Bamboo Scrimber’s Physical and Mechanical Properties in Comparison to Four Structural Timber Species

Abstract

Bamboo scrimber is a sustainable engineered material that overcomes natural round bamboo’s various weaknesses. This study compared the bamboo scrimber’s mechanical (strength, stiffness, and ductility) to timber. The results showed that scrimber’s physical and mechanical properties are comparable, even superior, to wood, especially in compression. Scrimber has a higher density than timber. Its drier equilibrium moisture content indicates that scrimber is more hydrophobic than timbers. The maximum crushing strength (σ_c//), compressive stress perpendicular-to-fiber at the proportional limit (σ_cp⊥) and that at the 0.04” deformation (σ_c0.04⊥), shear strength (τ_//), longitudinal compressive modulus of elasticity (Ec_//), lateral compressive modulus of elasticity (E_c⊥), and modulus of rigidity (G) of scrimber are higher than those of timbers. Both scrimber’s and timber’s flexural properties (modulus of rupture (σ_b) and flexural modulus of elasticity (E_b)) are comparable. On the contrary, the tensile strength parallel-to-fiber (σ_t//) of scrimber is weaker than that of timber. Scrimber is high ductility (μ > 6) when subjected to compression perpendicular-to-fiber, medium ductility (4 < μ ≤ 6) when subjected to compression parallel-to-fiber, and low ductility (brittle) when subjected to bending, shear, or tensile parallel-to-fiber. The higher ductility of scrimber may give an alarm and more time before failure than timbers. Timbers have brittle to lower ductility when receiving each kind of loading scheme. The ratio of shear modulus to strength (G/τ) and compression modulus to strength parallel-to-fiber (E_C∥/σ_C∥) strongly correlates with the ductility ratio. However, the ratio of the flexural modulus of elasticity to the modulus of rupture (E_b/σ_b) and the ratio of the modulus Young to compression stress perpendicular-to-fiber (E_c⊥/σ_cp⊥) do not strongly correlate to the ductility value.

1. Introduction

Awareness of sustainability toward the environment is rising, especially regarding materials for structural applications. This awareness and advances in technology encourage the development of sustainable engineering materials. Sustainable engineering materials can be produced using wood [1,2,3,4,5,6,7,8], bamboo [9,10,11,12,13,14,15,16,17,18], other lignocellulosic materials [19,20], or their composite with conventional material [21]. Engineered bamboo products overcome various weaknesses in natural round bamboo so that bamboo can be promoted to a high-class material. There are several types of engineered-bamboo products for structural members, i.e., laminated bamboo [22,23], bamboo board [24,25,26], and reconstituted densified bamboo [27,28]. Reconstituted densified bamboo, also known as bamboo scrimber, is a composite bamboo made by crushing bamboo culms into bundles and then pre-treating using heat treatment and immersing in adhesive before the hot pressing process [29,30]. Unlike laminated bamboo, where the layers are arranged parallel, the strips/bundles arranged in bamboo scrimber are irregular. There are no layer boundaries in bamboo scrimber [31]. Bamboo scrimber has several advantages, such as raw material efficiency (the percentage of the end product to the total raw material expended) of up to 80% [32], high resistance to biodegradation [33], higher heat storage and heat conduction properties and lower moisture absorption than timber [27], high density [34,35], and reliable mechanical strength. Bamboo scrimber is adequate for external applications such as deck flooring because of its extraordinary Janka hardness [32]. Bamboo scrimber structures could be combined using traditional furniture tenons and modern connections [36]. Sharma et al. [37] also mentioned that engineered products could create a standard homogenous and reduce variability among members.

In addition to its sustainability, the primary considerations in using bamboo and wood for structural members are the structural properties. In addition to strength and stiffness, ductility is an essential requirement and a preferable mechanical property in structural design. Ductile structures have a particular alternative of bearing capacity, i.e., large deformations occur before failure, that can be used to alert building occupants in case of an unexpected load. Regarding robustness, ductility is desirable because it is widely assumed that a flexible structure may be more reliable and robust than brittle structures. Kirkegaard et al. [38] evaluated the robustness of a timber truss structure. The result of the model with ductile timber behavior shows that the robustness index is higher for these ductile elements. The building designers shall consider the ductile behavior, especially for structures in seismic areas. The classifications of ductility can be found in Eurocode 8: EN-1998-1 [39]. The importance of ductility design in structures relates to dynamically loaded structures in their application, e.g., design structures for bridges [40], wall panels on CLT houses [41,42], and light-frame buildings [43].

Bamboo has good strength, stiffness, and ductility that enable it to be used in situations requiring the material’s ability to undergo considerable deformation [44]. Different researchers have expressed the ductility index on a different quantitative basis. It can be expressed as energy absorption [45], curvature ratio [46,47], rotation ratio [48], and displacement ratio [49]. Ductility, in general definition, is the ratio between the ultimate and yield displacement. However, the timber yield criterion needs to be clarified, as it is for example, for steel structures. There are also different methods for determining the yield point of timber structures. The value of yield displacement can vary up to 80% depending on the method used [50]. Although scrimber has a similar appearance and form to wood and wood-based products, its fundamental composition and ultimate fracture differ [51]. Experimental studies on its mechanical properties, including ductility and its comparison to structural wood, may significantly contribute to the scrimber utilization for green construction members. Further research is necessary to characterize scrimber structural properties.

