Abstract
An ultrasonic cavitation treatment was used to increase the permeability of phenol formaldehyde resin adhesive on the surface of bamboo strips, and two-ply laminated bamboo lumber was manufactured. The strain distribution along the bamboo interface under stretching conditions was investigated using the digital image correlation technique. The effect of the ultrasonic treatment on the bamboo strips in terms of the shear strain of the laminated bamboo lumber was investigated, and a finite-element analysis for the bonding interface was carried out to evaluate whether the deformation measurement could predict the mechanical behavior differences of the laminated bamboo lumber. The digital image correlation and shear strength results show that the strain was lower in the bamboo bonding interface after ultrasonic treatment and the shear strength was enhanced due to the increased adhesive penetration. The digital image correlation measurement and finite-element analysis simulation both showed that stress was more concentrated, and the strain value and strain zone width was higher in the carbonized bamboo bonding interface than in the bleached bamboo bonding interface. The finite-element analysis results appeared to be in agreement with the digital image correlation test results.
Download PDF
Full Article
Direct Strain Distribution and Finite-Element-Analysis Simulation of the Bonding Interface of Bamboo Laminated Lumber with Ultrasound-Treated Bamboo Strips
Mingjie Guan a,b,c,* and Zhiwei Huang a
An ultrasonic cavitation treatment was used to increase the permeability of phenol formaldehyde resin adhesive on the surface of bamboo strips, and two-ply laminated bamboo lumber was manufactured. The strain distribution along the bamboo interface under stretching conditions was investigated using the digital image correlation technique. The effect of the ultrasonic treatment on the bamboo strips in terms of the shear strain of the laminated bamboo lumber was investigated, and a finite-element analysis for the bonding interface was carried out to evaluate whether the deformation measurement could predict the mechanical behavior differences of the laminated bamboo lumber. The digital image correlation and shear strength results show that the strain was lower in the bamboo bonding interface after ultrasonic treatment and the shear strength was enhanced due to the increased adhesive penetration. The digital image correlation measurement and finite-element analysis simulation both showed that stress was more concentrated, and the strain value and strain zone width was higher in the carbonized bamboo bonding interface than in the bleached bamboo bonding interface. The finite-element analysis results appeared to be in agreement with the digital image correlation test results.
DOI: 10.15376/biores.17.1.890-907
Keywords: Ultrasonic treatment; Bamboo bonding interface; Digital image correlation; Finite-element analysis
Contact information: a: College of Materials Science and Engineering, Nanjing Forestry University, Nanjing, Jiangsu 210037 China; b: Bamboo Engineering and Technology Research Center of SFA, Nanjing Forestry University, Nanjing, Jiangsu 210037 China; c: Co-Innovation Center of Efficient Processing and Utilization of Forest Resources, Nanjing Forestry University Nanjing, Jiangsu 210037 China; *Corresponding author: mingjieguan@126.com
GRAPHICAL ABSTRACT
INTRODUCTION
Bamboo is a renewable natural green construction and building material that has the advantages of being environmentally friendly, sustainable, and has the capability to be recycled or reused. Bamboo-based composites, e.g., laminated bamboo lumber, bamboo scrimber, plybamboo, bamboo glulam, etc., can effectively alleviate the contradiction between the supply and demand of renewable materials in civil construction and the building industry (Li et al. 2013, 2015; Li et al., 2016; Xiao et al. 2017; Takeuchi et al. 2018; Chow et al. 2019; Li et al. 2019). Recently, laminated bamboo lumber has attracted increasing attention because of its excellent strength performance as a building material. Previous research has been primarily focused on the effect of the orientation direction on the mechanical properties and the effect of processing on the static bending properties and structural properties, physical-mechanical, and bonding performance, etc. (Xu et al. 2004; Correal and Ramirez 2010; Darzi et al. 2018; Bakar et al. 2019; Li et al. 2019; Rusch et al. 2019). Laminated bamboo lumber is bonded with adhesives, e.g., phenolic resin, so bond line and bonding interface characterization is important to ensure bonding performance (Correal and Ramirez 2010; Guan et al. 2014; Guan et al. 2016). It is necessary to improve adhesive penetration into the bamboo matrix. However, bamboo has a dense surface and an intrinsic lack of transverse tissue, e.g., wood rays (Liese 1998). Horizontal transmission only relies on pits in the cell walls, which may result in the shallow penetration of the adhesive into the bamboo and lead to unsatisfactory bonding performance (Ma 2009; Guan et al. 2014, 2016). Research has shown that micro-jets generated from ultrasonic cavitation may cause shock and a shear effect on the surface cells of bamboo strips. An ultrasonic treatment can enhance the permeability of the adhesive, relying on deeper and wider pits, and can, accordingly, increase the shear strength of laminated bamboo lumber (Guan et al. 2013).
