1 Introduction

Cross-laminated timber (CLT) is an engineered wood product that is made by gluing and nailing three or more layers of lumber or wood-based materials together at angles of 0° and 90° in accordance with ISO-16696-1 (2019). It is one of the most popular engineered wood panels of the 21st century (Li et al. 2020a). CLT was first developed in Europe and has been used in various countries including Australia, China, Japan, the Republic of Korea, and the United States (Davids et al. 2017; Navaratnam et al. 2020; Komatsu et al. 2021; Pang et al. 2021; Choi et al. 2021; Li et al. 2021a, b, c; Sikora et al. 2016). Various studies on CLTs have been conducted using local wood species, coniferous trees, broadleaf trees, and a mixture of wood species (Navaratnam et al. 2020; Li et al. 2020a, 2021a, b, c; Sikora et al. 2016; Sharifnia and Hindman 2017; Liao et al. 2017; Mohd Yusof et al. 2019; Aicher et al. 2016; Lu et al. 2019; Sciomenta et al. 2020). Various types of CLTs have been developed using local woods to reduce the cost of materials. In terms of general local lumber usage, the value of local lumber resources has been increased by CLT development (Li et al. 2021a). Additionally, suitable CLTs for various regions have been developed to reduce carbon emissions and costs associated with the long-distance transport of raw materials (Li et al. 2019).

Various types of CLTs have been developed to improve planar or rolling shear failure (RF), bending moments under out-of-plane load, and local lumber utilization (Wang et al. 2015). Among these types, more studies have been conducted on hybrid CLTs that utilize wood-based materials as the layer material (Li et al. 2020a; Choi et al. 2018, 2021; Wang et al. 2015, 2017; Davids et al. 2017; Gong et al. 2019; Song et al. 2022; Fujimoto et al. 2021; Pang et al. 2019). Laminated strand lumber (LSL) (Wang et al. 2015; Davids et al. 2017; Gong et al. 2019), laminated veneer lumber (LVL) (Wang et al. 2017), and plywood (Sikora et al. 2016; Choi et al. 2018; Fujimoto et al. 2021; Pang et al. 2019) have been used as wood-based materials. Additionally, hybrid CLTs with LSLs have been developed by researchers at the University of New Brunswick’s Wood Science and Technology Center in Canada (Wang et al. 2015). Meanwhile, Wang et al. (2015) have studied the mechanical properties of CLT with LSLs as the middle and outer layers. They showed that a CLT with LSL showed modulus of elasticity (MOE) and modulus of rupture (MOR) values that were 1.2–1.8 GPa and 8.5–12.8 MPa higher than those of conventional CLTs, respectively (Wang et al. 2015). To improve the bending properties, Wang et al. (2017) studied the mechanical properties of a CLT obtained with LVL as the middle and outer layers. When the outer layer was LVL, the MOE and MOR values increased by 1391 and 1.5 MPa respectively, compared to CLT. Fujimoto et al. (2021) investigated the bending and shear characteristics of a hybrid CLT, made with Korean larch plywood and Japanese cypress lumber. When plywood was used in place of wood, the MOE, MOR, and rolling shear strength values of the hybrid CLT were 49.0 MPa, 12.3 GPa, and 3.87 MPa, respectively. Song and Hong (2018) studied the bending properties of a general CLT with larch lumber. In the studies of Fujimoto et al. (2021) and Song and Hong (2018), plywood was found to increase the MOE and MOR values of a CLT by 0.8–1.7 GPa and 1.9–5.1 MPa, respectively. Choi et al. (2020) created ply-lam CLT with plywood layer materials and demonstrated that the CLT layer significantly improved the bending and shear properties of the studied hybrid CLT (Li et al. 2020a).

The number of studies comparing the bending, shear, and in-plane shear properties of engineered wood products (e.g., glued laminated timber (GLT) and CLT) has increased. Li et al. (2021b) investigated the bending characteristics of a CLT combined with a three-ply GLT, CLT, and bamboo-wood composite (B), which was fabricated from hem-fir lumber. The MOR values were 58.8, 47.3, 45.0, and 31.3 MPa for GLT, CLT, LBL, and BLB, respectively. The out-of-plane bending properties, shear characteristics, and joint performance of CLTs have also gained significant attention from researchers (Sikora et al. 2016; Shahnewaz et al. 2017; Wei et al. 2019). In this study, we compared the mechanical properties of ply-lam CLTs, which were composed of plywood as a layer material, with GLT and CLT. The bending properties of ply-lam CLTs predicted by beam theory were compared with the experimental values. The bending properties of the pilot-scale ply-lam CLT were also evaluated.

