Predicting mechanical degradation indicators of silver fir wooden strips using near infrared spectroscopy

Barré Jean-Baptiste, Bourrier Franck, Cécillon Lauric, Brancheriau Loïc, Bertrand David, Thévenon Marie-France, Rey Freddy. 2018. Predicting mechanical degradation indicators of silver fir wooden strips using near infrared spectroscopy. European Journal of Wood and Wood Products, 76 (1) : 43-55.

https://doi.org/10.1007/s00107-017-1209-4

Article de revue ; Article de revue à facteur d'impact

	Version Online first - Anglais Accès réservé aux personnels Cirad Utilisation soumise à autorisation de l'auteur ou du Cirad. s00107-017-1209-4.pdf Télécharger (2MB) \| Demander une copie
	Version publiée - Anglais Accès réservé aux personnels Cirad Utilisation soumise à autorisation de l'auteur ou du Cirad. s00107-017-1209-4-edite.pdf Télécharger (2MB) \| Demander une copie

Quartile : Q1, Sujet : MATERIALS SCIENCE, PAPER & WOOD / Quartile : Q2, Sujet : FORESTRY

Résumé : The management of ecological engineering structures making up a timber structure requires periodical evaluations, including the level of decay of the constituent parts of the timber structure. Methods exist to measure the level of decay in the laboratory or in the field. However, they are rarely suitable for the conditions of ecological engineering structures, or give partial information. The aim of this study was to predict two mechanical degradation indicators (DwMOE and DwMOR) of silver fir (Abies alba) wooden strips during microbial decomposition using near-infrared spectroscopy (NIRS). For 1.5 years, the degradation of 180 squared wooden strips, 30 mm wide and 500 mm long, buried in a greenhouse near Grenoble, France (Altitude: 200 m) was monitored. DwMOE and DwMOR were set from the normalized losses in modulus of elasticity (MOE) and modulus of rupture (MOR), two mechanical properties classically used for timber-structure design. A calibration set of 109 samples was selected to build two separate predictive models of DwMOE and DwMOR using partial least square regression. The NIR-based models applied to a validation set of 47 samples indicated good prediction performance. The model has a root mean square error of prediction (RMSEP) of 0.15 and a coefficient of determination (r2p) of 0.79. The DwMOR model has a RMSEP of 0.13 and a r2-value of 0.91. These results highlight the considerable potential of NIRS in assessing the extension of decay in wooden logs.

Mots-clés Agrovoc : bois, propriété mécanique, propriété technologique, biodégradation, spectroscopie infrarouge, Élasticité, méthode statistique, analyse de régression, Abies alba

Mots-clés géographiques Agrovoc : France

Mots-clés complémentaires : Essai non destructif

Classification Agris : K50 - Technologie des produits forestiers
J12 - Manutention, transport, stockage et conservation des produits forestiers

Champ stratégique Cirad : Axe 2 (2014-2018) - Valorisation de la biomasse

Auteurs et affiliations