It is known that the properties of bamboo engineered materials depends not only on the bamboo matrix itself but also on the properties of the bonding interface generated by mutual contacts with adhesives. The interface between the adhesive and bamboo plays a vital role in ensuring the performance of bamboo-based composite structure materials and their application (Guan et al. 2016). The strain distribution in the interface of wood composite materials influences the stress transmission and dispersion, which may influence the macro-performance of wood-based composites in terms of dimensional stability, durability, and shear strength (Serrano and Enquist 2005; Jeong et al. 2006; Gindl and Muller 2016). However, it is difficult to directly measure strain distribution in interfaces due to the thin character of the bond line, so contact-free measurement is usually used. Today, measurement of the bonding interface strain distribution of wood-based composites, e.g., plywood, wood adhesive bond line, wood block, and wood/adhesive interphase, is conducted by digital image correlation (DIC) (Serrano and Enquist 2005; Jeong et al. 2016; Guan et al. 2014; Gindl and Muller 2016). Relatively speaking, research on bamboo/based composites has employed DIC less often compared to wood-based composite research. Numerical simulation based on the finite element method has been frequently employed in strain simulations, for study of the bonding interface of wood particleboard, flooring, middle density fiberboard, timber, etc., or laminated Guadua bamboo lumber (Grigsby and Thumm 2004; Blanchet et al. 2005; Konnerth et al. 2006; Wong et al. 2003; He et al. 2012; Sebera et al. 2015; Takeuchi et al. 2018). However, previous research on bamboo or wood composites has primarily focused on macro-strain performance based on the assumption that the bonding quality is good, and the bond line adhesion has not changed. The effect of the bond line on the strain transmission is neglected, because its thickness is too thin to be considered in volume. However, the strain distribution of the bonding interface of laminated bamboo lumber made from ultrasound-treated bamboo strips is still unknown.
In this study, bamboo strips were treated via ultrasonic cavitation, and two-ply laminated bamboo lumber was made with a phenol-formaldehyde resin adhesive. The micro-scale strain distribution of the bonding interface was tested under tensile load conditions via DIC to understand the bamboo adhesion mechanism before and after the ultrasonic treatment. The effects of the ultrasonic treatment on the bonding shear strength of the laminated bamboo lumber were discussed. Finite-element analysis (FEA) simulation results were also presented for comparison with the experimental test result, so as to further understand or characterize the stress transmission function of the bonding interface and the bamboo bonding mechanism.
EXPERIMENTAL
Ultrasonic Treatment of the Bamboo Strips
Bleached and carbonized Moso bamboo (Phyllostachys pubescens) strips (4 years old) were obtained from the Zhejiang Bamboo Factory (Zhejiang, China), each measuring 150 mm × 20 mm × 5 mm. The strips were treated under water in an ultrasonic cell crushing machine. The ultrasonic treatment process for the bleached bamboo strips was carried out under the following conditions: a temperature of 60 °C, a treatment time of 60 min, and a power of 1200 W. The ultrasonic treatment for the carbonized bamboo strips was carried out under the following conditions: a temperature of 60 °C, a treatment time of 90 min, and a power of 1400 W. The bamboo strips were then dried under air-drying conditions and conditioned in a standard chamber (a temperature of 20 °C and a relative humidity of 65%) until a constant weight was reached.
Parameters Testing of the Bamboo Strips for the Finite-element Analysis (FEA) Evaluation
In order to conduct the FEA evaluation of the laminated bamboo lumber, the basic parameters of the bamboo strips before and after ultrasonic treatment were tested according to GB/T standard 15780 (1995). The longitudinal, radial, and tangential directions of bamboo were marked as L, R, and T, respectively. The parameters of the bamboo strips used for the FEA simulation included the elastic modulus in three directions (EL, ER, and ET), Poisson’s ratio in three directions (μLR, μLT, and μRT), and the shear modulus of bamboo in three directions (GLR, GLT, and GRT). These parameters were tested via tensile and calculated via DIC with 6 dog bone tensile samples using the size described in GB/T standard 15780 (1995). The test conditions of the DIC are shown in the following section (DIC measurement) with the same painted surface and 3D-camera system, loading steps, and calculation method.
Preparation of the Two-Ply Laminated Bamboo Lumber
The two-ply bleached and carbonized bamboo strips were laminated parallel to each other before and after ultrasonic treatment, and the assembly pattern was inner (the region close to the inner part of bamboo) to inner, i.e., low-density face to low-density face, as shown in Fig. 1.