2. Materials and Methods

2.1. Materials

Bamboo scrimber used in this study were commercially produced by Bambulogy, a building design company in Tangerang, Indonesia. The supplied scrimber had dimensions of 30 mm × 140 mm × 1870 mm. Scrimber used in this study was manufactured by mixing a few species of bamboo, including hitam (Gigantochloa atroviolacea), tali (Gigantochloa apus), mayan (Gigantochloa robusta), and gombong (Gigantochloa pseudoarundinacea), which have diameters range between 7 and 12 cm. The manufacturing procedures were as follows:

• Natural round bamboo was split into 3–8 parts and cut longitudinally into 2.5 m lengths.
• The outer skin and inner part were removed using an expanding machine, then crushed and flattened.
• The strands were carbonized for 2–3 h at 190 °C to remove extractive components in natural bamboo. After the carbonization process, the strands become less stiff and are easier to form the scrimber because their modulus decreases when more than 150 °C temperature is applied.
• The dry carbonized strands, having a moisture content of 0% (zero percent), were placed in the 65 °C temperature tunnel for 3.5 h until the moisture content increased and reached 0.5%.
• The adhesive used in this manufacturing process was phenol-formaldehyde (PF). The adhesion process was performed by dipping strands in PF adhesive solution (PF:water = 1:1) for 10 min, and then draining it.
• Strands coated with adhesive were dried at 70 °C for 2–4 h until they reached the target moisture content of 11%.
• After drying the resin-coated strands, the cold pressing process was carried out for 2 min with a compression load of 687 kN, equivalent to 2.625 MPa pressure.
• Next, blocks were formed in molds over 12 h using a hot curing machine with several temperatures in three parts of time, namely 110 °C for 3 h, 135 °C for 6 h, and 120 °C for 3 h. A 598–638 N load was applied to the beam during the molding process.
• Conditioning for three weeks, the last stage, releases the residual stress.

In addition to bamboo scrimber, the material used in this study included sawn lumber of red meranti (Shorea sp.), mahogany (Swietenia sp.), agathis (Agathis sp.), and pine (Pinus spp.) purchased from timber markets in Bogor, West Java-Indonesia as comparisons. Further, the bamboo scrimber and timbers were cut into specimens based on the standard requirement for testing of mechanical properties, ASTM D143 Standard Test Methods for Small Clear Specimens of Timber [52]. The specimen sizes are given in

Table 1. Experimental test methods for bamboo scrimber and timber.

Test Method	Direction	Number of Specimen (n)		Specimen Size (b cm × d cm × L cm)		Loading Rate (mm/min)
		Scrimber	Timber	Timber	Scrimber
Tension	Parallel-to-fiber	14	50	2.5 × 2.5 × 46 (Figure 1a)	2.5 × 2.5 × 46 (Figure 1a)	1.00
Compression	Parallel-to-fiber	9	50	2.5 × 2.5 × 10 (Figure 1b1)	3 × 3 × 20 (Figure 1b2)	0.30
Compression	Perpendicular-to-fiber	7	50	5 × 5 × 15 (Figure 1c1)	3 × 3 × 15 (Figure 1c2)	0.305
Shear	Parallel-to-fiber	7	50	5 × 5 × 6.3 (Figure 1d1)	3 × 5 × 6.3 (Figure 1d2)	0.6
Bending	Center point loading	10	50	2.5 × 2.5 × 41 (Figure 1e)	2.5 × 2.5 × 41 (Figure 1e)	1.3

and Figure 1. All specimens were air-dried in indoor environmental conditions for a month to reach equilibrium moisture content.

2.2. Methods

2.2.1. Physical Properties

The width (b) and depth (d) of small-clear size specimens were measured using a digital caliper with an accuracy of 0.01 mm. A digital caliper was employed to measure specimen length (L) when less than 150 mm; a ruler was used to measure the longer specimen. Specimen mass was weighed at the air-dry moisture content before the mechanical properties test (m₀). Not long after the mechanical properties test, the specimen’s stringy part was removed, then the mass was weighed (m₁). The specimens were placed in a 102 ± 3 °C oven for 48 h; after this, they were weighed every three hours and placed in the oven again until the mass was constant at three consecutive weighs (m₂). The moisture content (M_c), density (ρ), and specific gravity (G_b) were calculated following Equations (1a)–(1c).

$$M_c=\frac{m_1 - m_2}{m_2}\times 100\%$$

(1a)

$$\rho =\frac{ m_0}{V}=\frac{ m_0}{\mathit{Lbd}}$$

(1b)

$$G_b=\frac{\rho }{(1+M_c\%){\rho }_{\mathrm{w}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{r}}}$$

