Experimental investigation of engineered geopolymer composite for structural strengthening against blast loads

Shn Liu , Chunyun Liu ,*, Yifi Ho , Yi Zhng , Li Chn , Zhn Li

a Key Laboratory of Coast Civil Structure Safety of Ministry of Education, Tianjin University, Tianjin 300072, China

b Tianjin Key Laboratory of Prefabricated Buildings and Intelligent Construction, Hebei University of Technology, Tianjin 300400, China

c Institute of Defense Engineering, AMS, PLA, Beijing 100850, China

d Engineering Research Center of Safety and Protection of Explosion&Impact of Ministry of Education,Southeast University,Nanjing 211189,Jiangsu,China

e State Key Laboratory of Disaster Prevention&Mitigation of Explosion&Impact,Army Engineering University of the PLA,Nanjing 210007,Jiangsu,China

Keywords:Engineered geopolymer composites Fiber optimization Strengthening material Blast resistance Masonry wall Reinforced AAC panel Plain concrete slab

ABSTRACT The recent increase in blast/bombing incidents all over the world has pushed the development of effective strengthening approaches to enhance the blast resistance of existing civil infrastructures.Engineered geopolymer composite (EGC) is a promising material featured by eco-friendly, fast-setting and strain-hardening characteristics for emergent strengthening and construction.However, the fiber optimization for preparing EGC and its protective effect on structural elements under blast scenarios are uncertain.In this study, laboratory tests were firstly conducted to evaluate the effects of fiber types on the properties of EGC in terms of workability, dry shrinkage, and mechanical properties in compression,tension and flexure.The experimental results showed that EGC containing PE fiber exhibited suitable workability, acceptable dry shrinkage and superior mechanical properties compared with other types of fibers.After that, a series of field tests were carried out to evaluate the effectiveness of EGC retrofitting layer on the enhancement of blast performance of typical elements.The tests include autoclaved aerated concrete(AAC)masonry walls subjected to vented gas explosion,reinforced AAC panels subjected to TNT explosion and plain concrete slabs subjected to contact explosion.It was found that EGC could effectively enhance the blast resistance of structural elements in different scenarios.For AAC masonry walls and panels, with the existence of EGC, the integrity of specimens could be maintained, and their deflections and damage were significantly reduced.For plain concrete slabs, the EGC overlay could reduce the diameter and depth of the crater and spallation of specimens.

1.Introduction

Explosion occurrences caused by terrorist attacks or industrial accidents around the world have been reported recently, raising ongoing public concerns.Conventional building materials with brittle characteristics have weak blast resistance [1].Furthermore,material fragments brought by the reflection of tensile waves at the rear face could impose secondary harm and threaten people and facilities [2,3].However, the blast scenarios were often not considered during the design of existing buildings.Consequently,how to adequately strengthen the structures to enhance their resistance against blast loads and limit the loss of lives and property highlight increasing attention.

Existing investigations have demonstrated that strengthening materials such as polyurea [4], fiber reinforced polymer [5,6] and steel overlay [7] etc., can improve the blast resistance of the structural elements.It should be emphasized that the appropriate selection of strengthening materials depends on various parameters, including cost, performance and adaptability of geometry.In light of the above strengthening materials,the high cost of polyurea is a barrier to application.Fiber-reinforced polymers such as glass fibres, carbon fibres and basalt fibres, etc.have low deformation capacity [8], which may not provide sufficient strengthening to structural members with large deformation.Furthermore, fiberreinforced polymers tend to prematurely debond from the substrate before rupturing failure[9].For steel overlay,its application is limited by the geometric compatibility.

As a special type of high-performance fiber-reinforced cementitious composite, engineered cementitious composite (ECC) is characterized by strain hardening and multiple-cracking behavior[10].Thus, it is considered an innovative approach for enhancing the blast resistance of structural members.Li et al.[11] observed that the spalling damage in the ECC slab was less significant compared to the concrete slabs when subjected to a contact explosion.Xu et al.[12] investigated the dynamic response of ECC slabs under close-in blast loading,and demonstrated that ECC has a good potential for protective structures.In addition,only one study focused on the response of ECC overlay reinforced concrete panels subjected to near-field explosions [13].It was determined that the overlay thickness and retrofitting locations significantly impacted the strengthening effect.However, it should be noted that the setting time of ECC is at least 12 h due to the slow hydration reaction of cement.Although inorganic additives can speed up the setting and hardening processes, the early strength of sample remains low [14].The common emergent strengthening should be completed in 2-3 h, ECC obviously cannot guarantee its efficient implementation.

