Keywords

1 Introduction

As stated by Tobita et al. (2023), Ottawa F-65 sand was used as the standard sand of the “LEAP-ASIA-2019” exercise. This sand is a clean, poorly graded, whole grain silica sand, containing less than 0.5% fines by mass (Carey et al., 2019).

One of the main objectives of LEAP is to generate a large and high-quality database of physical models and element tests that would contribute to assess and validate the capabilities of the analytical tools to predict the mechanism and consequences of liquefaction. Researchers have made several efforts to establish and expand a laboratory test database to characterize the physical and mechanical properties of Ottawa F-65 sand, which has also been utilized in the LEAP-GWU-2015 and LEAP-UCD-2017 exercises (El Ghoraiby & Manzari (2018), El Ghoraiby et al., 2019; Carey et al., 2019; Parra Bastidas, 2016; Vasko, 2015).

In order to contribute to the understanding of the mechanical properties of Ottawa F-65 sand, a series of hollow cylinder dynamic torsional shear tests for four different relative densities (Dr = 50%, 60%, 70%, and 85%) were conducted in the installations of the Disaster Prevention Research Institute at Kyoto University. Although a comprehensive review of the test findings has already been published (Vargas et al., 2020), this paper focuses on the details of the model preparation and test results of the tests performed before the LEAP-ASIA-2019 workshop, which were made available to modelers for the calibration of constitutive models and simulation of element tests (Ueda et al., 2023).

2 Torsional Shear Apparatus

The torsional shear apparatus used in the tests presented in this paper is schematically shown in Fig. 4.1. The apparatus is capable of subjecting a hollow cylindrical sample with an inner diameter of 6 cm, outer diameter of 10 cm, and height of 10 cm to a wide range of stress paths for both monotonic and dynamic tests.

Fig. 4.1
A schematic of the torsional shear apparatus. It is labeled as vertical loading actuator, torsional loading, dial gauge, torque sensor, vertical load sensor, rotational displacement sensor, soil specimen, piping system, and pore pressure sensor.

Torsional shear apparatus at the Disaster Prevention Research Institute – Kyoto University. (Vargas et al., 2020)

Full size image

As shown in Fig. 4.1, the apparatus can apply vertical load and torque independently using pneumatic cylinders. Following the recommendation of Koseki et al. (2005), the load cell to measure vertical load and torque was placed inside the pressure cell to eliminate friction between the loading shaft and the bearing house.

To avoid non-uniform normal stresses across the sample, the same inner and outer pressures were applied at all test steps; also, since the tests were performed under isotropic conditions, no deviator stresses were applied.

Volumetric strains were measured using a low-capacity differential pressure transducer that measured changes in water volume in a burette connected to the back pressure (maintained at around 200 kPa) and the specimen. Rotational and axial deformations were measured using a potentiometer and a dial gauge, respectively. Table 4.1 summarizes the measured variables and the sensor characteristics.

Table 4.1 Sensor list
Full size table

3 Dynamic Torsional Shear Tests and Model Preparation

Prior to the LEAP-ASIA-2019 Workshop, 16 stress-controlled hollow cylinder dynamic torsional shear tests were conducted under isotropic conditions; Table 4.2 shows the characteristics of each test.

Table 4.2 Hollow cylinder dynamic torsional shear tests developed at Kyoto University
Full size table

The specimens were prepared taking as a reference the Japanese Standards JGS-0550-2009 (Japanese Geotechnical Society Standards, 2018a–1) and JGS-0551-2009 (Japanese Geotechnical Society Standards, 2018b–2).

3.1 Model Preparation: Air Pluviation and Chamber Setting

To obtain a fabric similar to the physical models (prepared for the “LEAP-ASIA-2019” exercise), the samples were prepared through the “air pluviation” technique, and, to guarantee the homogeneity of the sample and reduce the variability, the height of the pluviation tool was increased during the sample preparation, aiming to keep the dropping height constant along with the sample.

