Response surface methodology-based formulation optimization of silicate-modified PU/WG grout for controllable gelation times

Response surface methodology (RSM) was used to optimize the formulation of silicate-modified polyurethane/water glass (PU/WG) grouts. The effects of catalyst and chain extender contents on gelation time, maximum reaction temperature, and 7-day unconfined compressive strength (UCS) were systematically examined. Eight optimal formulations were obtained using a central composite design (CCD) combined with multi-response desirability function optimization to achieve target gelation times ranging from 120 to 960 s. Microstructural characterization using SEM, EDS, and FTIR was performed to examine the organic-inorganic hybrid structure of the optimized formulations. The results show that the quadratic response surface models were statistically reliable. The R² values ranged from 0.9633 to 0.9970, and the adequate precision values ranged from 15.9163 to 52.9906. The lack of fit was not significant for any response. These models can describe the nonlinear effects of catalyst and chain extender contents on gelation time, maximum reaction temperature, and 7-day UCS within the investigated formulation range. Catalyst content was the dominant factor controlling gelation time and maximum reaction temperature. Increasing catalyst content accelerated the reactions involving isocyanate groups, thereby shortening gelation time and increasing maximum reaction temperature. Meanwhile, chain extender content mainly regulated 7-day UCS through a nonlinear effect, with the optimal dosage concentrated within 4.608–4.880 wt%. Eight optimized formulations achieved measured gelation times of 129–993 s, while maintaining 7-day UCS values of 48.56–53.91 MPa. SEM/EDS and FTIR results reveal that the optimized grout forms a multiphase organic-inorganic hybrid structure composed of a PU-rich matrix, silicate-rich domains, and carbonate-rich crystallites, which contributes to the high 7-day UCS of the material.