Effects of Growth Ring Width, Height from Tree Base, and Loading Direction on Transverse Compression of Plantation Japanese Larch Wood

Abstract

This study aimed to investigate the effects of growth ring width, height from the tree base, and loading direction on the transverse compressive strength of Japanese larch wood, which is commonly used in wood structures in China. Plantation wood is often used to replace natural forest woods for reconstruction purposes, despite significant differences in properties (e.g., growth rings, density, strength) between them. The ends of transversely compressed wood members in such structures are prone to damage by breaking or crushing. A transverse compressive test was conducted following Chinese national standards, which revealed the following key findings. (1) There was a significant difference in the transverse compressive strength of wood with different growth ring widths (p < 0.05). The radial and slant compressive strength of wood increases with growth ring width, while the tangential compressive strength decreases as growth ring width increases. (2) The transverse compressive strength of wood decreases as the height from the tree base increases. The radial, tangential, and slant compressive strength at a lower height were 18.39%, 22.58%, and 18% higher than those at a greater height in the stem, respectively, with significant differences at the 0.05 level. (3) The load–displacement curve of Japanese larch wood under radial and slant compression follows a “three-segment” form. In contrast, the load–displacement curve of tangential compression is a continuous curve that drops sharply upon reaching its highest point. (4) There is a significant difference in the transverse compressive strength of Japanese larch wood in different loading directions when growth ring width and height from the tree base are constant (p < 0.05), which fall into order as tangential > radial > slant.

1. Introduction

Wood, as a very commonly used building material, has high practicality and excellent mechanical properties; it is widely applied throughout the construction industry [1]. Long-term load conditions have damaged many wooden components of ancient buildings. The uneven vertical settlement and horizontal displacement of wooden structures caused by transverse compression are particularly problematic [2,3].

Due to the natural anisotropy of wood, the mechanical properties of its radial, tangential, and axial parts differ substantially. The transverse compressive strength is the weakest among them [4,5,6,7]. Previous studies have shown that the radial and tangential compressive strength of wood are mostly equal or that the latter is slightly greater than the former [8,9,10,11]. Different loading directions not only affect the transverse compressive performance of wood but also have an impact on failure modes. Many recent studies have observed the macroscopic and microscopic failure characteristics of wood in both the radial and the tangential directions, in which radial compression has been shown to be mainly caused by cell collapse, while tangential compression is caused by growth ring buckling deformation and shear slippage between growth rings [4,12,13,14].

The load–displacement curves of wood under transverse compression also differ depending on the tree species and loading direction [8,9,14,15]. However, research on transverse compression has mostly centered on the performance, failure characteristics, and load–displacement curves in the radial and tangential directions, with relatively few studies on slant compression. In practice, wood members are not aways completely under transverse-grain radial and tangential compression. Slant compression failure phenomena are also common in wood structures. Therefore, studying the compression characteristics in different directions is crucial for the successful protection and repair of wood structures.

At present, the restoration of damaged wood structures is mainly based on plantation woods, the growth ring characteristics of which are significantly different from those of natural forest woods [16,17,18]. The growth ring width can be used to evaluate the physical and mechanical properties of wood [19,20,21,22,23,24]. Some studies have shown a negative correlation between the modulus of rupture (MOR) and modulus of elasticity (MOE) of wood, which can be used to predict bending performance [25,26,27]. Different positions even on the same log can also have very different physical properties [28,29,30]. To date, there are no research reports on the effects of growth ring width and tree height on the transverse compression performance of wood.

Previous studies have primarily focused on the effects of loading direction on the transverse compressive properties of wood, specifically in the radial and tangential directions. Limited research has been conducted on the effects of growth ring characteristics on the physical and mechanical properties of wood, but there have been no studies on transverse compressive strength and few on the effects of different trunk positions and heights. To address these gaps, we targeted problems with the practical application of wood structures by investigating the impact of different loading directions, growth ring widths, and heights from the tree base on the transverse compressive strength of Japanese larch wood. Our results may provide theoretical support for future wood-structure building repair techniques and material selection standards.

