Performance of Vacuum Consolidation Method in Peat Ground

Transcription

1 Int. Workshop on Geotechnics of Soft Soils-Theory and Practice. Vermeer, Schweiger, Karstunen & Cudny (eds.) 2003 VGE Performance of Vacuum Consolidation Method in Peat Ground Hirochika Hayashi, Jun ichi Nishikawa, Satoshi Nishimoto and Kengo Sawai Civil Engineering Research Institute of Hokkaido (CERI), Sapporo, JAPAN ABSTRACT: To examine the improvement effect of the vacuum consolidation method in peat ground, which is composed of highly organic soil, test construction of an embankment was conducted for an expressway construction project on peat ground. Although the peat ground of the site is extremely soft, the vacuum consolidation method enabled the construction of a 10-meter high embankment in a short period of time. After examining the test construction results, it was evident that the ratio of increments of the undrained shear strength to increments of the effective stress (Cu / v ) in the peat layer to which the vacuum consolidation method was applied had a higher value than that of the peat layer to which no countermeasure was applied. In addition, the actual values of the settlement correspond to the calculated values obtained using Barron s formula, which doesn t take well resistance into account. 1 INTRODUCTION The peat ground widely distributed throughout Hokkaido, Japan, is extremely soft problematic ground because it is highly organic soil with high natural water content. Therefore, when a road embankment is constructed on peat ground, countermeasures must be taken in advance. Although various design and construction methods for measures against peat ground have already been established in Japan, studies are continuously being carried out to develop better techniques. Under these circumstances, test embankment construction employing the vacuum consolidation method was conducted for an expressway construction project on peat ground. Vacuum consolidation is a construction method utilized to accelerate ground settlement and increase ground strength in a short period of time by discharging pore water contained in the ground by means of prefabricated vertical drains and vacuum pumps. There have only been a few cases in which the vacuum consolidation method has been applied to peat ground in Japan. Thus far, a design method that takes into consideration the engineering characteristics peculiar to peat ground has not been established. In concrete terms, the following problems must be solved: (1) Negative pressures seem to decrease deep in the ground; however, this has not been determined quantitatively. (2) A projection method for increased strength and settlement behavior taking into account the characteristics of peat ground has not yet been established.

2 2 TEST CONSTRUCTION CONDITIONS Test construction was performed at the Mihara-Bypass on National Highway Route 337 in the suburbs of Sapporo. At this site, it was necessary to construct a high embankment on peat ground. For this reason, the vacuum consolidation method (Fig. 1) was employed to assure stability and decrease residual settlement. Figure 1. Typical setup of vacuum consolidation method 2.1 Ground conditions The ground at the test site is a typical peaty soft ground commonly seen in Hokkaido (Fig. 2). The natural water content of the peat layer and that of the clayey peat layer underlying it range from 200% - 700% and the Dutch cone penetration resistance varies from 100 kn/m kn/m 2. These values are average for peat ground distributed in Hokkaido (Noto, 1991). The relationship between the consolidation yield stress and effective overburden pressure indicates that the ground is by and large normally consolidated. According to a permeability test conducted at the site, the coefficient of permeability of the fine sand layer between the upper and lower clay layers was cm/sec. Figure 2. Cross section of test construction site 2.2 Status of construction and field observations The vertical drain materials were laid out at intervals of 80 cm in a square shape and placed into the ground as deep as GL-19.9 m. Two vacuum pumps were installed for the ground improvement area of approximately 3,200 m 2. Ground behavior was observed using devices such as a settlement plate, a differential settlement gauge, a piezometer and a borehole inclinometer. Dutch cone penetration tests were performed to measure the ground strength after ground improvement. To obtain the strength distribution in this case, the Dutch cone penetration tests were also applied to the top and toe of the slope, in addition to the center of the embankment. Moreover, pressure meter tests in boreholes were subsequently conducted and an increase in the modulus of ground deformation was confirmed.

3 3 TEST RESULTS AND DISCUSSION 3.1 Test construction results Construction of the embankment commenced 21 days after the vacuum pumps were put into operation. Excluding the sand blanket, the depth of the embankment was approximately 10 meters. The average construction speed was 13 cm/day, which was very fast for embankment construction on peat ground (Fig. 3). However, the sufficient stability of the embankment was confirmed by the construction management. Moreover, no deformities, such as cracks on the crest of the embankment, that would indicate failure of the embankment, were observed, thus demonstrating the favorable improvement effect of this construction method for the safety of roads laid on peat ground. Observation of the settlement behavior clarified that settlement progresses even after stopping of the vacuum pumps, though the progress is gradual. By soil layer, the settlement of the underlying clayey peat layer lags behind that of the other layers. Regarding the excess pore water pressure in the peat layer before construction, a negative pressure of approximately 50 kn/m 2 occurred. The negative pressure in the lower soil layers, however, remained around the level of 20 kn/m 2 (Fig. 3). Figure 3. Changes of observed settlement and excess pore water pressure The borehole inclinometer measurement results immediately before construction, upon completion of construction, at the time of vacuum pump stoppage and 25 days after stopping of the vacuum pumps are shown in Fig.4. In the time period after the start of vacuum pump operation and before construction, lateral displacement took place toward the inside of the embankment between the peat layer and the upper clay layer. The peat layer showed a more conspicuous trend of displacement with a maximum value of approximately 15 cm. During construction of the

