Liquefaction Resistance and Internal Erosion Potential of Non-Plastic Silty Sand Jing-Wen CHEN 1, Wei F. LEE 2, Chun-Chi CHEN 3 1 Professor, Department of Civil Engineering, National Chen-Kung University (No.1, University Road, Tainan City 701, Taiwan) E-mail:geochen@mail.ncku.edu.tw 2 Research Associate Professor, National Taiwan University of Science and Technology ( ) E-mail: weilee@mail.ntust.edu.tw 3 Ph.D. Candidate, Department of Civil Engineering, National Chen-Kung University (No.1, University Road, Tainan City 701, Taiwan) E-mail: n6894108@mail.ncku.edu.tw This paper is to introduce research progress on liquefaction resistance and internal erosion features of non-plastic silty sand. A new sampling technique that applied to field allowing sensitive and high fines content silty sand material to be retrieved in sounding condition is described. Laboratory tests on the liquefaction resistance of non-plastic silty sand that emphasized on the influence of fines content percentages are conducted. The utilizing of a specially designed test device, Flexible Wall Pin Hole Test Device for the investigation of the internal erosion of non-plastic silty sand is also illustrated. It concludes that higher non-plastic fines content of silty sand would result in lower cyclic liquefaction resistance as well as higher internal erodibility. Moreover, the indicated trends would become more obvious when such non-plastic silty sand was subjected to disturbance. Results of this study is hoped that will improve the understanding of engineering behavior of non-plastic silty sand. Key Words : non-plastic silty sand, liquefaction resistance, internal erosion 1. INTRODUCTION Engineering properties of non-plastic silty sand have attracted great interest on the research to soil liquefaction and internal erosion induced ground failures. During the 1999 Chi-Chi earthquake, serious soil liquefaction damages were observed in central Taiwan including Wu-Feng, Nan-Tou, and Yuen-Lin areas. The post-earthquake study indicated that most soil liquefaction occurred in silty sand deposits with high fines content. In 2004 to 2005, several catastrophic subway construction failures occurred in Kaohsiung City, Taiwan. Results of forensic investigations indicated that piping failure of non-plastic silty sand is the dominate factor causing serious tunnel and excavation pit collapses. Moreover, the Tokyo Bay area also suffered from serious soil liquefaction damage during the 2011 Great East Japan earthquake. Preliminary reconnaissance also concluded that majority of liquefaction occurred in the reclaimed silty sand deposits. However, difficulties occurred in undisturbed sampling of sensitive non-plastic silty sand material with high fines content and high water content. Shelby tube and freezing techniques tend to cause loss of fines and disturbance during penetration and drainage processes. In an effort to investigate the engineering properties of non-plastic silty sand, authors have adopted - gel-p sampler is capable of preserving fines content and effectively reducing intrusion friction by introducing polymer gel along the sampling tube while it is penetrated into intact soil deposits. - ly introduced to acquire undisturbed soil specimens. Series of laboratory dynamic triaxial tests were then conducted to identify strength 1-573 -
properties of non- -plastic silty sand. It was found that higher non-plastic fines content of silty sand would result in lower cyclic liquefaction resistance as well as higher internal erodibility. Moreover, the indicated trends would become more obvious when such non-plastic silty sand was subjected to di ineering behavior of non-plastic silty sand. 2. GEL-PUSH SAMPLING TECHNIQUE Undisturbed sampling of high fines content silty sand faced several technical difficulties in the past. Con- content silty sand specimens because the excessive friction generated during penetration tends to cause serious disturbance to the specimens. Therefore, the Shelby tube sampling technique often results in an incomplete, poor qulity soil sample. Moreover, the ground freezing or tube freezing process those generally were used for preserving sampled would cause drifting of fines content and disturbance to sensitive micro structure during freezing and de-freezing. Serious fines content loss can occur when such freezing methodologies are used. The Gel-Push sampling technique was firstly developed to retrieve gravel material as an alternative replacing ground freezing method in Japan in 2004. The Gel-Push sampler was then introduced to Taiwan by the authors (Lee and Ishihara) in 2006 in an attempt to obtain undisturbed high fines content silty sand during the forensic investigation of a subway construction failure. It was modified to accommodate the thin wall tube inside the sampler to become a triple tube system. The newly developed sampler was designed