Testing the quality of bond in plywood by varying some adhesive properties

• Main Author: Calpaxidis, John C.
• Published: University of Aberdeen 1977
• Subjects: 674
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.450534

A major object of this study has been to examine some adhesive properties associated with increased viscosity to account for results which, favour higher viscosities. This aim has largely been fulfilled although some new questions have emerged for future investigation by experiment. A second major object of the study was to compare and contrast methods used to assess bond quality of plywood. Tests of bond quality have always yielded contradictory results. Examination of all existing methods could not be attempted so attention was focused on a selection of popular methods adopted as British, German, American and Canadian standards. A total of 1,460 plywood specimens were tested to failure following 5 test treatments (dry, vacuum-pressure, hot bath, boiling cycle and boil test) and 336 knife test specimens provided data for comparisons and suggestions. In the first experiment samples of UP adhesive, at 11 levels of viscosity/ viscosity from 1.5 to 14.8 Pas, were produced by aging uncatalyzed resin at 60C. Increased viscosity was associated with increased molecular weight as indicated by G.P.C. analysis and determinations of resin solids content. G.P.C. analysis revealed a shift from species of low to high molecular weight while resin solids content remained fairly constant. Increased viscosity was accompanied by increased contact angle and reactivity measured as gel-time on catalyzed samples at 100C. Differential Thermal Analysis of catalyzed samples of varying viscosity gave an exotherm with its peak at 88 1 C. The area below the exotherm was attributed to the exothermic catalytic reactions during curing of the adhesive. The thermograms revealed that the peaks, starting and finishing points, area below the exotherm and slopes of the exotherms were fairly constant and unaffected by viscosity. The conclusion was that regardless of viscosity the number of the condensation reactions, as represented by the magnitude of the exotherm, the time curing began and ended, and the duration of curing were similar. The decrease in gel-times, with- increasing viscosities, and results of the Differential Thermal Analysis suggests that the more viscous samples must have developed strongly cross-linked networks, at the end of curing. Plywood, made with UF adhesive samples aged to 11 levels of viscosity and Scots pine (Pinus sylvestris) veneer, showed better bond quality as viscosity rose to 6.69 Pas but beyond that bond quality deteriorated gradually. Bond quality was assessed by shear strength following dry, vacuum-pressure and hot-bath treatments, wood failure and knife test. The explanation proposed for these results is that, because of their/ their higher viscosity, reactivity (remaining liquid for a shorter time) and greater contact angle, the more viscous adhesive did not over-penetrate into the wood. Further, because they had developed a more extensive crosslinked network, as indicated by Differential Thermal Analysis, the more viscous adhesive had greater cohesive strength when fully cured. The second and third experiments extended the first by using PF adhesive and changing viscosity by prolonging reflux times at resin production to obtain different degrees of condensation and changing resin solids content to produce varying degrees of concentration. Prolonging reflux times raised viscosity from 0.085 to 0.8 Pas and also raised reactivity, the gel times falling from 105 to 35 minutes when measured on uncatalyzed adhesive samples at 100 C, and contact angles from 96.8 to 52.0 degrees when measured on glass slides. Resin solids content remained fairly constant during advancement of the PF adhesive to higher viscosities but G.P.C. analysis indicated a shift towards high molecular weight materials. The bond quality of plywood made with PP adhesive of varying viscosity, due to increased condensation, and African mahogany (Khaya ivorensis) veneer increased considerably in dry and wet shear strengths, wood failure percentages (following dry, vacuum-pressure, boiling-cycle and boil treatments) and knife test. The most likely explanation of improved bond quality seemed to lie in prevention of over-penetration, which was observed microscopically on transverse sections of the glue line to be severe in plywood made with the less condensed adhesive. However, PF adhesives also cure by poly-condensation reactions and in the more condensed PF adhesives must also have/ have had higher cohesive strength on completion of curing. Increased viscosities produced by raising resin solids content to varying concentrations were also closely associated with increased reactivity and contact angle. When resin solids content increased from .40 to 55 percent gel-time, measured on uncatalyzed samples at 100 C, decreased from 98 to 42 minutes while contact angle, measured on glass slides, increased from 25.3 to 57.9 degrees. The bond quality of plywood made with PF adhesive of 40, 45, 50 and 55 percent concentrations and African mahogany veneer improved as the concentration of solids rose to 50 percent but did not rise further at 55 percent concentration.