Physical and mechanical properties of two tropical wood ( Detarium macrocarpum and Piptadeniastrum africanum) and their potential as substitutes to traditionally used wood in Cameroon

Abstract

This study aimed to characterise two little-known woods and discuss their potential to replace those heavily exploited in Cameroon's forests. Amouk (Detarium macrocarpum) and Dabema (Piptadeniastrum africanum) were characterised for their physical and mechanical properties. Density, shrinkage, desorption kinetics, shear strength, tensile strength and modulus of elasticity (MOE) in compression and flexion were determined. A Weibull analysis was performed to predict their failure. It was found that Amouk and Dabéma can be classified as medium-heavy wood with low shrinkage and desorption, following Midilli's empirical model. The physical and mechanical characterisations show that the density, strength and MOE parameters of the woods are close to each other and enable them to be classified as medium-strength woods. The probability of failure in compression follows Weibull's law. These woods can, therefore, be used in structural applications such as construction, replacing common woods.

Keywords

characterisation alternative wood tropical woods technological properties

Introduction

In most Central African countries, and particularly in Cameroon, timber is a strategic resource for industrialisation¹ with a variety of species and more than 150² that could individually provide annual production of between 2.97 M m³ and more than 3 M m³ of logs³. According to the Food and Agriculture Organization of the United Nations, the forests of Central Africa cover an area of 200 million hectares,³ making up the second-largest block of tropical forests on the planet⁴ after the Amazon.² Cameroon is the second country in Central Africa with high timber potential,⁵ with 22.5 M hectares of tropical forest.⁶ It accounts for a large share of its formal production, with nearly 700,000 m³ of logs and a processing rate of almost 75% since 2016.⁷ Sawn timber dominates this industrial production in Central Africa, representing 80% of industrial production by volume in 2015⁷ and a transformation rate of around 70% since 2021.²

However, the timber sector in Cameroon and Central Africa makes a significant contribution to national economies⁸ but remains under-industrialised,⁹ as it is limited to primary processing products.⁸ Its contribution to gross domestic product and job creation¹⁰ is not negligible; it has often increased from 3.5% in 1989¹¹ to 6.7% in 1995 and 12% in 2000.¹⁰ It was estimated at 4% in 2016, with 9000 jobs created in logging and 6000 jobs in the timber industry.⁷ According to ITTO in 2007, the second and third wood processing industries generate several activities and products that create jobs.² With this in mind, the Libreville meeting in June 2018, as part of the African Wood Forum, resulted in a strategy for local wood processing by 2030, highlighting 100% local primary processing and 50% local secondary and tertiary processing.⁷ Despite the measures envisaged, these trends pose a severe problem for the industry, which remains focused on four key species,² Okoumé (Aucoumea klaineana) (36%), Sapelli (Entandrophragma cylindricum) (19%), Ayous (Triplochiton scleroxylon) (14%) and Tali (Erythrophleum suaveolens) (8%), which account for the bulk of large-scale production, estimated at around 70%,¹² and which are in danger of running out in the long term.¹³

Other ‘promotional’ but ‘lesser-known’ wood species are still underused, with technological properties similar to those prized and sufficiently available in the forest.¹⁴ Most of the wood species in the forests of Cameroon and other Central African countries suffer from insufficient knowledge of their technological properties,¹³ which leads to the intensive exploitation of traditionally used species and, consequently, their scarcity on the market. Therefore, to diversify the wood species harvested, investigations into the technological properties of lesser-known species could provide a large database of wood species classified according to the similarity of their characteristics. Tropical wood users would therefore have the possibility of choosing among a wide range of available species from a group with similar structural performance.¹⁵ It has been clearly demonstrated that the use of wood in engineering construction requires a good knowledge of its physical properties, such as moisture content, density, PSF, shrinkage and their mechanical properties, such as modulus of rupture (MOR) and longitudinal modulus of elasticity (MOE).¹⁶ In addition, the technological properties of wood are linked to the species’ environment.¹⁷ In Central Africa, and more specifically in Cameroon, Tshiamala-Tshibangu et al.¹⁸ were able to characterise the second-grade woods Essia (Petersianthus macrocarpus) and Mubala (Pentaclethra macrophylla) through the determination of their physicochemical and mechanical properties, thus guiding their use as structural wood. Their study highlighted the silica content, anisotropic ratio, volume shrinkage coefficient, density and axial cohesion characteristics such as compression and bending. However, these results are still insufficient. Mvogo¹⁹ uses non-destructive methods such as vibration MOE and MOR to group woods. Lissouk et al.²⁰ used multi-criteria analysis applied to glued laminated wood to obtain correlation coefficients that guided the different wood groupings. Nsouami et al.²¹ showed that the mechanical behaviour of wood is strongly influenced by density. In addition, Nyobe et al.,¹³ using Monte Carlo simulations, studied the influence of the mechanical strength of tropical woods on the optimal choice criteria for these woods.

Therefore, the literature shows that the use of physical and mechanical properties to elucidate the properties and establish a classification of lesser-known woods from the Congo Basin has only been poorly investigated.

Amouk (Detarium macrocarpum) and Dabema (Piptadeniastrum africanum) from the forests of the South Cameroon Region were chosen for this study, based on their net volume, estimated at around 105.212 m³ and 761.686 m³, respectively,⁷ and their availability in the forest.²²

This work aims to assess these woods’ physical and mechanical properties as the potential substitutes for the most widely used woods.

Methods

Collection and preparation of wood samples

Samples were taken at 1.30 m from the ground, corresponding to the average diameter measured at the breast height of the trees. The requirements recommended by the ASTM D 5536²³ were met in selecting four trees of each species, giving preference to those that did not present singularity defaults such as knots, epicormic shoots or grain slope. The minimum operating diameter (MOD) was calculated as a reference for selecting harvested trees according to the specific requirements of each species. By focusing on this methodology, sampling was more reliable, as described by researchers.^24,25 The trees selected by the certified MOD for each species were closer to mature individuals regarding stem length, free woody generative base and average stand volume.

For each tree harvested, 1.50 m long logs were cut at the base of the trunk between 0% and 25% of the useful length of the stem. The logs were then quarter-sawned to make removing the sapwood easier and obtaining blocks of wood containing no heartwood. The various quarters were sawn into planks using a circular saw, dried, and planned to final dimensions of 1500 mm (length) × 200 mm (width) × 20 mm (thickness). Thirty samples per tree were taken from the resulting planks before any selective sorting, for a total of 240 samples for 120 samples per species, avoiding areas with surface irregularities, knots and grain slopes. They were then cut and measured following the standards recommendations.^26,27 The sampling method, dimensions and shapes of the samples are shown in Figure 1.

Figure 1.

Sampling of the species studied: (a) Amouk (Detarium macrocarpum), (b) Dabema (Piptadeniastrum africanum), (c) log, (d) log quarter-lot, (e) board, (f) samples for physical and mechanical characterisation.

