Fracture parameters of alkali-activated aluminosilicate composites with ceramic precursor: durability aspects

Abstract

Four sets of alkali-activated aluminosilicate (AAAS) composites based on ceramic precursors were studied in terms of their characterization by mechanical fracture parameters as a basis for considerations of durability. AAAS composites made of brick powder as a precursor and alkaline activator with various silicate moduli (Ms = 0.8, 1.0, 1.2, 1.4, and 1.6) were investigated. The sets of AAAS composites differed in terms of the used filler: quartz sand or brick rubble. Two different precursor particle size ranges of 0–1 mm and 0–0.3 mm were used for both types of filler. The test specimens had nominal dimensions of 40 × 40 × 160 mm and were provided with a notch at midspan after 28 days of hardening. The notches were cut up to 1/3 of the height of the specimens. The specimens were subjected to three-point bending fracture tests during which force vs. deflection (F–d) and force vs. crack mouth opening displacement (F–CMOD) diagrams were recorded. Tensile strength ft,ID and specific fracture energy GF,ID values were identified using the inverse method based on a neural network ensemble. The obtained F–CMOD diagrams were subsequently evaluated using the double-K fracture model supported by the ft,ID and GF,ID values. The double-K model allows the quantification of two different levels of crack propagation: initiation, which corresponds to the beginning of stable crack growth, and unstable crack propagation.Four sets of alkali-activated aluminosilicate (AAAS) composites based on ceramic precursors were studied in terms of their characterization by mechanical fracture parameters as a basis for considerations of durability. AAAS composites made of brick powder as a precursor and alkaline activator with various silicate moduli (Ms = 0.8, 1.0, 1.2, 1.4, and 1.6) were investigated. The sets of AAAS composites differed in terms of the used filler: quartz sand or brick rubble. Two different precursor particle size ranges of 0–1 mm and 0–0.3 mm were used for both types of filler. The test specimens had nominal dimensions of 40 × 40 × 160 mm and were provided with a notch at midspan after 28 days of hardening. The notches were cut up to 1/3 of the height of the specimens. The specimens were subjected to three-point bending fracture tests during which force vs. deflection (F–d) and force vs. crack mouth opening displacement (F–CMOD) diagrams were recorded. Tensile strength ft,ID and specific fracture energy GF,ID values were identified using the inverse method based on a neural network ensemble. The obtained F–CMOD diagrams were subsequently evaluated using the double-K fracture model supported by the ft,ID and GF,ID values. The double-K model allows the quantification of two different levels of crack propagation: initiation, which corresponds to the beginning of stable crack growth, and unstable crack propagation.