Considering the impact of the number of cycles, strain amplitude, loading rate, wire diameter and other factors on the mechanical properties of SMA, the test was carried out at room temperature of 15 °C. The test conditions and loading regimes are shown in

Table 2. Test conditions and loading regimes for SMA wires.

Group No.	Number of Specimens	Specimen Designation	Gauge Length (mm)	SMA Diameter (mm)	Strain Amplitude (%)	Loading Rate (mm/min)	Cycle Number (Times)
1	12	SMA-0.5-2%-1 mm/min-10	200	0.5	2	1	10
		SMA-0.5-3%-1 mm/min-10	200	0.5	3	1	10
		SMA-0.5-4%-1 mm/min-10	200	0.5	4	1	10
		SMA-0.5-5%-1 mm/min-10	200	0.5	5	1	10
		SMA-0.5-6%-1 mm/min-10	200	0.5	6	1	10
		SMA-0.5-7%-1 mm/min-10	200	0.5	7	1	10
2	12	SMA-0.5-2%-1 mm/min-1	200	0.5	2	1	1
		SMA-0.5-3%-1 mm/min-1	200	0.5	3	1	1
		SMA-0.5-4%-1 mm/min-1	200	0.5	4	1	1
		SMA-0.5-5%-1 mm/min-1	200	0.5	5	1	1
		SMA-0.5-6%-1 mm/min-1	200	0.5	6	1	1
		SMA-0.5-7%-1 mm/min-1	200	0.5	7	1	1
		SMA-1-4%-1 mm/min-1	200	1	4	1	1
		SMA-1-5%-1 mm/min-1	200	1	5	1	1
		SMA-1-6%-1 mm/min-1	200	1	6	1	1
		SMA-1-7%-1 mm/min-1	200	1	7	1	1
		SMA-1-8%-1 mm/min-1	200	1	8	1	1
		SMA-1-10%-1 mm/min-1	200	1	10	1	1
3	10	SMA-0.5-7%-10 mm/min-1	200	0.5	7	10	1
		SMA-0.5-7%-15 mm/min-1	200	0.5	7	15	1
		SMA-0.5-7%-20 mm/min-1	200	0.5	7	20	1
		SMA-0.5-7%-25 mm/min-1	200	0.5	7	25	1
		SMA-0.5-7%-30 mm/min-1	200	0.5	7	30	1
		SMA-1-7%-10 mm/min-1	200	1	7	10	1
		SMA-1-7%-15 mm/min-1	200	1	7	15	1
		SMA-1-7%-20 mm/min-1	200	1	7	20	1
		SMA-1-7%-25 mm/min-1	200	1	7	25	1
		SMA-1-7%-30 mm/min-1	200	1	7	30	1
4	16	SMA-0.5-4%-1 mm/min-1	200	0.5	4	1	1
		SMA-0.5-5%-1 mm/min-1	200	0.5	5	1	1
		SMA-0.5-6%-1 mm/min-1	200	0.5	6	1	1
		SMA-0.5-7%-1 mm/min-1	200	0.5	7	1	1
		SMA-1-4%-1 mm/min-1	200	1	4	1	1
		SMA-1-5%-1 mm/min-1	200	1	5	1	1
		SMA-1-6%-1 mm/min-1	200	1	6	1	1
		SMA-1-7%-1 mm/min-1	200	1	7	1	1
5	6	SMA-0.5-LD-1 mm/min	200	0.5	-	1	-
		SMA-1-LD-1 mm/min	200	1	-	1	-

. Description of specimen names in the table: 0.5 and 1 represent the diameter of SMA wire, respectively; 2~7%, respectively, represents the maximum strain value (i.e., strain amplitude value) corresponding to the loading of SMA wire; 1 mm/min, etc., indicates the loading rate; 1 and 10 represent the number of loading and unloading cycles.

2.2. Mechanical Properties Test of SMA Wire under Temperature Reciprocation

The size parameters and loading regimes of SMA wires are shown in

Table 3. Test conditions and loading regimes for SMA wires for temperature cycling test.

