Negative skin friction from surface settlement measurements in model group tests: M. U. Ergun & D. Sonmez, Canadian Geotechnical Journal, 32(6), 1995, pp 1075–1079

Abstract

This paper presented the results of an experimental study on the negative skin friction of model pile groups. The authors should be commended on their well-organized research work and on a well-designed experimental setup. Since the outcome of this experimental program is totally dependent on settlement measurements, it is important to learn about the criteria used in selecting the locations where dial gauges were placed. The paper reports that sin and sout are the average surface settlements inside and outside a pile group, respectively. The point the discusser wishes to address is the following: are measured settlements representative of sand surface deflection? For spacing d = 2B ,4 B, and 5.5B, the authorsi Fig. 3 shows that most, if not all, of dial gauges used to measure settlement inside a given pile group are located near the center of the group and are more concentrated toward the first and second rows of piles. No reference was given in the paper to the order of pile installation. The authors have in fact realized that variation of the relative density of the sand may affect the angle of friction between the pile and the sand, which may impact the accuracy of the results of the phenomenon under consideration (negative skin friction). Furthermore, pile driving using a hammer in a confined sand layer where lateral movement is very limited may result in the upheaval of previously installed piles especially in closely spaced groups. This problem could be overcome if an orderly installation is followed to balance the behaviour of the group about the axes of symmetry. In such a case, each and every pile would contribute almost equally to the behaviour of the group. A balanced group behaviour can be closely monitored if dial gauges are placed very near to the locations of piles. Ghaly (1990) studied the behaviour of model screw anchors installed in sand. Although the anchors used in the experimental investigation were screwed into a sand deposit, which is believed to cause much less disturbance than driving piles by hammering, measurements of uplift loads of individual anchors using load cells showed that the order of installation and the location of the anchor within the group affected the magnitude of the measured uplift load. Figures 1 and 2 show X-ray photographs of a triangular group and 3 × 3 square group, respectively, after installation into a sand deposit has taken place (needless to mention that the sand deposit is not shown in the figures in order to make the anchors visible). Ghaly and Hanna (1994) recommended that an installation arrangement should be developed to provide consistency to the structure of the sand mass and to prevent the creation of highly stressed or densified