Experimental evaluation on the axial crushing performance of BFRP-bamboo winding composite hollow components

: In order to study the quasi-static axial crushing performance of BFRP-bamboo winding composite hollow components, considering the cloth ratio of BFRP (0%~14.68%) the number of BFRP layers (0 layer~4 layers) as an influencing factor, 20 tube specimens were designed for quasi-static compression tests. In this paper, the failure modes of the specimens under quasi-static axial compressive load are presented with the relevant load-displacement curves. The deformation types were carefully studied to evaluate the compressive crushing indicators of the specimens. The test results showed that when the cloth ratio of BFRP increased from 0 to 14.68%, the specific energy absorption (SEA), the mean crushing force (MCF) and the crushing load efficiency (CFE) increased to some extent, whilst the initial peak crushing force (PCF) did not show any trend. When compared with those of the bamboo winding hollow components (BT), the SEA, MCF and CFE of BFRP-bamboo winding composite hollow components with four layers of BFRP winding outside of BT (BBT4) increased by 87.53%, 194.37% and 255.59% respectively. Compared with other composite hollow components such as composite wrapped hollow components (CWT) and carbon reinforced composite hollow components (CRCT), BFRP-bamboo winding composite hollow components (BBT) showed superior crushing resistance while offering the advantages of light weight.


Introduction
Bamboo is known as a natural green building material [1,2] offering high strength-to-weight ratio, and plays a vital role in structural application [3,4].With the development of science and technology, many different kinds of engineered bamboo products were invented, such as laminated bamboo lumber (also named glued laminated bamboo), parallel bamboo strand lumber (also named scrimber bamboo), GluBam, bamboo winding composite material, bamboo chip composite material, bamboo plastic composite material, bamboo wood composite material, bamboo straw composite material and so on [3].Engineered bamboo products can partly replace traditional concrete, metal or plastic in different engineering area and our everyday life, which will effectively reduce the use of mineral products reducing the demand for high energy consumption products and eventually save resources and thus protect the environment [3][4][5][6][7].
As for bamboo winding composite material, it was invented firstly by Chinese scientists from Chinese Academy of Forestry and Nanjing Forestry University in 1950's, and bamboo curtain or bamboo mat was used as the winding element then [8].Ye et al. [9][10][11] has done a lot of research on bamboo winding composite tubes made of bamboo curtains.Li et al. [3,[12][13][14][15] invented many manufacturing methods for bamboo winding composites components with bamboo veneer or bamboo skin in 2008.Since 2008, Li et al. [3,[12][13][14][15] continued the research on bamboo winding composite materials, developing a series of products leading to obtaining more than 60 patents with different kinds of winding elements such as bamboo curtain (bamboo mat), bamboo veneer or bamboo skin and so on.The bamboo winding hollow components developed by his research team, in various diameters, can be applied in civil engineering, transportation, water conservancy, and other infrastructure construction fields, as shown in Basalt fiber reinforced polymer (BFRP) is an environmental-friendly building material [16].Wrapping BFRP on the outside of traditional materials has been shown to improve the durability of the material while enhancing its strength and ductility [17].Previous studies on fiber reinforced metal hollow components and fiber reinforced plastic hollow components have clearly demonstrated that they have excellent crushing resistance [18,19] which led to their wide usage in aerospace, automotive industry and other construction industries [20,21].
Researchers from across the world have done significant research on the quasi-static axial crushing properties of traditional metal hollow components or fiber reinforced composite hollow components.They explored the failure characteristics and crushing resistance of specimens by considering factors such as section shape [22][23][24], loading mode [25][26][27] and stacking direction [28].
No significant research has been reported on the quasi-static axial crushing performance of bamboo or winding wood circular hollow components.Li et al. [14] investigated the axial compressive behaviour of transversely reinforced bamboo winding tubes.Research has shown that bamboo winding hollow components, when reinforced transversely, exhibit significant improvements in axial compressive strength and ductility, making them suitable for structural applications, where high compressive strength is required [14].Jian et al. [15] have demonstrated that the winding angle of bamboo tubular sections plays a crucial role in their compression properties.By varying the winding angle, the structural integrity and energy absorption capacity of bamboo winding hollow components can be optimized; this clearly highlights the importance of careful design considerations in bamboo composite materials.Zhang et al. [29] carried out the ring stiffness test of winding bamboo composite pipe; the results showed that the ring stiffness of pipe at the connection is greater than that of the header pipe.Ma et al. [30] carried out the bending strength test of winding bamboo composite pipe.The bending strength difference between the two methods showed that the bending strength of the axial specimen much higher.Wei et al. [31] studied the axial compression properties of bamboo composite tubes, analyzed the stress-strain relationship based on the continuous strength design method.Gué lou et al. [32] compared the SEA and cost of wood tube and carbon tube, and concluded that the SEA of wood tube is approximately 1/3 rd of carbon tube, and the cost of wood tube is about 1/40 th of carbon tube.Gué lou et al. [33] prepared carbon fiber-wood based sandwich composite hollow components and carried out quasi-static compression tests.The results showed that the amount of absorbed energy changed linearly with the number of poplar veneers used.
Most of the previous studies focused on the quasi-static axial crushing performance of CFRP or winding GFRP composite hollow components, but not on the quasi-static axial crushing performance of BFRP-bamboo winding composite hollow components.This study aims to evaluate the axial crushing performance of BFRP-bamboo winding composite hollow components, fill the gap of axial compression performance of BFRP-bamboo winding composite hollow components, and promote its application in municipal engineering as well as civil engineering.