Number of Specimens	Specimen Designation	Gauge Length (mm)	SMA Diameter (mm)	Strain Amplitude (%)	Loading Rate (mm/min)	Number of Temperature Cycles (Times)
3	SMA-0.5-8%-1 mm/min-10	200	0.5	8	1	10
9	SMA-0.5-2%-1 mm/min-1	200	0.5	2	1	1
	SMA-0.5-5%-1 mm/min-1	200	0.5	5	1	1
	SMA-0.5-8%-1 mm/min-1	200	0.5	8	1	1

; the pretreatment and loading settings are shown in Figure 1. Due to the thin size of the SMA wire, it is difficult to observe in the figure. The SMA wire in Figure 1 is indicated by a bold red line, which does not represent the actual wire color and thickness. Figure 1a shows the layout of the pre-tensioning device. The SMA wire, jack, level gauge, etc. were all fixed on a special metal frame. The SMA wire was fixed by a fixture fixed on the jack and tensioned by the jack. The tensioning force was fed back by the force sensor at the top of the jack and displayed on the display screen. Figure 1b shows the arrangement of the experimental device. The SMA wire was also loaded through a jack and the tension force was displayed by a force sensor (the jack was located above the fixture, not shown in the figure). Both ends of the SMA wire were connected to the positive and negative poles of the power supply, and the temperature was increased by adjusting the applied current. A total of six test pieces were set up. The relationship between temperature and current is shown in Equation (1) [44]. The temperature was monitored in real time by the temperature sensor, and the force sensor was used to record the value of the resilience generated by the SMA wire at the corresponding temperature. Finally, the relationship between the resilience and the number of temperature cycles was recorded after ten temperature cycles. After the first four cycles, the relaxed SMA wire after the end of the cooling stage was tensioned straight, and then the last six cycles were applied for treating to obtain the pre-strain.

$$T=2.8+\frac{2771.6}{9.9\sqrt{0.5\pi }}\times e^{-2{\left(\frac{i-12.5}{9.9}\right)}^2}$$

(1)

where: T represents temperature (°C); i represents the current value (unit in amperes).

2.3. Mechanical Property Test of CFRP-SMA Composite

2.3.1. Specimen Design

The CFRP/SMA composite consisted of three layers: one layer of CFRP as the bottom layer, one layer of SMA wire as the intermediate layer and one layer of CFRP as the cover layer. The epoxy adhesive should be firmly bonded between each layer. The pre-strain of SMA wires was set to 4% [15,45], the number and diameter of SMA wires were considered in the test piece and four groups were defined, respectively: FRP/SMA-I-0.5-4% (five 0.5 mm SMA wires), FRP/SMA-I-0.5-10-4% (ten 0.5 mm SMA wires), FRP/SMA-I-0.5-15-4% (fifteen 0.5 mm SMA wires) and FRP/SMA-I-1-10-4% (ten 1 mm SMA wires). Figure 2 shows the geometric dimensions of each test piece, and

Table 4. Specimen design of CFRP/SMA composite material.

Specimen Designation	Numbers of SMA Wire	SMA Diameter (mm)	Strain Amplitude (%)	Composite Length (mm)
FRP/SMA-I-0.5-5-4%	5	0.5	4	250
FRP/SMA-I-0.5-10-4%	10	0.5	4	250
FRP/SMA-I-0.5-15-4%	15	0.5	4	250
FRP/SMA-I-1-10-4%	10	1	4	250

shows the design of each test piece. During the preparation of each test piece, the SMA wire should be straightened at 4% pre-strain level until the epoxy resin is fully cured.

