Wedge-splitting tests for tensile strength and fracture toughness of concrete. (February 2018)

• Record Type: Journal Article
• Authors: Guan, J.F.; Hu, X.Z.; Xie, C.P.; Li, Q.B.; Wu, Z.M.
• Is Part Of: Theoretical and applied fracture mechanics. Volume 93(2018)
• Page Start: 263
• Page End: 275
• Publication Date: 2018-02
• Subjects: Wedge-splitting (WS) -- Concrete -- Aggregate size -- Tensile strength -- Fracture toughness; Fracture mechanics -- Periodicals; 620.1126
• DOI: 10.1016/j.tafmec.2017.09.006
• Languages: English
• ISSNs: 0167-8442

Abstract

Highlights: The Boundary Effect Model was extended to WS geometry and relevant equations were derived. The fundamental fracture properties can be extrapolated from small WS specimens. The random aggregates distribution and micro-crack formation were considered by introducing a discrete number β . If the ratio of specimen size and maximum aggregate is small, β ≈ 1 is a good approximation. Quasi-brittle fracture of concrete can be fully predicted using the new Boundary Effect Model. Abstract: This study presents a new approach for determination of tensile strength f t and fracture toughness K IC of concrete using wedge-splitting (WS) specimens as neither property can be easily measured because of the heterogeneous material structures. The Boundary Effect Model (BEM) was extended to WS geometry and relevant equations were derived so that more concrete properties can be determined by the load and displacement measurements. The new model considers the influence of coarse aggregate structures on concrete fracture by linking the fictitious crack growth Δ a fic at peak load P max to the maximum aggregate size d max . The random aggregate distribution and inevitable variations in Δ a fic and P max are considered by a discrete number β . WS results of concrete specimens with d max from 10 mm to 25 mm and specimen size W from 150 mm to 1200 mm reported in literature are analyzed for a range of β from around 0.2 to 2.0. It is found that on average β ≈ 1 is a good approximation for common WS specimens with W / d max < 20. Then reliable f t and K IC, which is the asymptotic limit for a large structure with a long crack, can be obtained, regardless whether WS specimens are geometrically similar, or have the same size but different notch lengths. Finally, it is concluded that quasi-brittle fracture of concrete is fully predicted if both K IC and f t are known.