2 Materials and methods

2.1 Materials

Larch lumber (Japanese larch, Larix kaempferi (Lamb.) Carriere) of E9 and E10 grades and larch plywood made in the Republic of Korea were used as layers. Layer materials of visual grade 1 and mechanical grades E9 and E10 were used. The height (h), width (w), and length (l) dimensions of the lumber were 27, 140, and 2700 mm, respectively, and the moisture content of the layer material was 10.8 ± 1.4%. Two types of layer materials (E9, E10) were processed into specimens with h, w, and l dimensions of 18, 120, and 2400 mm, respectively. Larch plywood satisfying the structural first grade of the KS F 3113 standard (2013) was used. The h, w, and l dimensions of the larch plywood were 18.3, 1220, and 2440 mm, respectively. The moisture content and density values of the larch plywood were 8% and 0.57 g/cm3, respectively. Phenol resorcinol formaldehyde resin (PRF) (D-40) from Oshika Corporation (Tokyo, Japan) was used as an adhesive. The lumber layer of CLT and ply-lam CTLs was manufactured into a panel with a width of 1220 mm and a width direction butt joint using PRF adhesive.

2.2 Fabrication of GLT, CLT, and ply-lam CLTs

The basic properties and layer compositions in GLT, CLT, and ply-lam CLTs are summarized in Table 1. Representative photos of the samples are shown in Fig. 1. To simplify the material yield and manufacturing process, the GLT was manufactured in four layers. The larch lumber was prepared as a glued laminated board using PRF adhesive. Considering a single-sided application, PRF at 150 g/cm2 was used to prepare the glued laminated board and plywood. After applying the PRF, the material was cold pressed with a pressure of 10 kgf/cm2 that was maintained for 24 h. The prepared samples were then subjected to decompression curing for 7 days to obtain the GLT, CLT, and ply-lam CLTs. Twenty specimens of the four-ply GLT with h, w, and l dimensions of 90, 70, and 2400 mm, respectively, were prepared. CLT and ply-lam CLT specimens with h, w, and l dimensions of 90, 1200, and 2400 mm, respectively, were prepared with two panels per condition. The oven-dry density and moisture content values of the prepared specimens were measured as outlined in ASTM D2395 (2017). The measured values are summarized in Table 1. The specimens were cut as shown in Table 2, and their mechanical properties were evaluated.

Table 1 Basic properties and layer compositions of the GLT, CLT, and ply-lam CLTs
Full size table
Fig. 1
figure 1

Representative photographs of tested GLT, CLT, and ply-lam CLTs

Full size image
Table 2 Measurements and number of specimens used in the experiments
Full size table

2.3 Static tests

2.3.1 Four-point bending test

A four-point load bending test was performed on a test specimen in accordance with ASTM D 198 (Fig. 2) (2014). The lengths of the out-of- and in-plane bending specimens were 1890 and 1470 mm (span-to-depth ratio = 21), respectively. A displacement gage was installed between the load spans. Bending tests were performed using a 10-ton universal testing machine (KDPI-130-1 model) from Seoul, Republic of Korea. The loading rate was set such that failure occurred within 5 min. The MOE (GPa) and MOR (MPa) values were calculated using Eqs. (1) and (2), respectively.

Fig. 2
figure 2

Schematic of the four-point bending test. (h, thickness of specimen)

Full size image
$${\text{MOE}} = \frac{{23Pl^{3} }}{{108bh^{3} \Delta }}$$
(1)
$$MOR = \frac{{P_{{\max }} l}}{{bh^{2} }}$$
(2)

l, \(\varDelta\), b, h, and Pmax are the test span (mm), linear slope of the load–displacement response at 10–40% of the maximum load, specimen width (mm), specimen height (mm), and maximum load (kN), respectively.