2. Materials and Methods

2.1. Materials

Plantation Japanese larch (species: Larix kaempferi (Lamb.) Car), produced in Hubei, China, was utilized for this study. To minimize any difference between plants, we harvested a Japanese larch with a DBH of 30 cm and breast height of 1.3 m and cut it into nine segments from breast height. Parts with diameters below 20 cm were discarded. The remaining parts of the sample had a length of 1 m, a diameter class range of 20–31 cm, an average growth ring width of 3.74 mm, an age of 25a, and an initial water content of about 50%. The log was processed into sawn timber (Figure 1) and air dried until reaching a water content of about 12%. With reference to GB/T 1927.12-2021 Test methods for physical and mechanical properties of small clear wood specimens—Part 12: Determination of strength in compression perpendicular to grain, we designated a sample size of 20 mm × 20 mm × 30 mm and measured length along the grain (Figure 2).

2.1.1. Division of Growth Ring Width Samples

To analyze the relationship between the growth ring width and the transverse compressive strength of our wood sample, we grouped samples into equal numbers and divided the test pieces according to growth ring width as GRW1 < 4.3 mm, 4.3 mm < GRW2 < 6.3 mm, GRW3 > 6.3 mm. We used 4.3 mm and 6.3 mm as grouping boundaries because they are the upper and lower limits of the narrow growth ring region and the wide growth ring region, respectively. The middle part is the wider growth ring region. Using this value as the boundary is reasonably intuitive.

2.1.2. Division of Samples by Height from Tree Base

We selected the central plate (numbered A3, A5, A7, A9) and the adjacent tangential cut plate to prepare samples at designated heights from the tree base. The sample preparation position was required to fall on the same horizontal line with spacing in the height direction of 1 m. We divided samples into four groups according to heights from the tree base (H1, H2, H3, and H4 corresponding to 3 m, 5 m, 7 m, and 9 m, respectively) to investigate the effects on transverse compressive strength.

2.1.3. Division of Samples by Loading Direction

We took 20 radial (R), tangential (T), and slant (S; 35° < α < 45°) samples from each respective group of growth ring width (GRW) samples and height from the tree base (H) samples. We observed the effect of loading direction on the transverse compressive strength of the wood by intra-group comparison. The number and grouping of samples are listed in

Table 1. Number and grouping of samples. Note: GRW1 < 4.3 mm, 4.3 mm < GRW2 < 6.3 mm, GRW3 > 6.3 mm; H1, H2, H3, and H4 indicate heights from tree base of 3 m, 5 m, 7 m, and 9 m, respectively.

Loading Direction	Growth Ring Width/mm			Height/m
	GRW1	GRW2	GRW3	H1	H2	H3	H4
Radial	20	20	20	20	20	20	20
Tangential	20	20	20	20	20	20	20
Slant	20	20	20	20	20	20	20

.

2.2. Method

Determination of Growth Ring Width and Latewood Rate

We took a GRW sample from the same log section as described above, photographed the end face of the test piece (Figure 3a), and then imported it to ImageJ 1.53k to measure the growth ring width as the distance from the earlywood of the first complete growth ring to the latewood of the last complete growth ring. We took the average value of three measurements as the final recording (Figure 3b). We used AI to extract the latewood area; the ratio of this area to the end surface area of the test piece was recorded as the latewood rate (Figure 3c).

2.3. Test Method and Statistical Analyses

According to GB/T 1927.12-2021, we conducted a transverse compression test on a universal mechanical testing machine (Model: 5582, Instron Co., Ltd., Norwood, MA, USA) with load range of 0–100 kN, loading speed of 0.5 mm/min, and load of 90 kN. We used a camera to record the deformation process of the test samples throughout the loading process.

The transverse compressive strength can be calculated as follows:

$${\mathsf{\sigma }}_{\mathit{P},\mathit{W}}{=\mathit{F}}_{\mathit{P}}/\mathit{A}$$

(1)

where ${\mathsf{\sigma }}_{\mathit{P},\mathit{W}}$ is the transverse compressive strength of the sample with a moisture content of $\mathit{W}$ (MPa); ${\mathit{F}}_{\mathit{P}}$ and $\mathit{A}$ are proportional ultimate loads (N) and bearing areas (mm²), respectively.

