Abstract
Alnus acuminata, Vochysia ferruginea, and Vochysia guatemalensis are three low-density wood species used for reforestation in Costa Rica. The goal of this work was to study a thermo-hydro-mechanical densification process and test the characteristics of densified wood of these species. Twelve densifying treatments based on temperature, compression time, and use/no use of steam were tested. The variables of the densification process and the properties of the densified wood were determined. The results showed that the densification percentage was over 80% for wood of A. acuminata and over 70% for wood of V. ferruginea and V. guatemalensis. In the three species, the densification process was influenced by initial density. The influence of temperature during the densification process affected the heating rate and color change. An increase in the modulus of elasticity and modulus of rupture in static bending and in the hardness of the densified wood relative to the normal wood was observed, as well as a clear positive correlation of the properties with final density and maximum load, the latter being highly correlated with initial density. This showed that initial density was significant in the densification process and affects wood properties.
Download PDF
Full Article
Effect of Thermo-hydro-mechanical Densification on the Wood Properties of Three Short-rotation Forest Species in Costa Rica
Carolina Tenorio a and Roger Moya b,*
Alnus acuminata, Vochysia ferruginea, and Vochysia guatemalensis are three low-density wood species used for reforestation in Costa Rica. The goal of this work was to study a thermo-hydro-mechanical densification process and test the characteristics of densified wood of these species. Twelve densifying treatments based on temperature, compression time, and use/no use of steam were tested. The variables of the densification process and the properties of the densified wood were determined. The results showed that the densification percentage was over 80% for wood of A. acuminata and over 70% for wood of V. ferruginea and V. guatemalensis. In the three species, the densification process was influenced by initial density. The influence of temperature during the densification process affected the heating rate and color change. An increase in the modulus of elasticity and modulus of rupture in static bending and in the hardness of the densified wood relative to the normal wood was observed, as well as a clear positive correlation of the properties with final density and maximum load, the latter being highly correlated with initial density. This showed that initial density was significant in the densification process and affects wood properties.
Keywords: Tropical species; Hardwoods; Low density; Thermo-hydro-mechanical process
Contact information: a: Escuela de Ingeniería Forestal, Instituto Tecnológico de Costa Rica, Cartago, Costa Rica, 159-7050; b: Escuela de Ingeniería Forestal, Instituto Tecnológico de Costa Rica, Cartago, Costa Rica 159-7050; *Corresponding author: rmoya@itcr.ac.cr
INTRODUCTION
Wood supply for the industrial and construction sectors at the world level has decreased during the last few years (Toppinen et al. 2018). For this reason, the interest in fast-growing wood species has grown (Cossalter and Pye-Smith 2003). However, low density and limited mechanical properties are the main characteristics of those species because the trees are harvested while still young and the proportion of juvenile wood is dominant (Zobel and Van Buijtenen 1989; Kamke 2006; Cahyono et al. 2015). The high price of wood in Costa Rica led to the search for new opportunities to allow the replacement of high value timber species with less valuable or marginal species (Nölte et al. 2018).
Alnus acuminata, Vochysia ferruginea, and Vochysia guatemalensis are fast-growing hardwood species used in commercial reforestation programs in Costa Rica (Moya 2018). These are low-density, soft, and easy to work with wood species (Tenorio et al. 2016). However, the mechanical properties of these species are limited, which makes them unsuitable for structural uses, being limited to products of low structural demand and low commercial value markets (Serrano and Moya 2011). Due to these reasons, these species have been rarely studied and ignored by industries. Currently, studies can be found on the properties and behaviour of these and other reforestation species in industrial processes in Costa Rica (Moya et al. 2019). However, studies aiming at improving their mechanical properties are not available.
Many wood densification processes have been developed to improve the wood’s mechanical and physical properties (Fang et al. 2012). Such processes can increase the wood density in three ways: (i) mechanical compression by reducing the empty spaces; (ii) impregnation of the empty spaces with some substance, and (iii) combination of the previous processes (Fang et al. 2012). However, unlike densification by mechanical compression, chemical impregnation affects the natural and sustainable character of the wood, and is usually more expensive (Navi and Heger 2004). Mechanical compression combined with steam and heat, called thermo-hydro-mechanical (THM) densification, has been studied as an environmentally friendly alternative to increase the wood density and improve its mechanical properties, achieving enhanced dimensional stability without using chemical products (Bekhta et al. 2009; Büyüksari et al. 2012; Arruda and Del Menezzi 2013; Candan et al. 2013; Moya et al. 2013; Tu et al. 2014).