2.3.2 Shear test

To compare the shear strength of the prepared GLT, CLT, and ply-lam CLTs, we conducted out-of- and in-plane shear tests with center point loading according to the ANSI/APA PRG 320 standard (2019). The lengths of the out-of- and in-plane shear specimens were 450 and 350 mm, respectively, with a span-to-depth ratio of 5. The spans of the specimens were separated by a displacement gage, and a 10-ton universal testing machine was used for the shear tests (Fig. 3). The loading rate was chosen so that failure would occur within 5 min, and the shear strength of a specimen (fs) (MPa) was calculated using Eq. (3).

$${f}_{s}= \frac{3{P}_{max}}{4bh}$$
(3)

Pmax, b, and h are the maximum load (N), specimen width (mm), and specimen height (mm), respectively.

Fig. 3
figure 3

Schematic of the center-point shear test (h: thickness of specimen)

Full size image

2.3.3 Compressive strength test

The compressive strength of a specimen was tested according to ASTM D 198 (2014) using a 100-ton hydraulic compression strength tester from Dong ah test machine (Seoul, Republic of Korea) (Shin et al. 2011). The h, w, and l dimensions of a specimen were 90, 70, and 1200 mm, respectively, and the slenderness ratio (l/r; l-length of the column, r-the radius of gyration of the cross section) was ≥ 17. The loading rate was set to induce a failure within 5 min. The compression strength (fc) was calculated using Eq. (4), which considers the cross-sectional area (A) (mm2) and maximum load (Pmax) (N) used for a test specimen.

$${f}_{c}= \frac{{P}_{max}}{A}$$
(4)

2.3.4 Statistical analysis

The significance of the mechanical properties of the GLT, CLT, and ply-lam CLTs was verified using the SPSS 26 (V26.0.0) statistical software suite from IBM (NY, USA). The significance of a one-way analysis of variance (ANOVA) was determined using Duncan’s multiple test. Statistical significance was set at p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***). According to the ANOVA results, the mean values of the groups denoted by the same superscript letter (i.e., a, b, or c) did not differ significantly at the 0.05 significance level (Srivaro et al. 2022).

2.4 Theoretical bending properties of ply-lam CLTs

The effective bending stiffness (EIeff,s) (N/mm2) values of a CLT were calculated using Eqs. (5), (6), and (8), which are based on the shear analogy method in the Cross-Laminated Timber Handbook (2011), the composite theory method (K-method) used in the plywood industry (Fellmoser, and Blaβ 2004), and the gamma method in accordance with Eurocode 5 (2008), respectively (Li et al. 2020). The mechanical properties of the lumber and plywood samples, which were used to predict the bending properties of the CLT, are summarized in Table 3.

Table 3 Properties of larch lumber and plywood used for design strength
Full size table
$${EI}_{eff,s}= {B}_{A}+ {B}_{B}= \sum _{i=1}^{n}({E}_{i}{b}_{i}\frac{{{h}_{i}}^{3}}{12})+ \sum _{i=1}^{n}\left({E}_{i}{b}_{i}{h}_{i}{{z}_{i}}^{2}\right)$$
(5)

In Eq. (5), Ei, bi, hi, and zi are the MOE parallel to the ith layer major or minor strength direction (MPa), width (mm), height (mm), and distance from the centroid of the ith layer to the centroid of the cross section (mm), respectively.

$${EI}_{eff,k}= {E}_{0} {k}_{1}\frac{b{h}^{3}}{12}$$
(6)

k1 is the value of the composition factor, which is calculated using Eq. (7).

$${k}_{1}=1-\left[\left(1-\frac{{E}_{90}}{{E}_{0}}\right)\left(\frac{{a}_{m-2}^{3} -{a}_{m-4}^{3}+ \cdots \pm {a}_{1}^{3}}{{a}_{m}^{3}}\right)\right]$$
(7)

In Eq. (7), b, E0, h, and E90 are the width (mm), MOE parallel to the grain in the longitudinal layer (MPa)\(,\) height (mm)\(,\)and MOE perpendicular to the grain in the cross layer (\({E}_{90}= {E}_{0}/30)\) (MPa). \({a}_{m}, {a}_{m-2}, {a}_{m-4}\) is the layer thickness (m = total number of layers) (mm) (Fig. 4).