The moisture content and density of the sample were measured in accordance with Part 4 and Part 5 of GB/T 1927, respectively, and calculated as follows:

$$\mathrm{W}=\frac{{\mathit{m}}_1{-\mathit{m}}_0}{{\mathit{m}}_0}\times 100$$

(2)

where $\mathit{W}$ is the moisture content (%) of the sample, ${\mathit{m}}_1$ is the weight (g) of the sample during testing, and ${\mathit{m}}_0$ is the weight (g) of the sample when fully dry.

$${\rho }_W=\frac{m_W}{V_W}$$

(3)

where ${\rho }_W$, $m_W$, and $V_W$ are the air-dried density (g·cm⁻³), weight (g), and volume (cm³) at a moisture content of $W$, respectively.

Statistical analysis was conducted in Origin software and a Fisher test was conducted to verify the significance of the differences.

3. Results and Discussion

3.1. Effect of Growth Ring Width on Transverse Compressive Strength

The results for transverse compressive strength at different growth ring widths (Figure 4) showed that there was a significant difference in the transverse compressive strength of Japanese larch wood at different growth ring widths (p < 0.05). The radial compressive strength exhibited a positive correlation with the growth ring width, measuring 3.782 MPa, 4.108 MPa, and 4.555 MPa for various GRW samples. The transverse compressive strength of the GRW3 group increased by 20.43% and 10.88%, respectively, over that of GRW1 and GRW2. When subjected to slant compression, the transverse compressive strength of GRW1 was 1.915 MPa, that of GRW2 was 2.074 MPa, and that of GRW3 was 2.182 MPa. The relationship between transverse compressive strength and growth ring width was positive: a larger growth ring width yielded a greater transverse compressive strength. However, the regularity of tangential compression was found to be exactly the opposite of the former, wherein the transverse compressive strength decreased with an increase in growth ring width. The transverse compressive strength for growth ring widths from narrow to wide was 7.463 MPa, 6.802 MPa, and 6.217 MPa, respectively. The strength of GRW1 was 9.72% and 20.04% higher than that of GRW2 and GRW3, respectively.

We attribute the variation in transverse compressive strength among specimens with different growth ring widths to the varying proportion of latewood content. Latewood, the secondary xylem formed by cambium in the late growth season, has thick and compact cell walls, which contribute to its solid and dense appearance. The latewood rate of wood is an important factor in mechanical properties such as compressive strength, hardness, and shear strength [31,32]. Figure 5a shows the relationship between latewood rate and growth ring width. The latewood rate in our case decreased as growth ring width increased, as also observed by Wang Sidong [27]. The transverse compressive strength was fitted with the growth ring width and the latewood rate. As illustrated in Figure 4, when subjected to radial and slant loading, Japanese larch’s transverse compressive strength increased with growth ring width (p < 0.05). This is because the elastic stage of the load–displacement curve was only partially compressed by the earlywood at that point, so the transverse compressive strength was dependent only on the strength and content of the earlywood cells [33,34]. The latewood rates of GRW1, GRW2, and GRW3 decreased in turn as the earlywood percentages increased (Figure 5a). As a result, the transverse compressive strength of GRW3, with larger growth ring width, was higher than that of GRW1 or GRW2 (Figure 5b).

Both earlywood and latewood participate in tangential compression simultaneously [8,9]. The latewood cells, with narrow lumens and thick cell walls provide high mechanical strength and, thus, directly impact the transverse compressive strength of wood. A higher latewood rate remits greater transverse compressive strength, and there is an inverse relationship between latewood rate and growth ring width (Figure 5a). Consequently, the tangential compressive strength decreased as the growth ring width increased in our sample (p < 0.05).

3.2. Effect of Height from Tree Base on Transverse Compressive Strength

Table 2. Transverse radial compressive strength of Japanese larch with different heights from the tree base. See text for classification definitions. WC: moisture content; TCS: transverse compressive strength; SD: standard deviation; CV: coefficient of variation. Small letters “a” and “b” represent significant differences in the transverse compressive strength of wood at different heights (p < 0.05).

Group	Density(g·cm−3)	MC(%)	TCS(MPa)	SD	CV(%)
H1	0.515	11.27	3.901 a	0.511	13.11
H2	0.496	10.88	3.416 b	0.364	10.65
H3	0.480	11.34	3.365 b	0.355	10.54
H4	0.477	11.16	3.295 b	0.213	6.47

,

Table 3. Transverse tangential compressive strength of Japanese larch with different heights from the tree base. See text for classification definitions. WC: moisture content; TCS: transverse compressive strength; SD: standard deviation; CV: coefficient of variation. Small letters “a” and “b” represent significant differences in the transverse compressive strength of wood at different heights (p < 0.05).

