Ligações estruturais de peças de madeira por meio de adesivo em áreas reduzidas associado a elementos de pinos

Abstract

This work had the purpose of investigating both theoretically and experimentally the behavior of wood joints of reduced area in which an adhesive was used in conjunction with a mechanical connector (dowel), so as to describe their strength and their use as structural connection in general joints members, in joints of compound members and injoints of de trusses and frames. Air-dried eucalypt lumber was used in the experiments, of boards with 25 mm and 32 mm thick; it was later determined that they consisted of the species Eucalyptus grandis and Eucalyptus saligna. They were separated into two classis of specific gravity, from 550 to 750 kg/m3 or from 750 to 950 kg/m3. The adhesive used was a commercial resorcinol-formaldehyde resin. Compression tests were conducted in specimens built up of three wood pieces, glued together so that either the three pieces had their grains parallel to the direction of the applied load or the two lateral pieces were so placed that their fibers were perpendicular to the grain of the central piece. The nominal thickness of the lateral pieces was 24mm; the central piece was 24, 30 or 48mm thick. In addition to the adhesive, a central dowel, or a central dowel plus bolts of small diameters or a larger diameter, centrally driven bolt were also used. Statistical analysis indicated that the strength of the parallel-members joints were influenced by the thickness of the central piece and by the specific gravity of wood, but not by type of the mechanical connector. Rupture occurred suddenly, thus indicating fragility, but strength was high, when compared to the crushing strength of the central member. Efficiency was close to 100% when the hc/tc ratio between the length of the glue line and the thickness of the central member was of the order of 4 to 5. Significant losses of strength and stiffness were detected when the loads were applied off-center.Strain-gages were used to measure stresses along the glue line. The experimental results were close to those obtained by a numerical analysis. The mathematical formulation so developed gave good estimates of the rupture forces found on the laboratory. Shearstresses in the glue line given by analytical reasoning or by numerical analysis agreed closely, although the effect of stress concentrations at both ends of the glue line was more accentuated when numerical evaluation was used. Statistical analysis of data from orthogonally-glued members indicated that strength and stiffness were influenced by wood density but not by the presence of mechanical connectors or by the thickness of the central member. The strength of the joints was controlled by the low stiffness and low strength of the lateral members that were loaded perpendicularly to the grain. The analytical evaluations also agreed well with the experimental results. For eccentric loading, the average loss of strength varied from almost nil to about 30%, depending on joint model. The good results of glued joints assessed in this work, mainly the high strength of parallel-member joints, shows the potential of use of this solution for small and medium size structures.