Cyclic shear resistance model for Eurocode 8 consistent with the second-generation Eurocode 2

Abstract

The 1994 pre-standard ENV-Eurocode 8 followed the 1985 “Seismic Annex” of the CEB/FIP Model Code 1978 in reducing or eliminating the contribution of concrete to the shear resistance of higher ductility concrete members with axial compression less than 10% of the axial force resistance. The first-generation EN-Eurocode 8 tried to achieve the same end result for high ductility beams, within the constraints set by the elimination of the contribution of concrete to shear resistance in the first-generation EN-Eurocode 2. Comparison with over 1100 cyclic tests of shear-critical RC members shows that these code rules give on average a serious safety deficit, especially in members designed for ductility. The deficit is reduced in members designed for ductility using the first-generation EN-Eurocode 8 and turns into surplus for those designed for shear strength alone, but at the price of very high scatter. The strain-dependent monotonic shear resistance model in the 2018 draft of the second generation EN-Eurocode 2 agrees on average well with over 500 cyclic tests of RC members which failed in shear without yielding in flexure, but with lack-of-fit with respect to most variables affecting shear resistance. More important is its bias in the unsafe direction for almost 600 cyclic tests which led to shear failure after flexural yielding. The following modifications to this monotonic shear model give good average agreement with cyclic tests that led to shear failure before or after flexural yielding: (a) inelastic longitudinal strains at section mid-depth in plastic hinges are estimated using the “equal displacement rule”; (b) the transverse component of the strut force transferring a column’s axial compression from the compression zone of one end section to the diagonally opposite compression zone at the other end is added to the shear resistance; (c) the special model in Eurocode 2 for resistance to shear due to point loads near supports is used in squat walls or columns; (d) the concrete strength along the compression field is reduced by 37.5%, and (e) upper limits are set to the strength values of transverse reinforcement and concrete. In addition to filling a gap in Part 1 of Eurocode 8 concerning design of new concrete structures, the proposed approach can replace the current portfolio of (semi-)empirical models in Part 3 of the first-generation Eurocode 8 and in the draft of the second-generation one, which give the cyclic shear resistance before and after flexural yielding for four different combinations of failure mode and member type, without bias and with less scatter than the new, holistic approach.