(1c)

2.2.2. Mechanical Properties

Mechanical properties (including flexural, compression, tensile, and shear) were tested based on ASTM D143 [52] using a SATEC/Baldwin Universal Testing Machine (UTM) 30-ton capacity (SATEC/Baldwin, Grove City, PA, USA), equipped with periodically calibrated load cell, linear variable displacement transducers (LVDT), and multipurpose digital indicator (MPDI) data acquisition machine (installed by PT Testindo, Jakarta and calibrated by PT Global Quality Indonesia, Bandung, ID) [5], and UTM Instron type 3369 (PT Patochemi Murni Aditama, Jakarta, ID). The test methods, specimens, and parameters are summarized in Table 1. The flexural modulus of elasticity (E_b), modulus of rupture (σ_b), compression strength perpendicular-to-fiber (σ_c⊥), compression stress parallel-to-fiber (σ_C∥), the Young modulus parallel- and perpendicular-to-fiber (E_c∥ and E_c⊥), tensile strength parallel-to-fiber (σ_t∥), shear strength (τ_//), and shear modulus (G) were calculated following equations in

Table 2. Equations used to determine mechanical properties of timber and bamboo scrimber.

Mechanical Properties	Equations
Modulus of Elasticity (Eb)	${\mathit{E}}_{\mathit{b}}=\frac{\mathit{P}{\mathit{L}}^3}{4\Delta \mathit{b}{\mathit{h}}^3}$
Modulus of Rupture (SR)	${\sigma }_b=\frac{3P_{\mathit{max}} L}{2b{h}^2}$
Tension Parallel-to-fiber Strength (Ft∥)	${\sigma }_{t//}=\frac{P_{\mathit{max}}}{A}$
Compression Parallel-to-fiber Strength (Fc∥)	${\sigma }_{c//}=\frac{P_{\mathit{max}}}{A}$
Compression Perpendicular-to-fiber Strength at Proportional Limit (Fcp⊥)	${\sigma }_{\mathit{cp}\mathit{\perp }}=\frac{P_p}{A}$
Compression Perpendicular-to-fiber Strength at 0.04 inch (Fc0.04⊥)	${\sigma }_{c0.04\mathit{\perp }}=\frac{P_{0.04}}{A}$
Compression modulus (Ec)	$E=\frac{\sigma }{\epsilon }$
Shear Parallel-to-fiber Strength (Fs)	${\tau }_{//}=\frac{P_{\mathit{max}}}{A}$
Shear Modulus (G)	$G=\frac{\tau }{\gamma }$

. The elastic properties (including modulus of elasticity (E_b, E_c∥, and E_c⊥) and shear modulus (G)) were calculated as the slope of the linear (proportional) part of the load-deformation or stress-strain diagram.

2.2.3. Ductility Ratio

Ductility expresses the ratio between ultimate displacement (Δ_u) and yield displacement (Δ_y) (Equation (2)). The Δ_u and Δ_y are commonly obtained from tensile and compressive tests. However, this study measured ductility ratio (μ) through several mechanical property tests (i.e., tension parallel-to-fiber, compression parallel-to-fiber, compression perpendicular-to-fiber, shear parallel-to-fiber, and bending).

$$\mu =\frac{{\Delta }_u}{{\Delta }_y}$$

(2)

To estimate ductility, the determination of a yield point is necessary. Since the yield criterion for timber is not well-defined (as it is for example, for steel structures), different methods for determining the yield displacement for timber connections or structures exist (Figure 2). This timber yield point is known as the pseudo-yield point. Some well-known procedures utilized in this research are:

• Karacabeyli and Ceccotti (K&C) [53]: The yield point in this method is viewed as the point on the load-deformation curve equal to 50% of the maximum capacity (Figure 2a).
• Commonwealth Scientific and Industrial Research Organization (CSIRO) [54]: In this case, the yield point in this method is viewed as the point on the load-deformation curve corresponding to 40% of the maximum capacity. The 40% of the maximum deformation is adjusted by a factor of 1.25. The point on the load-deformation curve formed by the intersection of the projection line and the new coefficient of displacement value is viewed as the yield point, and the yield load is determined (Figure 2b).
• European Committee for Standardization (CEN) [55]: This method uses the secant and tangent lines of the two parts of the load-deformation curve to determine the yield point. The first line indicates the initial stiffness, calculated from 10% to 40% of the maximum load. The angle between the secant line and displacement axis is α. The slope of the second line is equivalent to one-sixth (1/6) of the slope of the second of the load-displacement curve. The yield point is resolved as the intersection of the first and second lines (Figure 2c).
• Yasumura and Kawai (Y&K) [56]: The secant line between 10% and 40% of the maximum load indicates the initial stiffness. A line connecting the data point of 40% and 90% of the peak load, called a chord line, is drawn. Then a line parallel to the chord line and tangent to the load-deformation curve is created. The last line represents the post-elastic area before the maximum load. The intersection point between the initial stiffness and tangent line is projected horizontally onto the load-displacement curve to obtain the yield point displacement (Figure 2d).
• Equivalent Energy Elastic-Plastic (EEEP) [57]: In this method, a bilinear curve represents an assembly’s perfect elastic-plastic curve (Figure 2e). The area under the load-displacement curve (W_failure) is assumed to be the same as the area beneath the bilinear curve. Initial stiffness in this method corresponds to the first straight line, which is defined as being between 0% and 40% of the peak load (K₄₀). Deformation at failure (∆_failure) is defined as deformation at 80% of maximum load. The following equation is used to calculate the yield load (P_y) is Equation (3).