Geopolymer,as a sustainable material,is a product of any active aluminosilicate materials under alkali activation [15].The geopolymer technology provides an eco-friendly solution to the utilization of industrial solid waste consisting of aluminosilicate (e.g.,fly ash, ground granulated blast furnace slag, silica fume and red mud,etc.)[16,17].Apart from its low carbon emissions and superior sulfate, acid, thermal, and freeze-thaw resistance [18-20], when the synthesized binder contains calcium-rich precursors, geopolymer is characterized by rapid setting and high early strength[21].Based on the micromechanical design, engineered geopolymer composite (EGC) exhibits better strain-hardening characteristics and energy absorption capacity than ECC [22,23].Thus,cement-free EGC is a promisingly eco-friendly, highly effective and high-performance solution for emergent strengthening under blast scenarios.However,the present studies[22-24]focus on the design of EGC in terms of mechanical and durability parameters at the material level.It is not clear how the EGC affects the response of structural elements, especially when subjected to dynamic loads.

Microfibers commonly used in the preparation of ECC, such as polyvinyl alcohol (PVA), polyethene (PE), and polypropylene (PP),are also suitable for EGC.Wang et al.[25]studied the effect of PVA content on the static and dynamic properties of EGC,and proposed the optimum dosage of PVA fiber in EGC.Lao et al.[26] prepared EGC with high tensile ductility (around 8%) using PE fiber.Almashhadani et al.[27] demonstrated the interfacial bonding between the PP fiber and the geopolymeric matrix is weaker when compared to that of PVA fiber.However, the precursor, activator types and mixture proportions adopted in existing studies are inconsistent.The complex material design presents significant challenges to the consistent properties of EGC [28].The optimal fibers used to prepare EGC with the same matrix have not been determined.Furthermore, compared to cement-based material under comparable compressive strength, geopolymer one has comparatively high tensile strength and more brittle behavior due to the distinct gel composites and denser microstructure [29].In addition, strong chemical bonds, weak frictional bond and small slip hardening coefficients were found between fibers and geopolymer matrix[22].According to the micromechanical theory,the matrix and fiber/matrix interaction affect the strain-hardening behavior of composite [30].Therefore, it may not be possible to extend the results from ECC to EGC empirically.

Given the lack of studies on tailoring EGC through different types of fibers and the blast-resistant characteristics of structural members strengthened by EGC, laboratory and field explosion experiments were conducted in this study, respectively.A variety of synthetic fibers were considered to explore the effect of fiber types on EGC properties in terms of workability, dry shrinkage, and compressive, tensile and flexural behavior.Subsequently, the blast response of typical structural elements retrofitted by EGC with preferred fiber were examined.The tests include autoclaved aerated concrete (AAC) masonry walls subjected to vented gas explosion, reinforced AAC panels subjected to near-field TNT explosion, and plain concrete slabs subjected to contact explosion.The excellent contribution of EGC to enhancing the blast resistance of structural elements was proven with respect to unstrengthened members as references.It helps to propose a solution for quickly strengthening structural elements against blast loading.

2.Laboratory experiments for tailoring EGC

2.1.Experimental preparation

2.1.1.Materials and mixture preparation

The precursor materials to synthesize geopolymer were low calcium class F fly ash(FA)and ground granulated blast furnace slag(GGBFS).Their chemical compositions are given in Table 1.Solid sodium silica(molar ratio of SiO2/Na2O=2.0) containing 21.7 wt%bonding water was used as alkali activator.Fine quartz sand with medium particle size of 214 μm was used as the aggregate.The particle size distributions of FA, GGBFS and fine quartz sand are shown in Fig.1.

As presented in Fig.2, four types of fibers were considered to prepare the EGC, including polyvinyl alcohol (PVA), polyethylene(PE), polypropylene (PP), and macro polypropylene (MPP).Among them, PVA, PE and PP microfibers are commonly used in ECC.At present study [31], MPP also got attention due to its low-cost,which can significantly enhance the ductility of composites.However, the limited fiber content (0.5 vol%) adopted in the existing study limits the ability of fiber bridging for cracks.Discussing the properties of composites with high content of MPP is meaningful.PVA with oil coating was supplied by Kuraray.PE,PP and MPP came from Chinese distributors.The physical and mechanical properties of fibers are tabulated in Table 2.

