Application of microscale methods to study the heat-induced delamination in engineered wood products bonded with one-component polyurethane adhesives

Abstract

One-component polyurethanes (1-c-PUR) are commonly used adhesives for the manufacture of cross-laminated timber (CLT). Typically, these adhesives do not govern the mechanical response of CLT under normal service temperatures. However, when subjected to heating from a fire, failures may transition from within the timber to the bond line interphase zones. CLT bonded with 1-c-PUR has been shown to be prone to heat induced delamination (HID), which may compromise its structural performance at elevated temperatures. Thermogravimetric analysis (TGA), differential scanning calorimetry (DSC), and dynamic thermo-mechanical analysis (DTMA) were used to study the thermal and thermo-mechanical responses of engineered wood products and components (i.e. adhesives and timber) and their interphase (i.e. bond line). Two commercially available 1-c-PUR adhesive films from the same manufacturer, two timber species (Norway Spruce and Radiata Pine, as sawdust or as veneers), and four different veneer shear lap combinations, were studied. Shear lap specimens bonded with a conventional 1-c-PUR adhesive consistently experienced bond line mechanical failure in DTMA at about 220-240 °C. The same adhesive films tested via DSC and TGA softened at 240-260 °C. Conversely, a different 1-c-PUR adhesive, formulated specifically for enhanced performance at elevated temperature, displayed no detectable softening in DSC and TGA, and shear lap specimens in DTMA consistently failed within the timber; even at elevated temperatures. The presented thermo-mechanical methods allow identification of failure modes and temperatures, whilst the thermal microscale methods (i.e. TGA, DSC, DTMA) assist in understanding the various factors contributing to failures.