Interface-level bond performance of Eucalyptus-rubberwood hybrid CLT: Effects of adhesive type and bonding pressure

This study investigates the interlaminar shear performance of hybrid Cross-Laminated Timber (CLT) interfaces fabricated from underutilized Asian hardwoods, Eucalyptus grandis and Hevea brasiliensis (Rubberwood), using polyurethane (PUR), phenol resorcinol formaldehyde (PRF), and melamine urea formaldehyde (MUF) adhesives. PUR was tested at 0.4, 0.6, and 1.0 MPa, while PRF and MUF were tested at 0.6 and 1.0 MPa, with 0.6 MPa included as a lower-than-recommended exploratory pressure; inferential pressure comparisons are therefore based on the balanced 0.6 versus 1.0 MPa subset. Three interface configurations, Rubberwood-Rubberwood (R-R), Eucalyptus-Eucalyptus (E-E), and hybrid Eucalyptus-Rubberwood (E-R), were evaluated via modified block shear tests to assess the bond-line shear performance of adhesive-bonded interfaces in CLT panels. The configurations achieved high interface-level shear strengths under laboratory conditions, with PUR at 1.0 MPa delivering the highest mean strengths in the homogeneous R-R (4.20 MPa) and E-E (3.19 MPa) interfaces and the lowest variability overall; in the hybrid E-R interface, PRF at 1.0 MPa achieved the single highest mean strength (3.71 MPa), though not significantly different from PUR at 1.0 MPa (3.49 MPa, Tukey p = 0.931). Three-way omnibus ANOVA identified a significant Interface × Adhesive interaction (p < 0.001, η² p = 0.399), confirming that adhesive performance is modulated by wood species pairing. Interface-specific decomposition confirmed adhesive dominance in homogeneous R-R and E-E interfaces, while the hybrid E-R interface was governed by both adhesive type and bonding pressure, with a non-significant but large-effect-size pressure-adhesive interaction. PUR predominantly exhibited wood failure, whereas MUF showed higher variability and glue-line failures in dense-wood interfaces. Analytic Hierarchy Process (AHP) multi-criteria analysis, supported by Monte Carlo weight sensitivity analysis (100,000 simulations), confirmed R-R PUR 1.0 MPa as the top-ranked configuration under both decision scenarios, retaining Rank-1 in 100.0% of simulations in both the mechanical and sustainability-prioritized cases. This result indicates strong ranking robustness with respect to moderate variation in criterion weights. The findings of this study demonstrate the promising interface-level bond performance of tropical hardwoods for CLT application, utilizing PUR at optimized pressures to obtain comparable results to softwood CLT systems.