Keywords

1 Introduction

The LEAP-ASIA-2019 project involved nine numerical simulation teams from different academic institutions and geotechnical companies from around the world; they participated in the modeling of some of the centrifuge model experiments performed at several research institutions. The simulation exercise consisted of the calibration of constitutive model parameters, Type-B predictions, and Type-C predictions. This chapter presents an overview of the results of the first phase (i.e., model calibration) of this exercise. The main objective of this phase was to provide the numerical simulation teams with the opportunity to calibrate their constitutive models, which will be used in the Type-B simulations, using the results of cyclic shear tests performed on Ottawa F-65 sand during the LEAP-2019 project.

For the calibration phase of constitutive models, a series of hollow cylinder torsional shear tests were performed at Kyoto University (KyU) for Ottawa F-65 sand with a relative density (Dr) of 50% and 60% under an initial effective confining stress of 100 kPa. Also, direct simple shear tests were performed at George Washington University (GWU) for Dr = 71% under an initial effective vertical stress of 100 kPa and Dr = 69% under 40 kPa.

The element tests mentioned above provided new datasets that complement the monotonic and cyclic triaxial shear tests reported by Vasko (2015) and Vasko et al. (2018), monotonic and cyclic simple shear tests by Bastidas (2016) and Bastidas et al. (2017), and cyclic triaxial tests by El Ghoraiby and Manzari (2018) and El Ghoraiby et al. (2019). These tests were previously made available to the numerical simulation teams that participated in the numerical simulation of the LEAP-2017 project. The new datasets were made available to all the numerical simulation teams that participated in the LEAP-2019 project via DesignSafe, as described below.

The timeline for this calibration phase of the LEAP-2019 project was as follows:

  1. 1.

    All the element test data were made available on DesignSafe to the numerical simulation teams by December 5, 2018. These are as follows:

  2. 2.

    The participating teams were requested to simulate a selected number of the provided test data and liquefaction strength curves that were obtained from cyclic direct simple shear tests and cyclic torsional shear tests. The critical tests to be simulated were the cyclic torsional shear test for Dr = 50% and 60% (under an initial effective confining stress of 100 kPa). It was required to compare the simulated stress paths and stress-strain responses to the experimental results reported by KyU. If time allowed, it was desirable to show the validity of constitutive models for the other experimental results having higher relative densities. The numerical simulation team submitted the results of their element test simulations and comparisons with those of the provided element tests in the form of a detailed report by January 11, 2019.

2 The Numerical Simulation Teams

Table 2.1 shows the numerical simulation teams who submitted their calibration reports and participated in the Type-B simulation exercise. The constitutive model and the analysis platform used by each numerical simulation team are also listed in the table. Mode-detailed information of each constitutive model and the numerical simulation techniques used by each simulation team are provided in separate papers (Tanaka et al., 2023; Hyodo & Ichii, 2023; Fasano et al., 2023; Qiu & Elgamal, 2023; Elbadawy & Zhou, 2023; Reyes et al., 2023; Wang et al., 2023).

Table 2.1 Numerical simulation teams
Full size table

3 Results of the Element Test Simulations

Figures 2.1, 2.2, 2.3 and 2.4 show a detailed comparison of the numerical simulations of the undrained cyclic torsional shear tests on Ottawa F-65 sand for Dr = 50% with different cyclic stress ratios (i.e., CSR = 0.19, 0.15, 0.13, and 0.10). The simulations are labeled Simulations 1 to 11. The numbers refer to the order of the simulation teams in the table presented above. The numerical simulation teams 1 and 2 belonging to the same organization used the same analysis platform with the same constitutive model, but they are distinguished because they carried out the calibration independently. It is also noted that the same predictor performed Simulations 8 and 9, but they are distinguished because different constitutive models were used in the simulations. Figures 2.5, 2.6, 2.7, 2.8 and 2.9 show a similar comparison of the numerical simulations of the undrained cyclic torsional shear tests for Dr = 60% with different cyclic stress ratios (i.e., CSR = 0.20, 0.18, 0.15, 0.13, and 0.12). The numerical simulation team 3 did not submit simulations for Dr = 60% with CSR of 0.12.

