Horizontal pullout resistance of concrete anchor blocks in sand backfill

Abstract

Anchored earth wall is one of the popular types of internally stabilized wall systems intended to retain soils vertically. Such walls are typically characterized by unreinforced incremental concrete facing panels tied back by horizontally laid tendons within and beyond potential Rankine or Coulomb failure plane. While one end of these tendons is connected to facing elements, the other end is connected to a cube-shaped concrete anchor block placed well within the passive zone. These anchor blocks withstand the mobilized active earth pressures via passive resistance. Some design manuals, e.g. NAVFAC (1982) suggests that the anchor blocks should be placed beyond a plane that makes an angle equal to angle of friction of the reinforced fill from the toe of the wall with horizontal. The manual also suggests that locating the anchor blocks in any other position between Rankine’s potential failure plane and this phi-plane will allow mobilization of a partial resistance. Ali, Bujang and Lee (2008) and Chonkar (2001) suggested that passive resistance of cube-shaped concrete anchor blocks is actually four times the Rankine’s passive resistance. However, they did not refer to any particular location of anchor blocks in the passive zone that would mobilise such passive resistance of the blocks. Jones (1996) and BS8006 (1995) have also suggested similar passive resistance for triangle shaped steel anchors. In order to investigate the effect of position of anchor blocks on their capacity of mobilizing passive resistance, a series of tests were carried out in a large tank made of Perspex and steel framing system. The test tank (1.2m x 0.90m x 0.90m) was filled with sand of three different fineness modulus (0.73, 1.5 and 2.5) at their individual maximum densities. 150mm x 150mm x 75mm anchor blocks were placed at mid-height of the sand backfill just on the border of 600, 450, 300 and 200 planes; 600 plane being the closest and 200 plane being the farthest from the potential Rankine failure plane. A soft yielding boundary was ensured at the front wall of the test tank in order to ensure mobilization of full active earth pressure on the wall. The concrete anchor block was then pulled by 3x1.8mm diameter steel wire over a frictionless pulley using incremental loading. The pullout resistance of an anchor block was determined as the force required for excessive displacement of the block. The tests were repeated three times for each of the twelve scenarios, i.e. tests were carried out at least 36 times for covering the scenarios of all the four locations of anchor blocks and three different types of sand backfill materials. From the tests, it was found that for soils having higher friction angle the passive resistance of anchor block was also higher for any particular location of the block in the passive zone. However, the enhancement of this resistance over Rankine’s passive resistance, denoted by C in this study, was not found to be equal to 4, as suggested by different researchers and code, for any of the twelve scenarios undertaken in this work. The C factor was found to be equal to 1 for all three types of soils when the anchor block was placed on the border of 600 plane, i.e. almost on the border of potential Rankine failure plane. When the block was placed on the border of 450 plane, the C factor was more and ranged from 1.66 to 2.21 in accordance with decreasing fineness modulus of the sand backfill used. The C factors for the block placed at 300 plane and 200 plane were almost identical and ranged from 2.39 to 3.48 again in accordance with decreasing fineness modulus of the sand backfill used. These results suggest that location of anchor block beyond 300 plane may not be effective and placing the anchor blocks on the border of 450 plane may be considered as a good compromise for catering many practical space constraints.