Fig. 4
figure 4

Schematic of the layer thickness (Fellmoser, and Blaβ 2004)

Full size image
$$EI_{{eff,G}} = ~\mathop \sum \nolimits_{{i = 1}}^{n} \left(E_{i} b_{i} \frac{{h_{i} ^{3} }}{{12}}\right) + ~\mathop \sum \nolimits_{{i = 1}}^{n} \left( {r_{i} E_{i} b_{i} h_{i} z_{i} ^{2} } \right)$$
(8)
$$r_{i} = \left( {1 + \frac{{\pi ^{2} E_{i} b_{i} h_{i} }}{{L_{{eff}}^{2} \frac{{\left( {G_{{90,j}} b_{j} } \right)}}{{h_{j} }}}}} \right)^{{ - 1}}$$
(9)

\({E}_{i}\), \({b}_{i}\), \({h}_{i}\), \({z}_{i}\), and \({r}_{i}\) are the MOE of the ith layer that is parallel to grain \({E}_{0}\) for the longitudinal layers and equal to unity for transverse layers (MPa), width (mm), height (mm), distance from the centroid of the ith layer to the centroid of the cross section (mm), and connection efficiency factor that is nonzero only for the longitudinal layer and equal to unity for the middle layer, respectively. \({L}_{eff}\), \({h}_{j}\), and \({G}_{80}\) are the effective length of the beam (mm), transverse layer connection to the central layer (mm), shear modulus in the plane perpendicular to the grain and rolling shear modulus (MPa) (= 0.4 MPa (Kim et al. 2022)), respectively.

The bending stiffness values based on the three theories were calculated in terms of MOE and MOR using Eqs. (10) and (11), respectively.

$$\text{M}\text{O}\text{E}= \frac{12{EI}_{eff}}{b{h}^{3}}$$
(10)
$$\text{M}\text{O}\text{R}=2.1 \left(\frac{6{\left({F}_{b}\frac{2{EI}_{eff}}{{E}_{1}h}\right)}_{eff}}{b{h}^{2}}\right)$$
(11)

Additionally, we intend to choose one of the three theories that can predict bending properties while considering the lumber and plywood composition of a given sample.

2.5 Pilot-scale ply-lam CLT

To compare the predicted and experimental bending properties of ply-lam CLTs that measured ≥ 3 m, including longitudinal plywood joints, pilot-scale fabrication of ply-lam CLT was conducted. The bending properties of the prepared ply-lam CLT were set at 10 and 30 MPa based on the bending properties of GLT, which is the most used engineered wood in the Republic of Korea (Li et al. 2022). Based on the predicted bending properties, mechanical-grade E12 lumber and structural first-grade plywood were the selected constituent materials. The h × w × l dimensions of the lumber and structural plywood samples were 30 × 140 × 3400 mm and 24 × 1220 × 2440 mm, respectively. The lumber was cut into specimens with h, w, and l dimensions of 24, 140, and 3198 mm, respectively, by surface grinding and length cutting. The lumber and plywood for ply-lam CLTs were prepared in the same way as that for the P-C. To connect plywood longitudinally, we selected the butt joint over half the lap and tongue and groove joints. The butt joint simplifies the fabrication process while retaining the bonding method-dependent bending properties of larch ply-lam CLT (P-A type) (Liao et al. 2017). Niederwestberg et al. (2018) evaluated the bending properties based on the longitudinal LSL position of a joint; the moment, resistance and bending stiffness of a joint were reduced by 15–23 kNm/m and 68–192 × 109 Nmm2/m, respectively, compared with joint-free specimens; they configured joint placements in plywood samples as shown in Fig. 5 (Niederwestberg et al. 2018). The h, w, and l dimensions of the P-C were 120, 1220, and 3198 mm, respectively (span-to-depth ratio = 16.7). The prepared specimen was subjected to a four-point load bending test as outlined in Sect. 2.3.1. The displacement gage (Fig. 2) was installed to measure the strain across the length of the specimen.

Fig. 5
figure 5

Butt joint position in the pilot-scale specimen

Full size image

3 Results and discussion

3.1 Bending test results

3.1.1 Bending properties of GLT, CLT, and ply-lam CLTs by static testing

The MOE and MOR values obtained for the GLT, CLT, and ply-lam CLTs are summarized in Table 4. The changes in the out-of- and in-plane MOE values with various loading directions were dependent on the type of material. The CLT and ply-lam CLTs exhibited significant out-of-plane MOE values, whereas the GLT showed a high in-plane MOE value, which is similar to the value reported by Li et al. (2022). The plywood used in CLTs and ply-lam CLTs showed a low in-plane MOE value because of the load along the cross section of the lumber. Compared with the value obtained for the GLT, the out-of-plane MOE values (F*** = 9.019) for the three ply-lam CLTs (i.e., P-A, P-B, and P-C) were improved by 11.5–29.5%. However, the MOE value of the P-D was 8.2% lower than that of the GLT. Additionally, the MOE values obtained for the three ply-lam CLTs were 11.6–23.4% higher than those for CLT. When larch lumber was used as the outer layer of ply-lam CLTs, higher MOE values were obtained, particularly for P-A and P-C.