The samples were synthesized according to Table 3, which indicated the mixed types of fibers.For example, EGC-PVA represents EGC specimens mixed with PVA fibers.From the perspective of toughness enhancement, the more fiber content, the better the performance of concrete.The volume fraction of microfibers (e.g.,PVA, PE and PP) was maintained at 2%, which is the optimal proportion that allows materials with ECC-like properties to have exceptional tensile strain capacity [32-34].For EGC-MPP, preliminary experiments showed that EGC reinforced with 2 vol%MPP achieved lower strength and toughness.Instead,the EGC reinforcedwith 3 vol% MPP exhibited adequate workability and better mechanical performance.Thus, 3 vol% MPP fibers were mixed in this study.

Table 1 Chemical compositions (wt%) of ingredients.

Fig.1.Particle size distribution of FA, GGBFS and fine quartz sand.

The mixtures were prepared in a 20 L mixer as follows.Firstly,dry ingredients including binder, activator and aggregate were agitated for 3 min to ensure adequate homogeneity.After that,water was added and mixed for 2 min, followed by adding and stirring 3 min with high speed for proper fiber distribution in geopolymer slurry.The mixture was then transferred into moulds described in subsection 2.1.2,followed by 30 s vibration to allow the air bubble to escape.All the samples were covered in plastic film and stored for 1 day in the ambient environment.Finally, the demolded specimens were cured at 20±2°C and relative humidity of 95% until the desired testing ages.

2.2.Test procedures, results and discussions

2.2.1.Workability

For practical application,the workability of EGC determines the ease of construction.The flow table test determined the workability per ASTM C1437-15 [35].The spread diameters in perpendicular directions were measured to obtain the averaged flow value.Two repetitions were performed.

Fig.3 illustrates the workability of matrix and EGCs with various fibers.As expected, the workability decreases with the addition of fiber since fiber dispersion in the slurry requires water wetting.By comparison, EGC-PE exhibits the lowest workability.This may be attributed to the largest specific surface area of PE fiber due to its high aspect ratio, even though it has a hydrophobic surface.The workability of EGC-PVA is slightly higher than that of EGC-PE.This may be attributed to the oil-based coating on PVA fibers, which reduces water demand.The flow diameter of EGC-PP is 195 mm.Since PP fibers have a lower surface energy than PVA and PE fibers,it means PP fibers are more hydrophobic [36,37].Thus, PP fibers require less water than PVA and PE fibers during blending.Macro fiber MPP caused the least reduction in the workability of the matrix due to its larger diameter and smaller aspect ratio,even with 3% volume fraction.

2.2.2.Drying shrinkage

Fig.2.Fiber appearance.

Table 2 Physical and mechanical properties of fibers.

Table 3 Mix proportions of specimens (by wt.).

Fig.3.Comparison of the workability of EGCs.

The drying shrinkage caused by moisture loss in hardened composites has a crucial impact on their durability and service performance.In addition to causing cracks and weakening the tensile capabilities of samples, when employed as a strengthening material, excessive drying shrinkage may also diminish the bonding properties between the overlay and the substrate [38].Three prism samples(40 mm×40 mm×160 mm)with embedded metal nails at both ends were prepared to measure the drying shrinkage for each group.The initial length L0was captured by a dial indicator comparator (accuracy: 0.001 mm) once the samples were demolded.The prisms were cured at 20 ± 2°C and 40 ± 5%relative humidity.The lengths of the prisms were measured at 1,7,14, 21 and 56 days, respectively, recorded as Lt.The drying shrinkage was calculated as per Eq.(1).

Fig.4.Comparison of the drying shrinkage of EGCs.

Fig.4 depicts the drying shrinkage of specimens at different curing ages.It can be seen that the drying shrinkage of all samples exhibits sharp raise during the initial curing stage and stabilizes until 14 days.Overall, samples with microfibers (e.g., PVA, PE and PP) exhibit significant decrease in shrinkage.After the specimens are cured for 56 days, the incorporation of PVA, PE and PP fibers results in the comparable reduction of drying shrinkage (1019 με,1047 με and 1112 με, respectively).It is indicated that microfiber type has little impact on the drying shrinkage of composites under the same volume content.Due to the limited drying shrinkage,cracks are not detected in the specimens [39].However, MPP can only reduce 14%drying shrinkage of the matrix due to its relatively large diameter and limited number of fibers.

