Compression performance of thin-walled square steel tube/bamboo plywood composite hollow columns with binding bars

Abstract

A new type of thin-walled steel tube/bamboo plywood hollow composite column with binding bars was developed in which transverse binding bars were used as reinforcement. The compression performance of 12 specimens was tested to examine the failure modes of the steel tube/bamboo plywood hollow composite column with binding bars, and the influences of the slenderness ratio, the net cross-sectional area of the bamboo plywood, the load eccentricity distance, and the binding bars on the ultimate load-bearing capacity were analyzed. The results indicated that the compression failure modes of the steel tube/bamboo plywood hollow composite column with binding bars were mainly bamboo plywood material failure and debonding failure of the adhesion interface; the ultimate bearing capacity of the steel tube/bamboo plywood hollow composite column with binding bars increased with an increasing net cross-sectional area of the bamboo plywood and decreased with an increasing slenderness ratio and eccentricity distance. Transverse binding bars were introduced to ensure the integrity of the composite column, to effectively reduce the debonding failure of the adhesion interface, to change the life-limiting damage mode, and to significantly improve the load-bearing capacity of the steel tube/bamboo plywood hollow composite column with binding bars. Finite element numerical simulations were performed to extend the influential factor analysis based on the test results. Furthermore, based on a non-linear regression analysis of the experimental data, a model for calculating the load-bearing capacity was formulated to determine the allowable compressive capacity of an steel tube/bamboo plywood hollow composite column with binding bar for engineering applications.

Keywords

bamboo plywood compression performance hollow composite column thin-walled steel tube transverse binding bar

Introduction

Bamboo-laminated composite engineering materials are among the most feasible applications of bamboo resources (Flander et al., 2009; Mahdavi et al., 2011). However, engineering applications of bamboo-laminated materials have focused on concrete-based construction; only recently has research been conducted on the use of bamboo-laminated materials as engineered structural materials (Li et al., 2014b, 2015; Xiao et al., 2010).Xiao (2009) suggested that bamboo has not been widely used in modern engineering structures primarily because of a lack of effective theoretical validation from solid mechanics, materials science, structural design, and experimental technology. In recent years, significant progress has been made in the study and application of bamboo composite materials and bamboo-laminated materials (Diwaker and Suresh 2013; He et al., 2015; Li et al., 2014b, 2015; Verma and Chariar, 2012, 2013; Xiao et al., 2010), including steel/bamboo plywood composite structural columns (Li et al., 2009, 2011, 2012, 2015c; Liu et al., 2013; Shen et al., 2011; Zhao et al., 2014, 2015). With the increasing shortage of engineered wood and the promotion of sustainable “green” construction materials, research into steel/bamboo composites has received significant attention worldwide. Liu et al. (2013) developed a steel/bamboo (curtain plywood) box composite column, and Li et al. (2014a) investigated a bamboo composite column reinforced with steel clamps. Li et al. (2009, 2011, 2012, 2015c) and Shen et al. (2011) developed a series of steel/bamboo plywood composite elements, and Xie (2012) explored a similar column application based on cold-formed thin-walled C-section steel and bamboo/wood composite plates. Zhao et al. (2014, 2015) developed a thin-walled steel tube/bamboo plywood hollow composite column. The thin-walled steel tube in the column served as the skeleton, making the composite suitable for industrial assembly. The thin-walled steel tube also increased the cross-sectional area of the composite column, thereby reducing the slenderness ratio and effectively decreasing the instability and failure performance of the column under pressure. Nevertheless, there have been few international reports of studies on steel tube/bamboo plywood composites used as structural bearing elements. The related studies on the mechanical behavior of bamboo composite columns have focused on their axial compressive performance; only a few studies have investigated their eccentric compressive performance. Recently, Li et al. (2015) studied the eccentric compression performance of parallel bamboo strand lumber columns, Huang et al. (2015) explored an ultimate-state-based model for the inelastic analysis of parallel bamboo strand columns of intermediate slenderness under eccentrically compressive loads, and Wei et al. (2016) studied the mechanical performance of bamboo glulam columns under eccentric loading.

Drawing on research conducted on concrete-filled steel tube columns with binding bars and the compressive properties of thin-walled steel tube/bamboo plywood hollow composite columns (SBCCs) (Cai and Sun, 2008; Zhao et al., 2014, 2015), this study investigated the use of transverse binding bars to reinforce SBCCs. Thin-walled steel tube/bamboo plywood composite columns with binding bars (SBCCBs) were developed to address the loss of effectiveness due to debonding failure and the low ultimate strength of regular SBCCs. Based on compression studies of 12 specimens, the compression failure modes of SBCCBs were examined, and the influences of the slenderness ratio, the net cross-sectional area of the bamboo plywood, the load eccentricity distance, and the binding bars on the ultimate load-bearing capacity were analyzed. Finite element numerical simulations were performed to extend the influential factor analysis based on the test results. Referring to the Chinese Code for the Design of Timber Structures GB 50005 (2003), a method of estimating the bearing capacity was proposed based on the experimental findings.

Experiments

Materials and design of the specimens

The bamboo plywood for the specimens was cut from the same batch of raw bamboo plywood (2440 mm × 1220 mm × 10 mm). The water content of the plywood was 9% (mass percentage). The transverse and longitudinal static bending strengths were 52 and 83 MPa, respectively, and the transverse and longitudinal elastic moduli were 7.4 × 10³ and 8.3 × 10³ MPa, respectively. The thin-walled steel tubes were Q235 galvanized seamless square steel tubes with an elastic modulus, yield strength, and ultimate strength of 2.05 × 10⁵, 260, and 340 MPa, respectively. Three sets of specimen dimensions, 40 mm×40 mm×1 mm, 60 mm×60 mm×1 mm, and 80 mm× 80 mm×1 mm, were selected based on the requirements for each type of specimen. The binding bars were Ø6 fully threaded screw rods with a yield strength of 260 MPa. Hilti HIT-RE 500-SD adhesive, which is suitable for binding metal, glass, and wood and shows excellent toughness and shock resistance, was used. The shrinkage ratio was below 1%, and in tests, the bonding strength between bamboo and bamboo and between bamboo and steel was found to be approximately 15.2–18.3 MPa.

