Uplift behaviour of cylindrical anchors in sand: Discussion

Abstract

Discussions 574 The Authors should be commended on their experimental and numerical studies on the uplift behaviour of cylindrical anchors in sand. The paper reports results of experiments conducted on a cylindrical anchor with corrugated surface. The height of the cylindrical anchor L = 250 mm, diameter D = 50 mm, and L/D = 5. The depth of the anchor was defined as the distance from the sand surface to the top surface of the cylinder. The Authors used the ratio Z/D to define shallow and deep anchors and to distinguish the two different behaviours. In conventional plate anchors, the ratio Z/D defines the depth of installation of the plate divided by plate diameter. For that matter, a plate anchor with Z/D = 0 will have a zero pullout load because the plate is only resting on the sand surface. The cylindrical anchors whose Z/D was zero (tests 1 and 14), displayed resistance against pullout. Obviously, this is due to the shaft friction. The Discusser, hence, feels that the analogy made by the Authors between the ratios Z/D for conventional plate anchors and cylindrical anchors may not be accurate. The incorporation of the embedded length of the cylinder into the ratio that defines shallow and deep cylindrical anchors should eliminate possible confusion. The Discusser believes a ratio of (Z+L)/D could be more representative and comparable with the definition used in the case of plate anchors. It is true that the value of the above ratio will not be greatly impacted if the length of the cylinder L is relatively small compared with the depth Z, but, on the other hand, (Z+L)/D could be significantly different if the anchor’s top surface is close to the sand surface and the cylinder is long. It is unclear what the Authors meant when they stated “In practice, it was found that L/D ratios ranging from 2 to 4.5 were the optimum underream dimensions to generate the adequate ratio between end resistance and shaft friction of the anchorage body.” Furthermore, the Authors selected a ratio of L/D = 5 for their tests, which is not within the range they specified to “generate the adequate ratio between end resistance and shaft friction.” Moreover, based on the figures resulting from the numerical analysis of the problem (see Figs. 10c and 11c), it seems that the failure surface of the sand was not only due to the top surface of the cylinder, but also because of the entire length of the cylinder as well. For instance, the deformation of the sand shown in Figs. 10c and 11c for Z/D = 4 and 16, respectively, can be approximated by the shapes shown in Figs. D1a and D1b for shallow and deep anchors, respectively. Figure D1a details the individual failure surfaces due to the plate (top surface of the cylinder) and the cylindrical shaft, for an anchor installed at Z/D = 4[ (Z+L)/D = 9]. Figure D1b details the individual failure surfaces due to the plate and the shaft, for an anchor installed at Z/D =1 6 [(Z+L)/D = 21]. The failure surfaces created by the plate in Fig. D1 resemble those reported by Ghaly and Hanna (1994) for shallow and deep anchors. It should be noted that the total failure zone confined by the outermost boundary surface originate from the bottom of the cylinder. This signifies the cylinder’s contribution to pullout load by displacing the sand and generating shearing resistance. This also indicates that the bearing resistance above the cylinder’s top surface is not the sole contributor to pullout resistance.