Fig. 2.1
A set of graphs depicts the comparison of measured simulations. a. Excess pore pressure ratio versus number of cyclic loading. b. Shear strain percentage. c. Shear strain versus mean effective stress. d. Shear stress versus shear strain.figure 1

Comparison of the numerical simulations of an undrained cyclic torsional shear test on Ottawa F-65 sand for Dr = 50%, CSR = 0.19. (a) Time history of excess pore pressure ratio, (b) Time history of shear strain, (c) Effective stress path, (d) Shear stress-shear strain relationship

Full size image
Fig. 2.2
A set of graphs depicts the comparison of measured simulations of the cyclic torsional test. a. Excess pore pressure ratio versus number of cyclic loading. b. Shear strain percentage. c. Shear strain versus mean effective stress. d. Shear stress versus shear.figure 2

Comparison of the numerical simulations of an undrained cyclic torsional shear test on Ottawa F-65 sand for Dr = 50%, CSR = 0.15. (a) Time history of excess pore pressure ratio, (b) Time history of shear strain, (c) Effective stress path, (d) Shear stress-shear strain relationship

Full size image
Fig. 2.3
2 sets of graphs. a. 12 graphs of excess pore pressure ratio versus number of cyclic loading depict simulations 1 to 11. All denote an increasing trend high at 1.00. b. Twelve graphs of shear strain % versus number of cycle loading. Simulation 10 denotes a decreasing trend low at (6.0, negative 5).figure 3

Comparison of the numerical simulations of an undrained cyclic torsional shear test on Ottawa F-65 sand for Dr = 50%, CSR = 0.13. (a) Time history of excess pore pressure ratio, (b) Time history of shear strain, (c) Effective stress path, (d) Shear stress-shear strain relationship

Full size image
Fig. 2.4
2 sets of graphs. a. 12 graphs of excess pore pressure ratio versus number of cyclic loading depict measured, simulations 1 to 11. All denote an increasing trend high at 1.00. b. 12 graphs of shear strain percent versus number of cycle loading. 10 denotes a decreasing trend (6.0, negative 5).figure 4

Comparison of the numerical simulations of an undrained cyclic torsional shear test on Ottawa F-65 sand for Dr = 50%, CSR = 0.10. (a) Time history of excess pore pressure ratio, (b) Time history of shear strain, (c) Effective stress path, (d) Shear stress-shear strain relationship

Full size image
Fig. 2.5
2 sets of graphs. a. 12 graphs of excess pore pressure ratio versus the number of cyclic loading depict measured, simulations 1 to 11. All denote an increasing trend high at 1.00. b. 12 graphs of shear strain percent versus number of cycle loading. It represents measured, simulations 1 to 11,figure 5

Comparison of the numerical simulations of an undrained cyclic torsional shear test on Ottawa F-65 sand for Dr = 60%, CSR = 0.20. (a) Time history of excess pore pressure ratio, (b) Time history of shear strain, (c) Effective stress path, (d) Shear stress-shear strain relationship

Full size image
Fig. 2.6
2 sets of graphs. a. 12 graphs of excess pore pressure ratio versus the number of cyclic loading depict measured, simulations 1 to 11. All denote an increasing trend high at 1.00. b. 12 graphs of shear strain percent versus number of cycle loading. It represents measured, simulations 1 to 11.figure 6

Comparison of the numerical simulations of an undrained cyclic torsional shear test on Ottawa F-65 sand for Dr = 60%, CSR = 0.18. (a) Time history of excess pore pressure ratio, (b) Time history of shear strain, (c) Effective stress path, (d) Shear stress-shear strain relationship

Full size image
Fig. 2.7
2 sets of graphs. a. 12 graphs of excess pore pressure ratio versus the number of cyclic loading depict measured, simulations 1 to 11. All denote an increasing trend high at 1.00. b. 12 graphs of shear strain percent versus number of cycle loading. It represents measured, simulations 1 to 11.figure 7

Comparison of the numerical simulations of an undrained cyclic torsional shear test on Ottawa F-65 sand for Dr = 60%, CSR = 0.15. (a) Time history of excess pore pressure ratio, (b) Time history of shear strain, (c) Effective stress path, (d) Shear stress-shear strain relationship

Full size image
Fig. 2.8
2 sets of graphs. a. 12 graphs of excess pore pressure ratio versus the number of cyclic loading depict measured, simulations 1 to 11. All denote an increasing trend high at 1.00. b. 12 graphs of shear strain percent versus number of cycle loading. It represents measured, simulations 1 to 11. 11 denotes a wide range.figure 8

Comparison of the numerical simulations of an undrained cyclic torsional shear test on Ottawa F-65 sand for Dr = 60%, CSR = 0.13. (a) Time history of excess pore pressure ratio, (b) Time history of shear strain, (c) Effective stress path, (d) Shear stress-shear strain relationship

Full size image
Fig. 2.9
2 sets of graphs. a. 12 graphs of excess pore pressure ratio versus the number of cyclic loading depict measured, simulations 1 to 11. All denote an increasing trend high at 1.00. b. 12 graphs of shear strain percent versus number of cycle loading. It represents measured, simulations 1 to 11.figure 9

Comparison of the numerical simulations of an undrained cyclic torsional shear test on Ottawa F-65 sand for Dr = 60%, CSR = 0.12. (a) Time history of excess pore pressure ratio, (b) Time history of shear strain, (c) Effective stress path, (d) Shear stress-shear strain relationship

Full size image

A review of Figs. 2.1, 2.2, 2.3, 2.4, 2.5, 2.6, 2.7, 2.8 and 2.9 indicates the following trends:

  1. 1.