To study the influences of the net cross-sectional area of the bamboo plywood, the slenderness ratio, and the load eccentricity distance on the eccentric compression performance of the SBCCBs, 12 specimens were designed: nine eccentric compression specimens (Figure 1(a)) and three axial compression specimens to serve as references (Figure 1(b)). The design parameters are given in Table 1, and a cross-sectional assembly diagram is shown in Figure 1(c). The cross section is simple, resulting in simple manufacturing and processing and making the composite suitable for industrial production. The locations of the installed binding bars were chosen based on previous experimental results regarding the compression debonding failure of SBCCs without binding bars (Zhao et al., 2014, 2015). Trapezoidal overhanging corbels of the same width as the eccentric compression specimens were glued onto one side of each specimen at both ends (Figure 1(a)); the thickness of the glued specimens was determined by the eccentricity distance, that is, the distance between the center of the load and the center of the SBCCB cross section (GB/T 50329, 2002). Four horizontal binding bars were installed in the upper, mid, and lower portions of each column, two cross bars were installed at the center, and one bar was installed at the end of the specimen in the same direction as the corbel. Because the elastic modulus of steel is much larger than that of bamboo, to prevent the steel tube from taking on load during the loading process and suffering buckling damage, the steel tube was shorter than the bamboo plywood by 15–20 mm from each end so that it would provide only inner support. The procedure for preparing the specimens was as follows (Figure 2): cutting the bamboo plywood in accordance with the size requirements → drilling the bamboo plywood and thin-walled steel tubes → treating the surfaces of the bamboo plywood and the thin-walled steel tubes → bonding → tightening and securing the binding bars → treating and curing after bonding.

Figure 1.

Square thin-walled steel tube/bamboo plywood composite hollow columns with binding bars: (a) eccentric compression specimens, (b) axial compression specimens, and (c) cross section of a specimen.

Table 1.

Design parameters of specimens.

Specimennumber	Cross section(mm × mm)	Slenderness ratio	Eccentricitydistance (mm)	Hollow section(mm × mm)	Sectional steelratio (%)	Binding bar spacing (mm)	Binding bar spacing ratio
ZZ1	80 × 80	25	0	40 × 40	1.62	200	2.5
ZZ2	100 × 100	30	0	60 × 60	1.83	400	4.0
ZZ3	120 × 120	35	0	80 × 80	1.95	600	5.0
ZP1	80 × 80	25	20	40 × 40	1.62	200	2.5
ZP2	100 × 100	25	30	60 × 60	1.83	300	3.0
ZP3	120 × 120	25	40	80 × 80	1.95	400	3.333
ZP4	80 × 80	30	30	40 × 40	1.62	250	3.125
ZP5	100 × 100	30	40	60 × 60	1.83	400	4.0
ZP6	120 × 120	30	20	80 × 80	1.95	500	4.17
ZP7	80 × 80	35	40	40 × 40	1.62	350	4.375
ZP8	100 × 100	35	20	60 × 60	1.83	450	4.5
ZP9	120 × 120	35	30	80 × 80	1.95	600	5.0

Figure 2.

Production process for specimens: (a) cutting bamboo plywood, (b) drilling holes, (c) bonding and assembly, and (d) tightening and curing.

Test methods

The equipment used for the loading tests is shown in Figure 3. The axial load was applied by a hydraulic jack through a load sensor at the lower end of the specimen, and the load was transferred to the specimen through one-way knife hinges on both ends. The lateral deformation was measured by six horizontal displacement gauges mounted on the adjacent sides of the specimen; the longitudinal deformation was measured by a vertical displacement gauge mounted on the steel knife hinge at the bottom end. A total of 18 strain gauges were arranged on two sides of the specimen (the side with the corbels and the opposite side) along the flexural direction. Before loading, preload adjustments and vertical corrections were performed on the specimens. Then, the load was applied in stages. During the initial phase of the loading process, the load was increased by 1/15–1/12 of the estimated load limit per stage. When the load reached 60% of the estimated load limit, the additional loading per stage was reduced to 1/25–1/20 of the estimated load limit. Each loading stage lasted 2–3 min. Once the loading had reached 85% of the estimated load limit, the loading was further increased slowly and continuously; the experiment was terminated when major cracking damage occurred or when the deformation began to increase rapidly. During the loading process, the pressure at the load sensor and the strain values for the specimen were recorded every 10 s, and the damage phenomena that occurred in the specimen during the process were observed simultaneously. The experimental tests were performed following the test methods specified in the GB/T 50329-2002 standard.

Figure 3.

(a) Measurement points and (b) photograph of test setup.

