1. Introduction
The wood of different wood species differs greatly from one species to another. Even within the same genus and family, there is a diversity of different wood species. Going further, even within the same wood species, its wood can differ depending on the type of soil, climatic conditions, geographical area, etc. For the majority of wood-processing technologies, and in finished or semifinished products, a homogeneity of the wood species used is needed [
1,
2]; otherwise, the produced products will have major deficiencies or will have reduced marketability. This homogeneity can be characterized from a structural point of view, including density [
3,
4], reaction to atmospheric humidity or water, color [
5] or the color difference between the sapwood and heartwood (mature wood) [
6], the presence/absence of defects, resistances, and surface roughness [
3], with all of these depending on the processing technology used and the final product that is obtained. It is known that the wood of the same wood species has a different structure. There are differences between the sapwood and heartwood, the annual rings of early and late wood, or the type of deciduous wood species. From this point of view, the most uniform species are those with small and uniform distributed pores, such as linden wood. There are fields in which only homogeneous woods are used due to variation in their density and structure [
1,
7], such as when casting molds in the shoe industry, in the textile industry, and in manual sculptures, and when casting molds of alloys of small shapes in the foundry industry.
In the field of manufacturing wood-based composite boards, such as LVL [
8], wood species with appropriate densities are needed; only in this way can chips be compatible with and compacted according to the technology used in this field. Also, the color of the wood species used in mixture chips, flakes, or chops must be similar and uniform; otherwise, the boards obtained will have uneven colors or a reduced esthetic appearance. In the field of restoration of cultural heritage objects with wooden supports, it is recommended to use the same wood species, especially species of a density equivalent to that found in the degraded object to be restored [
9,
10,
11]. Historically, cultural heritage objects were mainly made of linden wood due to the structural homogeneity of this species, as well as the homogeneity of its shrinkage/swelling in the two structural directions (the ratio of tangential shrinkage to radial shrinkage). When it is restored, the wood in degraded heritage objects has a low density, with the mass losses being 25–50% depending on the degree of degradation. This is why the replaced wood material must have a density equivalent to that of the object to be restored (much smaller than the initial one); otherwise, the product would be heavier and additional stresses would be induced. Consequently, if linden wood is used for restoration, it should be artificially degraded or torrefied to reduce its density to that appropriate for the wood in the restored object, or to use another species with a lower density, such as balsa wood or paulownia wood [
12,
13]. In the field of veneers and timbers, species that are very closed in terms of density, texture, pattern, and color are considered uniform, even with similarities inside the annual rings [
4,
8]. The investigation methods of wood homogeneity were different from one author to another, and included the use of X-rays [
4] or a combination of light transmittance and mechanical stiffness [
8]. The analyzed materials included bamboo veneers and LVL [
8], white fir, Aleppo pine, European beech, walnut, and white oak [
3].
Objectives. This paper aims to analyze the uniformity of the wood of three homogeneous wood species, namely linden, balsa, and paulownia, from a structural and coloristic point of view, with respect to their physical–mechanical properties, and their relationship with water (water absorption and shrinkage). In this study, linden wood is taken as the element of comparison, emphasizing the specific properties of paulownia wood, which is promoted as the replacement for linden wood in all possible uses. This paper begins with the premise that linden and balsa woods are unsuitable species for replacing very old wood from heritage objects, and that paulownia wood could be a much more advantageous solution.
2. Materials and Methods
Anatomical structure. The first element of the analysis of the homogeneity or differences between the three wood species (linden, balsa, and paulownia) was their anatomical structure, including a study of the cross-sections of these three deciduous species, and their inclusion in the category of species with typically annular or uniformly scattered pores [
1,
6,
7]. For this purpose, three cross-sections of 20 mm × 20 mm of the analyzed species were taken, as shown in
Figure 1.
Unit density. The density of the three wood species was determined as mass per unit volume, as defined by international standards in the field [
14,
15]. Because wood is found at different moisture content levels, for a good comparison of the three wood species, both the mass and the volume of 30 mm × 20 mm × 20 mm samples were determined at the same moisture content of 0% (oven-dried state). For this purpose, the samples were dried for 24 hours in an oven with a 105 °C temperature. Taking into account the volume of the rectangular parallelepiped, under which the samples’ density was determined, the wood density was calculated as follows (Equation (1)):
where
is the density at a wood moisture content of 0%;
m is the mass of the wooden specimen, in
g;
l is the length of the specimen, in mm;
b is the width of the specimen, in mm; and
g is the thickness of the specimen, in mm.
Ten samples were used for this test.
The CIEL*a*b* color of the 3 species. Immediately after cutting its log into lumber, paulownia wood darkens in color to a light gray. This coloring is observable on a small thickness of about 0.7–0.8 mm due to the oxidation of wood sap when it comes into contact with atmospheric air [
16,
17,
18]. This blackish color is kept as long as the wood lumber is not processed. The dark color will disappear when the timber surface is planned to a thickness of at least 0.7–0.8 mm, obtaining the real and beautiful color of paulownia wood surface [
19,
20,
21].