The preparation of FRP/SMA composite materials followed the following steps, and the physical flowchart of the production process is shown in Figure 3. (1) Cut the SMA wires into a length of approximately 400 mm for tensioning operation; (2) Use a pre-strain device to achieve a strain value of 4% for SMA wires, and then connect the SMA wires to a special tensioned steel frame. Steel frames were used to limit pre-strain wires and prevent loss of pre-strain during the manufacturing process. In addition, this framework helped to maintain uniform spacing between SMA wires; (3) Cut FRP unidirectional fabric (size 50 mm × 250 mm), place the bottom FRP on a plastic pad and evenly apply a thin layer of epoxy resin adhesive on it. Afterwards, flip the FRP cloth and apply pressure evenly using a roller to help the epoxy resin be fully absorbed; (4) Place the FRP sheet impregnated with colloid on a flat base plate, and then apply another layer of epoxy resin adhesive on it; (5) Place the tensioned steel frame connected with pre-strained SMA wires on epoxy resin adhesive and bottom FRP sheet. Add more epoxy resin adhesive to cover SMA wires and make the surface of the epoxy resin flat and smooth; (6) Use the roller to help the epoxy resin adhesive absorb and apply uniform pressure on the surface of the FRP substrate; (7) Lay the top layer FRP cloth on the epoxy resin adhesive, and then use a roller to apply uniform pressure along the surface of the FRP unidirectional cloth to help the epoxy resin be absorbed by the top layer FRP cloth; (8) Place the top pressing steel plate on the top layer of the FRP/SMA composite specimen to provide a uniform thickness for the specimen. Press the top pressing steel plate to remove excess epoxy adhesive, and maintain a uniform thickness of epoxy resin adhesive between the top FRP sheet and the bottom FRP sheet; (9) FRP/SMA composite specimens should be cured at room temperature for at least 7 days.

2.3.2. Loading Regimes

The test process was controlled by equal strain rate, and the test data were automatically recorded by the acquisition software of the universal testing machine, and then converted into the stress value of the test piece. Strain was collected by a TST3827E dynamic and static signal test and analysis system. The ambient temperature was 20 °C, the strain rate was kept constant during the loading process and the standard displacement rate was set at 2 mm/min, until the specimen was stretched to failure.

3. Results and Analysis

3.1. Analysis of Mechanical Properties of Single Material

3.1.1. Uniaxial Tensile Mechanical Properties of CFRP Sheet

Figure 4 shows the typical test results of monotonic tension to fracture of CFRP sheet specimens. At the loading rate of 2 mm/min, the stress-strain curves of the five CFRP specimens showed consistency. The stress-strain curve of CFRP sheet specimens was approximately a straight line from test loading to failure. Finally, each specimen underwent lateral fracture failure at a certain location. The specimen fractured along the transverse direction, with obvious brittle failure characteristics. The average value of tensile strength was about 3000 MPa, and the maximum tensile strain was 1.24%. According to the chord-line method, the elastic modulus of each CFRP sheet specimen was 247.7 GPa.

3.1.2. Uniaxial Tensile Mechanical Properties of SMA Wire

Figure 5 shows the typical test results of monotonic tension to fracture of SMA wires with different diameters. At the loading rate of 1 mm/min, the stress-strain curves of 0.5 mm and 1 mm wires showed consistency. The whole stress-strain curve of SMA wires could be divided into four stages from test loading to breaking:

• Linear elastic stage. From the beginning of loading of SMA wire to the stress reaching the beginning of phase transformation σ_s (about 100~120 MPa in the figure, the corresponding strain was approximately 1.5%), SMA was in the elastic stage; at this time, the theoretical deformation could be completely restored after the stress was removed;
• Yield stage. With the increase in tensile load, the tensile stress of SMA wire exceeded the initial stress of phase transformation σ_s, triggering the SMA to undergo the transformation from the twinning martensite phase to the de-twinning martensite phase. At this time, the stress of the SMA basically did not increase and the strain continued to increase. The yield ‘stress platform’ of the transformation section of 0.5 mm SMA wire was about 120 MPa, while the ‘stress platform’ corresponding to 1 mm SMA wire was stable between 100 and 110 MPa, and the strain of the corresponding transformation phase usually started from 1.5% to about 7%. The SMA wire would have residual deformation during unloading at the phase transformation stage, but this residual deformation could be recovered (the SMA wire will be heated to T > A_f austenite end temperature, which would induce the SMA wire to change from austenite to martensite, that is, the transformation from austenite to twin martensite);
• Strengthening stage. When the tensile stress of SMA wire exceeded the transformation end stress σ_f, the material recovered part of its resistance to deformation, and the stress-strain curve showed an obvious rising curve;
• Failure stage. After passing the highest point of the stress-strain curve, the section at the weak part of the SMA specimen underwent local necking, the deformation continued to increase and the stress decreased steadily until it was destroyed with fracture failure.