    The majority of the constitutive models are capable of reasonably capturing the overall trends of the measured time histories of excess pore pressure ratio and shear strain, effective stress paths, and stress-strain responses both for Dr = 50% and 60%.

  2. 2.

    Simulations 1–4: Since the constitutive model and the analysis platform are the same, the simulated results are similar to some extent. However, different responses are observed depending on the model parameters; there are many cases where the effective stress path does not reach the origin (i.e., complete liquefaction) in Simulations 1 and 2, but it almost reaches the origin in Simulations 3 and 4.

  3. 3.

    Simulations 5 and 6: Since the analysis platforms are different but the constitutive model is the same, the overall response tendency is very similar. The time history of the simulated excess pore water pressure shows that the pressure tends to rise rapidly at a certain stage, while it is relatively slow in the early stage of loading. This trend can also be seen in the simulated effective stress path.

  4. 4.

    Simulations 7 and 8: Although the constitutive model and the analysis platform are the same, the time history of the simulated excess pore water pressure, the effective stress path, and the associated strain development seem to be slightly different. This is probably due to the difference in the values of the model parameters used.

  5. 5.

    Simulation 10: When the excess pore pressure ratio increases to 0.8–0.9, large shear strains are generated, which is common to other simulations. However, after that, the strain tends to extend relatively slowly; the strain development is almost linear.

  6. 6.

    Simulation 11: As in the other simulations, the shear strain begins to develop when the excess pore pressure ratio exceeds 0.8–0.9. However, the development is not linear and tends to converge gradually; the brittle behavior, in which the strain increases rapidly, is suppressed compared to the other simulations.

Fig. 2.10
2 graphs depict the shear stress ratio versus the number of cyclic loads. The curves are labeled measured and simulations 1 to 11. All curves denote a decreasing trend, and simulation 10 low at (300, 0.10) and 2 low at (200, 0.10) respectively. The values are approximate.

Comparison of the simulated liquefaction strength curves by different numerical simulation teams with the experimental results reported by Ueda et al. (2018) and Vargas et al. (2020, 2023). (a) Dr = 50%, (b) Dr = 60%

Full size image

4 Liquefaction Resistance Curves

The simulated liquefaction resistance curves for γDA = 7.5% (i.e., the number of cycles required to reach a 7.5% double amplitude shear strain) are compared with the laboratory test results in Figs. 2.10a, b for Dr = 50% and 60%, respectively. The following trends are observed from the curves:

  1. 1.

    The majority of the constitutive models are capable of reasonably capturing the overall trends of the measured liquefaction resistance curves both for Dr = 50% and 60%; in particular, the liquefaction strength is accurately simulated for a cyclic stress ratio (CSR) of 0.149 and 0.174 for Dr = 50% and 60%, respectively.

  2. 2.

    Simulations 1–4: Since the constitutive model and the analysis platform are the same, the simulations show similar liquefaction resistance curves, although there are slight differences due to differences in the model parameters used. They can accurately simulate the experimental results even for low CSRs (i.e., a large number of cycles).

  3. 3.

    Simulations 5 and 6: Since the analysis platforms are different but the constitutive model is the same, the simulated liquefaction resistance curves are quite similar. The simulations are capable of reasonably simulating the experimental results, particularly in a relatively large CSR range.

  4. 4.

    Simulations 7 and 8: Although the constitutive model and the analysis platform are the same, the simulated liquefaction resistance curves look different; Simulation 8 shows steeper curves than the experimental curves, although both Simulations 7 and 8 can simulate the measured liquefaction strength for 20 cycles. The difference is probably due to the difference in the values of the model parameters used.

  5. 5.

    Simulation 10: The experimental curves are reasonably simulated over a wide range of CSRs, as in Simulations 1–4.

  6. 6.

    Simulations 9 and 11: The simulations show steeper curves than the experimental curves, although they can simulate the measured liquefaction strength for 10–20 cycles. It is unclear whether this is due to the characteristic of the constitutive models or the model parameters used.

5 Conclusions

This chapter presented a summary of the calibration exercises (i.e., element test simulations) submitted by nine numerical simulation teams that participated in the LEAP-ASIA-2019 prediction campaign. The objective of this element test simulation exercise was to assess the performance of the constitutive models used by the simulation teams for simulating the experimental results of a series of undrained stress-controlled cyclic torsional shear tests on Ottawa F-65 sand for two different relative densities (Dr = 50% and 60%). These simulations demonstrate that majority of the constitutive models are capable of reasonably capturing the measured liquefaction strength curves as well as the overall trends of the stress paths and stress-strain responses both for Dr = 50% and 60%. However, it appeared to be still left for future work to evaluate the validity of constitutive models in consideration of the variations in the laboratory test and/or numerical simulation results.