Wang et al. (2015) studied the bending properties of CLT using an LSL and pine lumber. When the second layer was changed to LSL, the MOE and MOR increased by 1,248 MPa and 8.58 MPa, respectively. Comparing P-A and CLT, the MOE and MOR increased by 19.1% and 24.3%, respectively. When LSL and plywood were applied as component materials of CLT, the increase in MOE and MOR in the fiber direction was consistent.

The in-plane MOE value obtained (F*** = 7.803) of GLT was the highest, followed by that of ply-lam CLTs and CLT. The average MOE values obtained for the GLT, ply-lam CLTs, and CLT were 13.5, 10.9–12.1, and 8.9 GPa, respectively. There were no significant differences in the MOE values found among the ply-lam CLTs in the significance test at the 5% level (F = 1.566, p = 0.229).

Table 4 Bending properties of the GLT, CLT, and ply-lam CLTs.
Full size table

The GLT and CLT showed higher out-of-plane MOR values than the ply-lam CLTs. The MOR values among the ply-lam CLTs were different, and the out-of-plane MOR value was high when lumber was used as the outer layer. Meanwhile, the in-plane MOR value was high when plywood was used as the outer layer. The out-of-plane MOR value (F*** = 11.957, p = 0.000) of GLT was the highest, followed by those of ply-lam CLTs and CLT. The trend for the out-of-plane MOR at the 5% lower limit was GLT (34.4 MPa) > P-B (32.4 MPa) > P-A (32.2 MPa) > P-C (31.6 MPa) > P-D (29.2 MPa) > CLT (26.1 MPa), which was similar to the average MOR values.

The in-plane MOR value (F*** = 9.99, p = 0.000) of GLT was the highest, followed by those of ply-lam CLTs and CLT. The significance test across the studied ply-lam CLTs exhibited no differences in their in-plane MOR values at the same level (F = 0.056, p = 0.982). The in-plane MOR values at the 5% lower limit for P-A, P-B, P-D, P-C, GLT, and CLT were 34.9, 33.9, 32.4, 31.8, 43.2, and 21.7 MPa, respectively. The difference in the in-plane MOR values for the studied ply-lam CLTs was 3.1 MPa. When plywood was utilized as a layer, the homogeneity of the material was guaranteed because the coefficients of variation for the ply-lam CLTs were lower than those for GLT and CLT.

Based on the MOE and MOR values, we believe that the MOE and MOR values obtained for P-A and P-C can be improved. Additionally, if plywood is used as the outer layer (e.g., P-B and P-D), we predict uniform bending characteristics. Specifically, the MOE and MOR values of P-D were similar or greater than those of regular CLT. Jang and Lee (2019) maximized the use of small larch logs by designing a CLT panel that consisted of bonded plywood and glued laminated board made from small-diameter lumber as the outer and inner layers, respectively. They concluded that the P-D method is acceptable for the fabrication of structural panels from low-mechanical-grade lumber and small-diameter logs (Jang and Lee 2019).

Niederwestberg et al. (2018) compared the bending properties of LSL and lumber (L) as CLT (5-C1), 5-B2 (LSL∥-L∥-LSL∥-L∥-LSL∥), and 5-A1 (L∥-LSL∥-L∥-LSL∥-L∥). The bending moment and stiffness of 5-B2 were the highest, followed by those of 5-A1 and 5-C1. In this study, the bending properties of P-C were the highest, followed by those of P-B and CLT. This is judged to be the difference between plywood and LSL applied as a layer. LSL is a composite structural lumber consisting of oriented wood stands (Wang et al. 2015), and plywood is glued together with adjacent layers with their wood grains rotated up to 90° to one another (Shreeranga et al. 2017); thus, plywood has lower bending properties than LSL. Accordingly, it is judged that when plywood is used as the outer layer, it acts as a factor impairing the bending properties.