Fig. 1. Assembled patterns of the laminated bamboo lumber, i.e., inner to inner face: laminated bleached bamboo lumber (LBBL); laminated carbonized bamboo lumber (LCBL); laminated ultrasonically bleached bamboo lumber (LUBL); and laminated ultrasonically carbonized bamboo lumber (LUCBL)
The phenol formaldehyde (PF) adhesive was self-made and used at a solid content of 50%. The specimens were all cured for 15 min in a Platen Vulcanizing Press at 1.2 MPa and an ambient temperature of 140 °C; the adhesive consumption was 140 g·m−2. After the curing process, the specimens were conditioned in a standard chamber (a temperature of 20 °C and a relative humidity of 65% until a constant weight was attained.
Characterization of the Bonding Interface and Bondline Thickness Calculation
A 30 μm × 3 mm × 3 mm cross-section was taken from the bleached and carbonized laminated bamboo lumber samples upon ultrasonic treatment and the creation of a control group. Each specimen was softened by soaking in water at room temperature for 7 d. The cross-sections were then cut and dehydrated with graded ethanol in gradient concentrations of 30%, 60%, 95%, and 100%. After desiccation, fixing, and gold spraying (E-1010, Hitachi Ion Sputter, Jeol, Japan), the cross-sections were observed via scanning electron microscopy (SEM) (Quanta 200, FEI, Hillsboro, OR) at 20 kV to evaluate the microstructure of the bonding interface and the adhesive bond line in the bamboo matrix. The thickness of the bond line was calculated according to Eq. 1,
(1)
where AT is the average value of the bond line thickness, µm; Ai is the objective area of the bonding interface (including the bond line), µm2; and xi is the objective length of the bonding interface, µm;. Five replicates were conducted for one specimen and calculated 30 times at different distances along the length of the bond line in one image.
Digital Image Correlation (DIC) Measurement
The shear strength specimens were processed experimentally with the following specifications: a 150 mm length, a 20 mm width, and a 10 mm thickness, with a shear strength test rabbet of 2.5 mm on both sides, which was in accordance with DIN EN standard 302-1 (2004) (Fig. 2).
Fig. 2. Geometry and specification of the test specimens for the DIC and test scene: (a) the speckle pattern specimen of laminated bamboo lumber; (b) the specifications applied to the DIN EN standard 302-1 (2004) specimen; and (c) the specimen loaded with cameras
The surfaces of all the samples were polished with sandpaper. A random white and black spray pattern was spray-painted onto the surfaces of the test objects to ensure that the dot in the black and white matrix could be focused, in which the objective zone image could be collected by the 3D-DIC system (Fig. 2). Samples were clamped at both ends and stretched along the sample-length direction. To maintain stable conditions for the 3D-DIC measurement, a 50-N pre-load was applied to the samples before measurement. The samples were then strained in 10 steps of 50 N for each step, with loading speed of 2 mm/min, so the final load was 550 N. The load was controlled in the elastic range. The micro-scale strain distribution of sample was collected by image and calculated. The data of the elastic modulus (E) and Poisson’s ratio (μ) in the fiber direction of the laminated bamboo lumber were obtained and were used as parameters in the following FEA evaluation.
Parameter Settings for the Finite-element Analysis (FEA) Model
The geometric model is shown in Fig. 3. It was assumed that the model was under tensile conditions. The movement of both ends of the sample was transversely restricted, and the only allowed movement was in the axial direction. A 550 N positive tensile load was loaded along the X-axis on the right-hand side of the model, and the gravity of the laminated bamboo lumber was neglected.
Fig. 3. Geometric model of the laminated bamboo lumber bondline
The elastic modulus, Poisson’s ratio, and shear modulus of the bamboo were selected based on the above tested parameters of the bamboo strips. The size of the geometric model was confirmed based on experiments, with a bamboo size of 5 mm and a total model length of 30 mm. The input data of the bond line thickness used for the model was calculated based on the bondline thickness from the SEM experimental investigation results, and the average bond line thickness was selected in the model.
Finite-element Analysis (FEA) Simulation Process
The upper and lower layers of the model were set as bamboo and the middle layer as adhesive. The properties of the bamboo and adhesive will be given in a following section.
The geometrical model was set up and then meshed into a finite number of units. Given that meshing could influence the calculation accuracy, the meshing quantity, density, and quality were taken into account (Guan et al. 2016). The model geometry was meshed with refinements in the bonding interface with mesh of 0.1 mm×0.35 mm, where strain and stress concentrations were expected (Fig. 4). At the left end of the model, the movements in the X-axis direction and in the Y-axis direction were restricted. A tensile load of 550 N was applied to the right end of the sample. All calculated analyses were conducted using ANSYS software (Version 10.0, Ansys, Inc., Canonsburg, PA). The FEA simulation results were compared with the DIC experimental results, and the effectiveness and error of the FEA simulation was considered.