A portable colorimeter of type Tes-135A, manufactured by Tess Electrical, Electronic Corp. (Taipei, Taiwan), was used for the color determination, using the CIEL*a*b* system. The CIEL*a*b* color space is one of the multiple color spaces (alongside CIEXYZ, CMYK, NCS, RGB, etc.) that was defined by the International Commission on Illumination (abbreviated as CIE) in 1976, which makes possible the transition from visual appreciation of colors to numerical quantification. It is based on human perception of colors as opposing colors, in which the group of red with green and the group of blue with yellow are usually opposing colors (
Figure 2).
This is why the CIEL*a*b* color space is quantified by three distinct parameters: L*, a*, and b*. In this color space, the letter “L*” represents lightness and has a value of 0 for black and 100 for white, with values in between being a series of shades of gray. The “a*” axis refers to the green–red opposition, with negative values toward green and positive values toward red. The “b*” axis quantifies the blue–yellow opposition, with negative values toward the blue zone and positive values toward the yellow zone [
5,
18].
For the color study, the three radial sections shown below were used (
Figure 3).
From
Figure 3, it can be seen visually that, roughly, the three species are similar in color, with paulownia wood being a little toward brown, and linden wood being a little toward reddish.
Water absorption after two hours of immersion. Water absorption after immersion in water for 2 h was determined for the absolutely dry samples, which were kept in the oven at 105 °C for 6 h [
22,
23,
24,
25]. The linden, balsa, and paulownia samples had the average dimensions of 20 mm × 20 mm × 30 mm, where the dimension of 30 mm was their length. The mass after immersion and the mass after drying in the oven were determined for the three wood species. Based on these masses, water absorption (
WA) was determined according to the following relationship (Equation (2)):
where
mw is the mass of the immersed specimen, in g, and
m0 is the mass of the oven-dried test piece, in g.
Seven specimens were used for this test.
Shrinkage in the radial/tangential direction and volume. Wood shrinkage was performed on the samples of the same types and dimensions as in the case of absorption. These samples were cut in such a way so that the radial and tangential directions were clearly identified [
26,
27]. In this test, with the help of an electronic caliper, the initial dimensions in the wet state and the final ones in the absolutely dry state were determined. Based on the masses of the samples in the wet and dry state, the wood moisture content was also determined.verifying whether the requirement of a certain moisture content above the fiber saturation point (over 30% moisture content) was fulfilled. If the moisture content of the samples did not touch 30% (
Figure 4), the necessary correction was performed; the value for 1% moisture content was determined, and then, this value was multiplied by 30%.
The total shrinkage (throughout the saturation moisture range) in the radial, tangential, longitudinal, and volumetric directions was determined based on the dimensions before and after drying according to the following specific relationships (Equations (3) and (4)):
where
Stl is the total longitudinal shrinkage, in %;
lmax is the maximum length of the specimen, after immersion in water, in mm;
lmin is the minimum length of the sample, after drying in the oven, in mm;
Stt is the total tangential shrinkage, in %;
tmax is the maximum thickness (on thickness) of the specimen, after immersion in water, in mm;
tmin is the minimum thickness of the sample, after drying in the oven, in mm;
Str is the total radial shrinkage, in %;
rmax is the maximum radial of the specimen, after immersion in water, in mm;
rmin is the minimum radial of the sample, after drying in the oven, in mm;
Stv is the total volumetric shrinkage, in %;
Vmax is the maximum volume of the specimen, after immersion in water, in mm; and
lmin is the minimum volume of the sample, after drying in the oven, in mm.
The partial volumetric shrinkage (only in a portion of the moisture fiber saturation), such as from the saturation point of the fiber to a certain MC1 value (
Figure 4), can be determined according to the following equation (Equation (5)):
where
Spv is the partial volumetric shrinkage, in %, and
V1 is the volume of wood corresponding to the MC1.
Modulus of resistance and modulus of elasticity to static bending conforming to ISO 13061-4: 2014 [
28]. Before testing, the samples were conditioned to obtain an average moisture content of 12%. The modulus of resistance (MOR) and modulus of elasticity (MOE) to static bending were determined according to the following general relationships (Equations (6) and (7)):
where
Fmax is the maximum force of breaking, in N;
l is the distance between supports, in mm;
b is the width of the specimen, in mm;
t is the thickness of the specimen, in mm;
F1 is the value of force at 10% from the maximum force, in N; F
2 is the value of force at 40% from the maximum force, in N;
a1 is deformation for force F
1, in mm; and
a2 is deformation for force F
2.
Ten samples were tested in order to obtain statistically significant values.