There are different THM densification processes, generally consisting of: (i) softening the wood structure, which can be attained at certain temperatures and moisture contents; (ii) compressing the wood, usually between two metal plates, and (iii) keeping the wood deformation obtained, by thermal modification in many cases (Sandberg et al. 2013). The THM densification process improves the natural properties of the wood and produces stable materials (Navi and Heger 2004; Sandberg et al. 2013). The heat treatment can improve resistance to decay (Huang et al. 2012), decrease hygroscopicity (Metsä-Kortelainen et al. 2006), and improve the dimensional stability (Esteves et al. 2007). The moisture induces a mechano-sorptive effect and further softens the wood, which enables mechanical compression of wood without cell wall fracture (Bao et al. 2017).
Most research on THM densification has focused on softwood species and closed systems, where the densification process takes place inside reactors (Navi and Heger 2004). Therefore, studies on hardwood tropical species, whose anatomical structures affect to a further extent the result of the process, are scarce (Navi and Heger 2004). This is because the compression properties of the wood depend on the frequency, size, and distribution of its anatomical structures (Darwis et al. 2017). For hardwoods, these structures are dominated by vessels, fibers, and radial parenchyma arranged in more complex matrixes (Gibson 2012) than the fiber tracheids of softwood species (Fratzl and Weinkamer 2007).
The aim of this study was to investigate the THM densification process and evaluate its effect on the characteristics of the densified wood of three fast-growing hardwood species used in commercial reforestation in Costa Rica: A. acuminata, V. ferruginea, and V. guatemalensis. The variables in the densifying process along with the physical and mechanical properties of the densified wood were determined.
EXPERIMENTAL
Origin and Characteristics of the Wood before Densification
The study tested the wood of Alnus acuminata, Vochysia ferruginea, and Vochysia guatemalensis from fast-growth plantations located in Cartago and Alajuela in Costa Rica. The trees used were approximately 8 years old, which normally implies a low heartwood content (Tenorio et al. 2016). Therefore, the wood used was mostly sapwood. Wood samples of 300 mm long × 70 mm wide × 20 mm thick of each species were prepared. Before densification, thickness, width, length, density, moisture content, and color were measured for each of the samples, in total were 240 samples per species (Table 1).
Table 1. General Characteristics of the Wood Before Densification
The values in parenthesis are standard deviations.
The density was calculated as the ratio of weight and volume determined by measuring the initial thickness, width, and length. The moisture content was calculated as the ratio between the initial weight and the oven-dry weight, expressed as a percentage according to ASTM D4422-16 (2016). For color measurements, a Mini Scan XE Plus spectrophotometer (HunterLab, Reston, VA, USA) was used along with the CIE L*a*b* system to measure the reflectance spectrum. The range of this measurement was from 400 to 700 nm with an opening at the point of measurement of 11 mm. For the observation of reflection, the specular component (SCI mode) was included at a 10º angle, which is normal for the specimen surface (D65/10), a field of vision of 2º (Standard observer, CIE 1931), and the standard illuminant D65 (corresponding to daylight in 6500 K). The Mini Scan XE Plus generated three parameters for each measurement, namely: L* (luminosity), a* (color trend from red to green), and b* (color trend from yellow to blue).
Densification Process
Three temperatures, 140 °C, 160 °C, and 180 °C for A. acuminata and V. guatemalensis and 140 °C, 150 °C, and 160 °C for V. ferruginea, two compression times of 10 and 15 min, and the application of water steam or just heat (as the initial stage before wood compression) were used in the densification process. In total, 12 treatments were tested with 20 specimens per densification treatment, resulting in 240 specimens per species (Table 2). The differences in temperature for V. ferruginea in comparison with the other two species are due to the fact that when carrying out the first tests with 180 °C the surface of the wood samples burned and, in some cases, parts adhered to the metal plates, making it impossible to evaluate the process, probably due to low thermostability of this species (Moya et al. 2017). Therefore, it was decided to work with different temperatures for this species.
The densification process was the same described in Tenorio and Moya (2019). The process consisted of three stages. Stage 1 was the steaming or heating stage, where steam was applied to half of the wood samples for 10 min, while the other half was only heat-treated. Stage 2 was the compression stage, where the wood samples were compressed perpendicular to the grain until reaching a target thickness of 9 mm (degree of compression of 55%) during 10 or 15 min. Stage 3 was the stabilization stage, where the samples were kept compressed and heated while unloaded for another 10 min. Throughout the process, the metal plates kept one of the three temperatures constantly.