3.2. Analysis of Factors Affecting Mechanical Properties of SMA Wire

3.2.1. Cyclic Loading Times

The stress-strain curve of SMA wire under 10 loading and unloading cycles is shown in Figure 6. It could be seen from this figure that the number of cycles had little effect on the mechanical properties of SMA wire at room temperature. The residual strain of SMA wire after 10 cycles of loading and unloading was basically the same as that of SMA wire after the first loading and unloading, and the material showed an obvious shape memory effect. The strain of the first loading and unloading of SMA started from the coordinate origin to the required strain amplitude, and then the unloading stress was zero to the corresponding residual strain value. The area of the hysteresis loop (energy dissipation capacity) was the largest. From the beginning of the second loading and unloading cycle, the area of the hysteresis loop tended to be stable. As the number of loading and unloading cycles increased, the hysteresis curve gradually became smooth. This was because the internal lattice of SMA wire tended to be stable after loading and unloading cycles, and the elastic modulus under the martensite phase also remained basically unchanged.

3.2.2. Strain Amplitude

The stress-strain curves of two SMA wires under one loading and unloading with incremental strain amplitude are shown in Figure 7. With the increase in strain amplitude, the stress of SMA wires increased in three stages, with a large increase under small strain amplitude (less than 1.5%). At medium strain amplitude (1.5~7%), the stress hardly increased. At large strain amplitudes (greater than 7%), SMA wires had entered the strengthening stage, showing stress growth similar to that of the first stage. It could also be seen from this figure that the strain amplitude of SMA wires was the most important factor affecting their energy consumption capacity. With the increase in strain amplitude, the martensitic phase transition occurred more fully and the area surrounding the hysteresis curve increased. With the increase in strain amplitude, the residual strain of SMA wires increased continuously, which was 0.5~1% smaller than the corresponding strain amplitude. This was because when the strain amplitude exceeded the elastic stage (generally 1~1.5%), the martensitic transformation from the twin martensite phase to de-twin martensite phase would occur, which could not be recovered when unloading, and the SMA wire must undergo the austenitic phase to de-twin martensite phase transformation through temperature rise to recover the deformation.

3.2.3. Loading Rate

Figure 8 shows the test results of stress-strain curves of SMA wires at different loading rates. At 7% strain amplitude, the properties of 0.5 mm and 1 mm wires were consistent: when the loading rate was low (less than 10 mm/min), the stress-strain curve had an obvious phase transition section ‘stress platform’, and the slope nearly remained horizontal. With the increase in loading rate (10–30 mm/min), the phase transition section of the SMA wire hysteresis loop curve gradually transited to a diagonally upward curve. Stress at the end of the SMA phase transition (twin to de-twin martensite phase) σ_f (stress of SMA wire entering the hardening section) increased with increasing loading rate.

3.2.4. Wire Diameter

Figure 9 shows the test results of stress-strain curves of SMA wires with different diameters under different strain amplitudes. At 1 mm/min loading rate, the properties of 0.5 mm and 1 mm wires were consistent: under four strain amplitudes (4~7%), the yield stress platform of 0.5 mm SMA wire phase transition section was slightly higher than that of 1 mm SMA wire, while the yield stress platform of 0.5 mm SMA wire phase transition section was about 120 MPa, while that of 1 mm SMA wire was stable between 100 MPa and 110 MPa.

3.3. Effect of Temperature Reciprocation on Resilience of SMA Wire

SMA fibers need to be in the austenite phase to generate resilience through shape memory characteristics, which corresponds to the need to maintain a higher treatment temperature. From the point of view of cost and construction, maintaining continuous high-temperature conditions is not conducive to practical application, and considering the reusability of SMA wires, this section establishes 8% pre-strain value under the martensite phase reciprocating from high to low temperature to explore the relationship between the resilience generated by SMA wires and temperature conditions.

3.3.1. Effect of Pre-Strain Value on Resilience

By unidirectional heating of SMA wires with different pre-strain values, the resilience produced after reaching the austenitic phase was studied. As shown in Figure 10, taking 2%, 5% and 8% pre-strain values as examples, the resilience was positively correlated with the initial pre-strain value. In the test, the SMA wire with 8% pre-strain value achieved the maximum resilience value in the austenite phase. Because the phase transition end temperature of the austenitic phase was about 100 °C, the effect of the phase transition process of SMA wire on its recovery stress could be neglected after exceeding this temperature, and the resilience value of SMA wire tended to be stable after the austenitic phase transition was completed.