Results and discussion

Failure modes and mechanism analysis

To investigate the failure process, deformation development characteristics, failure mode, and characteristic load, the 12 SBCCB specimens were subjected to compressive tests. Cracking, debonding, and local buckling failure were identified based on microscopic and macroscopic behaviors such as the cracking conditions during failure, the characteristics and locations of debonding and local buckling failures, the strain relaxation at the interfaces between the substrates and the load–displacement relationship. At the beginning of loading, the specimen was in an elastic working condition and showed no cracking. When the load reached 40%–50% of the estimated load limit, minor cracks appeared at the column ends or at the various adhesion interfaces at the ends of the column between the binding bars, along with intermittent short and faint cracking sounds. When the load reached 60%–70% of the estimated load limit, the specimen emitted louder cracking sounds; the ends of the specimen, the glue surfaces in the body of the column between the binding bars, and the bamboo material cracked; and the bamboo plywood bulged outward. Although the column body close to the binding bars remained relatively intact, the lateral and axial compression deformation increased significantly, and the constraining tension of the binding bars was tighter. When the load reached the limit, cracks developed rapidly in the vicinity of the binding bars, and the damage sounds were very loud. As the column reached its load-bearing capacity, the main damage modes were material crimping as well as glue failure at the specimen ends and between the binding bars. None of the binding bars showed buckling or breakage, effectively ensuring the integrity of the composite column and inhibit-ing debonding damage at the adhesion interfaces. Compared with the eccentric load testing results for SBCCs without binding bars from Zhao et al. (2015), these specimens with binding bars possessed larger linear elastic segments, and the material damage at the ends and between the binding bars was more severe. Because there was no up-and-down through-crack damage, the load limit and axial compression deformation increased significantly. Comparatively speaking, the damage to the SBCCs without binding bars predominantly consisted of debonding failure, whereas the damage to the SBCCBs reinforced with transverse binding bars predominantly manifested as material failure.

There were four types of damage modes: column-end debonding failure with material crimping damage, debonding failure with material crimping damage between the binding bars on the compression side, debonding failure between the binding bars on the tension side, and debonding failure with material crimping damage between the binding bars on all four sides of the column. (1) Damage at the ends: This type of damage manifested as bamboo or corbel glue failure at the column ends as the plywood material suffered crimping when close to the load limit. This type of damage was most prominent in specimens ZP1, ZP2, ZP3, and ZZ2 (Figure 4(a)); the other specimens also exhibited this type of damage to a certain degree. The column-end debonding damage and plywood material damage were mainly caused by the following factors: First, the axial compression and lateral deformation of the specimen due to the lateral tensile stress exceeded the adhesive strength of the glue between the different substrates of bamboo plywood. Second, the pressure on the load-bearing end surface of the column was not uniform; the local unevenness of the surface resulted in a shear force between the axial compression strips. Under the combined action of the lateral tensile and vertical shear forces, cracking damage could easily occur between the substrates and in the materials. Third, the axial compressive stiffness differed among the different pieces of material. Although all of the bamboo strips in the same composite column were cut from the same sheet in the same direction, bamboo is a remarkably anisotropic material, and different material properties can arise during the manufacturing, processing, and assembly of plywood. (2) Damage on the compression side of the column between the binding bars: Cracking damage first appeared on the compression side of the column between the binding bars. When the load reached a certain level, the plywood suddenly snapped and curved outward, and the lateral and axial deformations increased significantly. As the load approached the limit, glue failure occurred at the surface between the inner plywood and the thin steel tube, along with various degrees of body bulging due to glue failure on two adjacent sides. This damage mode was observed in specimens ZP4, ZP5, ZP7, and ZP9 (Figure 4(b)); the lateral tensile stress under the eccentric load exceeded the adhesive strength of the glue, and the axial stress exceeded the limiting compressive strength of the plywood. The columns typically exhibited large horizontal flexural deformations during the damage process; a larger slenderness ratio and a larger degree of eccentricity resulted in a larger degree of flexure. However, various situations occurred because of differences in the conditions at the termination of the experiment caused by glue failure and fracture damage. (3) Damage on the tension side of the column between the binding bars: Cracking damage appeared between the substrates on the tension side of the specimen. The crack gradually approached the horizontal binding bar until the experiment was terminated, but no damage appeared on the compression side of the column; material crimping damage occurred at the column ends. Specimen ZP6 was the only specimen that showed this type of damage (Figure 4(c)); the glue surface on the tension side of this specimen may have been weak, causing the lateral tensile stress to exceed the glue strength. (4) Damage on all four sides of the column between the binding bars: Cracking damage appeared on all four sides of the column between the binding bars. When the load reached a certain level, the plywood snapped and curved outward because of the small load eccentricity distance. This type of damage mode was observed in the axial compression specimens ZZ1 and ZZ3 and in the small eccentric compression specimen ZP8 (Figure 4(d)).

Figure 4.

Failure morphologies of specimens: (a) damage at the column ends, (b) damage on the compression side of the column, (c) damage on the tension side of the column, and (d) damage on all four sides of the column.

Influential factors

This experiment mainly considered the influence of the net cross-sectional area of the bamboo plywood, the slenderness ratio, and the load eccentricity distance on the ultimate compressive bearing capacity of SBCCBs. For the same net cross-sectional bamboo plywood area, the use of a hollow thin-walled steel tube can significantly decrease the slenderness ratio and load eccentricity of the specimen and consequently increase the ultimate bearing capacity; therefore, the effect of the hollowness can be indirectly reflected by the slenderness ratio and load eccentricity. The previously described damage phenomena observed in the experiment revealed four different types of damage among the 12 specimens, corresponding to different termination conditions for the experiments. It was therefore difficult to distinguish the impacts of the various factors by simply looking at the ultimate load; thus, the influential characteristics were analyzed based on a combination of the ultimate load and the curves of the load versus the column-end axial deformation (F–D curves).