Table 2. Treatments Used in the Densification Process of the Wood of Three Species
*This treatment code means 140 °C of heat treatment, 10 min of compression time, and with steam application
Evaluation of the Densification Process
To determine the maximum load and the internal heating rate of the wood, a temperature control probe that was placed at the centre of the sample thickness was used to monitor and record the load and temperature during the densification process. Additionally, each sample’s thickness was determined at the end of stages 2 and 3. After the densification process, the width, length, weight, color, density, and densification percentage were determined.
The same procedure used before the densification process was used for color measurements. The density was calculated as the ratio between the weight and volume of the wood sample after stage 3. For the volume calculation, the dimensions of stabilization thickness, length, and width were determined. The densification percentage was calculated as the ratio of the sample initial density and the density after densification.
The thickness determined in stages 2 and 3 helped determine the compression ratio and the spring back of wood thickness. The compression ratio was calculated as the ratio of the initial thickness and the compression thickness, expressed as a percentage. The spring back of the thickness of the densified wood was calculated as the absolute value of the ratio of compression thickness and stabilization thickness, expressed as a percentage.
Properties of the Densified Wood
In the evaluation of the wood properties, the thickness swelling was determined according to ASTM D4933-16 (2016). The wood samples were conditioned to 18% equilibrium moisture content, at 20 °C, and 85% relative humidity. Next, the thickness swelling was calculated as the ratio of the thickness after conditioning and the thickness before conditioning, expressed as a percentage. The mechanical properties, static bending, modulus of rupture, modulus of elasticity, and Janka hardness were determined following the ASTM D143-14 (2016) standard. A total amount of 20 specimens per species, per treatment, were prepared for each test. In total, 13 treatments were performed, 12 densification treatments and one un-densified wood treatment.
Microscopic Examination
For the microscopic examination, sections 10 mm × 50 mm × the densified thickness were taken from the densified specimens, and sections 10 mm × 50 mm × 20 mm were taken from the un-densified wood specimens. All samples were polished in their transversal sections using a sander (Struers, Tegramim 30 Model, Cleveland, OH, USA). The procedure is detailed in Tenorio et al. (2021).
After polishing, a microscope (Zeiss, Axioscope Model, Jena, Germany) was used to take the pictures of the anatomical structures of each specimen, using an Axicam 503 Color camera (Zeiss, Jena, Germany) in reflexion mode using a 10x and 50x lens. The un-densified specimens were photographed in reflexion mode using a 10x lens. The image editing ZEN program (Zen Pro, Zeiss, Version 2.3, Jena, Germany) was used.
Statistical Analysis
Compliance of the measured variables with the assumptions of normal distribution, homogeneity of variance, and outliers was verified. An analysis of variance was applied to verify the effect of the densification treatments in each one of the variables obtained during the densification process, in the physical properties, and mechanical properties per species (Table 2).
The Tukey test was used to determine the statistical differences between the means of the variables measured. A correlation analysis between the variables obtained during densification and the initial characteristics of the wood was performed. Correlation analysis was additionally performed between the physical and mechanical properties of the densified wood, and the variables obtained during the process of densification, independently for each species. The analysis of variance, the Tukey tests, and the correlation analysis were performed with the SAS software (SAS Institute Inc., v.4.11, Cary, NC, USA).
RESULTS
Evaluation of the Densification Process
Table 3 presents the results obtained as part of the evaluation of the densification process. A compression ratio of approximately 55% can be observed for the three species. No statistical differences were found between the treatments applied for any of the species.
Table 3. Characteristics of the Wood of Three Forest Species After Densification
Note: Different letters for each parameter represent statistical differences between different treatments (significance level 0.05)
The final density, the densification percentage, and the spring back for A. acuminata were 0.78 g/cm3, 84.49%, and 7.75% respectively, while no differences were observed between the densification treatments (Table 3).
For V. ferruginea the final density was 0.80 g/cm3 on average and the densification percentage was 76.08%, and no differences between the densification treatments were observed. The spring back was 14.26% with minimal differences between the treatments. Treatment 150-10-WT presented the highest value and 160-15-WT the lowest (Table 3). For V. guatemalensis, the final density was 0.68 g/cm3 on average; 180-10-ST presented the highest value, whereas 140-10-ST and 160-10-WT presented the lowest values. The densification percentage was 76.55%. Treatments 140-15-ST and 180-10-ST showed the highest values, while 160-15-ST had the lowest value. The spring back was 14.12%. Treatment 140-10-ST showed the highest percentage of spring back, while treatments 140-15-ST, 140-15-WT, and all treatments with temperature of 180 C presented the lowest averages (Table 3).