Following the air pluviation process, the top cap was placed, and negative pressure of approximately −15 kPa was applied to the model (between the membranes) to uphold the sample’s shape while removing the molds and placing the cell chamber. Figure 4.2 shows the sample after the placement of the top cap and the removal of molds. To minimize variability and guarantee uniformity among the experiments, precise measurements of the specimen dimensions and sand weight were taken subsequent to mold removal. It is noteworthy that even minor differences in weight or dimensions can considerably influence the estimated relative density.

Fig. 4.2
A photograph depicting the arrangement of hollow cylinders within the apparatus after the removal of molds.

Hollow cylinder model preparation – model after mold removal

Full size image

After confirming the model’s density, the outer chamber was carefully placed and fixed to prevent any leakage, as shown in Fig. 4.3.

Fig. 4.3
A photograph exhibits the placement of the outer chamber of the hollow cylinder, connected to the load cell.

Hollow cylinder model preparation – outer chamber setting

Full size image

3.2 Model Preparation: Saturation

Upon filling the outer and inner cells with water, an isotropic pre-consolidation of 20 kPa was applied to both the outer and inner cells to prevent sample deformations in subsequent steps; following this step, the vacuum pressure utilized to preserve the sample’s shape was released.

The saturation process began by flowing CO2 through the sample for around 15 min to promote the dissolution of gas bubbles trapped in the sand. Subsequently, degassed water was flowed through the sample until at least 1 liter of degassed water (roughly twice the sample’s volume) had passed through.

Once the necessary quantity of water had flowed through the sample, a back pressure of 200 kPa was applied to the sample (both inner and outer chambers). The saturation level was then verified by determining the Skempton B value, which was confirmed to be above 0.95 in all experiments for an increment of 20 kPa under isotropic and undrained conditions.

3.3 Model Consolidation and Testing

The specimen was isotropically consolidated to a total pressure of approximately 300 kPa (i.e., 100 kPa of effective stress) under drained conditions. Following the consolidation process, a stress-controlled cycling sinusoidal shear stress was applied to the sample in undrained conditions (Fig. 4.4 shows the applied target wave). As specified in the Japanese Standards (Japanese Geotechnical Society Standards, 2016), the load was applied until achieving at least 7.5% of double amplitude shear strain. Additionally, since the loads were applied using a stress-controlled mechanism, reliable information could not be obtained at very large strains.

Fig. 4.4
A line graph plots shear stress versus time. The line exhibits a sinusoidal pattern.

Cyclic sinusoidal shear stress applied to the sample. (Vargas et al., 2020)

Full size image

4 Test Results

The dynamic stress-strain characteristics of 16 hollow cylinder models were estimated using a torsional shear apparatus, as described in Sects. 4.2 and 4.3. Detailed test results (stress-strain characteristics of each model) have been included in the Appendix section.

Figure 4.5 shows the estimated liquefaction resistance curve (LRC) for γDA = 7.5% (i.e., the number of cycles required to reach a 7.5% DA shear strain); a clear trend and consistency among the tests is observed.

Fig. 4.5
A multi-line graph of cyclic stress ratio versus number of cyclic loads. The data points are D r equals 50%, 60%, 70%, and 85%. The regression lines for each data point exhibit a decreasing trend. This graph aims to depict the number of cycles required to achieve a 7.5% cyclic shear strain.

Liquefaction resistance curve of Ottawa F-65 sand (γDA = 7.5%). (Vargas et al., 2020)

Full size image

Similarly, Figs. 4.6, 4.7, and 4.8 show the LRC for γDA = 3.0%, γDA = 1.5%, and ru = 0.95.

Fig. 4.6
A scattered graph of cyclic stress ratio versus number of cyclic loads. The data points are D r equals 50%, 60%, 70%, and 85%. The data points are denser at 10 to 90 cyclic loads. The graph is designed to portray the cycle count necessary to reach a 3.0% cyclic shear strain.