The failure modes were analyzed after the out-of- and in-plane bending tests, and the results are shown in Fig. 6. The out-of-plane failure modes were classified into tensile, tensile and shear, and shear to highlight the differences in the modes depending on the compositions of the specimens. All the GLT specimens failed during the bending tests, and the CLT, P-A, and P-C specimens exhibited tensile and shear failure. The P-B specimen exhibited tensile and shear and shear failure values of 13 and 2 ea, respectively. When plywood was used as the outer layer, RF occurred between the veneers, which were arranged at the intersection that inhibited strength. The P-D specimen showed tensile, tensile and shear, and shear failure values of 9, 4 and 2 ea, respectively. When lumber was used continuously across two to four layers (P-D type), the tensile failure mode of the P-D specimen was identical to that of the GLT specimen.

Fig. 6
figure 6

Failure modes of the GLT, CLT, and ply-lam CLTs in the bending test

Full size image

The changes observed in a failure mode were closely related to the MOE and MOR values. We found that the ratio of shear failures existing in the material influenced the degradation rate of the bending qualities. Therefore, we concluded that the use of a layered material as the outer layer can ensure that the specimen would have better bending properties than plywood. When plywood was used as the outer layer to enable the use of low-mechanical-grade larch lumber, the specimen showed uniform bending properties. However, to avoid failure under shear of the outermost layer, it is necessary to consider both the thickness and shear properties of the layer.

3.1.2 Comparison of bending properties

The calculated MOE and MOR values obtained from the three beam theories were compared with MOE values from the out-of-plane bending test and MOR values at the 5% lower limit, respectively. These are presented in Figs. 7 and 8. The average MOE values were higher than the corresponding values predicted using the theories. Among the calculated MOE values, the values for MOEG were 15–30% lower than those for MOEC and MOES. The MOE values obtained for the P-C and P-D specimens and the calculated MOEC and MOES differed by 0.03 GPa. The MOE values by the k-method were identical, whereas the values by the shear analogy method were different.

Fig. 7
figure 7

Comparison of the design or calculated MOE values and the average MOE values obtained using the static bending test. (MOEa = average value of MOE by static test, MOEG = MOE value by gamma method, MOEC = MOE value by k-method, and MOES = MOE value by shear analogy method)

Full size image
Fig. 8
figure 8

Comparison of the MOR values obtained at 5% lower limit and the design or calculated MOR values. (MOR5% = MOR by static test at 5% lower limit, MORG = MOR value by gamma method, MORC = MOR value by k-method, and MORS = MOR value by shear analogy method)

Full size image

The MOR and MOE values exhibited similar trends. Based on beam theory, the MOE and MOR values were in the range of 1.3–2.4 GPa and 0.6–3.2 MPa, respectively, when the third layer in all specimens was changed. The type of layer showed a negligible impact on the bending ability of the specimens. Additionally, the calculated MOE and MOR values were lower than the 5% lower limit; thus, the three beam theories may be applied from a conservative view. The shear analogy method was most suitable for the ply-lam CLTs because it was able to verify any changes in the predicted value, which corresponded to changes in the composition of the inner layer.

3.2 Shear test results

The shear strengths of the GLT, CLT, and ply-lam CLT specimens were compared, and the results are shown in Fig. 9. Both the out-of- and in-plane shear strengths were GLT > ply-lam CLTs > CLT. The average values of the out-of- and in-plane shear strengths for GLT were comparable. However, the in-plane shear strength for the ply-lam CLTs was ~ 1.3 times higher than the out-of-plane shear strength. The out-of-plane shear strength (F*** = 31.234, p = 0.000) of GLT was the highest, followed by that of ply-lam CLTs and CLT. The order of strength was P-D > P-A = P-B = P-C. The result of the significance test for P-A, P-B, and P-C, which was performed at an equivalent level (F = 2.603, p = 0.086), indicated no significant differences at the 5% significance level. We confirmed the improved shear strength of the P-D sample with a continuous arrangement of lumber.

Niederwestberg et al. (2018) compared the out-of-plane shear strength of LSL and lumber as CLT (5-C1), 5-B2, and 5-A1. The shear strength results of this study are consistent with those reported by Niederwestberg et al. (2018). The shear strength of 5-B2 (3.6 MPa) was the highest, followed by 5-A1 (2.9 MPa) and CLT (5-C1: 1.3 MPa). The continuously arranged 5-A1 specimen showed failure at the interfacial glue bond and 5-B2 showed tensile and interfacial glue bond failure. Ply-lam CLTs showed that rolling shear failure was increased when the plywood was continuously arranged. Although there was no substantial difference in the shear strength results of P-A, P-B, and P-C, further research on shear stiffness is needed considering the type of failure mode.