Slenderness ratio and eccentricity

Specimens with the same net cross-sectional area of laminated bamboo were compared to investigate the influence of the other factors. The F–D curves of specimens with different slenderness ratios and eccentricity distances are shown in Figure 5 (the end of each curve represents the load and the corresponding axial deformation prior to termination of the experiment). The deformation curves show approximately linear behavior before the onset of cracking damage. In general, a larger slenderness ratio and a larger eccentricity distance correspond to a lower ultimate load-bearing capacity. This trend was particularly prominent in the first group (Figure 5(a): ZP1, ZP4, ZP7, and ZZ1). The second group (Figure 5(b): ZP2, ZP5, and ZZ2) and third group (Figure 5(c): ZP6, ZP9, and ZZ3) also showed this type of variation. Specimen ZP8 in the second group had a large slenderness ratio, but its eccentricity was small; therefore, the loads at the various stages after cracking were smaller than those for ZP2 but larger than those for ZP5. Specimen ZP3 in the third group had a large eccentricity but a small slenderness ratio; therefore, the loads at the various stages after cracking were generally smaller than those for ZP6 but larger than those for ZP9. Comparing the second group (Figure 5(b)) with the third group (Figure 5(c)) reveals an obvious reciprocal influence between the slenderness ratio and eccentricity. In each group, the load-bearing capacity of the axial compression specimen was the highest at the various stages of loading. The eccentric compressive failure was mainly controlled by the degree of lateral flexure. When the slenderness ratio was greater, the lateral deflection and additional moment were also greater, and the decrease in the bearing capacity was more significant. The ultimate bearing capacity of the composite column was not only related to the slenderness ratio and the eccentricity but was also affected by the bonding performance. According to Euler’s formula, an axial compression specimen with a larger slenderness ratio will have a smaller critical buckling load. The critical buckling load is much smaller with an initial load eccentricity than without one. According to the principle of the interaction between the bending moment and the axial compression load, a larger bending moment will result in a smaller axial compression load. A larger eccentricity causes the bending moment to increase, resulting in a smaller axial compression load. Overall, the test results were consistent with these principles. Furthermore, the influence of the hollow ratio is indirectly reflected in the slenderness ratio and eccentricity; increasing the hollow ratio decreases the slenderness ratio and eccentricity and increases the critical compressive load-bearing capacity of the specimen. For example, a specimen with a cross-sectional area of 100 × 100 and a hollow area of 60 × 60 yields a radius of gyration of the net cross-sectional area of i = 33.7; for a solid column with the same net cross-sectional area of 80 × 80, the radius of gyration would be i = 23.1. Therefore, for the same net cross-sectional area, the slenderness ratio of a hollow column is larger than that of a solid column.

Figure 5.

Comparison of F–D curves for SBCCBs with different slenderness ratios and eccentricity distances: (a) specimens ZP1, ZP4, ZP7, and ZZ1; (b) specimens ZP2, ZP5, ZP8, and ZZ2; and (c) specimens ZP3, ZP6, ZP9, and ZZ3.

Net cross-sectional area and eccentricity

For the comparison of specimens with the same slenderness ratio, the F–D curves of specimens with different net cross-sectional areas and eccentricities are shown in Figure 6. In general, when the net cross-sectional area of the bamboo plywood was larger, the load-bearing capacity of the specimen was higher; all three groups of specimens showed this variational trend. However, increasing the eccentricity also decreases the load-bearing capacity of a specimen. Therefore, the anti-lateral displacement stiffness of ZP3 in the first group (Figure 6(a)) and that of ZP9 in the third group (Figure 6(c)) were not the highest within their respective groups under the same load because of their higher eccentricity distances, even though they had the highest ultimate bearing capacities because they had the largest net cross-sectional areas of bamboo plywood. The net cross-sectional area ratio of the specimens in the first group was S_ZP1:S_ZP2:S_ZP3 = 1.0:1.33:1.67, and the ratios for the other two groups were the same; meanwhile, the ratios of the ultimate bearing capacities were N_ZP1:N_ZP2:N_ZP3 = 1:1.17:1.28 (Figure 6(a)), N_ZP3:N_ZP4: N_ZP5 = 1:1.16:1.47 (Figure 6(b)), and N_ZP6:N_ZP7: N_ZP8 = 1:1.32:1.39 (Figure 6(c)). Therefore, the ultimate bearing capacity is positively proportional to the net cross-sectional area, but the trend of variation is non-linear and is reciprocally related to the slenderness ratio and eccentricity distance.

Figure 6.

Comparison of F–D curves for SBCCBs with different net cross-sectional areas of bamboo plywood: (a) specimens ZP1, ZP2, and ZP3; (b) specimens ZP4, ZP5, and ZP6; and (c) specimens ZP7, ZP8, and ZP9.

Bamboo plywood has excellent compressive and flexural properties. Therefore, the load-bearing capacities of the SBCCB specimens increased as the net cross-sectional area of bamboo plywood increased when all other conditions remained the same. The specimens with relatively larger net cross-sectional bamboo plywood areas also had higher sectional steel ratios, as shown in Table 1. Even small differences in the sectional steel ratio may influence the bearing capacity and deformation of specimens. Therefore, the influence of the sectional steel ratio will require further study in the future.

Binding bar spacing

The influence of the binding bar spacing on the SBCCBs was directly related to their local buckling deformation. When the load reached the load limit, different degrees of local buckling occurred in the outer bamboo plywood around the column between the binding bars, as shown in Figure 7. The local buckling deformation exhibited a certain variational trend based on the ratio of the binding bar spacing (S) to the cross-sectional dimension (B) of the specimen. The local buckling relative to the initial positions of the binding bars decreased as this ratio (S/B) decreased; Figure 8 compares the local buckling results for the nine eccentric compression specimens. The local buckling deformations between the binding bars in specimens ZP1, ZP2, ZP3, and ZZ2, which had relatively small S/B values, were smaller than those of the other specimens; this was one of reasons that severe debonding damage occurred at the column ends. These results from the compression tests of the SBCCBs demonstrate that the transverse binding bars effectively inhibited debonding failure and altered the ultimate failure mode of the specimens, significantly improving the ultimate bearing capacity. The combination of the binding bars and the thin-walled steel tube generated a hooping effect similar to that of a steel stirrup in a reinforced concrete structure. A reasonable combination of binding bars and a thin-walled steel tube reduced the dependence on the strength of the bonded interfaces, thereby retarding debonding failure.