$$P_y=\left[{\Delta }_{\mathit{failure}}-\sqrt{{{\Delta }^2}_{\mathit{failure}}-\frac{2W_{\mathit{failure}}}{K}} \right]\times K$$

(3)

2.2.4. Analysis

The ductility ratio of all specimens tested is presented using descriptive statistics, including mean, standard deviation, minimum and maximum value. Simple linear regression analysis was conducted to analyze the relationship between ductility ratio and stiffness to strength ratio.

3. Results and Discussion

3.1. General Description of Bamboo Scrimber

The bamboo scrimber used in this study is a board similar to sawn timber. Scrimber has a smoother and flatter surface than timber. Its color is brown to dark brown due to the treatments during the manufacturing process. Unlike other common wood composites, scrimber has no layers, while the others, such as plywood and glued laminated timber/bamboo, have multi-layers. The fiber surface type of scrimber is stranded [58]. This stranded type gives a more natural look. Compared to timber, scrimber is not a lightweight material due to its high density, which may raise problems in transportation.

3.2. Physical Properties

Physical properties tested include moisture content (M_c), density (ρ), and specific gravity (G_b). Moisture content is crucial, especially on hygroscopic materials, because it affects their volume and mass, affecting other properties. The average moisture content of all specimens from this experimental study ranged between 8.14% and 15.46% (Figure 3), similar to the previous reports on wood [4,5,8], wood products [59,60], and bamboo [10,11,12,13,14,61] conducted in Bogor, West Java (ID). This equilibrium moisture content is wetter than bamboo in sub-tropical regions such as Coventry (UK) [15,16]. The scrimber’s equilibrium moisture content is (8.14 ± 0.34%), generally drier than wood. As hygroscopic material, scrimber’s air-dry moisture content is still higher than mortar [21]. Scrimber had the lowest moisture content, while pine had the highest one. Scrimber, as an engineered product, has undergone a series of processes, including drying, pressing, and gluing, which causes this material to be more hydrophobic than untreated wood. The bonding process using phenol-formaldehyde (PF) resin reduces the water absorption ability. PF resin adhesive is commonly applied for outdoor products, so it can excellently resist the environment’s humidity fluctuation.

The scrimber and timber’s average density and specific gravity are presented in Figure 4. The density varies between 0.46 and 0.98 g/cm³. The specific gravity (G_b) is determined based on oven-dried mass and air-dried volume. The G_b values range between 0.40 and 0.90. In both cases, agathis has the lowest value, while scrimber is the densest. Both scrimber and timber are hygroscopic materials that absorb water from their surrounding environment; thus, their masses relate to moisture content. The higher G_b value means the total hollow cavity volume decreases; this explains the scrimber’s lower moisture content than timber. The scrimber is denser than the other timbers. The lower moisture content of scrimber also indicates the less available void for free water [62].

3.3. Mechanical Properties

3.3.1. Tension Parallel-to-Fibers

The mean value for tension strength parallel-to-fiber of the scrimber is 34.27 MPa (

Table 3. Mechanical property test results for bamboo scrimber and timber specimens.

Mechanical Properties (MPa)	Scrimber (Mean ± s)	Agathis (Mean ± s)	Mahogany (Mean ± s)	Red Meranti (Mean ± s)	Pine (Mean ± s)
Flexural modulus of elasticity (Eb)	8525 ± 1275	6968 ± 618	8033 ± 1079 *	11002 ± 1379 *	7218 ± 1837 *
Modulus of rupture (σb)	71.14 ± 9.85	52.08 ± 4.49	70.47 ± 11.37 *	72.42 ± 12.15 *	55.15 ± 10.18 *
Tensile strength parallel-to-fiber (σt∥)	34.27 ± 16.37	61 ± 19.30	74.94 ± 21.21	128.48 ± 35.68	72.56 ± 28.68
Compressive strength parallel-to-fiber (σc∥)	64.85 ± 4.40	24.75 ± 2.86	30.67 ± 5.36	40.84 ± 6.46	29.07 ± 4.07
Compressive modulus of elasticity parallel-to-fiber (Ec∥)	5296 ± 577	1552 ± 275	2030 ± 376	2617 ± 452	2048 ± 335
Compressive stress perpendicular-to-fiber at proportional limit (σcp⊥)	19.60 ± 3.00	4.05 ± 0.88	7.76 ± 1.25	5.41 ± 0.75	5.11 ± 1.27
Compressive stress perpendicular-to-fiber at 0.04” deformation (σc0.04⊥)	21.31 ± 4.80	3.77 ± 0.65	7.08 ± 1.10	5.06 ± 0.65	5.02 ± 1.10
Compressive modulus of elasticity perpendicular-to-fiber (Ec⊥)	980.4 ± 58.2	194.7 ± 40.4	341.4 ± 46.8	243.3 ± 35.8	241.2 ± 50.0
Shear strength (τ∥)	11.15 ± 2.50	7.09 ± 1.48	9.88 ± 1.24	8.74 ± 1.41	8.48 ± 1.59
Shear modulus (G)	290.8 ± 66.3	206.6 ± 57.7	266.0 ± 57.5	261.6 ± 60.4	242.2 ± 66.3