4 embankment, lateral displacement occurred toward the outside of the embankment, but it was only about 25 cm at the maximum. This lateral displacement was less than the values in general for noncountermeasure layers. In addition, no behavior was observed that indicated progress in shear deformation after stopping of the vacuum pumps. Figure. 4. Depth distribution of lateral displacement Figure. 5. Depth distribution of negtive pressure 3.2 Behavior of negative pressures The depth distribution, taking the difference between the pore water pressure immediately before the start of construction (20th day after the start of vacuum pump operation) and the hydrostatic pressure as negative pressure is shown in Fig. 5. The design value of negative pressure is often set at 60 kn/m 2. However, the value has only been fulfilled for peat layers. In the deeper layers, only negative pressures of approximately one third of the design value were observed. This may be due to the effect of the fine sand layer between the upper and lower clay layers, but the exact cause remains unknown. In any case, this phenomenon suggests that the design value of negative pressure in deep layers should be decreased to a certain degree. 3.3 Evaluation of increases in strength Here increases in the ground strength after embankment construction are discussed. The increments of the undrained shear strength of the ground to the increments of effective stress (Cu / v ) by layer and ignition loss by layer were summarized (Fig. 6). In conjunction with this, the actual value (t =16-18 kn/m 3 ) was adopted as the site density of the embankment. The ground strength was calculated based on Dutch cone penetration tests carried out at the center of the embankment as well as the top and toe of the slope. The findings obtained from the embankment constructed on nearby noncountermeasure ground are also provided for comparison. Regarding the noncountermeasure ground, it was found that Cu / v is approximately 0.4 in the layers with ignition loss of more than 20%, a typical value for peat. In comparison with this, the layers other than inorganic clay to which the vacuum consolidation method was applied had larger values. This agrees with the test results from this project in which a high embankment was realized in a short period of time through the adoption of the vacuum consolidation method. It is thought that the reason the value of Cu / v differed from that of the noncountermeasure case was due to the difference in the deformation modes of the ground (Fig. 4). That is, it is believed that although in the area around the toe of the slope for construction of an embankment on noncountermeasure ground the shear deformation of the ground prevails over the consolidation deformation and an increase in strength by consolidation is not feasible, the vacuum consolidation method still provided a significant increase in strength by consolidation to the area. Moreover, in the inorganic clay layer, no difference was found in Cu / vfor the noncountermeasure embankment and that constructed using the vacuum consolidation method.

5 Fig. 6. Ignition loss and Cu /v Fig. 7. Ignition loss and E /v As with the case of increases in strength, a distinct relationship was observed between increases in ignition losses and the modulus of deformation (Fig. 7). An approximate expression for the relationship is as follows: E / v = 0.13 Lig (1) Where, E : increment of the deformation modulus of ground v : increment of effective stresses Lig : ignition loss 3.4 Evaluation of consolidation acceleration The settlement for the ground surface and each layer and its degree of consolidation are shown in Table 1. As for the surface settlement, the degree of consolidation was 84% upon the stopping of vacuum pump operation, indicating that as much as 70 cm of the residual settlement would be forecasted. Regarding the degree of consolidation by layer, more than 90% was obtained for the peat layer and the clayey peat layer upon the stopping of vacuum pump operation. However, it remained at 68% for the lower clay layer. Although it is common for the settlement speed of clay to be lower than that of peat ground, it is thought that the relatively low negative pressures of the lower clay soil were also influential. Table 1. Degrees of cosolidation for ground surface and each layer Next, a quantitative assessment of the effect of consolidation acceleration is attempted. The applicability of Barron s formula, which is used to analyze settlement through utilization of the vertical drain method, must first be examined. To calculate Barron s formula, the coefficient of consolidation Cv is required. For calculations for clay, the standard consolidation test results were used as they were. On the other hand, it is known that the settlement speed of peat is so high in comparison with that of clay that the calculated values do not agree with the actual settlement behavior even if the coefficient of consolidation Cv obtained from a standard consolidation test is used (Noto,1991). Therefore, Cv was inversely analyzed based on the settlement behavior measured at the embankment on the nearby noncountermeasure ground. The coefficient of consolidation Cv obtained by inverse analysis was seven times that obtained using the standard

6 consolidation test (Hayashi et al, 2002). This inversely analyzed coefficient of consolidation was then used for the calculation made employing Barron s formula. When comparing the actual values with the calculated values, the calculated results, which didn t take well resistance into account, and the actual values nearly correspond to each other (Fig. 8). This means that if the consolidation coefficient of peat layers is accurately set, the settlement behavior when the vacuum consolidation method is applied can be calculated using Barron s formula. Fig. 8. Comparison between observed and calculated settlment of peat and lower clay layer 4 CONCLUSIONS The findings from the test construction are summarized as follows: (1) Although it was an extremely soft peat ground, the vacuum consolidation method enabled the construction of a high embankment in a short period of time. (2) The Cu / v of the peat layer to which the vacuum consolidation method was applied showed a higher value than that of the noncountermeasure peat layer. (3) It is possible to estimate the modulus of deformation of the improved ground based on the ignition loss. (4) The actual values of the settlement nearly correspond to the calculated values obtained using Barron s formula, which doesn t take well resistance into account. (5) In the layers deeper than the peat ground on the surface, only negative pressures of about one third of the design value were observed. This indicates that the design value of negative pressure in deep layers should be decreased to a certain degree. ACKNOWLEDGEMENTS In conducting this test construction, we received generous support from those concerned at the Hokkaido Development Bureau, the Ministry of Land, Infrastructure and Transport. We would like to take this opportunity to express our immense gratitude to them. REFERENCES Hayashi H., Nishikawa J. & Egawa T Improvement effect of prefabricated vertical drain to peat ground. Proc. of 4 th International Conference on Ground Improvement Techniques: Noto S Peat ground engineering : Gihoudo Shuppan publication (in Japanese)