American Society of Civil Engineers

Direct Tension Pullout Bond Test: Experimental Results

by S. P. Tastani, (corresponding author), (Dr. Civil Engineer, Adjunct Lecturer, Dept. of Architecture, Demokritus Univ. of Thrace (DUTh), Xanthi 67100, Greece E-mail: and S. J. Pantazopoulou, M.ASCE, (Professor, Department of Civil Engineering, Laboratory of Reinforced Concrete, Demokritus Univ. of Thrace, Xanthi 67100, Greece. E-mail:

Journal of Structural Engineering, Vol. 136, No. 6, June 2010, pp. 731-743, (doi:

     Access full text
     Purchase Subscription
     Permissions for Reuse  

Document type: Journal Paper
Abstract: Results from 50 bond tests conducted using the so-called direct tension pullout specimen type are presented. Anchorages of steel bars with machined deformations were tested so as to enable a targeted study of the effect of rib height and related rib area on bond behavior; tests were conducted with or without the combined presence of external confinement over the embedded length. The novel specimen form presented in the paper was designed to simulate the state of stress arising in usual bar anchorages in the tension zones of flexural members (where both cover concrete and bar are stressed in tension). This development was motivated by the need to eliminate spurious influences of the test setup on specimen behavior, which are known to interfere with bond mechanics in conventional bond tests leading to unconservative estimates of bond strength and misleading interpretations as to the true parametric dependencies of the bond problem. Additional parameters studied in the experimental program were the development length, the cover thickness, the effect of confinement, and the tensile strength of concrete. Data reduction (local bond strength and slip estimates) was possible by fitting the exact solution of the differential equations describing the state of stress along the anchorage to the test measurements, while accounting for important phenomena such as yield penetration or debonding, and bond plastification. Next, the obtained values for local bond strength were used for calibration of the frictional analog for bond strength (a Mohr-Coulomb failure criterion). Milestone values for bond and slip were estimated with reference to the code limit-state model for bond.

ASCE Subject Headings:

Author Keywords:
Local bond
Bar diameter
Rib height
Direct tension
Pull out
Anchorage length
Splitting failure