Note: s = standard deviation, * The flexural test of mahogany, red meranti, and pine was conducted by Bahtiar et al. [4], and then we recalculated them.

). This value is lower than Kumar et al.’s report [35], which equals 111 MPa for scrimber. The tensile strength parallel-to-fiber of the scrimber is the weakest among other materials, while the strongest was red meranti (128.48 MPa). Bamboo scrimber used in this study was not flattened and crushed using an excellent machine for the defibering process but the simple manual one. It will cause the bamboo strips to not have well-maintained fiber direction longitudinally. At the same time, it is supposed to be horizontally maintained to decrease the mechanical properties, especially in tension parallel to the fiber. Huang et al. [36] reported that early defibering was mainly hand-made, using manual hammering and roll-in. However, the former method could cause mechanical damage to bamboo. Figure 5a shows a sample of the load-displacement curve in tension test parallel-to-fiber of each material. The load-displacement curves generally had a linear relationship with increasing load until the final failure. Thus, the proportional limit is near the ultimate load. Failures found in scrimber specimens are along with the bamboo fiber (Figure 5b).

This failure type of scrimber in tension parallel-to-fiber was splintering tension. The failure type found in this study differs from the experimental study by Huang [63] and Sharma [37]; the failure mode of engineered bamboo is similar to wood in that the tension test is brittle, with the fibers fractured. The difference in failure between scrimber and timber loaded in tension parallel-to-fiber is that the fibers separate in scrimber, then fracture as the final failure at maximum load. While in timber, the fiber did not separate but was immediately fractured.

3.3.2. Compression

The scrimber’s mean value of the compressive strength parallel-to-fiber is 64.85 MPa (Table 3). At the same time, the mean value of compressive stress perpendicular-to-fiber at the proportional limit and 0.04-inch deformations are 19.60 and 21.31 MPa. Both compression strength parallel-to-fiber and compression stress perpendicular-to-fiber values are lower than Li et al.’s study [64], which equals 100.9 and 52.8 MPa, respectively. This compressive property’s low value might be attributed to the scrimber’s lower density than that of Li et al. The scrimber’s density (ρ) in this study is 980 kg/m³, while the density in Li et al. is 1254 kg/m³ [64]. However, the compression strength of bamboo scrimber in this experimental study is equivalent to laminated bamboo studied by Sharma [37], with values of 77 MPa for parallel-to-fiber and 22 Mpa for perpendicular-to-fiber.

The compressive strength parallel-to-fiber is three times stronger than the compressive strength perpendicular-to-fiber because the bamboo fibers’ strength is higher in the parallel direction [35] than in the lateral direction. Scrimber’s compressive strengths in this study, both parallel- and perpendicular-to-fiber, are higher than those of timbers. Scrimber is an engineered material composed of compressed bamboo fiber bundles arranged in parallel. As a result, the scrimber can withstand a high load applied to the bamboo fibers (Figure 5a). Timber loaded in compression perpendicular-to-fiber cannot reach the ultimate load, while bamboo scrimber can reach the ultimate load (Figure 6a).

The scrimber’s compressive modulus of elasticities parallel and perpendicular-to-fiber are 5296 and 980 MPa. While the timber’s compressive modulus of elasticities parallel- and perpendicular-to-fiber range between 1552–2616 MPa and 195–341 MPa. Similar to the compressive strength value, the scrimber’s compressive modulus is lower than the one reported by Li et al. [65], which were reported as 14,160 MPa for parallel-to-fiber and 4313 MPa for perpendicular-to-fiber.

Figure 6a and Figure 7a show the load-displacement curves of the scrimber specimen under compression parallel- and perpendicular-to-fiber. Compared to the compression test parallel-to-fiber, all materials tested have a low compressive strength perpendicular-to-fiber. According to the load-displacement curve, the scrimber received twice the load with a similar displacement before failure to other materials tested. Therefore, scrimber has a higher compressive strength perpendicular and parallel-to-fiber than timber. The failure mode in compression parallel-to-fiber was buckling (Figure 6b).

In contrast, the bamboo fiber fracture is the failure mode of scrimber tested in compression perpendicular-to-fiber (Figure 7b). Li et al. [66] and Sharma et al. [37] also have similar buckling failure modes in laminated bamboo and scrimber. Fracture along the bamboo fiber in scrimber specimens indicates the bonding between bamboo fibers are the main weakness in terms of mechanical properties. Compression perpendicular-to-fiber failure is also found on the glue line.

