The objective of this research was to investigate the technical feasibility of producing plywood by green-gluing technology. 2.5 mm thick rotary cut veneers obtained from steamed maritime pine (Pinus pinaster Ait.) logs were used. Plywood was manufactured under vacuum with veneers in a wet state. Five-layer plywood panels were produced using a one-component polyurethane adhesive to evaluate the effects of two parameters, the wood moisture content and the amount of adhesive, on the mechanical properties of plywood. First, a full factorial experimental design was realised to study the bond performance. Shear tests were carried out in order to determine the bond quality and failure type. It was observed that the values of plywood panels' shear strength depend more on the wood moisture content than on the amount of adhesive applied. A model representing the shear strength of the panels according to the manufacturing parameters was established by using statistical analyses. Another part of this study focused on microscopic observations of the bondline in order to get information about the joint morphology and the degree of adhesive penetration into the wood structure. Results suggested that definite proportions of water (moisture content between 30 and 60 %) and adhesive have to be present on wood surface when gluing to create efficient bonds. These analyses were followed by the panel's characterisation in flexure to validate the process.

Three-layer polymeric diphenyl-methane-diisocyanate (pMDI)-bonded particleboards (PBs) were produced with different proportions of simulated recovered wood (rW) in the core layers (cLs) to assess the effect of rW on the formaldehyde (FA) release of PB. A pre-test was conducted on furniture and particle mixtures of rW to determine the range of expectable FA emission of rW. The FA content of the raw particle mixtures could be predicted from the contents of the individual raw material and did not change compared to the PB. FA content correlated strongly with PB-FA emission. It was possible to predict the maximum PB-FA contents, which should not be exceeded according to F**** and CARB 2 by linear regression models. At moderate and high total FA emission levels, the FA emission of the particle mixtures was approximately 60% higher than the emission of PB blocks. At low total FA level, the flask method and the gas analysis method exhibited different results with regard to the emissions from particles and their respective PBs.

No abstract