Group	Density(g·cm−3)	MC(%)	TCS(MPa)	SD	CV(%)
H1	0.507	11.29	7.730 a	0.451	10.65
H2	0.499	11.24	6.931 b	0.693	9.99
H3	0.491	11.32	6.852 b	0.423	6.17
H4	0.486	10.93	6.306 b	0.924	14.65

and

Table 4. Transverse slant compressive strength of Japanese larch with different heights from the tree base. See text for classification definitions. WC: moisture content; TCS: transverse compressive strength; SD: standard deviation; CV: coefficient of variation. Small letters “a” and “b” represent significant differences in the transverse compressive strength of wood at different heights (p < 0.05).

Group	Density(g·cm−3)	MC(%)	TCS(MPa)	SD	CV(%)
H1	0.508	10.94	2.556 a	0.204	7.97
H2	0.496	11.53	2.463 a	0.292	11.86
H3	0.486	11.23	2.410 a	0.184	6.47
H4	0.475	11.12	2.166 b	0.273	12.62

show the transverse radial, tangential, and slant compressive strength of Japanese larch wood at different heights from the tree base. As mentioned above, H1, H2, H3, and H4 correspond to heights from the tree base of 3 m, 5 m, 7 m, and 9 m, respectively. This experiment was conducted in a controlled laboratory environment with constant indoor temperature and humidity, so the impact of wood moisture content was negligible. Due to the narrow 1 m interval between the selected heights, there was only a slight difference in the transverse compressive strength between adjacent groups. However, through multiple comparisons between groups, we found a significant difference in the radial and tangential compressive strengths of H1, H2, and H3 compared to H4 at the 0.05 level (Table 2 and Table 3). The slant compressive strength differences of H1, H2, H3, and H4 were significant at the 0.05 level (Table 4). The radial, tangential, and slant compressive strengths of H1 were 18.39%, 22.58%, and 18% higher than those of H4, respectively. Thus, when utilizing Japanese larch for engineering structural purposes, careful consideration should be given to the height from the tree base to ensure optimal material effectiveness.

The primary cause of the above phenomenon lies in the disparity of wood properties between juvenile and mature wood. A log can be divided into two main parts considering the variations in properties and structure: juvenile and mature wood. Juvenile wood is formed when the initial cells of the cambium have not yet fully matured, in the early growth and development stage of the tree. It surrounds the pith center of the trunk and represents a small proportion of the overall trunk volume. In contrast, mature wood constitutes the majority and contains more fully developed cells with higher specific gravity values. Juvenile wood, located near the pith center, extends from the base of the trunk to the tip, whereas mature wood is more prevalent throughout the trunk. Juvenile wood has shorter tracheids or wood fibers and greater inclination angles of tracheids and microfibril angles of cell walls compared to mature wood. The density, stiffness, and strength of juvenile wood are inferior to those of mature wood, rendering it unsuitable for load-bearing applications [16,35,36]. Although the materials employed in this study came from the same level, most in group H1 (near the tree base) were mature wood, which is physically distant from the juvenile wood; the parts taken from H4 were closer to the juvenile wood, resulting in greater transverse compressive strength at lower heights.

The difference in wood properties between juvenile and mature woods contributes to the observed variations in transverse compressive strengths. However, the primary growth of trees also creates differences in wood properties. Wood near the top of the tree experiences a shorter growth cycle and exhibits inferior properties, including density, compared to wood found closer to the tree base [37]. This is consistent with the density results listed in Table 2, Table 3 and Table 4. Navigating these differences poses a challenge during wood harvesting and selection processes.

3.3. Effect of Loading Direction on the Transverse Compressive Strength

Wood is a porous biomass material characterized by anisotropy, owing to its diverse cellular and microstructural composition [38]. The distinctive tissue structure of wood is the main cause of its anisotropy. Most of the cells and tissues in wood are arranged longitudinally, parallel to the growth direction of the trunk, while tracheids and radial parenchyma cells exhibit a circular pattern perpendicular to the tree axis. Therefore, the mechanical responses of wood in the radial, tangential, and slant directions differ significantly. We analyzed the transverse compression failure characteristics, load–displacement curve, and mechanical strength in different loading directions as discussed below.