Color change in wood after densification was more pronounced in V. ferruginea with 5.23, followed by A. acuminata with 4.18, and V. guatemalensis with 2.48 on average (Table 3). Treatments with the temperature of 180 C for A. acuminata and V. guatemalensis, and 160 C for V. ferruginea presented higher color changes, while treatments with the temperature of 140 C for the three species presented the lowest color changes (Table 3).
Evaluation of the maximum load applied to wood in stage 2 of the densification process showed that greater load was applied to A. acuminata (201439 N) followed by V. ferruginea (175666 N) and next by V. guatemalensis (138323 N). A. acuminata and V. ferruginea showed no differences between the treatments, while V. guatemalensis treatment 140-15-WT presented the highest value and 140-10-WT and 160-10-WT the lowest values (Fig. 1 a through c).
Concerning the heating rate, in general, similar averages were observed for the three species: 3.69 °C/min for A. acuminata, 3.40 °C/min for V. ferruginea, and 3.92 °C/min for V. guatemalensis. In addition, densified wood with the highest temperatures, 180 C for A. acuminata and V. guatemalensis and 160 C in V. ferruginea, presented the highest heating rates. Moreover, the wood with 10 min compression time showed a higher heating rate than the wood under 15 min compression, while the wood densified without steam showed the highest values of heating rate compared to wood densified with steam (Fig. 1 d through f). As for A. acuminata wood, treatments 180-10-ST and 180-10-WT presented higher heating rate values, while 140-15-ST showed the lowest value (Fig. 1d). For V. guatemalensis and V. ferruginea, 180-10-WT and 160-10-WT showed the highest heating rate and 140-15-ST the lowest (Fig. 1 e and f).
Properties of Densified Wood
Statistical differences were observed between the treatments applied regarding the averages obtained for the properties evaluated in the densified wood (Table 4). The thickness swelling for A. acuminata showed an average of 41.71%. Treatment 160-15-ST presented the highest value, while treatment 140-10-ST was the lowest. For V. ferruginea, the average thickness swelling was 23.42% for all the densification treatments. Treatment 160-15-WT had the highest value and 140-10-WT the lowest. As for V. guatemalensis, treatment 140-15-ST had the highest thickness swelling average, and treatments 160-10-WT and 180-15-ST the lowest, for an average between the treatments of 20.73% (Table 4). For the three species the thickness swelling in the un-densified wood was less than 1%.
Fig. 1. Maximum load and heating rate during the THM densification process of the wood for Alnus acuminata (a and d), Vochysia ferruginea (b and e), and Vochysia guatemalensis (c and f). Different letters for each parameter represent statistical differences between different treatments (significance level 0.05)
With respect to mechanical properties, no differences in the modulus of elasticity (MOE) and modulus of rupture (MOR) in static bending were observed between treatments for the densified wood of A. acuminata (13.18 GPa and 134.22 MPa on average, respectively) (Table 4). Average MOE and MOR for densified wood of V. ferruginea was 14.11 GPa and 84.78 MPa, respectively. For MOE, treatment 160-15-WT presented the highest value and treatment 150-10-WT the lowest.
Table 4. Physical and Mechanical Properties of Densified Wood of Three Forest Species
Different letters for each parameter represent statistical differences between different treatments per specie (significance level 0.05)
All treatments increased MOE and MOR in flexion and hardness values in A. acuminate densified wood (Table 4). For densified wood of V. guatemalensis, treatment 160-10-ST showed the highest average MOE and treatments with low temperature (140-160 C) obtained the highest MOR in flexion (Table 4). In V. guatemalensis, the averages for un-densified wood were below the averages obtained for densified wood in the 12 treatments used, except to180-10-WT and 180-15-ST for MOE (Table 4).
For hardness test, the un-densified wood had the statistically lower value in V. ferruginea (Table 4), and there were few differences between the densification treatments. In the wood of V. guatemalensis there were many differences between the treatments. Treatment 140-15-ST had the highest value, and the un-densified wood had the statistically lowest average compared to densification treatments (Table 4).