Preparation of specimen and test setup 2.1 Preparation of specimen
The bamboo winding hollow components (BT) and BFRP-bamboo winding composite hollow components (BBT) were manually produced using the following procedure.The production process of BT ca be summarized as follows: (1) The plastic wrap was bound on the PVC pipe coated with release agent; (2) The glued bamboo sheets were wrapped on the PVC pipe; (3) The processed specimens were wrapped with plastic covering; (4) After the specimen was completely cured, the PVC pipe was pulled out and the plastic covering on the surface was removed; (5) Both ends of the specimen were polished flat.The production process diagram is shown in Fig. 2. BBT was produced by winding BFRP on the outer surface of BT.The fiber direction of bamboo was 0° and the fiber direction of BFRP was 90°.
The height of both BT and BBT was 126 mm.BT was made of 10 layers of bamboo sheets, with a single layer thickness of 0.4 mm.BBT1, BBT2, BBT3 and BBT4 were produced respectively by winding on the outer surface of BT with one layer, two layers, three layers and four layers of BFRP, resulting in different BFRP cloth ratios.The thickness of the single layer of BFRP was 0.15 mm.BFRP cloth ratio (ρF) is defined as the percentage of BFRP cross-sectional area in the cross-sectional area of the entire tube, and may be calculated using Eq. ( 1): where, ABFRP is the cross-sectional area of BFRP, Atube is the cross-sectional area of the entire tube.The current study considered 20 tube specimens, which were divided into 5 groups.The main parameters of each group of specimens are listed in Table 1.The specimen sizes are illustrated by taking BT and BBT1 as examples, as shown in Fig. 3.The loading work of the test was completed in Jiangning Training Center of Nanjing Forestry University.The quasi-static compression test was carried out using a 50 kN universal testing machine.The specimen was placed between two rigid platens.The upper platen was used to apply compression while the lower platen remained fixed.The upper platen moved downward at a speed of 3 mm/min until the set crushing displacement was reached.Two laser displacement meters were placed at the height of the lower platen to ensure that the laser emitted by the two laser displacement meters was on the same line and passed through the axis of the specimen.The displacement data was collected by TDS-530 data acquisition instrument, and the load was controlled by the universal testing machine.A still photo camera and a video camera were placed in front of the specimen to take pictures and record the whole process of specimen crushing.The diagram of quasi-static compression test setup is shown in Fig. 4.