Physical characteristics

Moisture content

The moisture content, calculated using equation 1, was determined by oven drying at 103°C until there was no further weight loss.^28,29 For this purpose, 20 samples of cubic shape with dimensions 20 × 20 × 20 mm³ were used.^31,32

\begin{matrix} MC (%) = \frac{M_{H} - M_{0}}{M_{0}} \times 100, \end{matrix}

(1)

where M_H is the mass of the wet sample in (g) and M₀ is the mass of the dry sample in (g).

Densities

Densities were determined following the standard.³¹ Two sets of 20 cubic samples measuring 20 × 20 × 20 mm³ were used by de Melo et al.³⁰ The volume of the samples was determined by measuring the specimens using a sensitive digital caliper with a resolution of 0.01 mm. The mass was then determined by weighing on a precision balance. Equation (2) was used to determine the wet density with the first set of samples.

\begin{matrix} ρ_{H} = \frac{M_{H}}{V_{H}} \end{matrix}

(2)

Equation (3) were used to determine the anhydrous (ρ_o) and basal (ρ_b) densities, respectively, using the second set of samples. The density (ρ₁₂) at 12% equilibrium moisture was then determined by equation (4).³²

\begin{matrix} ρ_{0} = \frac{M_{0}}{V_{0}}, ρ_{b} = \frac{M_{0}}{V_{S}} \end{matrix}

(3)

\begin{matrix} ρ_{12} = ρ_{H} [\frac{112}{100 + H}] [\frac{100 - r_{v} (H_{s} - H)}{100 - r_{v} (H_{s} - 12)}], \end{matrix}

(4)

with ρ_H the wet density (g/cm³), ρ_o the anhydrous density (g/cm³), ρ_b the basal density (g/cm³), and ρ₁₂ the density at 12% moisture (g/cm³), M_H the mass of the wet sample in (g), V_H its volume in cm³, M₀ the mass of the anhydrous sample in (g) and V₀ its volume in cm³, V_S the saturated volume in cm³, H the moisture content in percentage at the time of measurement, H_S the moisture in percentage corresponding to the fibre saturation point (FSP) and rv the volumetric shrinkage coefficient in percentage.

Fibre saturation point

The FSP corresponds to the limit of moisture content at which the cell walls of the wood material no longer absorb water.³³ According to the literature, the moisture content at FSP varies from 22% to 44% and depends on the wood species.²⁸ The FSP was determined on 20 specimens of 50 × 50 × 10 mm³ ± 0.5 mm³ each. The specimens are those used to determine the various shrinkages, and the data collection method is the same as the one described for the shrinkage calculation. During the drying phase, radial and tangential measurements were taken at regular intervals to plot shrinkage curves as a function of moisture content R = f (H), corresponding to FSP.²⁸

Shrinkage

Shrinkage was measured per the standard.³⁴ Twenty samples measuring 50 × 50 × 10 mm³ ± 0.5 mm³ were used by de Melo et al.³⁰ The samples were immersed in distilled water until the fibres were saturated. The dimensions (R_S and T_S) in the radial (R) and tangential (T) directions were measured using a caliper. The samples were air-dried, and then oven-dried by slowly increasing the temperature to 103 ± 2°C. The consistency of the mass after 24 h indicates the anhydrous state in which the dimensions R₀ and T₀ were measured. The total shrinkage and shrinkage coefficients R_R and R_T; r_R and r_T in the radial and tangential directions were determined using equation (5). During the drying phase, radial and tangential dimensions were taken regularly.

\begin{matrix} R_{i} = \frac{i_{S} - i_{0}}{i_{S}} \times 100, r_{i} = \frac{R_{i}}{H_{s}}, \end{matrix}

(5)

With i (R, T), R_S, R₀, T_S and T₀ the dimensions of the saturated and anhydrous samples in the radial (R) and tangential (T) directions, respectively, H_S the moisture content at the FSP.

For volumetric shrinkage, the volume of the saturated sample V_S after immersion in distilled water was determined. The sample was air-dried and then oven-dried to an anhydrous state to determine the volume V₀. The total volume shrinkage R_V and the volume shrinkage coefficient r_v are determined by equation (6).³⁵

\begin{matrix} R_{V} = \frac{V_{S} - V_{O}}{V_{S}} \times 100, r_{v} = \frac{R_{v}}{H_{s}}, \end{matrix}

(6)

where V_S and V₀ are the volumes of the samples in the saturated and anhydrous states, respectively; H_S is the percentage of moisture content corresponding to the FSP, determined by plotting the shrinkage curve against the moisture curve.

Desorption kinetics

Desorption kinetics were determined by an empirical experimental approach³⁶ on 40 samples of dimensions 50 × 50 × 10 mm³. The ambient temperature of the laboratory varied between 25°C and 28°C and that of the Memmert oven model UF110 was set at 103°C.³⁷ The desorbed moisture content W (%) over a period of time (t) was determined by equation (7).

\begin{matrix} W % = \frac{M_{t} - M_{i}}{M i} \times 100, \end{matrix}

(7)

With M_t_, the mass of the sample at time t, M_i, the mass of the sample in the anhydrous state, while the drying rates (M_R) and the mathematical models derived from the study of the drying kinetics of the two woods were determined. The correlation between the various experimental curves obtained and the desorption models was established. Tables 1 and 2 show the symbols and abbreviations used in this section.

Table 1.

Specification of symbols used.

Symbols	Meaning
MC	Moisture content (%)
Mi	Anhydrous mass (g)
M_t	Saturated mass (g)
M_h	Wet mass (g)
Mt	Mass at time t (g)
R	Perfect gas constant (kJ/mol k)
$M R_{p r e, i}$	Predicted moisture ratio
$M R_{e x p, i}$	Experimental moisture ratio

Table 2.

Specification of abbreviations used.

Abbreviations	Meaning
RMSE	Root mean square errors
SSE	Mean errors of squares
N	Number of observations
P	Number of constants
t	Drying time
r²	Coefficient of determination
T	Drying temperature
a, b, c, ko, k1, k, g, h	Model constants
M_R	Drying ratio
n	Positive integer, essential coefficient for models

The experiments made it possible to determine the desorbed water content of the species studied and the corresponding desorption curves using the steaming method following the standard.³⁸ This method involves soaking wood samples in distilled water for a long period until the mass stabilises after three successive weighs. They were then removed and weighed using a 0.001 g precision digital balance to obtain the masses (M_t) at the time (t) before being introduced into the oven. After getting the initial masses (M_h), they were introduced into a Memmert UF110 oven with a precision of ±5°C set at a temperature of 103°C.³⁹ The samples were reweighed at 15-min intervals until the mass was constant. One of the control samples was taken and weighed at a reduced time to determine the wet mass at the time (t). This sample's dry mass (m_o) was then measured after obtaining a current mass for the defined time (t). The desorption ratio (M_R) is determined from the mass at time M_t, the wet mass M_h and the mass of the material in the anhydrous state M_i according to equation (8).

\begin{matrix} M_{R} = \frac{M_{t} - M_{h}}{M_{i} - M_{h}} . \end{matrix}

(8)

The parameters of the different models were identified using non-linear regression and fitted using Matlab R2020b software. The effectiveness of the different mathematical models was evaluated using statistical criteria such as the correlation coefficient R2, the squared error (SSE) and the root mean square error (RMSE) calculated from equations (9) and (10).

\begin{matrix} RMSE = {[\frac{1}{N} \sum_{i = 1}^{N} {({MR}_{pre, i} - {MR}_{\exp, i})}^{2}]}^{\frac{1}{2}} \end{matrix}

(9)

\begin{matrix} SSE = \frac{1}{N - P} [\sum_{i = 1}^{N} {({MR}_{pre, i} - {MR}_{\exp, i})}^{2}], \end{matrix}

(10)

where

{MR}_{pre, i}

and

{MR}_{\exp, i}

are the predicted and the experimental moisture content for the ith observation, respectively. N is the number of samples and P is the number of constants. Some of the mathematical models tested are shown in Table 3.