3.3.3. Shear

The shear strength of scrimber describes the bonding behavior between strands and adhesive [67]. Inter-fiber bonds greatly influence the shear strength within the bundle of fibrils of the bamboo culm. The shear strength value for scrimber ranges between 7.37 and 14.11 MPa, with a mean value of 11.15 MPa (Table 3). This mean value is similar to Kumar et al.’s [35] study, which equals 11.89 MPa for scrimber with slightly higher density (1.05 g/cm³) than this study (0.98 g/cm³). Scrimber has a higher shear strength than the timber. The shear strength value of the timber specimen ranges between 7.09 and 9.88 MPa. The highest shear strength value of timber was mahogany. Like the shear strength value, the highest shear modulus (G) of all specimens tested in this experimental study belonged to the scrimber, which equals 290 MPa. The failure type of scrimber can be seen in Figure 8b. Cracks along the main shear line split the specimen into two parts.

3.3.4. Bending

The scrimber’s mean values for modulus of elasticity and modulus of rupture are 8525 MPa and 71.14 MPa, respectively (Table 3). This elastic modulus value is comparable to Li et al.’s [64] report, which equals 9199 MPa for a scrimber with high density (1.25 g/cm³). However, the elastic modulus of scrimber in this experimental study is higher than Moso bamboo studied by Chen et al. [44], which is equal to 10,470 MPa. However, the scrimber’s modulus of rupture is lower than Moso bamboo, which equals 145.71 MPa.

The mean values for elastic and rupture modulus of timber are 6.97–11.00 GPa and 52.08–72.42 MPa. Red meranti had the highest value for elastic and rupture modulus of all specimens tested. Figure 9a shows the load-displacement curve of scrimber under flexural test. It shows that the scrimber specimen has reached maximum load with a relatively small displacement compared to other materials tested.

The failure of scrimber appeared as a small crack in the middle part of the specimens (Figure 9b). This typical pattern of failure is commonly named ‘simple tension’ according to Bodig and Jayne [68]. Simple tension is not a common bending failure, especially in high-density wood. A similar failure mode was also found by Zou et al. [31]; the failure of all scrimber specimens occurred at the bottom of the mid-span. The typical failure mode is fiber fracture caused by tensile force.

3.4. Ductility Ratio

Eurocode 8 [39] defines ductility as the “ratio of the ultimate deformation and the deformation at the end of elastic behavior”. Although the definition of ductility is well-defined, the definition of yield point has yet to reach an international agreement [69]. Four different methods to determine yield points were used in this study, as mentioned above.

Table 4. Descriptive statistic of ductility ratio.

Materials		Methods
		K&C			CSIRO			CEN			Y&K			EEEP
		n	Mean	s	n	Mean	s	n	Mean	s	n	Mean	s	n	Mean	s
Bamboo scrimber	σb	10	2.7	0.4	10	2.6	0.3	10	1.5	0.2	10	2.2	0.4	10	1.53	0.18
	σc∥	9	3.5	0.4	9	4.3	0.5							9	1.78	0.20
	σc⊥	7	9.5	3.7	7	10.2	4.8	7	6.7	2.9	7	7.0	2.3	7	5.15	1.90
	τ∥	7	1.5	0.2	7	1.8	0.3							5	1.24	0.51
	σt∥	14	2.0	0.3	14	1.9	0.3
Agathis	σb	50	3.3	0.3	50	2.6	0.3	50	1.9	0.2	50	2.8	0.4
	σc∥	50	3.1	1.1	50	2.9	1.0
	σc⊥	50	3.2	0.5	50	3.2	0.5	50	1.8	0.4	48	2.4	0.7
	τ∥	50	2.4	0.5	50	2.4	0.5
	σt∥	50	1.8	0.3	50	1.8	0.5
Mahogany	σb	50	3.3	0.6	50	3.4	0.6	50	2.0	0.4	50	2.9	0.5
	σc∥	50	2.7	1.2	50	2.4	1.1
	σc⊥	50	3.8	1.2	50	3.7	1.2	50	2.2	1.0	48	2.8	2.0
	τ∥	50	2.5	0.4	50	2.4	0.4
	σt∥	50	2.0	0.3	50	2.0	0.3
Red Meranti	σb	50	3.7	0.6	50	3.7	0.6	50	2.2	0.4	50	3.2	0.5
	σc∥	50	2.6	0.7	50	2.4	0.7
	σc⊥	50	2.7	0.3	50	2.7	0.3	50	1.4	0.2	45	1.9	0.4
	τ∥	51	2.3	0.4	51	2.2	0.5
	σt∥	50	2.1	0.4	50	2.0	0.5
Pine	σb	50	3.6	0.6	50	3.7	0.6	50	2.2	0.5	50	3.2	0.6
	σc∥	50	2.7	0.8	50	2.5	0.7
	σc⊥	50	3.3	0.6	50	3.3	0.6	50	1.9	0.5	47	2.3	0.7
	τ∥	50	2.8	0.9	50	2.8	0.9
	σt∥	50	2.0	0.4	50	1.9	0.5

