Evaluation of Elongation Criteria and Friction Loss in Ground Anchors

"This report describes a research project for evaluating the applicability of the widely accepted 80 percent criterion for elongation of ground anchors to anchors with unbonded lengths exceeding 100 feet. This issue is driven by several recent projects involving ground anchors for landslide stabilization in which a significant percentage of the anchors did not meet the criterion that requires measured elongation during proof load testing of at least 80 percent of the theoretical elastic elongation. The projects involved anchor unbonded lengths in the range of 85 to 220 feet, which is outside the range traditionally used in practice, although anchors of this length are being used more frequently for landslide stabilization. The principal objective of this research is to address whether the widely accepted criterion of 80 percent elongation is applicable for such applications, and whether other factors affect the ability of anchors to meet the criterion.Analytical methods for predicting transfer of load along the length of steel strand due to friction loss are used routinely in the prestressed concrete industry. These analytical expressions provide a rational framework for quantifying changes in load due to friction along the length of a ground anchor in terms of a ‘wobble coefficient’ (K) as defined in Aalami (2004). Values of the wobble coefficient for ground anchors can only be determined by back-calculating from load tests, i.e., fit the value of K to the appropriate analytical expression based on the known test load and measured percent elongation. The analytical basis is first developed and shown to provide a tool for evaluating results of anchor load tests to determine the magnitude of expected elongation as a function of unbonded length. Next, a database of anchor load tests is used to back-calculate values of K for anchors with unbonded lengths in the range typically used in geotechnical applications (< 100 feet). These values of K are then used to calculate expected friction loss for anchors with high unbonded lengths (100 to 250 feet) to evaluate whether the 80 percent criterion is reasonable."