Table 3.

Mathematical models tested.

Authors	Functions/models	Reference
Newton and Lawis	$f (t) = e x p (- k t)$	^36,37,39
Page	$f (t) = e x p (- k t^{n})$
Henderson and Pabis	$f (t) = a \exp (- k t)$
Logarithmic	$f (t) = a e x p (- k t) + b t$
Two terms	$f (t) = a e x p (- k_{0} t) + b e x p (- k_{1} t)$
Midilli	$f (t) = a e x p (- k t^{n}) + b t$
Verma et al.	$f (t) = a e x p (- k t) + (1 - a) e x p (- g \times t)$
Peleg	$f (t) = 1 - [t / (a + b t)]$
Aghbashlo	$f (t) = e x p (- k t / (1 + a t))$

Mechanical characterisation

Axial compression test

The axial compression test was carried out using an M&O universal machine (Figure 2) with calibrated dynamometric rings with a capacity of 30 kN.⁴² Twenty samples measuring 20 × 20 × 60 mm³ were used to carry out a test according to the standard⁴⁰ described and used by de Melo et al.³⁰ It consists of positioning sample 11 between supports 4 and activating handwheel 7 until the sample breaks (Figure 2(a)). The stress at break σ_CH and MOE E_CH were determined by equations (11) and (12). Then, equations (13) and (14) were used to calculate the failure stress (σ_C12) and MOE (E_C12) at 12% mc from σ_CH and E_CH.³²

\begin{matrix} σ_{C H} = \frac{P}{S} \end{matrix}

(11)

E_{C H} = \frac{σ_{C H}}{ε}

(12)

\begin{matrix} σ_{C 12} = σ_{C H} [1 + C (H - 12)] \end{matrix}

(13)

\begin{matrix} E_{C 12} = E_{CH} [1 + C (H - 12)], \end{matrix}

(14)

where σ_CH is the fracture stress in MPa, P is the maximum load (N), S is the cross-sectional area of the specimen (mm²), E_CH is the Young's modulus (MPa), ɛ the strain, σ_C12 and E_C12 and the Young's modulus in compression at 12% mc, C the compression strength correction coefficient and H is the mc in percentage (%) corresponding to the FSP.

Figure 2.

Diagram of the compression (a) and bending (b) test.⁴²

Static three-point bending test

The three-point bending test and the samples were carried out following the standard⁴¹ used by Ndapeu et al.⁴² Twenty specimens³¹ of dimensions 20 × 20 × 360 mm³ were tested on a universal tensile testing machine (Figure 2(b)), the principle of which has been described previously. The ultimate tensile stress σ_fH is determined by equation (15); and then the apparent MOE E_fH is determined from the force (P) and strain (f) curves in equation (16). The breaking stress σ_f12 and the MOE E_f12 at 12% moisture content are deduced from σ_fH and E_fH by equations (17) and (18).

\begin{matrix} σ_{fH} = \frac{3 PL}{2 {bh}^{2}} \end{matrix}

(15)

E_{fH} = \frac{L^{3}}{4 {bh}^{3}} (\frac{Δ P}{Δ f})

(16)

\begin{matrix} σ_{f 12} = σ_{fH} [1 + C_{f} (H - 12)] \end{matrix}

(17)

\begin{matrix} E_{f 12} = E_{fH} [1 + C_{f} (H - 12)], \end{matrix}

(18)

where P is the bending force (N), f is the displacement induced by force (mm), the ratio (ΔP/Δf) is taken in the linear region of the bending curve, L is the length between supports (mm), b and h, respectively, are the thickness and height of the sample section (mm), Cf is the bending strength correction coefficient and H is the percentage moisture content (%) corresponding to the FSP.

Tensile strength test

Tensile strength was determined using 20 dog-bone-shaped specimens³¹ of dimensions 20 × 20 × 150 mm³ each (Figure 3(a)) as defined by ASTM D143-94.²⁶ The load is applied progressively until the specimen breaks (Figures 3(b), (c) and (d)). The value of the tensile stress at failure σ_th is given by equation (19), which is the ratio between the maximum load (P) and the most loaded section (S) of the specimen, while the failure stress at 12% mc σ_t12 is deduced from σ_th by equation (20).

\begin{matrix} σ_{tH} = \frac{P}{S} \end{matrix}

(19)

\begin{matrix} σ_{t 12} = σ_{tH} [1 + C (H - 12)], \end{matrix}

(20)

Figure 3.

Tensile procedure: (a) tensile specimens, (b) schematic principle of the longitudinal tensile test, (c) specimen before testing and (d) specimen after testing.

Where σ_tH is the tensile strength,⁴² σ_t12 is the tensile strength at 12% moisture, C is the tensile strength correction coefficient and H is the moisture, in percent, corresponding to the FSP.

Longitudinal shear test

The shear test was performed on prismatic specimens (Figure 4(a)), complying with the standard⁴³ as described by Saoud et al.⁴⁴ Twenty specimens were tested, and the forces were applied until failure (Figure 4(d), (c) and (d)). The shear stress γ_H of moisture H was determined using equation (21).

Figure 4.

Shear test procedure: (a) shear specimens, (b) shear principle diagram, (c) specimen before testing and (d) specimen after testing.

The shear stress at 12% mc γ_H12 is deduced from γ_H by equation (22).

\begin{matrix} γ_{H} = \frac{P}{S} \end{matrix}

(21)

\begin{matrix} γ_{H 12} = γ_{H} [1 + C (H - 12)], \end{matrix}

(22)

where γ_H the shear stress (MPa), P is the shear force in Newton (N), S is the sheared section (mm²), γ_H12 is the shear stress in (MPa) at 12% mc, C is the strength correction coefficient and H the moisture content, in percent, corresponding to the FSP.