Note: The ductility ratio for bamboo scrimber and wood subjected to: σ_b = static bending σ_t∥ = tension parallel-to-fiber, σ_c∥ = compression parallel-to-fiber, σ_c⊥= compression perpendicular-to-fiber, τ_∥ = shear parallel-to-fiber.

lists the descriptive statistics of ductility ratio (μ) determined from various procedures to obtain yield points. For tension, shear, and compression parallel-to-fiber tests, the yield point procedure based on CEN and Y&K methods cannot be determined, although these methods are bilinear methods, which should help balance the yield load according to the shape of the curve [50]. In this experimental study, the K_10–40 line is located off the load-displacement curve causing the yield point to be undetermined for the examples shown in Figure 10. In contrast, in bending and compression perpendicular-to-fiber, the yield points can be obtained using four methods (i.e., K&C, CSIRO, CEN, and Y&K). Jorrisen and Fragiacomo [70] reported that for statically determined structures, ductility results in large local displacements such as compression perpendicular-to-fiber at the supports, compression at a certain angle in the wood-working joint, or large deformation in connection. However, yield points obtained using the EEEP method in timber also cannot be determined because, in timber, load-displacement curves do not have enough slip. The slip in this method tends to be very important to obtain deformation at failure (∆_failure), defined as deformation at 80%. This behavior is also found in the tension parallel-to-fiber test of scrimber. It can be seen that scrimber, as material construction occurs, slips before it fails.

Figure 11 shows a representative diagram of the position of the yield point in all mechanical tests of bamboo scrimber. In scrimber, the yield load estimated using CEN and EEEP methods was significantly higher, and the lowest yield load value was estimated using CSIRO or K&C methods. However, the CEN and EEEP yield points are outside the load-displacement curve. The point of intersection between the initial stiffness and the tangent with a slope equal to one-sixth of the initial stiffness determines the yield point estimated using the CEN method. However, it is because the displacement at yield is not directly related to the actual behavior. In other words, CEN and EEEP procedures do not reflect the point on the actual load-displacement curve of the specimen. In contrast to the Y&K method, which projected the yield point horizontally onto the load-displacement curve.

Therefore, the given yield point reflects the actual point, unlike the CEN method [71]. The Y&K method, according to Munoz et al. [50], provided a better estimate of the yield load and can be well balanced by a second independent slope representing the elastic back zone. The intersection of the slope and its projection on the load-deformation curve provides the actual yield point.

The ductility ratio ranges in this experimental study were 0.03–19.74 and can be classified from low to high according to Eurocode 8 [39]. In contrast, according to Smith [72], it can be classified from brittle to high ductility (Table 5). The mean ductility ratio value of bamboo scrimber ranged between low to high, while in timber, the overall mean values of ductility ratio obtained from various mechanical properties are low. The ductility ratio of bamboo scrimber subjected to compression had the highest μ than timber, which can be classified as low to medium for compression parallel-to-fiber and medium to high for compression perpendicular-to-fiber. Although bamboo scrimber had the highest μ in compression, bamboo scrimber overall had the lowest μ of all specimens tested, other than compression, which can be classified as low to medium ductility. This indicates that the displacement or plastic area before failure in bamboo scrimber is smaller than in timber. This behavior can be seen from the load-displacement curve of bamboo scrimber in the bending test, in which after reaching the yield point, the load continues to increase with a relatively small displacement and then reaches the maximum load (P_max), causing the specimen’s failure.

Bamboo scrimber specimen reach their maximum load before timber specimens (e.g., pine) reach the maximum load (Figure 12). Chen et al. [29] studied the flexural ductility of Moso bamboo, which was 3.06 times higher than wood. The mean flexural ductility factor of Moso bamboo is 6.48, while the flexural ductility of bamboo scrimber in this experimental study range between 2.25 and 3.42. The contrast to the result in this study could be caused by bamboo as a raw material of bamboo scrimber that has undergone a compression process under high temperature and pressure [65], so it becomes denser than the untreated one. Obataya et al. [73] pointed out that the excellent bending ductility of bamboo is attributed to the combination of fiber-rich outer and compressible inner parts. In contrast, in bamboo scrimber, the outer part of natural round bamboo has been removed. However, the ductility ratio of bamboo scrimber in this study is comparable to bamboo scrimber strengthening timber beams with CFRP/wooden pin anchorage reported by Chen et al. [74], in which the value ranged between 1.08 and 3.35.