2.2.3.Compressive behavior

The compressive tests were conducted according to ASTM C109-16[40]on the samples with 50 mm×50 mm×50 mm dimensions.The loading equipment automatically captured the axial displacement during the test.A total of three samples were investigated for each mix.

Fig.5 depicts typical stress-displacement curves of EGC under compression.It is observed that the matrix exhibits brittle failure after peak stress while EGCs exhibit post-peak residual resistance.EGC-PE exhibits superior compressive ductility compared to EGC containing other microfibers.It may be attributed to the activated fiber bridging capacities after matrix cracking.Moreover,EGC-MPP shows the highest post-peak ductility due to the high energy dissipation capacity caused by large geometrical shape and fiber content.

After fiber inclusion in the matrix,the compressive strengths of EGC-PVA, EGC-PE and EGC-PP are improved by 11%, 18% and 2%,respectively.Similar results were reported by Nematollahi et al.[41] who observed that micro fibers have a positive effect on the strength of geopolymer composite.It may be attributed to the strong frictional and chemical bond between micro fibers and geopolymer matrix [41].However, the excessive macro fibers of EGC-MPP are prone to entangle together,resulting in more defects in the specimens.Therefore,the compressive strengths of EGC-MPP are smaller than those of the matrix.

2.2.4.Tensile behavior

Fig.5.Typical stress-displacement curves of UTHGC under compression.

Fig.6.Direct tension test setup and specifications of specimens.

The tensile behavior was evaluated according to JSCE Recommendations[42].Fig.6 shows the geometric size of the dog-boneshaped sample and direct tension test setup.The test was conducted with a displacement control of 0.5 mm/min using a universal testing machine of 5 kN load capacity.The average displacements captured by LVDTs (Linear Variable Displacement Transducers) were considered in calculating the strain.Five repetitions were conducted for each mix of specimens.

Fig.7 displays the tensile stress-strain curves of EGCs and corresponding cracking patterns after failure.EGC-PE exhibits exceptional strain hardening properties with ultimate tensile strains over 4%, whereas the tensile strain capacity for EGC-PVA is limited(approximately 2%).The tensile properties of the specimens can also be somewhat reflected in the cracking characteristics.Notably,saturation cracking behavior in EGC-PE was considerably remarkable,demonstrating that composite incorporating PE fiber achieves outstanding ductility.

The fiber/matrix interface properties play crucial roles in the strain hardening property of composites [43].PE fiber has a hydrophobic surface and a negligible chemical bond with the matrix.In contrast, the surface of PVA has inevitable chemical bond even though its hydrophilic surface is coated with oil [30].Relatively high chemical bond between fiber and matrix is prone to causing fiber rupture and diminishing the bridging stiffness.The absence of fiber slippage-induced interfacial frictional work causes the bridging fibers to be less energy-absorbent[30].The factors above could lead to relatively low complementary energy in EGC-PVA,which cannot activate more new cracks at the crack tip [44].Thus,the number of cracks in EGC-PVA is less than that of EGC-PE.

In contrast, EGC-PP and EGC-MPP display tension-softening characteristics.The stress-strain curves of them drop promptly after initial cracking.After that, the activated bridging fibers cause the stress to increase to a certain extent, although it is still below the initial cracking stress.It is demonstrated that the stress required for a newly formed crack is higher than the maximum fiber bridging stress.Consequently, ECC-PP and ECC-MPP had single localized widening cracks after failure and the decline of loadcarrying capacity.As expected, the tensile strength of EGCs shows a similar trend to the above strain capacity,in descending order of EGC-PE, EGC-PVA, EGC-MPP and EGC-PP.

2.2.5.Flexural behavior

Fig.7.Comparison of tensile stress-strain curves of EGCs: (a) EGC-PVA; (b) EGC-PE; (c) EGC-PP; (d) EGC-MPP.

Fig.8.Diagrammatic drawing of bending test.