Weibull analysis

Weibull analysis is highly appreciated when a considerable scatter in test results is observed.⁴⁵ It predicts microstructural causes and the probability of material failure and survival.⁴⁶ The two-parameter Weibull statistical distribution has the advantage of being expressed as a function⁴⁷ that describes the failure probability of tropical woods with high accuracy.⁴⁸ The density function of the two-parameter Weibull distribution is given by equation 23.⁴⁹ The cumulative probability of failure is obtained by equation (24) integrating the density function.⁵⁰

f (x) = \frac{β}{α} {(\frac{x}{α})}^{β - 1} e^{{(\frac{x}{α})}^{β}}

(23)

f (x) = 1 - e^{{(\frac{x}{α})}^{β}} .

(24)

The value of the Weibull parameter (m) is descriptive of the factors that cause a material to break. Indeed, when m < 20, the failure of the material is due solely to the stress of the material during the test.⁵¹ This parameter m is deduced from the graphical resolution of affine equation 25.

\ln (- \ln (1 - P_{r})) = m \ln (X) - m \ln (X_{0})

(25)

Considered as an equation of the type

Y = m x + 1

with

Y = \ln (- \ln (1 - P_{r})), m x = m \ln (σ), b = m \ln (σ_{0})

.⁵²

The probability of rupture (equation 26)

P_{r} = \frac{i}{N + 1}

(26)

is estimated using the mean rank approximation (Herd Johnson approximation),⁵³ where f(x) is the density of the two-parameter Weibull distribution, β is the shape parameter of the distribution (β = m), x is the independent variable, α is the shape parameter of the distribution, Pr is the probability of failure, m is the Weibull parameter, X and X₀ are the experimental and normalisation magnitudes (MPa) respectively (X and X₀ are E and E₀ for modulus, σ and σ₀ for stress), i is the sample rank, and N is the size of the distribution.

Results and discussion

Physical properties

Moisture content

Moisture content (mc) is a critical characteristic for determining the properties of tropical woods. Table 4 shows that the moisture content of the Amouk wood (24.71 ± 2.70) was significantly higher than that of Dabema (22.65 ± 4.09). Indeed, the difference in variation estimated at 2.06% found for these two woods could be justified by the very slow water loss from this wood during air drying due to its sap's very sticky and viscous nature. According to the literature, the results align with the French standard principles,²⁹ which recommends moisture contents below the FSP (25–30%) for testing clear and flawless wood specimens.⁵⁴

Table 4.

Moisture content of the woods studied.

Species	Amouk	Dabema	p-value
MC (%)	24.71 ± 2.70	22.65 ± 4.09	<.001
Cv (%)	10.92	18.05	<.001

Densities

Table 5 shows the results indicate no significant difference in density between the two studied wood species (p-value = .925). In addition, the small variations of around 1%, 2% and 5% for the density average ρ₁₂, ρo and ρb, respectively, of the species studied (Amouk and Dabema) also show that there is no difference between the two woods. these results are close to previous studies made by Gérard et al.³⁵ and Tropix7 cirad.⁵⁵ These results suggest that the species studied wood can be considered to be of average resistance and, therefore, can be used in construction engineering, flooring, and joinery, as suggested by several authors.^{13,19,55–57}

Table 5.

Average densities and coefficient of variation (in bracket) of Amouk and Dabema and obtained value from literature review.

Density	Amouk	Dabema
$ρ_{12}$	593.20 ± 25.48 (4.29)	597.05 ± 44.24 (7.40)
$ρ_{b}$	503.26 ± 19.68 (3.91)	529.64 ± 18.63 (3.52)
$ρ_{0}$	560.02 ± 25.59 (4.57)	568.17 ± 18.85 (3.32)
³⁵	590–800	590
²²	660–700	////

Fibre saturation point

The hygroscopic behaviour of the species studied is shown in Figure 5. Water diffusion of the species studied took place in two phases: a rapid absorption phase (Phase 1) up to (FSP) and a stabilisation phase (Phase 2). The average (FSP) values obtained were 24.1% for Amouk and 26.7% for Dabema. These values are close to those reported in the literature (24% and 27%, respectively),^35,55 with a marginal difference of around 1% compared to those reported by Gérard et al.³⁵ and Tropix7 Cirad.⁵⁵

Figure 5.

Shrinkage curves as a function of moisture content.

Shrinkage and shrinkage coefficients

The shrinkage tests indicate no significant difference between the species studied for radial and volume shrinkage, except for tangential shrinkage where this is noticeable (Table 6). It can be seen that both species have very low radial shrinkage, but deformations underwater load in the radial direction will be less significant. According to these results and certain authors these wood species could qualify as medium-shrinkage species.^55,58 It is also important to note that the very low values of shrinkage coefficients in the radial direction obtained (0.25 ± 0.03 and 0.29 ± 0.01, respectively, for Amouk and Dabema) could be described as homogeneous and dimensionally stable species.^35,55 Based on these results, the studied species could be recommended for work in which dimensional changes are less significant, such as interior joinery, flooring, panelling and heavy structural work (e.g., heavy frames, staircases, bonded laminates, vehicle or container bottoms).

Table 6.

Average values for the various shrinkages of Amouk and Dabema wood species.

	Direction of shrinkage	Amoukmean (%)	Dabemamean (%)	p-value
Shrinkage	Rr	3.50 ± 0.57	3.80 ± 0.36	.155 <.001 .868
	Rt	7.50 ± 1.23	7.80 ± 1.29
	Rv	10.09 ± 2.19	14.71 ± 0.83
Shrinkage coefficients	r_r	0.25 ± 0.03	0.29 ± 0.01
	r_t	0.63 ± 0.08	0.29 ± 0.01
	v_v	0.42 ± 0.09	0.61 ± 0.03

Desorption kinetics

The results for desorbed moisture content showed that Amouk (123.55 ± 9.73) wood loses more and significantly water (p-value˂ .001) due to its higher desorbed moisture content than Dabema (90.64 ± 6.14) as found elsewhere by others authors.⁵⁹ It was found the mass of Amouk wood takes longer to stabilise at around 600 min (Figures 6(a) and 7) while the mass of Dabema wood stabilises at around 450 min (Figures 6(b) and 7). It was found that the modelling of desorption kinetics as a function of time for the single temperature of 103°C for the species studied was similar. At the start of the process, we also observed rapid drying, which is not new as it is well-known from the work of other authors.^60,61 Table 7 shows that Midilli's model (f (t) = a × exp (−k × tn) + b × t) tested using Matlab software shows the best correlation for both woods. It could, therefore, be chosen as the reference model that best predicts the moisture behaviour of the species studied during drying.

Figure 6.

Modelling of wood desorption kinetics, (a) Amouk and (b) Dabema.

Figure 7.

Experimental curves for Amouk and Dabema species.

Table 7.

Desorption kinetics parameters used for the hygrometric behaviour of the species studied.