Material properties
In order to obtain the basic material properties of bamboo sheet and BFRP, dumb-bell shaped tensile specimens of bamboo sheet and BFRP were prepared, and seven repeated tests were conducted on each specimen following ASTM D3039/D3039M-14 [34].The quasi-static tensile tests were conducted using a 50 kN universal testing machine with a loading speed of 1 mm/min.The tensile stress-strain curves of bamboo sheet and BFRP along grain direction were obtained from the test and the material properties of bamboo sheet and BFRP were calculated from the obtained data.According to the data (Table 2), the tensile strength and elastic modulus of bamboo sheet were 150.17MPa and 26 GPa, respectively, whilst those for BFRP were 987.95MPa and 115 GPa, respectively.The mechanical properties of the adhesive were provided by the manufacturer; the specified minimum values for various strength parameters are as follows: tensile strength 30 MPa, flexural strength 40 MPa, tensile shear strength 10 MPa, and compressive elastic strength 1.0×10 3 MPa [35].

Method of calculation
The evaluation indicators of energy absorption obtained from the test are as follows: The initial peak crushing force (PCF) is the maximum force that first appears in the initial stage of crushing [27], which can be easily obtained from the load-displacement curve.
The total energy absorption (EA) is defined as the total energy dissipation of the structure during the deformation process [36], which can be calculated by integrating the area below the loaddisplacement curve as follows: The specific energy absorption (SEA) represents the energy absorption per unit mass of the structure.SEA can be defined as the ratio of EA to the total mass (m) of the tube [37], as follows: The mean crushing force (MCF) is the average value of the crushing force during axial loading [38].It is defined as the ratio of EA to d, which is given by the following formula: The crush force efficiency (CFE) is usually used to evaluate the consistency of the load, expressed as follows: where, F(x) is the compression load, d is crushing displacement, dx is the differential of crushing displacement, m is the total mass of specimen.

Test results
The test results are listed in Table 3.

Failure modes
In the process from the initial load to 70% of the initial peak load, there was no obvious crack and deformation on the surface of BT and BBT specimens.When the load reached almost to 95% of the initial peak load, BT and BBT specimens were gradually condensed and bulged outward.When the load reached the initial peak load, the bamboo fiber cracked, the BFRP was folded with audible noise of fiber fracture.With the compression process continuing, the cracks gradually expanded, and the specimen was destroyed.
According to the test process, in general, there were three failure modes for BT and BBT, Fig. 5.
Failure mode I: Opening.The upper bamboo fiber cracked and progressed to the lower end, causing full-length cracks as shown in Fig. 5 (a).Such failure mode mainly occurred in BT and BBT1.The damaged BT looked like an open "petal".As for BBT1, in addition to bamboo fiber cracks, the BFRP on the outside was also torn into irregular sheets.
Failure mode II: Overall buckling.The specimen was folded and buckled at several places, resulting in overall crushing failure, as shown in Fig. 5 (b).This kind of failure mode mainly occurred in BBT2, BBT3 and BBT4.The specimen folded in many places along the height, and the fiber folding was the most obvious in BBT4.
Failure mode III: Local buckling.Only occurred in BBT2-1.The specimen was asymmetrically inclined, the upper end was partially depressed causing the bamboo fiber to be bent and folded inward.Eventually the lower bamboo fiber was broken and the BFRP was torn, as shown in Fig. 5 (c).