3.3.1. Failure Characteristics

Figure 6 illustrates the failure condition of the Japanese larch specimen under radial compression, exhibiting significant compression deformation of the earlywood as there was no obvious deformation of the latewood part. The edge of the specimen initially formed an “S” shape due to the shrinkage of the earlywood and then was completely compacted though remaining intact. Figure 7 shows the ring interior in the transition zone between earlywood and latewood. The upper portion of the images reveals darker latewood characterized by thicker walls and smaller lumens. Conversely, the lower portion of the images displays lighter earlywood distinguished by larger cell lumens and thinner walls. A comparison of the images before and after the test reveals damage primarily concentrated in the earlywood region, with buckling and collapse of earlywood cell walls and lumens, accompanied by a decrease in intercellular spacing, which contributed to a reduction in specimen height. Japanese larch belongs to the genus Larix of Pinaceae and is a typical softwood, so the lumens of latewood tracheids are smaller compared to those of the earlywood, while their cell walls are thicker. Thus, when external force is applied, the latter is more prone to deformation and damage than the former [39].

We did not observe any obvious compression phenomena in the tangential direction; however, there were varying degrees of tracheid breakage at the top and bottom of the specimen. Tracheid buckling was observed in the middle section along the growth ring layer, forming “U”- or “V”-shaped layers. At the junction of the growth ring between the earlywood and latewood, several cracks were observed on the end face indicative of a loss of bearing capacity (Figure 8). Figure 9 shows the failure of the ring boundary. In the lower portion of the images, the latewood is darker due to its thicker walls and smaller lumens, while the upper portion displays lighter earlywood from the next ring with larger cell lumens and thinner walls. Before the test, a crack was observed in the earlywood; after the test, there was a slight decrease in the distance between cells in the compressed earlywood. However, the cells in the earlywood were not completely crushed or deformed as was the case under radial compression. This can be attributed to the vulnerability of earlywood cells and the parallel arrangement of early- and latewood when subjected to tangential compression, which enabled their cells to collectively participate in the compression and effectively inhibit deformation [40,41]. Boding [13], Kennedy [42], and Tabarsa and Chui [15] similarly found that the whole cell layer of latewood bends under tangential compression. This is consistent with the fiber buckling of our specimen along the growth ring.

The failure characteristics of the Japanese larch specimen under slant compression and radial compression were similar, as the earlywood part was compressed and underwent significant compression deformation until it became compacted. Shear slippage was observed at the junction of the early- and latewood of the growth ring, which caused the edge part to tilt to one side as longitudinal cracks formed at the top (Figure 10). Figure 11 shows inside the ring in the transition zone between earlywood and latewood. The left side of the image shows darker latewood with thicker walls and smaller lumens, while the right side displays lighter earlywood with larger cell lumens and thinner walls. The photo on the left was taken before the test and the one on the right was taken after, revealing damage in the earlywood. The cell walls exhibit irregular shapes, which we attribute to the crushing and buckling of earlywood cells that deflected due to shearing.

3.3.2. Load–Displacement Curve and Compressive Strength

Figure 12 presents the load–displacement curve of Japanese larch under transverse compression. Both radial and slant compression exhibit a “three-segment” load–displacement curve that includes an elastic stage (I), a plastic stage (II), and a compression/compaction stage (III) [12,15,43]. The height of the load–displacement curves for wood under radial and slant compression were consistent, which is in line with the failure patterns we observed. In practical applications, slant compression can be approximated as radial compression for many components. The load–displacement curve for tangential compression does not show a “platform zone” and instead follows a continuous curve that abruptly drops at its highest point, which is consistent with previous findings [44].