Fig. 9
figure 9

Shear strength of the GLT, CLT, and ply-lam CLTs based on the compositions of the layers by the static test

Full size image

The in-plane shear strength (F***= 7.314, p = 0.000) for all the specimens except CLT was at the same level. They followed the order GLT = ply-lam CLTs > CLT. Therefore, the significance of GLT and ply-lam CLTs was verified. Consequently, there was no significant difference at the 5% significance level (F = 0.580, p = 0.68). Based on the out-of-plane and in-plane shear strength results, the shear strength of the ply-lam CLTs followed the order P-D > P-B = P-C = P-A.

We concluded that the placement of plywood and lumber in the third layer had no significant effect on the shear strength of the ply-lam CLTs. Nonetheless, we found that continuous stacking of lumber as an inner layer in the P-D was preferred for the intersectional or continuous arrangement of plywood as an inner layer that enhanced the out-of-plane shear strength.

Fig. 10
figure 10

Shear failure modes of the GLT, CLT, and ply-lam CLTs

Full size image

The shear failure modes of the specimens were investigated, and the results are shown in Fig. 10. The planar shear failure modes were divided into two types: bending failure due to shear (BF) and RF. The bending failure of the GLT and P-D specimens was completely caused by the shear failure mode. CLT (BF = 3 ea, RF = 12 ea), P-A (BF = 8 ea, RF = 7 ea), P-B (BF = 13 ea, RF = 2 ea), and P-C (BF = 8 ea, RF = 7 ea) failed because of a combination of bending failure due to BF and RF. The ply-lam CLTs showed a weaker RF mode compared with the CLT. When plywood was arranged in the outermost layer (P-B and P-D), RF was not observed in the layers. Meanwhile, when plywood was arranged in 2, 3, and 4 layers (P-A, P-B, and P-D), RF was observed. According to the in-plane shear failure mechanism, all the specimens experienced bending failure due to shear.

3.3 Compression test results

The compressive strength and failure modes of the GLT, CLT, and ply-lam CLTs are summarized in Table 6 and Fig. 11, respectively. The compression strength (F*** = 53.987, p = 0.000) of GLT was the highest, followed by those of ply-lam CLTs and CLT. The compression strength of ply-lam CLTs decreased as the plywood content in the layered composition was increased. The significance verification test on the compression strength of the ply-lam CLTs (F = 0.676, p = 0.516) indicated no difference at the 5% significance level. Therefore, we concluded that it is possible to increase the compression strength by using plywood as the third layer.

The failure modes followed the order P-A > P-B > GLT > P-D > P-C > CLT due to the 5% lower limit (Table 5). The coefficients of variation (COV) were 5.6%, 3.1%, and 10% for P-A, P-B, and GLT, respectively. As the ratio of the fiber of the raw material arranged in the perpendicular direction increases, the compressive strength decreases. Based on the average compressive strength, GLT, ply-lam CLTs, and CLT showed the highest strength in that order. On the other hand, because of comparing the 5% lower limit value, the compression strength of P-A and P-B was found to be higher than that of GLT. As shown in the COV results, the COV of the compression strength of ply-lam CLTs is small, ensuring the homogeneity of the material. Accordingly, the 5% lower limit value is higher than GLT. Therefore, the failure modes obtained for P-A and P-B with 5% lower compression strength were greater than those for GLT. The P-A and P-B specimens can be utilized as compression members. However, further research on their buckling and compression characteristics is required before they can be utilized as suitable building columns.

Table 5 Compression strength values for the general GLT, CLT, and ply-lam CLTs.
Full size table

All the compression failure modes of ply-lam CLT were comparable. The layer material under compression load first failed because of load concentration at knots and defects, which was followed by a transfer to the plywood, where the failure occurred along the veneer placed at the intersection. In particular, we confirmed that when the plywood was arranged in the outer layer, the delamination of the plywood caused by the compressive load was reduced.