Figure 7.

Local buckling under compression.

Figure 8.

Influence of binding bar spacing on the local buckling.

Strain and stress analysis

The factors that influence the cracking damage to and deformation of a specimen are numerous. Furthermore, the damage modes of different specimens are quite different. Even when the basic damage modes are similar, the strains measured at similar locations on different specimens are still quite different; however, there is a similar general trend of variation. Using specimens ZP2 and ZP4 as examples to illustrate the strain development characteristics, Figure 9 shows the load–strain curves measured at various locations on the specimens (Ai represents the three strain gauges on the corbel side at the center of the specimen, and Bi represents the three central strain gauges on the side without the corbels). The results show that the strains on the corbel side and the back of the specimen both increased with increasing load; the strain on the corbel side was negative and compressive, whereas that on the back side was positive and tensile, which is consistent with the structural eccentric load force state. The strain on the compression side was larger than that on the tension side; the termination of the experiment was caused by glue failure of the bamboo plywood substrates and material damage on the corbel side.

Figure 9.

Strain–force curves at the center of the column: (a) ZP2 and (b) ZP4.

The eccentric compression of nine SBCCs without horizontal binding bars was tested by Zhao et al. (2015). Compared with the parameters of the specimens tested in this study, the slenderness ratios and eccentricity distances in that study were smaller. To determine the influence of the binding bars on the load-bearing capacity of the SBCCBs, the average limiting stresses from the two sets of experiments were compared. Table 2 shows the comparison of the limiting stresses of the SBCC specimens (without binding bars) and the SBCCB specimens (with binding bars). Table 2 shows that the average limiting stress of the specimens without binding bars was 15.76 MPa, whereas that of the specimens with binding bars was 19.18 MPa, higher by 21.7% compared with that of the specimens without binding bars. The above comparison does not consider the influence of the slenderness ratio or the load eccentricity distance; under identical conditions in terms of these parameters, the magnitude of the improvement could be even greater. From the above data and the observed experimental phenomena, it can be concluded that installing horizontal binding bars can effectively limit the debonding damage suffered by SBCCBs, change the life-limiting damage mode, and increase the ultimate bearing capacity.

Table 2.

Comparison of the limiting stresses.

Specimen number		Slenderness ratio	Eccentricity distance (mm)	Measured ultimateload (kN)	Net cross-sectional area (mm²)	Compressive stress (MPa)	Mean value (MPa)Standard deviation (MPa)
Specimens without binding bars(Zhao, 2015)	Z1	17.3	0	58.0	3200	18.13	15.76 1.57
	Z2	17.3	20	54.5	3200	17.03
	Z3	17.3	30	43.5	3200	13.59
	Z4	23.1	20	48.5	3200	15.16
	Z5	28.7	20	46.5	3200	14.53
	Z6	17.3	20	84.0	4800	17.50
	Z7	17.3	20	102.5	6000	17.08
	Z8	17.3	20	90.0	6400	14.06
	Z9	17.3	20	142.5	9600	14.84
Specimens with binding bars(this study)	ZZ1	25	0	109.3	4800	22.77	19.18 2.07
	ZZ2	30	0	132.1	6400	20.64
	ZZ3	35	0	159.0	8000	19.88
	ZP1	25	20	105.5	4800	21.98
	ZP2	25	30	123.3	6400	19.26
	ZP3	25	40	135.2	8000	16.90
	ZP4	30	30	99.9	4800	20.83
	ZP5	30	40	115.4	6400	18.02
	ZP6	30	20	147.3	8000	18.41
	ZP7	35	40	87.7	4800	18.26
	ZP8	35	20	115.4	6400	18.02
	ZP9	35	30	121.2	8000	15.15

Numerical analysis

Relatively few SBCCB specimens were tested under compression in these experiments. To further study the compressive load-bearing capacity of SBCCBs, the finite element software package ABAQUS was used to simulate the test results and to extend the findings regarding the analyzed parameters.

Material parameters

A specimen is composed of three types of materials, and the stress–strain models for these materials were determined first. According to the results of our material property tests, the stress–strain curve for the bamboo plywood is shown in Figure 10. The longitudinal and transverse tensile and compressive curves have the same form; the only difference is the locations of the intersections denoted by point a and point b, as shown in Table 3. The longitudinal load-bearing capacity of bamboo plywood is both compressive and tensile (side compression), involving the along-the-grain compressive and tensile elastic moduli and strengths. The transverse direction is mainly tensile, involving only the cross-grain tensile elastic modulus and strength. Poisson’s ratio in all three directions has a value of 0.3.

Figure 10.

Stress–strain curve for bamboo plywood.

Table 3.

Characteristic parameters of the stress–strain relationship for bamboo plywood.

Characteristic parameter	Elastic modulus (MPa)	Elasticity limit f_p (MPa)	Yield strengthf_y (MPa)	Yield strainε_y (%)	Limit strainε_u (%)
Along-the-grain compression	8807	0.9f_y	24	0.35	0.6
Along-the-grain tension	10,866	1.0f_y	47	1.03	1.03
Cross-grain tension	2600	1.0f_y	10	0.27	0.27

The thin-walled steel tube is typically under stress from three directions. However, because the steel walls are quite thin, the radial stress is much smaller than the stress in the other two directions; thus, the stress analysis can be performed in two directions (Figure 11(b)). The stress–strain curve of the steel material is shown in Figure 12. Thin-walled steel tubes of the Q235 type were used in this study; such a tube has an elastic modulus of 205 GPa, an elastic strength of 235 MPa, a yield strength of 260 MPa, an ultimate strength of 340 MPa, a yield strain of 0.2%, a plastic strain of 0.5%, a maximum ultimate strain of 1%, and a Poisson’s ratio of 0.24. The stress–strain model for a thin-walled steel tube is the same as that for a steel rod.