Authors	Functions	Species	SSE	R-Square	k	n	A	b	k₀	k₁	c	g	h
Newton and Lawis	$f (t) = e x p (- k \times t)$	Dabema	0.5255	0.9074	0.004307
Newton and Lawis	$f (t) = e x p (- k \times t)$	Amouk	0.6431	0.8864	0.003396
Page	$f (t) = e x p (- k \times t^{n})$	Dabema	0.007121	0.9987	2.04E−05	1.962
Page	$f (t) = e x p (- k \times t^{n})$	Amouk	0.01355	0.9976	8.78E−06	2.036
Henderson and Pabis	$f (t) = a \times e x p (- k \times t)$	Dabema	0.3022	0.9467	0.005153		1.226
Henderson and Pabis	$f (t) = a \times e x p (- k \times t)$	Amouk	0.3904	0.931	0.004105		1.221
Logarithmic	$f (t) = a \times e x p (- k \times t) + b \times t$	Dabema	0.1494	0.9737	0.00388		1.162	−2.77E−04
Logarithmic	$f (t) = a \times e x p (- k \times t) + b \times t$	Amouk	0.1018	0.982	0.002239		1.121	−5.62E−04
Two-term	$f (t) = a \times e x p (- k_{0} \times t) + b \times e x p (- k_{1} \times t)$	Dabema	0.03357	0.9941	0.01033		71.43	−70.46	−0.1618	0.01054
Two-term		Amouk	0.3934	0.9305	0.004015		1.185	0.01117	−0.1909	0.003951
Midilli	$f (t) = a \times e x p (- k \times t^{n}) + b \times t$	Dabema	0.006153	0.9989	2.81E−05	1.903	1.01	−1.56E−05
Midilli	$f (t) = a \times e x p (- k \times t^{n}) + b \times t$	Amouk	0.006823	0.9988	1.49E−05	1.931	1.002	−6.47E−05
Ferma et al	$f (t) = a \times e x p (- k \times t) + (1 - a) \times e x p (- g \times t)$	Dabema	0.03698	0.9935	0.01009		37.72					0.01046
Ferma et al		Amouk	0.07455	0.9868	0.008173		56.78					0.008367
Modify Anderson and Pabis	$f (t) = a \times e x p (- k \times t) + b \times e x p (- g \times t) + c \times e x p (- h)$ *t)	Dabema	0.1345	0.9763	0.000815		−0.7094	0.0789			1.747	0.002594	2.62E−03
Modify Anderson and Pabis		Amouk	0.04357	0.9923	0.009119		21.68	−21.64			0.9637	0.01008	0.02289
Peleg	$f (t) = 1 - [t / (a + b \times t)]$	Dabema	0.2647	0.9534			305.8	0.42
Peleg	$f (t) = 1 - [t / (a + b \times t)]$	Amouk	0.2127	0.9624			548	0.00157
Aghbashlo	$f (t) = e x p (- k \times t / (1 + a \times t))$	Dabema	0.06949	0.9878	0.002372		−0.00155
Aghbashlo	$f (t) = e x p (- k \times t / (1 + a \times t))$	Amouk	0.03468	0.9939			−0.00154

Mechanical properties

Axial compression and static three-point bending test

It was that Amouk shows a plastic deformation (Figure 8(a)) after the linear elastic phase compared with Dabema (Figure 8(b)), which gives it a ductile character. In addition, both timbers showed abrupt failures during the bending test (Figure 8(b)) just after the plastic limit. This indicates that these timbers have brittle bending behaviour (Figure 8(b)). Generally, both woods showed behaviour close to a linear phase corresponding to the elastic zone, a slow failure phase in compression by crushing and a rapid failure phase in bending.

Figure 8.

Mechanical behaviour of the species studied in (a) compression and (b) flexion.

The mean values of MOR and MOE for Dabema are significantly higher than those for Amouk in both cases (Table 8). Although these results are dispersive for certain values, they are close to those found in the literature.^35,55 Amouk has a lower compressive strength (44.07 ± 4.11 MPa) than Dabéma (68.60 ± 6.97 MPa), which could be influenced by its high moisture desorption rate (123.55% ± 9.73) compared to Dabema (90.64% ± 6.14). The longitudinal MOE of Dabema in the present study is 14,638 ± 1556.58 MPa, close to the 15,190 Mpa obtained by Tropix7 cirad⁵⁵ and far from 12,300 Mpa found by Gérard et al.³⁵ This difference could be due to the high concentration of stresses applied to a single central point, as described in the principle of the three-point static bending test used in this study.⁶² In addition, the variability of the sampling area, growing conditions and the non-homogeneity of the species studied within the same species could also explain these differences.^33,63

Table 8.

Modulus of elasticity and modulus of rupture for compression and bending tests.

	Properties	Amoukmean (MPa)	Dabemamean (MPa)	p-value
Axial compression test	$σ_{c}$	37.01 ± 3.45(9.32)	67.93 ± 6.89(10.14)	<.001<.001
	$σ_{c 12}$	55.83 ± 5.20(14.06)	68.60 ± 6.97(10.16)
	Ec	880.39 ± 82.48(9.37)	1258.57 ± 185.78(14.76)
	Ec₁₂	1187.10 ± 111.22(12.63)	1384.12 ± 204.32(14.76)
Static bending test	$σ_{f}$	80.96 ± 6.64(8.20)	88.73 ± 8.17(9.21)	<.001.0499
	$σ_{f 12}$	100.39 ± 8.23(8.20)	102.04 ± 9.40(9.21)
	E_f	9455.13 ± 1361.20(14.40)	13,554.21 ± 1441.28(10.63)
	E_f12	14,031.41 ± 1501.00(10.69)	14,638.55 ± 1556.58(10.63)

The lower moisture content of Dabema (22.65%) compared with Amouk (24.71%) also contributed to the improvement in its mechanical performance. Young's and rupture modulus strongly depend on this parameter.⁶⁴ According to the literature, a decreasing trend with increasing moisture content is observed for Young's moduli of tropical woods in all orthotropic directions.³³ This relevant observation is justified in this study for Amouk. These results are in line with the work reported by Mvondo et al.,³¹ which gives an overview of the literature data for Young's moduli as a function of moisture of several Cameroonian tropical wood species Bilinga, Iroko and Tali below the FSP. On the other hand, the very low values of MOE and MOR for Amouk subjected to compression show that this wood becomes more ductile as moisture content (MC) increases. Because of their very similar properties, the bending and compression failure stresses at 12% (MC) would allow these species to be classified as medium-stress wood.⁶⁴

Tensile and shear strength test

It was found that be seen that the maximum tensile and shear strengths and stresses at the break of Dabema wood are higher than those of Amouk wood (Table 9). This can be explained by its low MC. This theory of the influence of the MC of hardwoods in the Congo Basin has been justified by several authors.^1,32 The values obtained for the tensile test are close to those obtained in the work of Mvondo et al.³¹ for Bilinga (Nauclea Diderrichii) and Iroko (Milicia excelsa) woods.

Table 9.

Average values for tensile strength (σ_t) and shear strength (γ_LT).

