Splice tests of plain steel bars in concrete

Abstract

Fifteen splice specimens reinforced with plain steel bars, including three specimens instrumented with both steel and concrete strain gauges, were tested under monotonically applied four-point loading to develop a database of reliable bond test results and contribute to the development of a reliability based bond provision for plain steel bars to evaluate historical concrete structures. The maximum applied load for the specimens and their observed failure behaviour are reported. In addition to that, a strain compatibility analysis, average bond stress distribution, and flexural section analysis within the lap splice length of the instrumented specimens are also reported.

All of the specimens failed in bond within the lap splice length. The load capacity of two specimens reinforced with plain steel bars was 60% of the reported load resistance of specimens with identical geometry and reinforced with deformed bars. The CEB-FIP Model Code provisions for average bond stress of plain steel bars underestimated the maximum applied load recorded for the tested specimens by 16% on average. An empirically derived equation to predict the bond capacity of plain steel bars was determined to be proportional to both the splice length and the nominal bar diameter.

Observed cracks in the shear spans remained vertical and suggest the development of arch action within this region. The formation of a large crack at one end of the lap splice length and a review of the load versus deflection behaviour indicated a sudden bond failure of the specimens. Removal of concrete cover at the ends of the lap splice length following testing of the specimens showed evidence of slip of the lapped bars.

Instrumented splice specimens provided evidence of bond loss within the lap splice region. As-measured steel strains were higher than those measured for the surrounding concrete due to a loss of strain compatibility. The average bond stress distribution within the lap splice length became more uniform as the applied load approached the maximum applied load. The flexural analysis calculated based on concrete strains above the neutral axis and steel strain provided a reasonable estimate of specimen capacity.