Friction Loss in Tieback Anchors used for Landslide Stabilization

"Tieback anchors are now used routinely for landslide stabilization in the USA and abroad. It is generally understood by design engineers that friction loss over the unbonded length of a tieback anchor is inherent to the system, and such loss is permitted as long as it does not exceed a certain threshold, typically 20 percent of the post-tensioning load. However, recent experience from projects utilizing tieback anchors with unbonded lengths in excess of 150 feet has demonstrated that this criterion may be difficult to achieve, particularly in certain ground conditions. Furthermore, anchor friction loss, even if below the 20-percent threshold, has important implications for the stability of the slope that are often ignored in practice. This paper presents the results of a study in which friction loss is quantified from load test data for projects in a variety of ground conditions by means of a “wobble coefficient.” Factors that may influence the wobble coefficient are examined, and recommendations for addressing this issue in practice are presented. A recent instrumented load-test program for a landslide stabilization project in Southern California is examined, demonstrating that with appropriate monitoring controls, uncertainty with regard to friction loss can be minimized. IntroductionSteel-strand tieback anchors have been used by the geotechnical community to support shoring systems for many decades, and more recently as a means to stabilize landslides. For strand encased in grease-filled plastic sheathing, which is intended to allow the strand to elongate freely, some load can still be transferred by friction between the strand and the grease and between the grease and sheathing. The sheathing then transfers load to the surrounding grout and, in the case of ground anchors, to the surrounding soil or rock. The resulting decrease in load over the length of the unbonded zone is known as “friction loss.”A widely used criterion for ground anchor acceptance based on load testing is that the measured anchor elongation must achieve a minimum value of 80 percent of the theoretical elastic elongation, or in other words the actual elongation can be up to 20 percent less than the theoretical value. This ensures that the majority of the load applied at the anchor head is reaching the bond zone behind the assumed failure surface, since load shed within the failure mass does not contribute to stability. This criterion acknowledges that friction loss is an inherent feature of the anchor system which cannot be eliminated completely, only minimized."