Properties		Amoukmean (MPa)	Dabemamean (MPa)	p-value
Tensile test	$σ_{t}$	57.28 ± 5.18 (9.04)	97.00 ± 8.81 (9.08)	<.001
Tensile test	$σ_{t 12}$	85.97 ± 7.80 (9.07)	112.50 ± 8.85 (7.87)	<.001
Shear strength test	$γ_{L T}$	6.00 ± 0.54 (9.00)	12.00 ± 1.09 (9.08)	<.001
Shear strength test	$γ_{L T 12}$	9.05 ± 0.82 (9.06)	13.92 ± 1.09 (7.83)	<.001

Weibull distribution

Figure 9(a) and 9(c) shows the compression linearisation curves for the two species studied and the parameters of the two-factor Weibull analysis. The Weibull parameters (Table 10) after the compression test of the species studied show that the failure of these orthotropic materials follows a Weibull law. Similar results have been observed for various tropical hardwood species in the literature.⁵³ The Weibull modulus (m) of between 4.34 and 9.87 shows that the material has a fracture rate comparable to that of Andiroba,⁶⁵ Azobé⁶⁶ and Gmelina arborea, which is an angiosperm of the Lamiaceae family and the genus Gmelina.⁴⁸ In view of these results, the woods studied could be an alternative to these species for construction applications such as parquetry, staircases, joinery, and in the furniture industry for cabinetmaking, turning, veneering and structural purposes,²² Standard.⁶⁴ However, the use of wood for structural applications is envisaged based on knowledge of its mechanical properties.³³ In addition to fracture parameters, we have the density at 12% (MC), a relevant indicator for the classification of wood according to the standard⁶⁴ used by Eurocode 5 in structural applications.

Figure 9.

Weibull compression parameters of the species studied for (a) stress, (b) Young's modulus, (c) probability of failure for stress, (d) probability of failure for Young's modulus.

Table 10.

Weibull parameters of the species studied.

	Compressive stress (MPa)				Young's modulus (MPa)
Species studied	m	b	σ₀	R²	M	B	E₀	R²
Amouk	8.28	34.69	65.86	0.97	9.87	70.26	1234.65	0.96
Dabema	4.78	20.50	72.87	0.95	4.34	32.37	1734.59	0.95

The very low Weibull modulus for both wood species (m < 20) reflects the fact that specimen failure is solely due to specimen compressive stress during testing and not crack propagation, showing the potential of Amouk and Dabema for engineering purposes.⁴⁸ Figure 9(b) and 9(d) show Herd Johnson fracture probability curves, from which it is evident that the curves represent slopes that decrease with increasing stress.^48,53

Classification of the properties of central African woods into similar groups

The physical and mechanical characteristics of the species studied make it possible to consider possible substitution by other woods with similar properties. Table 11 was drawn up to compare some previous results available in the literature^13,15,67 with Amouk wood (D. macrocarpum) and Dabema wood (P. africanum). The EN 338 standard⁶⁴ has made it easier to decide which species could be considered substitutes for species heavily exploited in wood-processing units and dense forests in Central Africa.⁶⁷ Such approaches have been addressed by other authors such as Lissouck et al.,¹ Nyobe et al.,¹³ Sout et al.,⁶⁸ and Akter et al.⁶⁹ who have characterised orthotropic materials for the design of structural elements made from engineered wood products. Cirad can envisage areas of use for these two woods in construction applications such as parquetry, staircases and joinery; for the furniture industry such as cabinet making, turning, and veneering; and for structural purposes.²²

Table 11.

Characteristic properties of Amouk (Detarium macrocarpum) and Dabema (Piptadeniastrum africanum) as well as the proposed alternative woods species for substitution.

Wood with similar properties	Density(g/cm³)	Coefficient of volumic shrinkage (%)	Total radial shrinkage (%)	Total tangential shrinkage(%)	FSP(%)	MORcompression12% MC(MPa)	MORbending12% MC(MPa)	MOEbending12% MC(MPa)	Reference
Acajou	0.57	0.39	3.7	5.5	28	46	77	11,820	^22,35
Amouk	0.66	0.38	3.8	5.4	24	55	99	13,100
Assaméla	0.70	0.50	3.2	5.9	20	64	93	13,140
Bete	0.66	0.44	4.6	7.4	28	60	110	13,620
Bilinga	0.76	0.55	4.5	7.5	25	63	95	14,660
Bossé	0.60	0.45	4.1	6.8	31	55	95	12,650
Dabema	0.70	0.55	3.8	8.5	27	57	98	15,190
Dibetou	0.60	0.43	3.7	5.8	27	47	72	10,460
Doussié	0.80	0.44	3.0	4.4	19	74	124	17,020
Fraké	0.54	0.42	4.3	6.1	28	47	80	11,750
Framiré	0.50	0.37	3.6	5.2	27	44	71	11,350
Iroko	0.64	0.44	3.5	5.4	23	54	87	12,840
Kosipo	0.69	0.42	4.8	6.7	32	53	87	11,190
Kotibé	0.76	0.50	5.1	7.5	30	67	120	13,020
Movingui	0.73	0.54	3.6	5.8	23	64	116	14,740
Padouk	0.79	0.44	3.2	5.0	21	65	116	15,870
Sapelli	0.69	0.47	5.0	7.2	29	62	102	13,960
Sipo	0.62	0.42	4.6	6.4	30	56	91	13,240
Zingana	0.79	0.56	8.8	11.0	30	62	110	17,520
Amouk	0.59	0.42	3.5	7.5	24.1	55.83	100.39	14,031.41	Present study
Dabema	0.60	0.61	3.8	7.80	26.7	68.60	102.04	14,638.55	Present study

Conclusion

This study was carried out to improve the knowledge of the physical and mechanical properties of tropical woods from Central Africa as alternatives to the most widely exploited woods. Mechanical and physical characterisations were carried out on samples of Amouk and Dabema wood. Physical properties such as 12% (mc), anhydrous densities and volume shrinkage were determined, while mechanical properties were tested in flexure, compression, tension and shear. A statistical Weibull analysis of the compressive properties was also carried out. The physical properties obtained classify these woods as medium shrinkage species and, therefore, as medium-heavy. Mechanical properties such as MOR and axial MOE showed that the two species have similar properties, enabling them to be considered as medium-strength woods according to the criteria of EN 338 used by Eurocode 5. These properties suggest that these two species could replace Iroko and Sapelli, commonly used and overexploited in the Congo Basin for structural, joinery and cabinet-making applications. Statistically, the fracture properties obtained in compression can be predicted by Weibull's law, and the fracture observed in Amouk (D. macrocarpum) and Dabema (P. africanum) is due solely to the stress exerted on the samples.

Footnotes

Acknowledgements

The authors are grateful to Nicodem Rodrigue Tagne Sikame, Fabien Kenmogne and Fonyuy Godwin Banyuy for their contribution to the review of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Lissouck

Raïssa Ngono Mvondo

René Ateba

, et al.

Investigation of green-glued laminated timber from the Congo Basin: durability, mechanical strength and variability trends of the bondlines.

Southern Forests 2022; 84: 225–241.

Eba’a Atyi

Hiol Hiol

Lescuyer

, et al. Bogor, Indonésie : CIFOR Les Forêts Du Bassin Du Congo : État Des Rorêts 2021. 2021.

