Tensile behavior of carbon textile concrete composite captured using a probabilistic multiscale multiple cracking model

• Published in: Composite structures Vol. 277; p. 114624
• Main Authors: Vořechovský, Miroslav, Li, Yingxiong, Rypl, Rostislav, Chudoba, Rostislav
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.12.2021
• Subjects: Brittle-matrix composite; Matrix fragmentation; Probabilistic modeling; Strain hardening; Textile reinforced concrete; Uniaxial tensile test; Probabilistic modeling; Matrix fragmentation; Uniaxial tensile test; Strain hardening; Brittle-matrix composite; Textile reinforced concrete
• ISSN: 0263-8223; 1879-1085
• DOI: 10.1016/j.compstruct.2021.114624

The fast growing amount of high-performance textile fabrics made of carbon, glass and basalt yarns open up enormous design scope for cementitious composites, that needs to be sorted and graded using novel characterization procedures. A combined experimental and numerical characterization method of the composite tensile behavior is presented here for concrete reinforced with non-impregnated carbon textile fabrics. Because of the heterogeneous structure of the bond layer, which consists of a non-impregnated multifilament yarn partially penetrated by the cement paste, this material combination poses particularly high demands on a realistic modeling approach. It is well known, that the tensile response of strain-hardening brittle-matrix composites is primarily shaped by the multiple cracking process of the matrix. To account also for the local damage developing in the heterogeneous bond layer ahead of a crack bridge during the multiple cracking process, we combine a probabilistic crack bridge model with a generalized version of a crack tracing algorithm reflecting the random matrix strength. The resulting probabilistic multiscale multiple cracking model has been calibrated and validated using a consistent series of crack bridge and composite tensile tests by comparing the predicted and measured stress–strain curves and crack spacing histories. The efficient, flexible and realistic modeling concept can serve as a basis for the development of rapid characterization methods and more economic and reliable design rules.