SciELO - Scientific Electronic Library Online

vol.33 issue4Evaluation of the woods of Eucalyptus camaldulensis Dehnh and Eucalyptus urophylla S.T. Blake During machining tests for use by the furniture industryInfluences of experimental plot forms and statistical model analysis in Eucalyptus spp Clone selections author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Revista Árvore

On-line version ISSN 1806-9088


SANTOS, Ricardo Balleirini dos; GOMIDE, José Lívio  and  SOUSA, Leonardo Chagas de. Prediction of wood and kraft pulp quality by near infrared spectroscopy (NIRS). Rev. Árvore [online]. 2009, vol.33, n.4, pp.759-767. ISSN 1806-9088.

The objective of this study was to establish mathematical calibration models for near infrared spectra (NIRS) to predict wood quality and kraft pulp characteristics for Eucalyptus. Two different wood chip moisture contents were used (33% and 10%) and 116 kraft pulp samples were measured (kappa number 13-18). For more humid chips, the following correlations were established: 97% for wood basic density, 84% for extractive content and 93% for lignin content. For drier chips, the correlations were: 97% for wood basic density, 92% for extractive content and 90% for lignin content. For pulp quality prediction, NIRS spectra were obtained from handsheets and from the pulp in the original form (disaggregated). NIRS models for handsheets presented 94% correlation for kappa number, 93% for pulp viscosity, 90% for screened yield and 86% for hexenuronic acid content. Models for disaggregated pulp presented the following correlations: 90% for kappa number, 91% for pulp viscosity, 88% for screened yield and 85% for hexenuronic acid content. These results demonstrate the viability and precision of the NIRS technique to estimate wood quality and pulp characteristics.

Keywords : NIRS; wood; kraft pulp.

        · abstract in Portuguese     · text in Portuguese     · Portuguese ( pdf epdf )


Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License