Fig. 11
figure 11

Compression failure mode of the GLT, CLT, and ply-lam CLTs

Full size image

After comparing the bending properties, shear strength, and compression strength, we were able to confirm the better performance of ply-lam CLTs compared with CLT, indicating their suitability as structural panels. Additionally, the MOE values of the ply-lam CLTs were superior to those of GLT. The MOR values at the 5% lower limit for the ply-lam CLTs were 2–5.2 MPa lower than those for GLT. The compression strength values at the 5% lower limit for P-A and P-B were 0.5–1 MPa higher than those for GLT. Meanwhile, the P-C and P-D values were 3.6–4.2 MPa lower than those of GLT. P-C was chosen as the representative type of ply-lam CLT because it did not show yielding of the raw materials and shear failure in its outer layer.

3.4 Bending properties of the pilot-scale ply-lam CLT by static test and theoretical method

To fabricate panels with lengths ≥ 2400 mm, the plywood was bonded longitudinally. The degradation effect caused by the longitudinal joint does not take into consideration the bending properties predicted using the shear analogy method. Therefore, we compared the characteristics of the longitudinal plywood joint-equipped pilot-scale beam. The predicted bending characteristics based on the mechanical grade of the layered materials are summarized in Tables 6 and 7. The layered materials of E12 and E14 grades satisfied the established bending performance standards at 10 GPa and 30 MPa. Kim (2020) studied the adhesive strength and bending performance of CLT layers made of larch (Larix kaempferi) and pine (Pinus densiflora). A total of 650 larch lumber samples were used, and the MOE values were measured and classified into mechanical grades. Among them, the E12 and E14 grades accounted for 20% and 10%, respectively; thus, we conclude that it is appropriate to use the E12 grade layer material with a high layer material composition rate.

Table 6 Prediction of the bending properties of P-C according to mechanical grades of lumber
Full size table

The bending characteristics of the 3198-mm pilot-scale P-C are presented in Table 7. The average MOE and MOR values were 11,045 and 41.89 MPa, respectively, which are comparable with the values predicted using the shear analogy method. Thus, the shear analogy method is useful in estimating the bending properties for the configuration illustrated in Fig. 5. Additionally, the pilot-scale P-C presented the same failure mode of tension and shear.

Table 7 Bending characteristics of the pilot-scale P-C.
Full size table

4 Conclusion

We compared the mechanical properties of ply-lam CLTs, which were composed of plywood as a layer material, with GLT and CLT. The bending properties of ply-lam CLTs predicted by beam theory were compared with the experimental values. The bending properties of the pilot-scale ply-lam CLT were also evaluated. The key results are summarized as follows.

  1. (1)

    Based on the out-of- and in-plane bending properties of GLT, CLT, and ply-lam CLTs, the MOE values of the P-A, P-B, and P-C ply-lam CLTs were the highest, followed by those of GLT and CLT. The MOR values of GLT were the highest, followed by those of ply-lam CLTs and CLT. The bending properties of the ply-lam CLTs with plywood as a layer were superior to those of CLT; thus, they can be used appropriately as panels. Furthermore, the P-D type is made up of a continuous stack of layered materials; hence, it can be used to fabricate small-diameter wood and low-grade materials.

  2. (2)

    The out-of- and in-plane shear strengths of the GLT, CLT, and ply-lam CLTs were compared, and showed similar in-plane shear strengths to GLT, whereas CLT and ply-lam CLTs showed higher in-plane shear strengths than out-of-plane shear strength. The in-plane shear strengths of the ply-lam CLTs was similar to that of GLT. Ply-lam CLTs in panel form can be applied as beam members for GLT.

  3. (3)

    The compression strength at the 5% lower limit was in the following order: P-A > P-B > GLT > CLT > P-D > P-C. Two types (P-A and P-B) of ply-lam CLTs were stronger than the GLT.

  4. (4)

    From a conservative point of view, the shear analogy method is most suitable among the various methods used for predicting the bending properties of ply-lam CLTs. The bending properties (MOE = 11 GPa and MOR = 41.89 MPa) and the predicted bending performances (MOE = 10 GPa and MOR = 30 MPa) of the pilot-scale P-C specimen containing plywood joints were compared, and the bending properties obtained for all specimens passed the predicted bending performance standards.

  5. (5)

    The ply-lam CLTs showed higher strength than CLT; however, they showed poorer strength than GLT. Their strength performances were better than those of CLT. Additional studies on joint and design concepts are needed to apply ply-lam CLTs as structural materials in buildings.