Figure 11.

Stress states of a thin-walled steel tube: (a) three-dimensional stress state and (b) two-dimensional stress state.

Figure 12.

Stress–strain curve for a thin-walled steel tube.

The structural model of the adhesive layers is shown in Figure 13. The actual crack tip coincides with the crack tip damage location. Under the assumption that the maximum damage stress occurs at the crack tip, the constitutive relation for the model of the cohesive zone is the corresponding relation between the damage stress and strain at the crack tip. This relation is a mixed bilinear constitutive relation for mixed-mode fracture (Figure 14). Based on our test results, the limiting pure shear strength of the cohesive zone at a steel–bamboo interface is σ₁ = σ₂ = τ₁₃ = τ₂₃= 6.37MPa, whereas that of the cohesive zone at a bamboo–bamboo interface is σ₁ = σ₂ = τ₁₃= τ₂₃= 5.36 MPa. The limiting pure tensile strength of the cohesive zone at a steel–bamboo interface is σ_c = σ₃ = 1.51 MPa, whereas that of the cohesive zone at a bamboo–bamboo interface is σ_c = σ₃ = 1.15 MPa.

Figure 13.

Model of a cohesive zone.

Figure 14.

Bilinear stress–strain curve for mixed-mode fracture.

Numerical model and analysis results

The slenderness ratio, cross-sectional area, and eccentricity of the specimens were expanded into three factors and three levels, resulting in the construction of a total of 27 three-dimensional finite element models of eccentric compression specimens (including models corresponding to the nine test specimens) for a full influential factor analysis. The hinge bearing was simulated using a “coupled” contact relation based on boundary conditions and constraints. The off-center loading was achieved using the load shifting method. By means of the coupling contact method, the top end of the specimen was kept consistent with the loading-point deformation. A three-dimensional, eight-node reduced integration element scheme (C3D8R) was used in the model. The loading was controlled through displacement. The limiting state was considered to occur when the magnitude of the load increment was ≤0.5 kN, and the parameters for analysis were extracted in this state.

The response contour plots in the overall limiting state for the various components of the numerical model for test specimen ZP8 are shown in Figure 15. An examination of the loading conditions of the bamboo plywood and the thin-walled steel tube reveals that the stress in the steel tube is 10–15 times the stress in the bamboo plywood. The reason for this is that the binding bars and adhesives have the effect of requiring the thin-walled steel tube and the bamboo plywood to exhibit consistent deformation. However, the elastic modulus of the steel tube is higher than that of the bamboo plywood; thus, under the same deformation conditions, the stress in the thin-walled steel tube is greater than that in the bamboo plywood. As the specimen approaches the limiting state, the ratio of the stress in the thin-walled steel tube to that in the bamboo plywood decreases, and the strain in the bamboo plywood is significantly greater than that of the thin-walled steel tube, especially once the compressed bamboo plywood reaches or exceeds its limit strain, indicating that the bamboo plywood is already damaged. The maximum stress and strain in the bamboo plywood occur at the ends of the structure, which is consistent with the crushing damage observed at the ends of the specimens in the experiment. The maximum stress in the steel tube occurs in the middle section, which is consistent with the occurrence of bulging failure in the middle sections of some specimens during the experiments. The maximum stress on the binding bars is 168.6 MPa, indicating that the tensile capacity of the binding bars is not fully utilized. The binding bars in the specimens never showed fracture deformation during the experiment; thus, the results of the numerical simulation are generally consistent with the experimental results.

Figure 15.

Limiting-state maps for specimen ZP8: (a) stress in the bamboo plywood, (b) strain in the bamboo plywood, (c) stress in the thin-walled steel tube, (b) strain in the thin-walled steel tube, and (e) stress in the binding bars.

The load–displacement curves from the simulations are plotted and compared with the experimentally measured curves in Figure 16. The numerical results for the nine eccentric compression specimens are reasonably consistent with the measured curves. The error on the ultimate load-bearing capacity is within 16% (a comparison of the simulated and measured values of the ultimate bearing capacity is presented in Table 4). Thus, the finite element models for the tested specimens are reliable.

Figure 16.

Comparison of numerically simulated results and experimentally measured results for eccentric compression specimens: (a) ZP1, (b) ZP2, (c) ZP3, (d) ZP4, (e) ZP5, (f) ZP6, (g) ZP7, (h) ZP8, and (i) ZP9.

Table 4.

Comparison of the measured values, numerically simulated values, and calculated values of the ultimate load-bearing capacity.

Specimen number	Measured values N_u₁ (kN)	Numerically simulated values N_u₂ (kN)	Calculated values N_u₃ (kN)	Bamboo plywood contributionN_u₄ (kN)	(N_u₁ − N_u₂)/N_u₁	(N_u₁ − N_u₃)/N_u₁	N_u ₄/N_u₂
ZZ1	109.3	−	101.6	82.2	−	0.070	0.809
ZZ2	132.1	−	130.7	102.9	−	0.011	0.787
ZZ3	159.0	−	155.1	120.4	−	0.025	0.776
ZP1	105.5	101.51	91.5	73.8	0.038	0.133	0.806
ZP2	123.3	125.53	122.4	96.4	−0.019	0.007	0.786
ZP3	135.2	144.80	153.1	118.8	−0.071	−0.132	0.776
ZP4	99.9	102.38	81.7	66.0	−0.025	0.182	0.808
ZP5	115.4	105.45	110.5	87.0	0.086	0.042	0.787
ZP6	147.3	149.70	154.6	120.0	0.016	−0.050	0.776
ZP7	87.7	81.80	72.6	58.6	0.067	0.172	0.807
ZP8	115.4	126.34	112.4	88.6	−0.096	0.026	0.788
ZP9	121.2	140.52	139.6	108.4	−0.159	−0.152	0.776