As a strengthening material, the flexural behavior of EGC has a significant impact on the mechanical response and energy absorption of the structure.Four-point bending tests were carried out to evaluate the flexural strength and ductility of EGC,as presented in Fig.8.The test machine was the same as the tension test mentioned above.The displacement rate used was 2 mm/min.LVDT mounted on the samples measured the midspan deformations.The ability of the material to withstand significant deformation while maintaining its load-bearing capacity is represented by ductility index (DI).It can be calculated using Eq.(2).

where δPrepresents the deflection at peak load (mm), and δFrepresents deflection corresponding to first-crack load (mm).

Fig.9 displays the flexure stress versus midspan deflection curves of specimens.The composites maintain or gain load capacity with increasing mid-span displacement after initial cracking,except EGC-PP.In comparison, EGC-PE outperforms EGC-PVA and EGC-MPP in terms of flexure strength and ductility.The fiber bridging capacity discussed in subsection 2.2.4 may account for the performance difference between EGC-PE and EGC-PVA.

It is interesting to note that EGC-MPP possesses excellent ductility with a DI of 20.2.Due to the limited specimen thickness,MPP fiber with a 20 mm length is orientated parallel to the length of the specimen.In addition,the MPP fiber features high maximum elongation and low elastic modulus, according to Table 2.The factors above may result in the excellent ductility of EGC-MPP.However,further research ought to be explored on the larger specimen size with MPP fibers.

Fig.9.Comparison of flexural behavior of EGCs.

Based on the results discussed above, EGC-PE outperformed EGCs containing other fibers in terms of overall evaluation including workability,dry shrinkage,and compressive,tensile,and bending properties.Thus, PE fibers are recommended for the optimal performance of EGC and are considered in field blast tests as reported in the following sections.

3.Examination of EGC retrofitting on blast response of typical structural elements

In this study, the effectiveness of EGC prepared by PE fiber was proven by a series of field tests,including AAC masonry walls(built by AAC blocks) subjected to vented gas explosion, reinforced AAC panels subjected to near-field TNT explosion and plain concrete slabs subjected to contact explosion.

3.1.AAC masonry walls subjected to vented gas explosion

3.1.1.Test specimens

The dimension of the AAC masonry wall constructed by AAC block was 3 m × 2 m × 0.15 m.The AAC block had a density of 580 kg/m3.Its compressive and tensile strengths were 3.44 MPa and 0.53 MPa[45],while those of EGC were 58.1 MPa and 3.3 MPa,respectively.The ultimate tensile strain of EGC was about 3%.Fig.10 shows the preparation processes of samples.The AAC blocks(600 mm × 250 mm × 150 mm) were installed in the running pattern on the RC frame to design a one-way wall.Generally,increasing the surface roughness of substrate is usually performed to enhance the bond between the strengthening layer and the substrate.However,due to the low density and strength of AAC,the above method may cause damage to this material and reduce its integrity.Thus, in this study, the polyvinyl alcohol based adhesive was applied on the wall surface for a strengthened AAC masonry wall to enhance the bond between the substrate and EGC.Finally,a 5-mm thick fresh EGC was smeared and troweled on the wall surface and cured at least 28 days before testing.Its thickness only accounts for 3% of the wall thickness, this small difference has a negligible influence on the dynamic response of test planes.

3.1.2.Test program

Fig.11 shows the arrangement of the equipment for vented gas explosion.The testing system consisted of an RC chamber, AAC masonry wall specimens, gas systems, and data-acquisition devices.The detailed design and installation of RC frame and testing procedure were reported by Li et al.[46,47].A 5 mm-thick glass panel was installed on the vent window to control the appropriate gas explosion loads [48,49].A pressure sensor was mounted near the vent window to record the explosion load inside the chamber.The sliding door connected the gas concentration detection system to the chamber.As shown in Fig.11(b),the mid-span displacement of the wall was recorded by LVDT.A high-speed camera(frame rate:1000 FPS)was positioned at a distance of about 15 m from the wall to monitor the failure processes of wall specimens.

Fig.10.Preparation processes of samples.

Fig.11.The arrangement of the equipment on the RC chamber and wall: (a) Back view; (b) Front view.

This study conducted two tests under 12.5% methane concentration and the same venting pressures, as shown in Table 4.

