Alberto García-Iruela

Sorption/desorption hysteresis revisited. Sorption properties of Pinus pinea L. analysed by the parallel exponential kinetics and Kelvin-Voigt models

Abstract The hygroscopic properties of Pinus pinea L. wood at 35 and 50°C were investigated by the dynamic vapour sorption (DVS) technique. The sorption kinetic behaviour was studied through the parallel exponential kinetics (PEK) model, which is subdivided into a fast and a slow process. The parameters obtained by PEK were interpreted based on the Kelvin-Voigt (KV) model to determine elasticity and viscosity values of the wood cell wall. The PEK data perfectly fit the experimental data. The temperature-dependent transition between the fast and slow processes is fluent. The slow process contributes more to the total hysteresis of sorption. The kinetic properties varied in relation to the type of cycle and the temperature. The moduli of elasticity and viscosity were higher in the slow process than in the fast one. In both processes, the moduli showed a decreasing tendency in relation to relative humidity.

Wood anatomy of Tetraclinis articulata from its natural distribution area in southeast Spain

For the first time, the wood anatomy of Tetraclinis articulata (Vahl) Masters has been studied using representative samples from its natural distribution area in Spain, in Sierra de Cartagena (Region of Murcia). Mature wood was collected from five individuals representative of the forest stand and their anatomy was compared with other genera of the Cupressaceae. Axial tracheids without helical thickenings, low homogeneous rays, cupressoid pits and the absence of normal axial resin canals are characteristic features of this monotypic genus, as they are of most other Cupressaceae genera. An obvious warty layer separates this wood from the genera sharing its territory (Cupressus and Juniperus) and its semi-spherical, slightly anastomosed warts distinguish it from other, geographically distant genera (Actinostrobus and Callitris). The presence of traumatic axial resin canals is reported for the first time and supports the occurrence of this feature outside the Pinaceae. The wood anatomical diversity within the clade comprising Tetraclinis, Microbiota and Platycladus, as reconstructed by molecular analysis, is discussed.

Comparative wood anatomy of Juniperus from Macaronesia

The wood anatomy of the three species of Juniperus occurring in Macaronesia is compared for the first time using representative samples of each species collected in its natural region of provenance: J. cedrus Webb & Berthel and J. phoenicea L. var. canariensis Guyot, in the Canary Islands, and J. brevifolia (Seub.) Antoine, in the Azores. The three species are anatomically similar, although some qualitative differences were observed: distribution of axial parenchyma very scarce in J. phoenicea compared with the other two species, presence of crassulae only in J. phoenicea, presence of torus extensions and notches on pit borders in the radial walls of J. brevifolia, and ray parenchyma end walls slightly nodular in J. cedrus as opposed to very nodular in J. phoenicea and J. brevifolia. In addition, the biometry of tracheid pit diameter in the radial walls, ray height in number of cells, and largest and smallest diameters of cross-field pits shows differences for a significance level of 95%.

Study of the influence of the physical properties of particleboard type P2 on the internal bond of panels using artificial neural networks

Abstract Particleboard panels are normally manufactured in three layers of different sized particles of wood. One of the most important properties of particleboard is internal bond. This study determined the thickness and the physical properties of swelling, water absorption and density of 300 type P2 particleboards, as well as tensile strength perpendicular to the plane of the board, to examine the influence of these physical properties on internal bond of panels. To study the influence on internal bond, a multilayer perceptron artificial neural network was designed using the hyperbolic tangent sigmoid as the transfer function. The artificial neural network designed is capable of explaining at least 82% of the variability of the samples, with no significant differences between the experimental values and those obtained through the network for a significance level of 95%. The neural network proposed is suitable for studying the influence of the physical properties on internal bond, revealing a decrease in internal bond as panel thickness increases. A slight increase in internal bond was observed as swelling and absorption increase to values close to the mean, followed by a decrease. In relation to density, internal bond increases to values of about 700 kg/m3, then decreases.

Application of artificial neural networks as a predictive method to differentiate the wood of Pinus sylvestris L. and Pinus nigra Arn subsp. salzmannii (Dunal) Franco

The wood structure of conifers in general and the Pinus genus in particular makes species differentiation by traditional qualitative or quantitative methods complicated or even impossible at times. Pinus sylvestris L. and Pinus nigra Arn subsp. salzmannii (Dunal) Franco are a clear example of this because they cannot be differentiated by traditional methods. However, correctly identifying these species is very important in some cases as they are extensively used in a large variety of fields because of their wide distribution range in the forests of Europe and Asia. Using trees selected from the same forest to minimise the influence of site and performing a biometric study of 10 growth rings from the same climate period, a feedforward multilayer perceptron network trained by the resilient backpropagation algorithm was designed to determine whether the network could be used to differentiate these species with a high degree of probability. The artificial neural network achieved 90.4% accuracy in the training set, 81.6% in the validation set and 81.2% in the testing set. This result justifies the use of this tool for wood identification at anatomical level.

Estimating the Uncertainty of a Multilayer Perceptron Using the Monte Carlo Method

Artificial neural networks have become a powerful modeling tool. However, although they obtain an output with very good accuracy, they provide no information about the uncertainty of the network or its coverage intervals. This study describes the application of the Monte Carlo method to obtain the output uncertainty and coverage intervals of a particular type of artificial neural network: the multilayer perceptron.