Fig. 4. FEA meshing at the bondline interface
Tensile Shear Strength of the Laminated Bamboo Lumber
Four kinds of laminated bamboo lumber specimens were manufactured for tensile shear measurements according to DIN EN standard 302-1 (2004) and were fixed to the testing machine using pin-ended attachments. The specimen size was 150 mm × 20 mm × 10 mm, and an overlap with a width of 10 mm was defined in the middle of the section center in the test pieces with thick glue lines. There were 10 replicates.
RESULTS AND DISCUSSION
Parameters of the Bamboo Strips for Finite-element Analysis (FEA)
The parameters of the bamboo strips used for the FEA simulation are shown in Table 1.
Table 1. Basic Parameters for the Finite-element Analysis (FEA) Model of the Laminated Bamboo Lumber
Scanning Electron Microscopy (SEM) Morphology and Bond Line Thickness
Images of the laminated bamboo lumbers and the morphology of the bonding interface of the laminated lumber bamboos are shown in Figs. 5 and 6. In addition, the bondline thickness results are displayed in Table 2, as well as the bonding line values, which were used for the FEA evaluation. As shown in Table 2, the increment of bondline thickness of LBBLU was more than 1.80 times that of LBBL, and the increment of bondline thickness of LCBLU was more than 1.23 times that of LCBL. Therefore, ultrasonic treatment on bamboo strip can distinctly increase the bondline thickness of laminated bamboo lumber.
Figure 5 shows the presence of an intact bondline in the cross-sections of the laminated bamboo lumber. Since bamboo does not have transverse tissue, e.g., wood rays, the pit of the cell wall is the only way for adhesive to penetrate. Therefore, it is difficult for adhesives to penetrate a bamboo block, and is distributed on the bamboo surface, especially between the cell walls and in the broken cell cavities formed during sample preparation. As a result, the thickness of the bond line is relatively small and straight. However, after ultrasonic treatment, the bond line was relatively larger and curved. More adhesive was able to penetrate the bamboo matrix since the pits in cell wall had been partly broken up under ultrasonic treatment.
Fig. 5. SEM images of the LBBL, LUBBL, LCBL, and LUCBL samples
Figure 6 shows the presence of the bonding interface in the parenchyma area in the cross-sections of the laminated bamboo lumber. After the ultrasonic treatment, the boundary between the adhesive and the treated bamboo was not clear, and the deformation of the parenchyma cells near the bonding interface was more pronounced than the deformation in the untreated bamboo. The increased penetration of the adhesive enhanced the mechanical interlocking of the interface between the adhesive and bamboo, which led to an increase in the shear strength of the laminated bamboo lumber (Guan et al. 2016). As a result, the adhesive not only was distributed in the broken cell cavities on the surface during the phenolic resin preparation, but it also penetrated the farther parenchyma cells through an extended pit channel, leading to the increase in the bond line thickness (Table 2). The bondline value from the SEM measurement was used for the FEA model.
Fig. 6. Bonding interface morphology of the LBBL, LUBBL, LCBL, and LUCBL samples
Table 2. Bondline Thicknesses for the Finite-element Analysis (FEA) Model
Strain Distribution of the Bonding Interface via Digital Image Correlation (DIC)
The DIC test results showed the strain distribution of the bonding interface of four kinds of laminated bamboo lumbers at 550 N (Figs. 7 and 8). Thus, the specimens at a 550 N tensile load were used for investigating the differences in strain distribution. The positive and negative strain value meant tensile and compressive strain, respectively.
Fig. 7. Strain distribution of the LBBL and LUBBL samples (left is the strain area and right is the calculated value)
As shown in Fig. 7, the DIC measurements showed that the strain area (within the green surrounded area) of the LBBL and LUBBL samples almost covered half of the thickness of the sample; the thickness of the calculated strain zone was 4.0 to 4.5 mm. The higher strain occurred at the ends of the overlapping area with a small volume. In the bonding interface region of the LBBL, the strain was -6×10-4 to 1×10-3, while the data changed to -4×10-4 to 0.8×10-3 in the LUBBL samples. Generally, the absolute value of the compressive strain and tensile strain of the LUBBL sample decreased. The lower the interface strain, the higher the interface strength, which demonstrated that the ultrasonic treatment could increase the shear strength of the laminated bamboo lumber. In addition, it showed that the higher shear strength not only relied on the higher mechanical interlock due to the greater adhesive penetration but also on the lower interface stress.
As shown in Fig. 8, the ultrasonic treatment did not cause much change in the strain distribution area across the bamboo bonding interface, coinciding with the previous study by Guan et al. (2016), in which the penetration depth of the adhesive into the bamboo surface was found to be relatively small.