3.1.3.Results and discussions

Fig.12 shows the overpressure-time histories acting on AAC masonry walls under vented gas explosion.With the same gas concentration, the peak pressure of W0 and W5 is 10 kPa and 14 kPa, respectively.Accordingly, the impulses, representing the cumulative effect of pressure on time, of W0 and W5 are 1.12 kPa∙ms and 2.07 kPa∙ms,respectively.In other words,the additionof EGC overlay resulted in a more significant load in vented gas explosion test.It might be attributed to the weak blast resistance of the unstrengthened AAC masonry wall,which was easily damaged during the rise of the pressure of the explosion inside the chamber.The damage and cracks provided more area for pressure release.Therefore, the peak pressure and impulse collected in W0 are lower.By comparison, the specimen strengthened by EGC has higher blast resistance.The blast pressure was totally released through the cracked vent window.

Table 4 Test program.

Fig.12.Comparison of overpressure-time histories of AAC masonry wall under vented gas explosion.

The failure process and modes of the tested specimens are shown in Figs.13 and 14, respectively.It can be seen in Fig.13(a)that one-way bending was the predominant failure mechanism,and W0 showed a fracture in the mid-height region, followed by the collapse of the fractured specimen.The wall fragments were up to 5 m from the RC chamber, as shown in Fig.14(a).

In contrast,W5 resisted the larger pressure of the gas explosion and appeared only a few cracks.The pressure and fire could only escape through the vented window.It is indicated that the structural stiffness of the wall is enhanced by EGC overlay.After the explosion, negligible damage occurred on the internal surface,while a few cracks appeared on the middle of the exterior surface as shown in Figs.14(b) and 14(c).

Fig.15 shows the mid-span displacement-time history of W5 specimen under vented gas explosion.It can be seen that the residual mid-span displacement of W5 is only 1.5 mm.It may be attributed to the high energy dissipation capacity of EGC by bridging cracks, even at high strain rates[50].

3.2.Reinforced AAC panels subjected to near-field TNT explosion

3.2.1.Test specimens

In this study,unstrengthened and strengthened reinforced AAC panels were fabricated and named P0 and P5, respectively.P represents the panel, the number after the letter represents the thickness of EGC.The P0 with 2420 mm × 600 mm × 100 mm is schematically shown in Fig.16, where double layers of 5 mm diameter orthogonal rebars were arranged in the panels.The yield strength of rebars was 860 MPa.The AAC cover thickness was 25 mm.The density of AAC was 680 kg/m3,and its compressive and tensile strengths were 4.5 MPa and 0.55 MPa, respectively.For P5,the thickness of the EGC overlay on the bottom side of panel was 5 mm.It should be noted that this thickness accounts for only 5%of that for AAC panel.The difference in panel thickness hardly affects the response of the panel.The mechanical parameters of EGC are same to that in subsection 3.1.1.The strengthening process was carried out as in subsection 3.1.1.

3.2.2.Testing program

As illustrated in Fig.17, the testing system consisted of specimens, box-like steel supports, TNT charge, overpressure sensors and LVDTs.The box-like steel supports were described by Su et al.[51].The span of the testing specimen was 2 m.The cylinder TNT charge was suspended above the center of the panel, and its axial direction was parallel to the width of the panel.The incident and reflected overpressure sensors (sampling frequency: 1 MHz) were positioned on the midline of the width of the panel,equal distance from the TNT and the panel center.They were employed to record the pressure of the specimen center.LVDT was installed to record the quarter-span deformation of panel, respectively.After explosion, the residual mid-span and quarter-span deformations of the specimens were measured manually.

Fig.14.Failure modes of AAC masonry walls after vented gas explosion(a)W0;(b)W5(internal face); (c) W5 (external face).

Fig.15.Mid-span displacement-time history of W5 specimen under vented gas explosion.

Fig.16.Geometry and reinforcement of test panel.

The test schemes are summarized in Table 5, where the specimens were tested under the explosion with TNT charge weighing 2 kg.The standoff distances (R) and the corresponding scaled distances (Z) were 1.4 m and 1.11 m/kg1/3(in which, Z = R/W1/3, W is the charge weight of the TNT).This explosion test with Z less than 1.2 m/kg1/3is considered a near-field explosion,as per ASCE 59-11[52].

