Cross-laminated thick larch veneers with groove-and-tenon splicing: how joint position and layer layout drive panel strength

Introduction

Small-diameter, fast-growing logs are abundant in plantation forestry but are awkward to turn into large structural elements. Sawmills and panel plants face a familiar problem: a log of 300 mm diameter yields short, narrow pieces, while construction markets want long, wide and stiff panels. A team of eight authors from the Research Institute of Wood Industry of the Chinese Academy of Forestry, the Beijing Key Laboratory of Wood Science and Engineering at Beijing Forestry University, and the company Penglai Zhengtai Wood Industry Co., Ltd. (Yantai) reports a possible route in a paper published in the journal Forests in January 2025.

The paper introduces a new engineered wood product, which the authors call cross-laminated-thick veneers (CLTV). Thick rotary-cut veneers of larch are spliced lengthwise with a groove-and-tenon joint (described as infinite splicing technology) and then laminated crosswise, in the manner of cross-laminated timber. The stated aims are to develop the splicing method, to produce large panels from small-diameter, fast-growing larch, and to find out how the position of the splice, the orientation of the grain in each layer and the combination of veneer quality levels affect compressive and bending properties.

For the portal's readers the paper is relevant for two reasons: it shows how the radial position in a log is used as a sorting criterion for veneer stiffness, and it demonstrates one possible way of making use of small logs. It is a laboratory study with small test panels; the numbers below are the authors' own and the portal has not reproduced them.

Research methodology

Raw material and adhesive. The logs were Gmelin larch (Larix gmelinii) of about 300 mm diameter at breast height from Hohhot, Inner Mongolia. A polyurethane adhesive supplied by BASF SE (viscosity about 4500 MPa s at 25 degrees Celsius, pH 5 to 6) was used for bonding.

Veneer preparation. Veneers of 7.5 mm thickness were rotary cut at an initial moisture content of about 20 percent. They were then flattened and stress-relieved by hot, humid pressing with superheated steam. The veneers were sorted into three levels according to their radial position in the log: Level 1 close to the heartwood (density 0.51 g/cm3), Level 2 from the middle (0.49 g/cm3) and Level 3 from the outer sapwood (0.45 g/cm3). Three-point bending tests on veneer strips (300 x 50 x 7.5 mm, 50 per level, according to the Chinese standard GB/T 17657-2022) gave the following results:

The MOR of Level 2 is given as 62.82 MPa in the table but as 62.62 MPa in the text. Because Level 2 showed large scatter (standard deviation of its MOR was 19.42 MPa), only Levels 1 and 3 were used for the panels.

Panel fabrication. Five-layer panels of 600 x 300 mm were made, with 300 g/m2 polyurethane, an open assembly time of 40 to 50 minutes and cold pressing at 1 MPa for 3 hours. In total 19 CLTV types were produced. Two splice patterns were used: pattern A with one central joint, and pattern B with joints 150 mm from each edge. Layers could be oriented parallel (written as //) or perpendicular (written as a perpendicular sign) to the panel length. The authors also describe a large panel of 12 m x 3.25 m x 105 mm made from 1.2 m x 3.25 m veneers spliced into the perpendicular layers, with 3.25 m square veneers forming the parallel layers.

Testing. Compression was tested on specimens of 35 x 37 x 140 mm according to GB/T 1927.11-2022, and bending on specimens of 50 x 37 x 600 mm in three-point bending according to GB/T 17657-2022, with six specimens per condition. Statistics were done in SPSS with one-way ANOVA, and transverse sections were examined with an ultra-depth-of-field microscope at 900 times magnification.

Key findings

Compression. Compressive strength rose with the number of layers whose grain runs parallel to the load: 13.01, 20.11, 26.93 and 31.85 MPa for one to four parallel layers, which are increases of 54.50, 106.94 and 144.73 percent relative to a single parallel layer. Using the better veneer also helped. A panel made entirely of Level 1 veneer (CLTV-5I) reached 26.67 and 20.08 MPa in the two compression directions reported, which was 7.98 and 14.09 percent above the same construction made entirely of Level 3 veneer (24.70 and 17.60 MPa). The coefficient of variation dropped as more Level 1 layers were used (for example 10.33 and 8.77 percent for the all-Level-3 five-layer type, 6.89 and 5.28 percent for a three-layer Level 3 type). Splice position did not significantly affect compressive strength under the tested specimen size.

Bending, splice position. In bending, the position of the joints mattered greatly. With Level 1 veneer in an orthogonal layout, the authors report that moving from central joints in all layers (AAA) to dispersed joints with joints at the surface layers (BAB) raised the bending strength (MOR) in steps of 161.44 percent (AAA to BBB), 109.36 percent (BBB to ABA) and 46.61 percent (ABA to BAB). Moving joints in the perpendicular surface layers to below the load point lowered the bending stiffness (MOE) by 29.42 percent.

Bending, layer arrangement. The highest bending values were 38.96 MPa MOR and 7,286.31 MPa MOE for a layout with four parallel layers in the pattern parallel-parallel-perpendicular-parallel-parallel. The lowest were 3.52 MPa MOR and 478.27 MPa MOE for a layout with one parallel layer. Compared with a layout of three perpendicular layers in the middle, the four-parallel-layer pattern gave 284.33 percent higher MOR and 49.02 percent higher MOE. Placing the parallel layers on the outer faces was better than concentrating them in the core: a panel with parallel layers in the middle (CLTV-14) was 1.39 times lower in MOR and 1.46 times lower in MOE than a panel (CLTV-5) with the same number of parallel layers on the outside.