Load-displacement curves
The load-displacement curves of the specimens are shown in Fig. 6.The corresponding deformation modes are also added to each figure.
Fig. 6 (a) shows the load-displacement curves and deformation mode of BT.The deformation mode is shown by taking BT-2 as an example.BT-3 and BT-4 were similar to BT-2, and BT-1 data was not successfully collected.During the initial stage of crushing, the load increased with the increase of displacement.In the middle crushing stage, the crushing load fluctuated between 5 and10 kN, and the area enclosed below the load-displacement curve was relatively large, indicating the specimen absorbed most of the energy during this stage.In the late crushing stage, the load carrying capacity decreased, with the crushing displacement becoming smaller and the energy absorption becoming less.In the initial stage of crushing, bamboo sheets started cracking along the grain direction from the upper end.With the loading platen moving down, the specimen became flaky due to delamination cracking.Finally, the layer bundle broke, and the specimen was damaged.
Fig. 6 (b) shows the load-displacement curves and deformation mode of BBT1.The deformation mode is shown by taking BBT1-2 as an example.Other specimens in the same group were similar to BBT1-2.The shape of load-displacement curve of BBT1 was similar to that of BT.The main difference between BT and BBT1 is that the former had a large initial peak load, while the latter absorbed more energy when it was damaged.In the initial stage of crushing, the bamboo fiber in the upper end broke in a brittle manner and the BFRP started being folded.Then the bamboo sheets became flaky due to delamination cracking and the BFRP was torn.Finally, a mixed failure mode was formed.6 (c).The load-displacement curve of BBT2-1 showed three-stage linearity, while the rest two curves were not completely similar to the load-displacement curves of BT or BBT1 neither.Their differences were mainly shown in that the load carrying capacity of BBT2-2 and BBT2-3 had been greatly improved in the late crushing stage, which was due to the BFRP circumferential constraint on the radial fracture of bamboo fibers.Bamboo fibers were then forced to bend and fold inward, finally the bamboo fragments were compacted and thus improved the bearing capacity.As for BBT2-1, the crushing started from the oblique crack at the upper end, then a large plastic deformation occurred, thus it had a low load carrying capacity and a large plastic displacement; the energy absorbed by the specimen crushing was large.In general, BBT2 absorbed more energy than BT and BBT1 when damaged.
Fig. 6 (d) shows the load-displacement curves and deformation mode of BBT3.The deformation mode is shown by taking BBT3-4 as an example.Other specimens in the same group were similar to BBT3-4.The load-displacement curves of BBT3 were mostly similar to that of BBT2-3, with some difference in the middle crushing stage; the crushing load of the former fluctuated greatly and absorbed more energy.BBT3 started crushing from the upper end followed by plastic deformation.Meanwhile, specimens folded in an asymmetric mode during crushing.Finally, a mixed folding deformation mode was formed.
Fig. 6 (e) shows the load-displacement curves and deformation mode of BBT4.The deformation mode is shown by taking BBT4-4 as an example, which was similar to other specimens in the same group.The load-displacement curves of BBT4 were generally similar to that of BBT3.The main difference was that the load carrying capacity of the former was greatly improved and more energy was absorbed in the late crushing stage.BBT4 crushed from the upper end, then plastic deformation occurred, and specimens folded in an asymmetric mode during crushing.Finally, a mixed folding deformation mode was formed.It is worth noting that, compared with BBT2 or BBT3, BBT4 had the largest number of folded layers and the maximum energy absorption.

Evaluation indicators analysis
The average-load-displacement curves of each group of specimens are shown in Fig. 7 (a), showing that the average PCF and average MCF of each group of specimens were different.The MCF from BT to BBT3 increased significantly.The MCF of BBT1 was 64.69% higher than that of BT, the MCF of BBT2 was 42.58% higher than that of BBT1, the MCF of BBT3 was 22.13% higher than that of BBT2, and the MCF of BBT4 and BBT3 were almost the same.In the fragment compaction stage, the average load carrying capacity of BBT2, BBT3 and BBT4 had been greatly improved.The average-energy-displacement curves of each group of specimens are shown in Fig. 7 (b).It is obvious that before the crushing displacement reached approximately 81 mm, the order of energy absorption under any displacement was BBT3> BBT2> BBT4> BBT1> BT, and the curves of BBT2 and BBT4 were extremely close.When the crushing displacement reached 81mm, the energy absorption  Based on the obtained results, the relationships between MCF and SEA with ρF may be expressed by Eq. ( 6) and Eq. ( 7), respectively: where, MCF0 is the mean crushing force of BT specimens; SEA0 is the specific energy absorption of BT specimens.MCF and SEA both increased gradually with the increase of ρF, with the increase rate of MCF obviously faster than that of SEA.When the ρF reached 11.64%, i.e., 3 layers of BFRP, the increase rate of MCF and SEA remained unchanged, as shown in Fig. 9.  Table 4 lists the data obtained for BT and BBT specimens as part of the current research with those of composite wrapped hollow components (CWT) and carbon reinforced composite hollow components (CRCT) reported by Praveen [39] and Huang [40].It is obvious that for comparable dimensions, the mass of BT, BBT1, BBT2 and BBT3 are less than that of CWT and the mass of all BT and BBT specimens in the current research are significantly lower than those of CRCT.The SEA of CWT is lower than that of BBT, but the SEA of CRCT is higher than that of BBT and CWT.The CFE of both CWT and CRCT are higher than those of BT, but the CFE of BBT2, BBT3 and BBT4 are higher than those of CWT and CRCT.In summary, the specimens tested in the current study are the lightest.In terms of energy absorption, the SEA of BBT was higher than that of CWT, while the CFE of BBT specimens, especially BBT2, BBT3 and BBT4, are significantly higher than their counterparts.All these parameters clearly shows that BBT offers excellent crushing resistance.Note: t is the wall thickness of tube; m is the mass of the tube; a is from literature [37]; b is from literature [38].