3.2.3.Results and discussions

The overpressure-time history curves of each blast scenario are presented in Fig.18.Table 6 shows the detailed blast data in terms of the peak value of overpressure and corresponding impulse.It can be seen that the explosion test has inevitable errors.However,this is within permissible limits.

Fig.19 shows the post-blast damage of panels.For P0,the typical bending failure occurred in the specimen, and severe cracks and fragmentations were detected on both sides.On the lateral side,the steel bars in the specimen were visible and bent to some extent.It indicates that the poor mechanical strength of AAC makes it impossible to withstand the given explosion pressure.By comparison, P5 exhibits superior blast resistance.A few longitudinal and transverse cracks were observed on the top side.However, on the bottom side, only the micrometer cracks were detected.It implies that the EGC overlay consumes more explosion energy, and the fiber bridging of EGC also limits the derivation and propagation of cracks of panel.

Fig.20 shows the quarter-span deflection-time histories of the control and strengthened specimens.The detailed deflection parameters are summarized in Table 7.It can be seen that the maximum quarter-span deflection of P0 is up to 93 mm,while that of P5 is only 41 mm.Furthermore, the mid-span and quarter-span residual deflections are significantly decreased by 63% and 67%after strengthening, respectively.The effectiveness of EGC overlay on enhancing the blast resistance of reinforced AAC panel under near-field TNT explosion is demonstrated.

3.3.Plain concrete slabs subjected to contact explosion

3.3.1.Test specimens

Contact explosion tests are introduced in this section.The propagation effect of the blast wave predominates the impact of contact explosion on the members [53].The longitudinal and transverse steels in the reinforced concrete component normally modify the blast wave propagation and restrict the concrete deformation, complicating the interpretation of results.On the other hand, in this experiment, whether steels are added or not is not a key parameter, so the validity of the results is not affected.Therefore,the TNTcontact explosion tests were carried out for plain concrete slabs to avoid the influence of steel and more straightforwardly study the strengthening effect of EGC.

Fig.17.The test setup of AAC panel subjected to near-field TNT explosion.

Table 5 Test schemes.

Fig.18.Overpressure-time history curves of test specimens.

Table 6 The peak overpressure and impulse of mid-span of specimens.

Three test specimens were fabricated in the experimental program, including one unstrengthened plain concrete slab and two EGC-strengthened concrete slabs.The dimensions of the unstrengthened slab were 1500 mm×1500 mm×200 mm.EGCs were available in two strength grades and strengthened on the blast rear side of samples.The preparation process was similar to that in Section 2.1.1.The basic mechanical properties of concrete and EGCs are shown in Table 8.45 and 52 denote EGCs with compressive strength of 45 MPa and 52 MPa, respectively.

For strengthened specimens, it is worth mentioning that the smooth surface between the strengthening material and substrates may cause debonding failure under dynamic loading [13], which further reduces the protective effect of the overlay.Thus, the slab surfaces were roughened by steel wire gauzes before the concrete hardened, as shown in Fig.21.After slabs being cured at ambient temperature for 28 days,the polyvinyl alcohol based adhesive was applied on the concrete surface.And then the 10 mm-thick EGC layers were manually retrofitted on the rear face of concrete slabs and cured for another 28 days.

3.3.2.Test program

As illustrated in Fig.22, the specimen was freely placed on the steel support.The TNT charges were placed at the center of the slab.Table 9 shows the test program.CS0 represents unstrengthened concrete slab.CS45 and CS52 represent that the concrete slabs are strengthened by 45 MPa and 52 MPa EGCs, respectively.

3.3.3.Results and discussions

The failure modes of specimens under contact explosion are displayed in Fig.23, and the corresponding failure parameters are summarized in Table 10.It can be seen that EGC overlay can significantly enhance the blast resistance of plain concrete slab,which is related to the strength of EGC.

Fig.19.Post-blast damage of panels: (a) P0 (top side); (b) P0 (bottom side); (c) P0(lateral side); (d) P5 (top side); (e) P5 (bottom side); (f) P5 (lateral side).