Generally, the lowest ductility ratio of all mechanical properties tested was found in the tension parallel-to-fiber test. In contrast, the highest ductility ratio was found in the bending test for timber and compression parallel to the fiber of bamboo scrimber. In the bending test, the specimen will undergo deflection before the specimen failure. While in the tension parallel-to-fiber test, the specimen had little or no deflection and immediately reached failure. The same behavior was also found in the CLT connection studied by Ceallaigh and Harte [75], that shear forces are more ductile than tensile. For connection subjected to a shearing force, significant deformation occurs before failure occurs after the yield point. Connections in tension forces are less ductile than in shear forces, and the load will drop significantly after reaching the maximum load. Similar behavior is also found in this study, shown in Figure 5a. The ductility behavior of bamboo scrimber in compression perpendicular-to-fiber can be categorized as high. Meanwhile, the ductility behavior in compression parallel-to-fiber is medium. However, bamboo scrimber subjected to bending, tension, and shear is low ductility or brittle. Smith and Asiz [76] concluded that typical range responses for components in structural timber systems illustrated in Jorissen and Fragiacomo’s [70] report are brittle for tension and bending members and ductile for compression both parallel and perpendicular-to-fiber.

Table 5. Classification of ductility ratio.

Classification	Ductility Ratio [72]	Ductility Ratio [77]
Brittle	μ ≤ 2	−
Low ductility	2 < μ ≤ 4	μ ≤ 4
Moderate ductility	4 < μ ≤ 6	4 < μ ≤ 6
High ductility	μ > 6	μ > 6

Table 5. Classification of ductility ratio.

Classification	Ductility Ratio [72]	Ductility Ratio [77]
Brittle	μ ≤ 2	−
Low ductility	2 < μ ≤ 4	μ ≤ 4
Moderate ductility	4 < μ ≤ 6	4 < μ ≤ 6
High ductility	μ > 6	μ > 6

Bamboo scrimber is high ductility (μ > 6) when subjected to compression perpendicular-to-fiber, medium ductility (4 < μ ≤ 6) when subjected to compression parallel-to-fiber, and low ductility (brittle) when subjected to bending, shear, or tensile parallel-to-fiber. The higher bamboo scrimber’s ductility may give an alert and more time before failure than timbers. Timbers have brittle to lower ductility when receiving each kind of loading scheme.

3.5. Relationship between Ductility and Ratio of Stiffness to Strength

The correlations between the parameters describing the ratio of stiffness to strength with ductility are illustrated in Figure 13. The determination coefficients of linear regression (R²) assumed values vary from 0.0011 to 0.2933 for timber specimens and 0.0017 to 0.7007 for scrimber specimens. The low value of R² in the correlation between E_b/σ_b and E_c⊥/σ_cp⊥ with ductility ratio for both timber and scrimber means that the ratio of stiffness to strength, in this case for E_b/σ_b and E_c⊥/σ_cp⊥, is independent of the ductility. While the ratio of shear modulus to shear strength (G/τ_//) and compression modulus to strength modulus parallel-to-fiber (E_c∥/σ_c∥) significantly correlate to its ductility.

As mentioned above, red meranti has the highest E_b and σ_b value among all tested specimens. Red meranti also has the highest E_b/S_R value. Figure 13a shows that the ductility ratio of bamboo scrimber and timber has a positive correlation to E_b/S_R value. In compression, both parallel and perpendicular-to-fiber, the ratio between compression modulus and compression strength is negatively correlated to the ductility ratio of timber specimens, while in bamboo scrimber, vice versa. This kind of behavior also can be seen in the ratio between shear modulus and shear strength (G/τ_//), which is negatively correlated to the ductility ratio of bamboo scrimber specimens. On the contrary, the correlation between timber’s (G/τ_//) and ductility ratio is positive.

4. Conclusions

Bamboo scrimber has a higher density than timber. Its drier equilibrium moisture contents indicate that scrimber is more hydrophobic than timbers. The maximum crushing strength (σ_c//), compressive stress perpendicular-to-fiber at the proportional limit (σ_cp⊥) and at the 0.04” deformation (σ_c0.04⊥), shear strength (τ_//), compressive longitudinal modulus of elasticity (Ec_//), compressive lateral modulus of elasticity (E_c⊥), and modulus of rigidity (G) of scrimber are higher than those of timbers. Both scrimber’s and timber’s flexural properties (modulus of rupture (σ_b) and flexural modulus of elasticity (E_b)) are comparable. On the contrary, the tensile strength parallel-to-fiber (σ_t//) of scrimber is weaker than that of timber.

Bamboo scrimber is high ductility (μ > 6) when subjected to compression perpendicular-to-fiber, medium ductility (4 < μ ≤ 6) when subjected to compression parallel-to-fiber, and low ductility (brittle) when subjected to bending, shear, or tensile parallel-to-fiber. The higher scrimber’s ductility may give an alert and more time before failure than timbers. Timbers have brittle to low ductility when receiving each kind of loading scheme. The ratio of shear modulus to shear strength (G/τ_//) and compression modulus to strength modulus parallel-to-fiber (E_c∥/σ_C∥) strongly correlate with the ductility ratio. However, the ratio of elasticity to rupture modulus (E_b/σ_b) and the ratio of compression perpendicular-to-fiber modulus to strength (E_c⊥/σ_cp⊥) are independent of the ductility value.