3.3.2. Effect of Reciprocating Temperature on Maximum Resilience

Due to the single-pass, double-pass and whole-process shape memory effect of SMA wire, after the cooling stage of the cycle, the pre-tensioned length under the pre-martensitic phase had a small recovery, and after the heating stage, this part of the recovery length would cause a certain loss of the resilience generated by SMA wire in the austenitic state. As mentioned in Section 2.2, the relaxation of SMA wire after the cooling stage would be tensioned in the first four cycles; from Figure 11, it could be observed that the resilience of SMA wire heated after the first four times of tightening had increased, and the resilience of SMA wire gradually stabilized under the next six cycles without any treatment.

Table 5. Maximum resilience under reciprocal temperature cycling.

Cycle-Index	Maximum Resilience (MPa)
	Specimen 1	Specimen 2	Specimen 3	Average	Predictive Value	Predictive Value/Test Average
0	462.420	486.624	471.656	473.566	474.802	1.003
1	466.242	490.458	482.929	479.876	478.650	0.997
2	470.063	497.554	495.668	487.762	485.842	0.996
3	475.159	504.898	507.159	495.738	497.668	1.004
4	492.159	510.738	514.649	505.849	505.023	0.998
5	480.471	502.496	503.649	495.539	495.778	1.000
6	473.114	493.528	489.171	485.271	484.496	0.998
7	467.554	489.515	475.286	477.452	477.928	1.001
8	463.006	484.859	467.713	471.859	474.408	1.005
9	463.732	483.394	466.057	471.061	472.419	1.003
10	464.554	485.942	468.057	472.851	471.216	0.997
Average of predictive value/test average						1.000
COV						9.181 × 10−4

shows the maximum resilience of three specimens under 0 to 10 reciprocating temperature cycles under existing tests, and the mean values are given in column 5 of the Table 5. Lorentz function coupling analysis was carried out based on 11 mean data, and the predictive function model of the maximum resilience on the number of reciprocating temperature cycles was obtained, as shown in Equation (2). Figure 12a compares the mean value of the test data with that of the predictive model, and the error values are also given in Table 5. The absolute maximum error was 0.005. The mean value of the ratio of the predictive model to the test mean was 1.000 and the variance was 9.181 × 10⁻⁴, which verified the accuracy of the regression prediction model. On this basis, the prediction model of Equation (2) was used to predict the maximum resilience of SMA wires after 30 cycles. The prediction results are shown in Figure 12b. It could be seen that the resilience value of SMA wire could still remain at a stable level after 30 cycles of reciprocating temperature.

$${\sigma }_f={\sigma }_0+\left(2\frac{A}{\pi }\right)\left[\frac{w}{w^2+4{\left(x-x_c\right)}^2}\right]$$

(2)

where: σ_f is the predicted value of the maximum resilience; x is the cycle-index of reciprocating temperature; σ₀ is 468.06 ± 1.34; The value of x_c is 3.93 ± 0.07; The value of w is 3.70 ± 0.35; The value of A is 215.38 ± 22.89; The fitting function determination coefficient value R² reaches 97.7%, and the matching degree is high.

3.4. Analysis of Mechanical Properties of CFRP/SMA Composites

3.4.1. Failure Mode

Figure 13 shows the failure mode of the CFRP/SMA composite specimens. It can be seen from the figure that when the number of SMA wires (five) was relatively small, the failure of the specimen was represented by the bursting failure of the middle part of the CFRP sheet. With the increase in the number of SMA wires, the CFRP sheet would undergo longitudinal splitting failure when the specimen was damaged. The reason was that the bonding interface between SMA wires and CFRP increased and an obvious weak layer appeared, resulting in the fiber bundles in the CFRP sheet not jointly bearing the tensile force, causing damage between the fiber bundles. In addition, the composite specimen with a diameter of 1 mm SMA wire also showed a similar failure phenomenon. It should be noted here that, indeed, the linear expansion coefficient and stiffness of the two materials, CFRP and SMA wire, are not the same, which will lead to a separation trend between the two. However, through the use of the epoxy resin, the two are prevented from separating under strong bonding stress. There was no obvious separation phenomenon in the specimen, so the bonding method used in this paper is feasible and can ensure that both work together.