Liquefaction resistance curve of Ottawa F-65 sand (γDA = 3.0%). (Vargas et al., 2020)

Full size image
Fig. 4.7
A scattered graph of cyclic stress ratio versus number of cyclic loads. The data points are D r equals 50%, 60%, 70%, and 85%. The data points are denser at 5 to 70 cyclic loads. The graph is designed to portray the cycle count necessary to reach a 1.5% cyclic shear strain.

Liquefaction resistance curve of Ottawa F-65 sand (γDA = 1.5%). (Vargas et al., 2020)

Full size image
Fig. 4.8
A scattered graph of cyclic stress ratio versus number of cyclic loads. The data points are D r equals 50%, 60%, 70%, and 85%. The data points are denser at 5 to 80 cyclic loads. The graph is designed to portray the cycle count necessary to reach a 0.95% cyclic shear strain.

Liquefaction resistance curve of Ottawa F-65 sand (ru = 0.95). (Vargas et al., 2020)

Full size image

Additionally, as seen in Fig. 4.9, it has been found that the friction angle ϕf value remained constant for all relative densities (ϕf ≈ 35°). On the other hand, the phase transformation angle ϕp increased with the relative density; values of 20.2°, 22.6°, 24.0°, and 25.0° were obtained for relative densities of 50%, 60%, 70%, and 85%, respectively.

Fig. 4.9
4 multi-line graph plots shear stress t versus mean effective stress for D r equals 50%, 60%, 70%, and 85%. In the first two graphs, the paths exhibit a spiraling pattern within the range of negative 20 to 20 on the y-axis. The paths in the third and fourth graphs spiral within the range of negative 35 to 35 on the y-axis.

(a) Stress path for Model Dr50_3. (b) Stress path for Model Dr60_5. (c) Stress path for Model Dr70_3. (d) Stress path for Model Dr80_2

Full size image

5 Repeatability of Test Results

To ensure the consistency of the tests, a repeatability test (Dr70_3) was performed; this repeatability test was prepared and tested under the same conditions as test Dr70_2.

Figure 4.10 shows a comparison between the test results of Dr70_2 and Dr70_3 (repeatability test), including the development of excess pore water pressure ratio (EPWP ratio), the development of shear strain, and the corresponding stress path. The results indicate a good agreement between the tests, demonstrating that the tests are repeatable and consistent with each other.

Fig. 4.10
3 double line graph of E P W P ratio, shear strain, and shear stress versus number of cyclic loading and mean effective stress. The lines are D r 70 underscore 2 and 70 underscore 3. In 1 and 2 the graph has a fluctuating trend. In 3 graph the path spirals between negative 25 and 25 on the y-axis.

(a) Development of EPWP – Models Dr70_2 and Dr70_3. (b) Development of shear strain – Models Dr70_2 and Dr70_3. (c) Stress path – Models Dr70_2 and Dr70_3. (Vargas et al., 2020)

Full size image

6 Conclusions

This paper presents a series of hollow cylinder dynamic torsional shear tests performed to contribute to the understanding of the mechanical properties of Ottawa F-65 sand. The tests were developed at the installations of the Disaster Prevention Research Institute (DPRI) at Kyoto University, as part of the “LEAP-ASIA-2019” activities.

  • Based on 16 isotropically consolidated tests, the stress-strain characteristics of Ottawa F-65 sand were investigated under 4 different densities and a wide range of shear stress ratios.

  • Based on the stress paths, it was shown that the friction angle ϕf remained constant for all relative densities (ϕf ≈ 35°); meanwhile, the phase transformation angle ϕp increased with the relative density.

  • The good agreement in the repeatability tests shows that the tests are repeatable and consistent among each other; therefore, the provided test results are ready to be used in the calibration of constitutive models, simulation of element tests, V&V exercises/research, etc.