to allow polymer lubricant to seep into the tube wall while the tube was penetrated into the soil by hydraulic pressure. Figure 1 shows the schematic drawings of the Gel-Push sampler at different stages of sampling on process. As shown in the figure, the outer tube is designed to secure the borehole and to keep the penetration rod and piston fixed in alignment during penetration. The middle tube acts as the guiding tube to push sampler into soil. As the sampling process starts, polymer gel is squeezed out from the chamber and seep into both outer side of the guiding tube and the inner side of the thin wall tube. The sampler is also designed with a cutter which is attached to the guiding tube to allow smooth penetration, and a catcher fixed at the bottom of the thin wall tube to keep soil specimen falling out during uplifting. The polymer gel only contaminates limited superficial portions of the specimen because a very small amount of polymer gel is applied. Figure 2 shows that the silty sand specimen was obtained using Gel-Push technique. 3. SITE INFORMATION The high fines content silty sand exists extensively over central to southern parts of western Taiwan. The formation of such unique geological material is a result of rapid weathering and abrading processes. Figure 3 shows the Scanning Electron Microscope (SEM) image of fines particles that obtained from various studied sites. As shown in the figure, fines particles of such silty sand material have angular to sub-angular shapes. This evidence clearly indicates that such soils have almost no plasticity. Figures 4 and 5 show the location map and soil profile information of the studied site (HH01). The studied site in located in Hsin Hwa, Tainan, Taiwan. This site was selected because the widespread soil liquefaction was observed during a magnitude 6.4 earthquake that occurred in 2010 (Figure 5). As depicted in Figure 5, a silty sand layer located between 2m to 10m below the ground surface contains high fines content ranging from 10% to more than 50%. A total of four boreholes were drilled. Gel-Push sampling was conducted in three boreholes, and conventional Shelby tube sampling was conducted in the remaining borehole for purposes of comparison. 4. LABORATORY TESTS Two laboratory tests were conducted to preliminarily investigate the engineering properties of non-plastic 2-574 -
silty sand. Cyclic triaxial tests were conducted to investigate the dynamic properties of the non-plastic silty sand. Tests were performed on both undisturbed soil samples which obtained by the Gel-Push sampling technique and bulkily remolded samples. Effect of disturbance and fine contents are two major investigated factors. (FWPH) Test Device, developed in this study was used to observe the phenomena of internal erosion for remolded non-plastic silty sand with different fines contents under various confining pressures. (1) Cyclic Triaxial Test Figure 6 shows some typical test results from cyclic triaxial tests. As shown in the figure, the Gel-Push specimen is with high cyclic resistance and produces larger yielding strain than those of remolded specimen with the same density and deviator stress. The results of all cyclic triaxial tests are shown in Figure 7, as illustrated in the figure, undisturbed soil specimens have higher cyclic strengths than those of remolded specimens with the same fines contents and the same void ratios. Under the same void ratio condition, specimen with higher fines contents tends to have smaller cyclic strength. This phenomenon becomes more noticeable for the remolded specimens. Furthermore, results of the cyclic triaxial tests on undisturbed specimens were converted into field cyclic resistance ratio by taking the cyclic stress ratios at number of cycle of 20. Field standard penetration test N-values at locations of sampling were also converted to corrected blow count, N1-60, accordingly. Comparison of the test results with semi-empirical charts proposed by Youd and Idirss (1997) is shown in Figure 8. As shown in the figure, test results are quite deviated from the proposed curves. It was found that specimens with higher non-plastic fines content result in lower cyclic liquefaction resistance. Results of the cyclic triaxial tests verify that soil liquefaction can even occur in silty sand deposits with high non-plastic fines contents. Moreover, such non-plastic silty sand deposits would have less liquefaction resistance when such deposits subjected to disturbances. Fig. 1 Schematic drawings of Gel-Push sampling technique Fig. 2 Undisturbed sensitive silty sand specimen retrieved using Gel-Push sampling technique Fig. 3 Fines particles SEM images of studied silty sand 3-575 -