Japanese larch is a typical softwood, characterized by a sharp transition from earlywood to latewood and a distinct growth ring boundary. During radial and slant compression, earlywood cells easily deform due to the large size of their lumens and thin walls. As shown in Figure 12a,c, the elastic stage of the load–displacement curve reflects the elastic deformation process of the earlywood cells, which have relatively weak walls. After the failure of the first layer of cells, the load–displacement curve reaches an inflection point as the material progresses from the elastic stage to the plastic stage. The plastic stage is a relatively flat part of the load–displacement curve, which can be referred to as a “platform stage” or “plateau”. This stage shows a plastic yield phenomenon: the load increases slowly, but the displacement increases sharply as the earlywood cells are constantly crushed [39]. Once all the earlywood cells are crushed, the load–displacement curve reaches the “compaction stage”. The load increases sharply while the deformation is relatively slow at this point, which reflects the elastic deformation of latewood cells participating in the pressure acting on the earlywood. The load–displacement curve for tangential compression does not show a “three-segment” pattern due to the joint participation of early- and latewood cells, which prevent significant plastic deformation (such as earlywood cell compression loss or collapse). The bearing capacity was lost after the onset of buckling, however, at which point the load–displacement curve dropped sharply (Figure 12b). This curved shape mimics the load–displacement curve of wood compression in the longitudinal direction.

Table 5. Transverse compressive strength of wood in different loading directions. Note: GRW1 < 4.3 mm, 4.3 mm < GRW2 < 6.3 mm, GRW3 > 6.3 mm; H1, H2, H3, and H4 indicate the height from the tree base, which is 3 m, 5 m, 7 m, and 9 m, respectively; “∗∗” represents a significance level of 0.05.

Loading Direction	Growth Ring Width(mm)			Tree Height(m)
	GRW1	GRW2	GRW3	H1	H2	H3	H4
	Strength(MPa)			Strength(MPa)
Radial	3.782	4.108	4.555	3.901	3.416	3.365	3.295
Tangential	7.463	6.802	6.217	7.730	6.931	6.852	6.306
Slant	1.915	2.074	2.182	2.556	2.463	2.411	2.166
Conspicuousness Significance	∗∗	∗∗	∗∗	∗∗	∗∗	∗∗	∗∗

shows the transverse compressive strength of Japanese larch in different loading directions with the same growth ring width and height from the tree base. The transverse compressive strength of wood in GRW1 was highest in the tangential direction, 97.33% and 289.71% higher than that in the radial and slant directions, respectively. The tangential compressive strength of wood in GRW2 was 65.58% and 227.97% higher than that in the radial and slant directions, respectively. Similarly, the transverse compressive strength of wood in GRW3 was also highest in the tangential direction, at 36.49%, and 184.92% higher than that in the radial and slant directions, respectively, with significant differences at the 0.05 level. At the same height from the tree base, there were also significant differences in the transverse compressive strengths of the three loading directions (p < 0.05) falling into order as tangential > radial > slant. During the initial stages of tangential compression, the participation of latewood cells with narrow cell lumens, thick cell walls, and high mechanical strength contributed to higher tangential compressive strength compared to radial and slant compression. Conversely, the angle between the growth ring direction and the loading direction caused the vertical load to disperse along a certain slope, making the specimen prone to shear failure along the junction of the growth ring boundaries. Thus, the slant compressive strength is lower than the radial compressive strength [8].

4. Conclusions

We used plantation Japanese larch wood as a research object in this study to observe the effects of growth ring width, height from the tree base, and loading direction on transverse compressive strength. Our conclusions can be summarized as follows.

(1)

There was a significant difference in the transverse compressive strength of wood with different growth ring widths (p < 0.05). Samples with the widest growth rings showed a 20.43% and 10.88% greater resistance to radial compression than the samples with narrower and the narrowest growth rings, respectively. The relationship between slant compressive strength and growth ring width was positive: greater growth ring width yielded higher transverse compressive strength. The trend for tangential compression was the opposite, however, as transverse compressive strength decreased as growth ring width increased. Samples with the narrowest growth rings showed 9.72% and 20.04% higher transverse compressive strength compared to samples with wider and the widest growth rings, respectively.

(2)

The transverse compressive strength decreased with increasing height from the tree base. The radial, tangential, and slant compressive strengths at lower heights were 18.39%, 22.58%, and 18% higher than those at greater heights on the stem, respectively, with significant differences at the 0.05 level.

(3)

The load–deformation curves exhibited distinct shapes depending on the type of test piece. Japanese larch wood under radial and slant compression displayed a three-segment load–displacement curve, while tangential compression created a continuous curve that abruptly dropped after reaching its highest point.

(4)

Under the same growth ring width and height from the tree base, the average radial, tangential, and slant compressive strength values of Japanese larch wood showed significant differences (p < 0.05) falling into order as tangential > radial > slant. In the selection of engineering structural materials, tangential compression should be appropriately considered.