Conclusion
A total of 20 tube specimens were produced, including BT and four types of BBT.All specimens were subjected to quasi-static axial compression by varying the cloth ratio of BFRP as an influencing factor.Moreover, the failure modes, load-displacement curves and crushing resistance of all tube specimens were analyzed.Following are the main conclusions that can be summarized based on analysis of obtained results: (1) The quasi-static axial compression failure modes of BT and BBT were divided into three types including opening, overall buckling and local buckling.
(2) The SEA and MCF of the tube specimens increased with the increase of BFRP cloth ratio during failure.When the BFRP cloth ratio reached 11.64%, i.e., 3 layers of BFRP were wrapped outside, the increase rate of MCF and SEA remained unchanged.Compared with BT, the SEA, MCF and CFE of BBT4 increased by 87.53%, 194.37% and 255.59% respectively.Compared with other composite hollow components such as CWT and CRCT, BBT showed excellent crushing resistance despite being significantly lighter than the other materials.
Di is the inner diameter of the tube; H is the tube height; tb is the thickness of one layer of bamboo sheet; tf is the thickness of one layer of BFRP; nb is the number of bamboo layers; nf is the number of BFRP layers; ρF is the BFRP cloth ratio.

Fig. 6
Fig.6 (c)  shows the load-displacement curves and deformation mode of BBT2.The deformation mode is shown by taking BBT2-1 and BBT2-3 as examples.BBT2-2 was similar to BBT2-3, and BBT2-4 data was not successfully collected.Two essentially different shapes of load-displacement curves are shown in Fig.6 (c).The load-displacement curve of BBT2-1 showed three-stage linearity, while the rest two curves were not completely similar to the load-displacement curves of BT or BBT1 neither.

8 .
(a) PCF and MCF (b) CFE and SEA Fig. Histogram of energy absorption evaluation indicators of all groups of specimens.

Table 2 .
Tensile strength and tensile modulus of elasticity of specimens for material characteristics

Table 3 .
Evaluation indicators of all groups of specimens Inc is the increase rate of energy absorption evaluation indicators compared with BT.
BBT3 and BBT2 signficantly increased at almost the same time.After the crushing displacement exceeded 81mm, the order of energy absorption under any displacement was BBT4>BBT3>BBT2>BBT1>BT, indicating that the BBT4 had the best energy absorption capacity in the late crushing stage.The data of the evaluation indicators are listed in Table2and are graphically shown in Fig.8.It can be seen from the results that MCF, SEA and CFE generally increased with the increase of the number of BFRP layers, while PCF did not follow this trend.The PCF of BBT1, BBT2, BBT3 and BBT4 was 18.40%, 30.64%, 10.73% and 17.21% lower than those of BT, respectively.Compared with BT, the CFE of BBT1, BBT2, BBT3 and BBT4 increased by 101.82%, 238.38%, 221.24% and 255.59% respectively.

Table 4 .
Comparison between data from this paper and existed studies