Veneer quality level. Changing the layers of a particular layout from Level 1 to Level 3 veneer reduced MOR by 52.22 percent, with a smaller effect on MOE; in another layout (BABAB) the effect of Level 3 layers was slight and not significant. The authors report that the veneer-level factors interacted: combined design choices could raise MOR by up to 2.03 times and MOE by 1.46 times.

Stiffness gap. The measured MOE of the finished CLTV panels (averages of 5,376.79 and 5,152.79 MPa for two Level 1 and Level 3 configurations) was well below the MOE of the individual veneers (5,908 to 9,779 MPa), which the authors treat as a reason for further optimisation.

Standard comparison and large panel. The authors compare one CLTV type (CLTV-4) with a Chinese forestry standard for five-layer panels of grade TDT9, citing a requirement of 5,000 MPa MOE and 15.1 MPa MOR. The text read by the portal does not give a pass or fail statement for that comparison. The 12 m x 3.25 m x 105 mm panel is described as having been fabricated, but no mechanical test results for it were reported in the text.

Practical applications for sawmills

Portal interpretation (these are not claims made by the paper):

• Radial position as a sorting rule. The veneer results show a clear stiffness gradient from near the heartwood (MOE about 9,800 MPa) to the outer sapwood (about 5,900 MPa). Operations that peel or saw larch could keep track of where in the log each piece comes from, and the paper's results suggest a use for each grade: stiffer veneer where bending demands are high, lower grades where they are not.
• A use for small-diameter logs. Plants that currently receive small logs of about 300 mm could regard lengthwise splicing as a way of getting around length and width limits. The paper demonstrates the concept on 600 x 300 mm test panels and describes a large panel, but gives no throughput, yield or cost figures, so no business case can be derived.
• Joint location is a design variable. Splice position affected bending performance much more than compression. A producer of spliced products would need to treat the joint layout as part of the product specification, not as a manufacturing detail.
• Layer layout matters more than raw material in some cases. Placing parallel layers at the faces gave better bending values than concentrating them at the core, with the same amount of wood. Product development could therefore focus on layout before changing the raw material.
• Compare with sawn-lumber routes. Whether CLTV competes with glulam or cross-laminated timber depends on cost, yield and equipment, none of which are covered by the study.

Limitations

The paper has no dedicated limitations section. The following are stated or evident from the text:

• Only larch was used, in five-layer panels of 19 configurations, with six specimens per condition. The statistical power for comparing so many configurations is limited.
• The measured panel MOE stayed below the veneer MOE; the authors state that optimisation is needed to reach higher values.
• The compression specimens were 140 mm thick, shorter than the 150 mm veneers, so splice effects could not be evaluated in compression.
• No mechanical test results are reported for the large 12 m x 3.25 m panel.
• The authors mention future work on the ratio between overall panel size and splice position.
• The text has a few inconsistencies, such as the Level 2 MOR value (62.62 MPa in the text versus 62.82 MPa in the table), and the mapping of specimens to configurations in the tables relies on the text, which is not always consistent.
• Figures 1 to 14 (load-displacement curves, failure modes, microscope images, the bending trend chart and the photographs of the large panel) could not be read, and the supplementary tables S1 to S5 with the ANOVA results were not available. Statements that rely on them are not verified here.
• Some cited references seem not to support the statements to which they are attached in the introduction. This does not affect the experimental results, but a reader should check the references.

Conclusions

The study shows, in a laboratory setting, that cross-laminating spliced thick larch veneers can produce panels whose compressive and bending properties depend strongly on how many layers run parallel to the load, where the joints are placed and which veneer quality level is used. Compressive strength rose by up to about 145 percent with more parallel layers, and bending performance could be raised by up to about 2 times in MOR and 1.5 times in MOE through combined design choices. The authors consider continuous large-scale production by groove-and-tenon splicing feasible and present CLTV as a low-carbon way of using small-diameter timber. The panel stiffness reached so far is below the stiffness of the veneers, and results for the large panel are missing, so the concept remains at an early stage.

Original research

• Authors: Yuxin Yang, Juan Hu, Xinguang Ning, Yahui Zhang, Yingqi He, Yingchun Gong, Wenji Yu, Yuxiang Huang
• Title: Structure Effects on Mechanical Properties of a Novel Engineered Wood Product: Cross-Laminated-Thick Veneers Based on Infinite Splicing Technology
• Journal: Forests, volume 16, article 181
• Year: 2025 (published 19 January 2025)
• DOI: 10.3390/f16010181
• Affiliations: Research Institute of Wood Industry, Chinese Academy of Forestry, Beijing; Beijing Key Laboratory of Wood Science and Engineering, Beijing Forestry University; Penglai Zhengtai Wood Industry Co., Ltd., Yantai, China
• Licence: open access, CC BY 4.0, as stated in the article PDF

This article is an independent summary prepared by the portal; the original paper should be consulted for full methods, tables and figures.