Pullout Performance of Fully and Partially Bonded Retrofit Anchors in Low-Strength Concrete

Gurbuz T., İlki A.

ACI STRUCTURAL JOURNAL, vol.108, no.1, pp.61-70, 2011 (SCI-Expanded) identifier

  • Publication Type: Article / Article
  • Volume: 108 Issue: 1
  • Publication Date: 2011
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.61-70
  • Istanbul Technical University Affiliated: Yes


In the seismic retrofit of reinforced concrete structures, steel reinforcing bars are often used as post-installed chemical anchors to integrate new structural members to the existing structural system. The efficiency of conventional fully bonded post-installed anchors is significantly reduced when they are anchored into low-strength concrete due to premature cone-type failure of low-strength concrete around the anchor In this experimental study, pullout behaviors of fully and partially bonded chemical anchors post-installed in low-strength concrete are investigated. The test parameters are the embedment length, the bonded length of the anchors, and the surface conditions of the anchor holes in terms of humidity and dust. The pullout behavior of fully and partially bonded anchors is examined through the load-displacement curves, modes of failure, and anchorage strengths of specimens. The test results show that the pullout performance of partially bonded, post-installed chemical anchors embedded in low-strength concrete is significantly better than conventional fully bonded post-installed chemical anchors. Therefore, the partially bonded, anchoring of deformed bars appears to be a promising method for anchorages in seismic retrofit applications, 'particularly in the case of existing structures with low-strength concrete. Furthermore, a model is proposed for predicting the pullout strength of anchors. that fail due to a combination of concrete cone and bond failures. The predictions of the proposed model are compared with experimental data obtained by other researchers, and a good correlation is found between the predictions and test results.