Compressive strength conforming to ISO 13061-17 [
28,
29]. Before testing, the samples were conditioned in order to have up to 12% moisture content. The compressive strength parallel to the grain used in the work was obtained in the moment of breaking when the layers of wood in the samples were sheared or crushed, and the force dropped suddenly.
Therefore, the resistance to compression parallel to the wood grain was determined as the ratio between the maximum breaking force and the area of the breaking surface (Equation (8)):
where
σc is the compressive strength, in N/mm
2;
Fmax is the maximum force of breakage, in N; and
b1 and
b2 are the widths of the specimen in tangential and radial directions, in mm.
Ten replicates of each type of specimen were used for this test.
Brinell hardness conforming to EN 1534:2003 [
30]. Brinell hardness quantifies the resistance of a wood surface when a metal ball tries to penetrate and destroy it. The punch that is pressed onto the surface of the wood is created by a ball with a diameter of 10 mm, and, to highlight the diameter of the mark, it is necessary to arrange a copy paper between the punch and the wooden surface. This resistance is higher on the transverse section, compared to the longitudinal–radial and longitudinal–tangential sections, but, due to the fact that the transverse section is less used in industrial practice of wood processing, only the radial and tangential sections are used in research. Brinell hardness was calculated as follows (Equation (9)):
where
P is the applied force, related to wood density, in N;
D is the diameter of the punch ball, in mm; and
d is the diameter of the trace left on the wood, in mm.
Ten valid samples were used for testing each wood species.
Ecological aspects of the three species. All forests, through the trees that they have in their area, have several functions, among which are hydrological, soil protection, esthetic, recreational, ecological, and production functions [
31]. The ecological function of forests is related to physiology, namely the fact that humans need a certain amount of oxygen for breathing. Oxygen is released into the atmosphere by forest vegetation (about 1.3 t of oxygen per ton of woody mass, in a day) through its chlorophyll assimilation process. A mature tree produces about 1.7 kg of oxygen/day, which means the oxygen requirement of a human for three days [
32]. Also, in the same process of tree photosynthesis, a large amount of carbon dioxide is consumed, approximatively 1.8 t per one cubic meter of wood. On hot days, a hectare of a highly productive forest consumes an average of 220–280 kg of carbon dioxide per day. Wood production of the three forestry species in this study is different: linden at 350 m
3/ha, balsa at 550 m
3/ha, and paulownia at 950 m
3/ha. The volume of timber obtained after the first 3–4 years of paulownia is 0.3–0.8 m
3/tree, i.e., 200–500 m
3/ha. The medium age of exploitation is 75 years for linden, 50 years for balsa, and 15 years for paulownia [
33,
34]. To establish the ecological factor for all three species, the unitary coefficient for linden forests was considered, which was used as a reference for comparison with the other two forest species (paulownia and balsa).
The comparative analysis of the three species. The method of comparative analysis of the 3 wood species took linden wood as the element of comparison because it is currently considered the most homogeneous species, especially from the point of view of the ratio of tangential shrinkage to radial shrinkage. Several analysis criteria were used, namely anatomical structure, density, resistance, color, shrinkage and swelling, and the ratio of tangential shrinkage to radial shrinkage. All the analyzed characteristics for paulownia and balsa woods were compared to those of linden to obtain positive/negative percentages, using the following relationship (Equation (10)):
where
Cb,p is the characteristic of balsa/paulownia wood, and
Cl is the characteristic of linden wood.
Based on the all obtained values of characteristics, it was cumulatively determined which of the two species (balsa and paulownia) is more appropriate as a possible replacement for linden wood in its uses.
Statistical analysis. Standard deviation and arithmetic mean were calculated for each group of tests. Also, the regression tendences and the Pearson coefficients R2 were identified on the graphs obtained using Microsoft Excel 2019 (Microsoft Corp., Redmond, WA, USA). With the help of the statistical program Minitab 18 (Penn State University, State College, PA, USA) and its generated graphs, the mean, standard deviation, p-value, and Anderson–Darling coefficient of all tested values were found, with a confidence interval of 95% or an error of 0.05.
4. Discussion
To be able to have a conclusive discussion regarding the results obtained in this work, all results are grouped in
Table 7. Also, this table includes the comparison values obtained for paulownia and balsa wood to those obtained for linden wood, corresponding to one of the relationships (Equation (10)).
The first analysis of the results shown in
Table 7 was a comparison between the characteristics of the three species, namely how close the characteristics of paulownia and balsa are to those of linden wood. It was clearly observed that paulownia wood was much more appropriate as a replacement for linden wood with a total of 526.6 points, thus being 4.4 times more appropriate than balsa wood.