This article mainly examines the trend of variation of the load-bearing capacity of a composite column. To verify the experimentally observed variations due to the main influential factors, the ultimate bearing capacities were numerically analyzed for 27 modeled specimens. As shown in Figure 17, when other conditions were kept constant, the bearing capacity of the specimens significantly increased with an increasing cross-sectional area. As the eccentricity increased, the bearing capacity significantly decreased; as the slenderness ratio increased, the bearing capacity showed a decreasing trend. The variational trends observed for the influential factors discussed above are consistent with the experimental results. However, the variation in the bearing capacity with respect to the slenderness ratio was quite small; we carefully analyzed the reason for this phenomenon. In addition to the possibility of error in the numerical analysis, the main reason for this finding is that this set of designed specimens had relatively small slenderness ratios. According to the definition provided in the Chinese standard specifications for timber structures, a column with a slenderness ratio of less than 50 is classified as a short column, and the influence of its slenderness is relatively small.

Figure 17.

The influences of individual factors on the numerical results: (a) cross-sectional area, (b) eccentricity ratio, and (c) slenderness ratio.

Model for calculating the load-bearing capacity

The analysis presented above shows that the load-bearing capacity of an SBCCB increases as the net cross-sectional area of the bamboo plywood increases but decreases as the slenderness ratio and eccentricity increase. The hollow nature of the specimens reduces their slenderness ratios and improves their critical compressive load-bearing capacities. Therefore, the significant improvement in the critical compressive load-bearing capacity that is achieved by virtue of the hollowness of a column is reflected in its slenderness ratio. Thus, the effects of the area of the thin-walled steel tube, which can be equivalently expressed in terms of the net cross-sectional bamboo plywood area, were considered in this study. The transverse binding bars in SBCCBs inhibit debonding failure, change the life-limiting damage mode, and significantly improve the load-bearing capacity; however, these effects are reflected in the regression parameters, and therefore, the binding bars were not considered as a separate factor. To develop guidelines for engineering applications, a model was formulated for computing the allowable load-bearing capacity for an engineering design based on the experimental data collected with regard to the ultimate load-bearing capacity. To allow the theoretical model to be used to compute both the axial and eccentric compressive load-bearing capacities, it was assumed that the effects of the slenderness ratio and eccentricity on the ultimate load-bearing capacity were independent of one another. That is, the interaction between the slenderness ratio and the eccentricity was not considered. Thus, the total reduction coefficient was taken to be the load-bearing capacity $ϕ$ , which is the product of two influence coefficients—one associated with the eccentricity, $ϕ_{e}$ , and one associated with the slenderness ratio (stability factor), $ϕ_{λ}$ —as shown in equation (1)

N_{u} = ϕ f_{b} A_{b} = ϕ_{λ} ϕ_{e} f_{b} (A_{b}^{n} + \frac{f_{s}}{f_{b}} A_{s})

(1)

ϕ_{λ} = \frac{1}{1 + {(\frac{λ}{C_{1}})}^{2}}

(2)

ϕ_{e} = C_{2} \exp (C_{3} \cdot \frac{e_{d}}{B})

(3)

λ = \frac{l}{i}

(4)

In the above equations, $N_{u}$ is the allowable load-bearing capacity of the SBCCB; $f_{b}$ is the allowable value (16.2 MPa) of the compressive strength along the primary fiber direction of the bamboo mat plywood, as experimentally determined by Xiao et al. (2012); $A_{b}$ is the cross-sectional area of the compressed bamboo plywood; $A_{b}^{n}$ is the net cross-sectional area of the bamboo plywood; $f_{s}$ is the strength of the steel (235 MPa); $A_{s}$ is the cross-sectional area of the thin-walled steel tube; i is the radius of gyration of the net cross-sectional area of the composite column; $l$ is the length of the specimen; $e_{d}$ is the eccentricity distance; B is the length of one side of the cross section; and C₁, C₂, and C₃ are the fitting parameters.

Based on the general trends observed for the variables, the boundary conditions, and the Chinese Code for the Design of Timber Structures GB 50005 (2003), $ϕ_{e}$ and $ϕ_{λ}$ were formulated as shown in equations (3) and (4), and a non-linear regression analysis was performed to obtain the following fitting parameters: C₁ = 61.76, C₂ = 1.23, and C₃ = –0.42. Table 4 compares the results estimated from the fitted equation with the test results. There is good overall agreement between the measured and calculated values, with relative errors of less than 18.2%. In addition, the bamboo plywood is found to contribute 77.6%–80.9% of the compressive bearing capacity, with an average of 79.1%; consequently, the constraining effect of the binding bars contributes 21.9% of the compressive bearing capacity.

Conclusion

This study experimental investigated the compressive performance of SBCCB specimens. The following conclusions were drawn:

Four eccentric compression damage modes were observed for the SBCCB specimens: debonding failure with crimping damage to the bamboo plywood material at the ends of the column, debonding failure with material crimping damage between the binding bars on the compression side, debonding failure with material crimping damage between the binding bars on the tension side, and debonding failure with material crimping damage between the binding bars on all four sides of the column. Material damage was the main failure mode.

The load-bearing capacity of SBCCBs generally increases with as the net cross-sectional area of the bamboo plywood increases and decreases with increasing slenderness ratio and eccentricity distance; there are reciprocal influences among these three factors. The hollow thin-walled steel tube in an SBCCB reduces its slenderness ratio and eccentricity ratio, consequently improving the critical compressive load-bearing capacity.