Significant crater and spallation were observed for the plain concrete slab under the contact explosion of 0.4 kg TNT charge.For contact explosion, the stress wave propagation effect predominates.The blast wave exerts a sudden and intense pressure on the concrete surface, causing the concrete to experience high compressive stresses.As the blast wave propagates to the opposite side of the concrete,it reflects and forms a tensile wave.The crater is generated due to the excessive compressive stress wave that damages the concrete material,while the spalling is induced due to the reflected tensile stress wave.As the tensile strength of concrete is much less than its compressive strength,the diameter and depth of spallation are more severe than those of the crater [54].The diameter and depth of the crater on the front side were 36 mm and 8.5 mm, while those of spalling on the rear side were 67 mm and 10 mm, respectively.In addition, the concrete cracks divided the specimen into four parts.

Fig.20.Deflection-time history of specimen.

Table 7 Maximum and residual displacements of test specimens.

Table 8 The mechanical properties of concrete and EGCs.

Fig.21.The rough surface of slab.

Fig.22.Testing arrangement.

Table 9 Test program.

By comparison, the failure modes of strengthened specimens are significantly altered under the same explosion condition.As shown in Fig.23(b)(i.e.,CS45),the diameter and depth of the crater decreased to 33 mm and 5.5 mm.A portion of cracks was distributed radially on the center of the specimen,and long cracks parallel to the edge length direction also were detected near the supports.On the rear side, spalling was also induced with 64 mm diameter,and the corresponding coating was partly damaged.Moreover,multiple radial cracks could be observed on the surface of EGC.Since the bond between EGC and concrete is strong enough, they can work together when the blast wave propagates inside the specimen.This ensures that the explosive stress is effectively transmitted to EGC.In addition, due to the strain-hardening and multi-crack characteristics of EGC under tensile stress,it can help to absorb more explosive energy under blast loading by having larger deformation.Therefore, the EGC overlay can effectively reducing the damage degree of concrete slabs under explosion conditions.

Fig.23.Post-blast damage of panels: (a) CS0; (b) CS45; (c) CS52.

When the compressive strength of EGC overlay increased from 45 to 52 MPa, the cratering diameter dropped to 31 mm.The spalling caused by the tensile reflected wave was almost extinguished.Significant reduction in the number of microcracks was observed on the rear side.It might be attributed to the ability of fiber bridging of crack being positively correlated with the sample strength to a certain extent [41].With the increment of matrix strength, the denser microstructure of the matrix consequently enhances the adhesion properties (e.g., the chemical bonding and frictional bonding)of fiber/matrix[55].The peak bridging stress of fiber increases and the crack opening decrease[55].Consequently,EGC overlay with higher strength has better energy absorption capacity.

4.Conclusions and outlooks

This is the first study to experimentally explore the feasibility of EGC in strengthening structural elements against blast loads.Through comprehensive evaluation with consideration of different types of fibers, PE fibers were found to outperform other types offibers in terms of compressive, tensile and bending behavior.Furthermore, EGC-PE has acceptable workability and qualified dry shrinkage.Thus,PE fibers were selected to prepare EGC.

Table 10 Comparison of test results.

The effectiveness of EGC in strengthening typical structural elements against blast loading was demonstrated by a series of field tests, consisting of AAC masonry walls subjected to vented gas explosion, reinforced AAC panels subjected to near-field TNT explosion and plain concrete slabs subjected to contact explosion.After strengthening by a thin EGC layer, the failure mode of structural element was altered and the damage was significantly reduced.The strengthened AAC masonry wall remains intact after the gas explosion.For strengthened reinforced AAC panel,the midspan residual deflection was significantly decreased by 63%.For plain concrete slab strengthened by higher strength EGC,the crater depth dropped up 47% and the spalling caused by the tensile reflected wave was negligible.The marked effectiveness of EGC retrofitting should be attributed to the strong bond between EGC and substrate and the multiple cracking and strain-hardening characteristics of EGC.More energy was dissipated through EGC layer applied on the rear side of the element.Therefore, EGC strengthening method has great prospects for emergency reinforcement of structures under explosion loading.

While EGC has been demonstrated to be an excellent blast retrofitting material, it has been observed that only limited load equivalents, strengthening layer thickness, strength grade and strengthening schemes are considered.A systematic research program should be envisioned, as well as design criteria.In addition, this study uses the EGC manually, which limits construction efficiency.Based on its rheology and setting features, its shootability should be realized in the future.It is crucial for the effective building of emergency repairs and robust engineering.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This research was supported by National Natural Science Foundation of China (Grant Nos.51908188 and 51938011).