Paulownia wood had a density lower than that of linden by 34%, but higher than that of balsa wood by 1.62 times. It was observed that the experimental value of balsa wood density was closest to that obtained by other researchers [
1], with the difference only 1.6% lower. In the case of linden wood, the difference between the two values was higher than 9.7%, and in the case of paulownia wood, the difference was higher than 21.6%. From the point of view of density, the three analyzed species are different as the differences are greater than 5–10% between them, paulownia wood being more appropriate as a replacement for linden [
35,
36]. Some authors [
17] found a wood density of about 300 kg/m
3 for paulownia, which was very close to the value of 304 kg/m
3 of this research.
The analysis of color parameters with regard to the natural color of planed wood showed that from a color point of view, the three wood species are very appropriate (L = 21.8–25.1; a = −(35.2–37.6); and b = 3.6–4), with the points of all three species being positioned in the fourth trigonometric quadrant. Appropriate values were also found by other authors [
5,
18], and a similar value of paulownia was obtained by others [
17].
The water absorption of these three wood species was dependent on their porosity; it increased with a decrease in the density of these species. The exception to this rule was paulownia wood, which had a lower absorption than linden wood. The explanation for this phenomenon could be attributed to the microscopic structure of this wood species (
Figure 1) and the fact that it has pores filled with cellulosic formations that prevent the penetration of water into the wood.
As a general trend, the wood shrinkage of these three wood species decreased with a decrease in density (regardless of whether the direction of analysis was radial or tangential), with balsa wood having the lowest shrinkage (1.4% in the tangential direction and 1.6% in the radial direction) [
37]. The values of the ratio of tangential shrinkage to radial shrinkage showed that these three wood species are uniform, with balsa wood and linden wood having the lowest shrinkage coefficients.
The Brinell hardness of the three wood species had values corresponding to the values reported in other research studies in the field [
38,
39]), but there were also small differences between them. For example, the Brinell hardness of balsa wood was only 22.8% lower than that of paulownia wood, although the density difference between the two wood species was 74.4%.
The compressive strength of these three species (
Figure 17) was different and depended on density, with a higher compressive resistance (52.9 N/mm
2) being obtained for linden, which had a higher wood density (461 kg/m
3). As a general trend, from the point of view of compression strength, paulownia wood has a value closer to linden wood. The values obtained in this research were slightly higher than those reported in the literature [
1,
40], being higher by 15.3% for linden, 8.4% for paulownia, and 8.2% for balsa wood.
The bending strength (MOR) was in accordance with the density of the wood species, with higher densities corresponding to higher resistances. The dependence relationship between the densities and resistances of these three species (
Figure 19) was approximated as linear; from this point of view, paulownia wood has superior characteristics compared to the other two species. Differences between the values obtained in this research and those reported in the literature [
1,
41] were also identified; in this case the difference was about 7.4% for linden wood. The modulus of elasticity (MOE) maintained the same trend as the modulus of resistance (MOR), showing a decrease along with a decrease in density. However, the differences between the values of the three species have increased, as the modulus of elasticity of paulownia wood (4852 N/mm
2) becoming closer to that of linden wood (6782 N/mm
2), while the modulus of elasticity of balsa wood (927 N/mm
2) moving further away. The values obtained in this work differed slightly from those reported in the literature [
1]. For instance, there was a large difference of about 4.72 times in the case of balsa wood. In the direction of what was mentioned above, in the reference [
23], an MOR value of 111 N/mm
2 and an MOE value of 11,479 N/mm
2 were found for beech wood, but with a much higher density than the species analyzed in this study, of 727 kg/m
3.
Based on the results shown in
Table 7, due to the different rates of growth, paulownia and balsa wood had superior ecologically characteristics compared to linden wood, being about five times higher [
30,
33]. From an ecological point of view, the results showed that paulownia and balsa wood are superior to linden wood but are very similar to each other, with the difference between balsa and paulownia of only 5.4%.
The quality coefficient (related to both compressive strength and modulus of resistance) of paulownia wood was better than balsa and linden. From this point of view, paulownia wood could be considered a strong species that is comparable to the strongest wood species in the word.
5. Conclusions
Generally speaking, the three deciduous woody species are as appropriate as possible, having well-observed similarities, and, because that, they can be considered homogenous.
The closest to linden is paulownia wood, which can replace linden in its uses in modeling, restoration, conservation, sculpture, etc.
Ecologically, balsa wood and paulownia wood are better than linden wood due to their faster growth.
The mechanical properties (compressive strength, MOR, MOE, and Brinell hardness) of the three species are consistent with their density.
From the point of view of quality coefficient, paulownia wood is the best species, compared to linden and balsa woods.
In a broader context, this research study falls into the area of knowledge of wood species with a low density and a good structural homogeneity.
Among the limitations of this research study and possible new areas of investigation there are the behavior of these species during thermal treatments at temperatures above 200 °C, in a nitrogen environment, or when there is no access to oxygen.