Transverse binding bars are used to ensure the composite effect of the column, to effectively reduce debonding failure, to change the life-limiting damage mode, and to significantly improve the load-bearing capacity of SBCCBs. The average compressive stress of test specimens with binding bars was found to be 21.7% greater than that of specimens without binding bars.

The results of numerical simulations of the individual influences of the slenderness ratio, load eccentricity, and cross-sectional size of an SBCCB specimen on its compressive performance were found to be essentially consistent with the results of the experimental tests. These findings served as a theoretical basis for constructing models for load-bearing capacity calculations.

The research results indicate that bamboo plywood and a thin-walled steel tube can be formed into a well-integrated composite column for providing structural support. The tested SBCCBs exhibited excellent compressive performance, indicating that bamboo plywood can be used as a modern structural material. A model for calculating the allowable bearing capacity to serve as a guideline for engineering applications was formulated based on the experimental data.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors gratefully acknowledge the financial support provided by the Foundation of the State Key Laboratory of Subtropical Building Science, South China University of Technology (grant no. 2015ZB26), and the Natural Science Foundation of Hunan Province (grant no. 2017JJ3302).

References

Cai

Sun

(2008) Experimental investigation onL-shaped concrete-filled steel tube stub columns with binding bars under axial load. China Civil Engineering Journal 41(9): 14–20.

Diwaker

Suresh

(2013) Composite bamboo column elements for structural applications. In: Proceedings of UKIERI congress on innovations in concrete construction, Jalandhar, India, 5–8 March, No. UCC-495.

Flander

Rovers

(2009) One laminated bamboo-frame house per hectare per year. Construction and Building Materials 23(1): 210–218.

GB/T 50329 (2002) Standard for Methods Testing of Timber Structures. Beijing, China: Architecture and Building Press.

GB 50005 (2003) Code for Design of Timber Structures. Beijing, China: Architecture and Building Press.

Sun

. (2015) Experimental investigations on mechanical properties and column buckling behavior of structural bamboo. Structural Design of Tall and Special Buildings 24(7): 491–503.

Huang

Bian

Huang

. (2015) An ultimate-state-based-model for inelastic analysis of intermediate slenderness PSB columns under eccentrically compressive load. Construction and Building Materials 94: 306–314.

Deeks

. (2015a) Eccentric compression performance of parallel bamboo strand lumber columns. Bioresources 10(4): 7065–7080.

Zhang

. (2015b) Mechanical performance of laminated bamboo column under axial compression. Composites Part B: Engineering 79: 374–382.

10.

Chen

Meng

(2014a) Compressive strength test on bamboo column reinforced by clamps. China Forestry Science and Technology 28(4): 113–115.

11.

Shan

Shen

(2015c) Bending resistance ofI-section bamboo–steel composite beams utilizing adhesive bonding. Thin-Walled Structures 89: 17–24.

12.

Shen

Shan

(2011) Experimental study on shear behavior of I-section bamboo-steel composite beams. Journal of Building Structures 32(7): 80–86.

13.

Shen

Shan

(2012) Flexural behavior of lightweight bamboo–steel composite slabs. Thin-Walled Structures 53: 83–90.

14.

Zhang

Shen

(2009) Experimental study on flexural behavior of profiled steel sheet-bamboo composite slabs. Journal of Building Structures 30(2): 176–181.

15.

Xiao

Wang

(2014b) Experimental studies of light-weight wood frame shear walls with ply-bamboo sheathing panels. Journal of Building Structures 34(9): 142–149.

16.

Liu

Zhou

(2013) Study on compression resistance of bamboo building structural column. Journal of Central South University of Forestry & Technology 33(1): 104–108.

17.

Mahdavi

Clouston

Arwade

(2011) Development of laminated bamboo lumber: review of processing, performance, and economical considerations. Journal of Materials in Civil Engineering 23(7): 1036–1042.

18.

Shen

Zhang

(2011) Research advances of bamboo-steel composite structural members. Journal of Computational and Theoretical Nanoscience 4(8): 2963–2967.

19.

Verma

Chariar

(2012) Development of layered laminate bamboo composite and their mechanical properties. Composites Part B: Engineering 43(3): 1063–1069.

20.

Verma

Chariar

(2013) Stiffness and strength analysis of four layered laminate bamboo composite at macroscopic scale. Composites Part B: Engineering 45(1): 369–376.

21.

Wei

Zhou

Yuan

(2016) Mechanical performance of glulam bamboo columns under eccentric loading. Acta Materiae Compositae Sinica 33(2): 379–385.

22.

Xiao

Shan

Chen

(2009) Development of laminated bamboo modern structures. In: Proceedings of the 11th international conference on non-conventional materials and technologies, Bath, 6–9 September.

23.

Xiao

Yang

Shan

(2012) Experimental research on mechanical properties of Glubam. Journal of Building Structures 33(11): 150–157.

24.

Xiao

Zhou

Shan

(2010) Design and construction of modern bamboo bridges. Journal of Bridge Engineering 15(5): 533–541.

25.

Xie

Zhang

Jiang

. (2012) Experimental study on performance of steel-bamboo composite column under axial compression. Engineering Mechanics 29: 221–225.

26.

Zhao

Cao

Zhou

. (2015) Eccentric compression experimental study on square thin-walled steel tube/bamboo-plywood composite hollow column. Journal of Building Materials 18(4): 601–607.

27.

Zhao

Zhang

Zhou

. (2014) Axial compression behavior of square thin-walled steel tube-laminated bamboo composite hollow columns. Transactions of the Chinese Society of Agricultural Engineering 30(6): 37–45.