Fig. 4 Location map of the studied site HH01 Fig. 5 Soil profile and photos of the studied site (HH01) (a) undisturbed specimens (b) remolded specimens Fig. 6 Typical results of cyclic triaxial tests (a) undisturbed specimens Fig. 7 Summary of cyclic triaxial tests (b) remolded specimens (2) Flexible Wall Pin Hole Test In order to investigate internal erosion properties of the non-plastic silty sand, the Flexible Wall Pin Hole Test Device, was developed by combining concepts both of the conventional Pin Hole test and of the triaxial test. the conventional Pin Hole test is limited in condition with a constant hydraulic gradient and to be conducted in a rigid wall chamber. Movever, the confining pressure is not able to adjust to simulate field conditions. FWPH test device was designed to take advantages of both the specimen preparation and pressure control functions of the triaxial test so that confining pressure and internal erosion pressure can be controlled during testing. Figure 9 shows the details of FWPH device. It was modified from regular triaxial test chamber. The loading rod was adapted to accommodate a pin with 1mm diameter and allow controlled erosion pressure 4-576 -
to flow through. The chamber and top caps of FWPH device were also modified to have spaces for the erosion pin and still with the sealing for water tight. The lower cap was designed to allow fines to be drained out with the erosion flow. The FWPH test begins with applying cell and back pressures to the specimen. After completing both saturation and consolidation, the back pressure is then increased to the internal erosion pressure and in proportions to cell pressure until large amount of fines bleeding out of the specimen or 80% of original cell pressure is reached. Critical erosion pressure, Ue,cr, is defined as the internal erosion pressure when large amount of fines are observed and bottom of the specimen starts to cave in as shown in Figure 10. For each stage of internal erosion pressure applied, pore water drained from the specimen is collected to examine the level of internal erosion. Table 1 summarizes results of FWPH tests. Test specimens were prepared with two different relative densities to account for loose and dense states with different fines contents ranging from 0% to 40%. As depicted in the table, a non-plastic silty sand at loose a state has much higher internal erosion potential than that at its dense state. Moreover, when higher confining pressures are maintained, a non-plastic silty sand has less internal erosion potential. Most importantly, the higher the fines contents of the non-plastic silty sand, the higher the internal erosion potential is clearly observed. Fig. 8 Cyclic resistance ratio versus corrected blow count for non-plastic silty sand Fig. 9 Details of the Flexible Wall Pin Hole Test Device Fig. 10 Details of the Flexible Wall Pin Hole Test Device c ' (kpa) 100 150 200 Table 1 Summary of FWPH test results Dr Ue,cr 60 % 85 % 0 % c ' c ' 20 % c ' c ' 40 % c ' c ' 0 % c ' c ' 20 % c ' c ' 40 % c ' c ' 0 % c ' c ' 20 % c ' c ' 40 % c ' c ' F C 5. CONCLISIONS The Gel-Push technique has been proven to be a better and more reliable sampling measure for retrieving the good quality non-plastic silty sand specimens. The triple tube system and polymer gel lining appear to be able 5-577 -
to effectively reduce sampling disturbance due to the wall friction. Results of cyclic triaxial tests on non-plastic silty sand indicate that, for specimens with the same void ratios, silty sand with higher fines contents tends to have smaller cyclic strength. This phenomenon becomes much more noticeable on the remolded soil specimens. Moreover, it was found that a higher non-plastic fines content of silty sand results in lower cyclic liquefaction resistance. Such non-plastic silty sand deposits have less liquefaction resistance when subjected to disturbance. FWPH test results show that internal erosion potential of non-plastic silty sand is affected by its density and confining pressure. Non-plastic silty sand at a loose state has much higher internal erosion potential than that at its dense state. Moreover, when higher confining pressures are maintained, non-plastic silty sand has less internal erosion potential. Most of all, the higher the fines contents of the non-plastic silty sand, the higher the internal erosion potential. This paper reports the preliminary progress of research efforts in investigating engineering properties of non-plastic silty sand. It includes the application of newly developed sampling technique, cyclic triaxial tests, and FWPH tests. Results of rties of non-plastic silty sand material. More efforts should be paid to similar research to generalize and to verify the findings and to obtain more knowledge on engineering behaviors of non-plastic silty sand. REFERENCES 1) S Seed, H. B. Soil liquefaction and cyclic mobility evaluation for level ground during earthquakes, Journal of the Geotechnical Engineering Division, ASCE, Vol. 105 GT2, pp. 201-255. 1979 2) Technical Report NCEER-97-0022 (1997). 6-578 -