Ngono

Raïssa

Oum Lissouck

, et al.

Investigation on mechanical and thermal properties related to hygroscopicity of two African hardwoods.

Wood Material Science and Engineering 2021; 0: 1–12. https://doi.org/10.1080/17480272.2021.1967447

Diop

Vedrine

. “The Forest Taxation and Log Export Ban Effect on Deforestation : Evidence from Cameroon *.”. 2020.

Karsenty

. Géopolitique Des Forêts d’Afrique Centrale. Hérodote N° 2020; 179: 108–129.

Abdou

Habibou

Vasseur

, et al. Valorisation et promotion des essences forestières camerounaises Peu Ou non utilisées. In: 5èmes Journées Du GDR 3544 « Sciences Du Bois ». Bordeaux: I2 M Bordeaux, 2016, pp.1–4.

Temgoua

, et al.

Land use and land cover dynamics in the Melap forest reserve, west Cameroon: implications for sustainable management.

Geology, Ecology, and Landscapes 2022; 6: 305–315.

FRMi. Vision Stratégique et Industrialisation de La Filiere Bois Dans Les 6 Pays Du Bassin Du Congo Horizon 2030. 2018.

Mfomo

, et al.

Carbonization techniques and wood species influence quality attributes of charcoals produced from industrial sawmill residues in eastern Cameroon.

Bois et Forets des Tropiques 2020; 345: 63–72.

10.

de Wasseige

Tadoum

Eba’a Atyi

, et al. The Forests of the Congo Basin - Forests and Climate Change . 2015. http://www.observatoire-comifac.net/.

11.

Forgha

Aquilas

. The impact of timber exports on economic growth in Cameroon: an econometric investigation. Asian Journal of Economic Modelling 2015; 3: 46–60.

12.

Eba’a

, et al. Lutte Contre La Déforestation Importée et Engagements En Faveur de La Zéro Déforestation. Les Forêts du Bassin du Congo: État des Forêts 1998; 2021: 225–251.

13.

Perry

Bediang

. Etude La Filière Bois Au Cameroun : Identification Des Interventions Porteuses d’emplois. Rapport Final Du Ministère de l’Emploi de La Formation Professionnelle. Yaounde. 2009.

14.

Nyobe

Lissouck

Ayina Ohandja

, et al.

Variability of the mechanical strength of Congo Basin timbers.

Wood Material Science and Engineering 2022a; 17: 210–220.

15.

Biologie

Haag

. “Anatomisch-Strukturelle Und Topochemische Untersuchungen Zur Charakterisierung Der Holzeigenschaften Neu Eingeführter Handelshölzer (Lesser Known Species).” 2020.

16.

Mvogo

Ohandja

LMA

Minsili

, et al. Mechanical grading of structural timber and species conservation in the forest of the Congo Basin. African Journal of Environmental Economics and Management 2020; 8: 1–15.

17.

Boussougou

, et al. “Étude de l ‘ Influence Des Conditions Hydriques Sur Le Comportement Mécanique Différé Des Bois Tropicaux : Expérimentation et Modélisation To Cite This Version : HAL Id : Hal-03440580 Etude de l ‘Influence Des Conditions Hydriques Sur Le Comportement Méc.” 2021.

18.

Agnès

Cédric

Olivier

, et al. “Modèles Prédictifs Du Comportement Physico-Mécanique de Pièces de Bois Pour La Conception Mécanique.”: 20–22. 2018.

19.

Tshiamala-Tshibangu

Foudjet

Tchoffo

, et al.

Valorisation des bois secondaires du cameroun par la connaissance de leursteneur en silice, ainsi que de leurs propriétés tant physiques que mécaniques: cas de essia et de mubala.

Holzforschung 1993; 47: 62–67.

20.

Mvogo

. “Regroupement Mécanique Par Méthode Vibratoire Des Bois Du Bassin Du Congo.” : 1–165. 2008.

21.

Lissouk

Pommier

Taillandier

, et al. “Ingénierie Du Bois Comme Outil de Conservation de La Diversité Des Essences Ligneuses Des Forêts Tropicales : Cas Du Bassin … Ingénierie Du Bois Comme Outil de Conservation de La Diversité Des Essences Ligneuses Des Forêts Tropicales : Cas Du Bassin Du.” XXXe Rencontres AUGC-IBPSA (November 2016): 0–10. 2012.

22.

Nsouami

Boussougou

Bastidas-Arteaga

, et al.

Effects of long-term loading on moabi wood beams in the tropical environment of Gabon: variability in properties and effects of exposure conditions on mechanical properties in 3-point bending tests.

Procedia Structural Integrity 2021; 37: 576–581.

23.

Cirad. Description du bois. Tropix 2011; 7: 1–4.

24.

ASTM D 5536. Standard practice for sampling forest trees for determination of clear wood. Methods 2017: 1–10.

25.

Mvolo

Stewart

Koubaa

. Comparison between static modulus of elasticity, non-destructive testing moduli of elasticity and stress-wave speed in white spruce and lodgepole pine wood. Wood Material Science and Engineering 2022; 17: 345–355.

26.

Reis

, et al. Physico-Mechanical properties of the wood of freijó, Cordia goeldiana (boraginacea). Produced in a Multi-Stratified Agroforestry System in the Southwestern Amazon. 2021; 51: 171–180.

27.

ASTM D143-94. Standard test methods for small clear specimens of timber. American Society for Testing and Materials - ASTM. Annual Book of ASTM 1994; 94: 31.

28.

NF B51-003, NF. “Conditions Générales d’essais: Essais Physiques et Mécaniques.” (ISSN 0336-3931): 155–58. 1985.

29.

Fouotsa

, et al. “Contribution to the Study of Variations of Physical Properties of Pericopsis Elata with Respect to Different Stages of Growth Contribution to the Study of Variations of Physical Properties of Pericopsis Elata with Respect to Different Stages of Growth.” (April 2020). 2019.

30.

NF B51-004. “Bois- Détermination de l’humidité.” Afnor: 159–61. 1985a.

31.

Mvondo

RRN

, et al.

Influence of water content on the mechanical and chemical properties of tropical wood species.

Results in Physics 2017; 7: 2096–2103.

32.

Rodolfo de Melo

, et al.

Ultrasound to determine physical-mechanical properties of Eucalyptus camaldulensis wood.

Wood Material Science and Engineering 2021; 16: 407–413.

33.

NF B51-005. “Bois - Détermination de La Masse Volumique.” Afnor: 11. 1985.

34.

Engonga Edzang

, et al.

Comparative studies of three tropical wood Species under compressive cyclic loading and moisture content changes.

Wood Material Science and Engineering 2021; 16: 196–203.

35.

Guitard

. Mécanique Du Matériau Bois et Composites. Cépaduès 1987.

36.

NF B51-006. “Bois - Détermination Du Retrait.” Afnor: 9. 1985.

37.

Gérard

Edi Kouassi

Daigremont

, et al.