3.4.2. Uniaxial Tensile Properties

Figure 14 shows the stress-strain curve of CFRP/SMA composite specimens under uniaxial stress. Under the loading rate of 2 mm/min, the stress-strain curves of four groups of CFRP/SMA composite specimens were approximately a straight line. After failure, the specimens still had certain residual strength, which was mainly due to the contribution of SMA wire. Compared to single CFRP specimens, the fracture strain of CFRP/SMA composite specimens has increased by 10.5–32.3%. With the increase in the number of SMA wires, the maximum fracture strain increased to a certain extent. For example, the average maximum fracture strain of FRP/SMA-0.5-5 specimen group was 1.41%, while the average maximum fracture strain of FRP/SMA-0.5-15 specimen group could reach 1.64%. The ultimate tensile strength of the test pieces had also been significantly improved. For example, the average ultimate tensile strength of the FRP/SMA-0.5-5 test piece group and the FRP/SMA-0.5-10 test piece group was 2496 MPa and 2568 MPa, while the average ultimate tensile strength of the FRP/SMA-0.5-15 test piece group could reach 2989 MPa. However, with the increase in the diameter of SMA wire, the average maximum breaking strain (1.37%) and the average ultimate tensile strength (2031 MPa) did not increase significantly, which was due to the early destruction of the colloid on the weak surface between the SMA wire and CFRP sheet. The elastic modulus of the FRP/SMA composite specimens can be obtained according to the chord-line method, and the test results are shown in

Table 6. FRP/SMA composite specimen test results.

Specimen Designation	Average Ultimate Tensile Strength (MPa)	Elastic Modulus (GPa)	Fracture Strain (%)
FRP/SMA-I-0.5-5-4%	2496	177	1.41
FRP/SMA-I-0.5-10-4%	2568	184.7	1.39
FRP/SMA-I-0.5-15-4%	2989	182.3	1.64
FRP/SMA-I-1-10-4%	2031	148.2	1.37

.

4. Conclusions and Recommendations

In order to introduce pre-stress into FRP sheets through heating recovery using SMA wire, this paper conducted a series of experiments to explore the basic mechanical properties of FRP/SMA composite materials and the recoverable properties of SMA wire. Based on the experimental results, the following conclusions have been obtained:

(1) At normal temperature, the stress-strain curve of SMA wire showed four stages, namely the linear elastic stage, yield stage, strengthening stage and failure stage. As the diameter of SMA wire increases, its yield strength slightly decreases. The yield strength is usually 120 MPa for 0.5 mm diameter SMA wire, while for 1.0 mm wire, the yield stress is approximately 100 MPa. The fracture strain at failure could reach 7%, which had excellent deformation properties.

(2) The maximum resilience value provided by SMA wire increased with the increase in pre-strain value in a certain range, and the resilience value tended to be stable after the completion of the high-temperature austenite transformation. In the reciprocating temperature cycle test, SMA wire with 8% pre-strain value in the low temperature martensitic state tension test could provide a maximum resilience of 514 MPa after heating to the austenitic state. According to the verified prediction model, the SMA wire could still provide stable resilience after 30 cycles.

(3) The basic mechanical test results of CFRP/SMA composite specimens showed that increasing the amount of wires (volume ratio) could improve the maximum fracture strain and ultimate tensile strength of CFRP/SMA composite specimens, and the more wires were added, the greater the residual strength after fracture. The diameter of the fiber could significantly reduce the maximum fracture strain and ultimate tensile strength of the FRP/SMA composite specimen.

This paper conducted some exploratory basic research on the recoverability of SMA wire and the basic mechanical properties of the composite material formed by CFRP sheet and SMA wire, but the research results are still very limited. Further research is expected to be needed on important issues such as the joint working performance of composite materials under temperature, the bonding performance between SMA wire and CFRP sheet and the durability and fatigue performance of composite materials under various environmental conditions.