Relationship Between Variables
Regarding the relationships between the initial characteristics of the wood and the variables obtained in the densification process, the initial density had the highest number of correlations with the variables obtained in the wood densification process for all three wood types (Table 5). In A. acuminata, the initial thickness of the wood presented a positive correlation with the degree of compression, and a negative correlation with the heating rate (Table 5). The initial density was related to almost all the variables of the densified process, except for color change. A positive correlation was observed between the initial density, final density, the maximum load, and the spring back. A negative correlation was observed with the compression ratio, the densification percentage, and the heating rate (Table 5). An aspect to highlight in A. acuminata is that the initial moisture content did not show any relationship with the variables obtained in the densification process (Table 5).
For the V. ferruginea densified wood, the initial thickness showed a positive correlation with the compression ratio. Initial density correlated to almost all the variables in the densification process, except for the heating rate. A positive correlation was observed with the final density, color change, maximum load, and spring back. A negative correlation was shown with the compression ratio and the densification percentage. The initial moisture content correlated negatively with color change and positively with the heating rate (Table 5).
For the V. guatemalensis wood, the initial thickness showed positive correlation with the final density and negative correlation with the heating rate. The initial density correlated to almost all the variables in the process of densification except color change and correlated negatively with all the remaining variables. Moisture content correlated positively with the densification percentage and with the heating rate (Table 5).
As for the relationship between the variables obtained in the densification process and the wood properties, the final density presented the greatest number of correlations with the wood properties for the three species (Table 6). In A. acuminata the thickness swelling and the MOE in static bending, showed a positive correlation with the compression ratio, the final density, and the densification percentage, and a negative correlation with the spring back. As for the MOR in static bending and hardness, both correlated positively with the final density and maximum load (Table 6).
Table 5. Correlation Coefficients Between the Initial Characteristics of the Wood and the Variables Obtained in the THM Densification Process
**: Statistically significant level 0.1; *: statistically significant level 0.05; ns: not significant
Table 6. Correlation Coefficients Between the Variables Obtained in the THM Densification Process and the Densified Wood Properties
**: Statistically significant level 0.1; *: statistically significant 0.05; ns: not significant
For V. ferruginea, the thickness swelling had a negative correlation with the compression ratio, final density, maximum load, and spring back. The MOE in static bending had a positive correlation with the final density and the densification percentage, and a negative correlation with the spring back.
The MOR in static bending had a positive correlation with the final density, maximum load, and spring back, and a negative correlation with the densification percentage. The hardness had a positive correlation with the densification percentage and maximum load (Table 6).
In the densified wood of V. guatemalensis the thickness swelling only showed a negative correlation with the heating rate. The MOE in static bending had a positive correlation with the final density and the maximum load, and a negative correlation with the heating rate. The MOR in static bending showed a positive correlation with the final density and maximum load, and a negative correlation with the compression ratio, the densification percentage, and the heating rate. Hardness had a positive correlation with the final density and the maximum load, and there was not any correlation found with other variables (Table 6).
Microscopic Evaluation
Cross-sections of the un-densified and densified specimens’ anatomy features of the three species are presented in Fig. 2. The three species presented solitary and multiple diffuse vessels (Fig. 2a, e, and i). According to previous anatomical description in the same species (Moya et al. 2019), A. acuminata presents a pore frequency of 16 pores/mm2 with small diameters (75 µm) and short lengths (150 µm). V. ferruginea presents 2.84 pores/mm2 of medium diameter (145 µm) and length (346 µm). V. guatemalensis has 2.88 pores/mm2, with regular size diameters (169 µm) and length (339 μm). Fibers in the three species are irregularly arranged in rows perpendicular to the growth rings or parallel to ray parenchyma, which is typical of non-stored fibers. The rows are crooked due to the presence of large vessels. A. acuminata formed more uniform fiber rows than the other two species. A. acuminata features finer rays than those of the other two species, which are multi-serial, showing 44 to 8 series per ray (Moya et al. 2019).
In the cross-sections of densified wood of the three species, the effect of the densification process can be perceived in the size of their anatomical structures and the differences in the vessels. The vessels tended to collapse and flatten completely in the case of A. acuminata (Fig. 2d). For the other two species, the vessels tended to collapse but not flatten completely (Fig. 2h and i). For the three species, the vessels formed a row horizontal or perpendicular to the application of the compression strength (Fig. 2b, f, and j). The deformation occurring in the vessels caused the close rays to collapse and lose their original shape in the three species. Most fibers tended to make an “S” shape, especially those that were close to the collapsed vessels (Fig. 2c, d, g, h, k, and l). In the case of the rays of A. acuminata, the frequency of the waves was greater than those of V. guatemalensis and V. ferruginea.