Synthèse Sur Les Caractéristiques Technologiques de Référence Des Prinvipeaux Bois Commerciaux Africains.

Cirad-Forêt 1998: 1–185.

38.

Ndapeu

, et al. “Etude Expérimentale de La Cinétique de Séchage Des Coques de Noix de Coco (Nucifera) Du Cameroun.” 2013: 822–30. 2013.

39.

Ganou

BMK

. “Construction En Brique de Terre Comprimée et Granulats Biosourcés : Une Solution Pour Un Habitat Durable à Douala Par Thèse de Doctorat l ‘ Université de Liège (ULiège) et l ‘ Université de Douala (UDou Ala) Préambule “ Notre Vie Vaut Ce Qu ‘ Elle Nou.” Thèse de Doctorat Ph.D: 216. 2021.

40.

NF P94-050. 1995. “Détermination de La Teneur En Eau Pondérale Des Matériaux - Méthode Par Étuvage.” : 7P.

41.

Koungang

BMG

, et al. Physical, water diffusion and micro-structural analysis of ‘Canarium Schweinfurthii engl.’. Materials Sciences and Applications 2020; 11: 626–643.

42.

Ndapeu D , Demze N, Sikame T, Ganou

Njeugna

. Characterization of a brake lining composite based on aiele fruit cores (Canarium Schweinfurthii) and palm kernel fibers (Elaeis Guineensis) with a urea-formaldehyde matrix. International Journal of Scientific & Engineering Research 2020; 11(5).

43.

NF B51-007. “Essai de Compression Axiale.” Afnor: 173–75. 1985.

44.

NF B51-008. “Bois - Essai de Flexion Statique - Détermination de La Résistance à La Flexion Statique de Petites Éprouvettes sans Défaut.” Afnor. 2017.

45.

NF B51-017. “Bois - Traction Parallèle Aux Fibres - Détermination de La Résistance à La Rupture En Traction Parallèle Au Fil Du Bois de Petites Éprouvettes sans Défaut.” Afnor. 1988.

46.

Saoud

, et al.

Technological characterization and correlation between mechanical and physical properties of the wood of Thuja (Tectracanalis articulata) of the Khemissat region in Morocco.

ARPN Journal of Engineering and Applied Sciences 2018; 13: 632–637.

47.

Maache

, et al.

Characterization of a novel natural cellulosic fiber from Juncus effusus.

Carbohydrate Polymers 2017; 171: 163–172.

48.

Roy

, et al.

Improvement in mechanical properties of jute fibres through mild alkali treatment as demonstrated by utilisation of the Weibull distribution model.

Bioresource Technology 2012; 107: 222–228.

49.

Barbosa

, et al.

Analysis of the fatigue life estimators of the materials using small samples.

J Strain Analysis 2018: 1–12.

50.

Iwuoha

Seim

Onyekwelu

. Mechanical properties of Gmelina arborea for engineering design. Construction and Building Materials 2021; 288: 123123.

51.

Djeghader

Redjel

. Effect of water absorption on the Weibull distribution of fatigue test in jute-reinforced polyester composite materials. Advanced Composites Letters 2019; 28: 1–11.

52.

Naresh

Shankar

Velmurugan

. Reliability analysis of tensile strengths using Weibull distribution in glass/epoxy and carbon/epoxy composites. Composites Part B 2017: 1–38.

53.

Redjel

Remadnia

Chaplain

. “Utilisation Du Modèle Probabiliste de Weibull à La Caractérisation de l ‘ Aspect Aléatoire de La Rupture En Traction de Panneaux En Bois à Lamelles Orientées Abstract :” In 23eme Congrès Français de Mécanique, lille, 1–10. 2017.

54.

Sodoke

Toubal

. Hygrothermal effects on fatigue behavior of quasi-isotropic flax/epoxy composites using principal component analysis. Journal of Materials Science 2016; 51: 10793–10805.

55.

Torabi

Ghouli

Marșavina

, et al.

Mixed mode fracture behavior of short-particle engineered wood.

Theoretical and Applied Fracture Mechanics 2021; 115: 103054.

56.

Ozigi

Jimoh

Rahmon

, et al. “Investigation on Some Physical and Mechanical Properties of Apa (Afzelia bipindensis) Timber Grown in Kwara State, Nigeria.” (March). 2018.

57.

Tropix7 cirad. “Tropix7 Cirad.”. 2011.

58.

Lissouck

Pommier

Castéra

. “Ingénierie Du Bois Comme Outil de Conservation de La Diversité Des Essences Ligneuses Des Forêts Tropicales : Cas Du Bassin Du Congo.” (November 2016). 2012.

59.

Mvondo

RRN

Damfeu

Meukam

, et al.

Influence of moisture content on the thermophysical properties of tropical wood Species.

Heat and Mass Transfer/Waerme- und Stoffuebertragung 2020; 56: 1365–1378.

60.

Langbour

Gérard

Guibal

, et al. Caractérisation technologique et valorisation En Bois d’oeuvre Du Pin l’Alep (Pinus halepensis) de La Régon Provence-Alpes-Côte d’Azur. forêt méditerranéenne t. XXXII 2008; 1985: 263–270.

61.

Nsouandélé

Bonoma

Simo Tagne

, et al.

Determination of the diffusion coefficient of water in the tropical woods | determination Du coefficient de diffusion de l’eau Dans Les Bois Tropicaux.

Physical and Chemical News 2010; 54: 61–67.

62.

Simo-Tagne

Chinenye Ndukwu

Ndi Azese

. Experimental modelling of a solar dryer for wood fuel in Epinal (France). Modelling—International Open Access Journal of Modelling in Engineering Science 2020; 1: 39–52.

63.

Ali

, et al.

Hydrothermal modification of wood : a review.

polymers 2021; 13: 1–18.

64.

Vörös

Németh

Báder

. The effect of different moisture contents on selected mechanical properties of wood. Gradus 2019; 6: 75–81.

65.

Brito

, et al. Physical and mechanical properties of thermally modified short-rotation wood in a closed system. International Wood Products Journal 2023; 14: 94–101.

66.

BS EN 338. Structural timber-strength classes. British Standards 2003: 1–10.

67.

De Melo

Pellicane

De Souza

. Goodness-of-Fit analysis on wood properties data from six Brazilian tropical hardwoods. Wood Science and Technology 2000; 34: 83–97.

68.

Kuilen

JWGvd

Blass

. Mechanical properties of azobe (Lophira alata). Original Arbeiten 2005; 5: 1–10.

69.

Cunha

, et al.

Mechanical characterization of Iroko wood using small specimens.

Buildings 2021; 116: 11.

70.

Sout

, et al.

Mechanical qualification of green glued laminated timbers from the Congo Basin towards preservation of forest species diversity.

Journal of Tropical Forest Science 2022; 34: 358–369.

71.

Akter

Binder

Bader

. Moisture and short-term time-dependent behavior of Norway spruce clear wood under compression perpendicular to the grain and rolling shear. Wood Material Science and Engineering 2023; 18: 580–593.