Queenstown, New Zealand, is a known seismically-active area. A major soil type overlying the flat lands in and around the city is the varved micaceous clayey silt / sandy silt laminated lake sediment overlying the schist bedrock. In-situ (boreholes, CPTu, sCPTu, DMT, Vs and Vp downhole testing) and laboratory investigations (including X-Ray Diffraction mineralogy tests) were carried out on these soils to evaluate the liquefaction potential in the area of interest. However, despite the comprehensive list of assessment methods used, the factual data are not converging at one definitive conclusion and engineering judgement was needed to better interpret the body of data. To clarify the matter, further samples have been collected and monotonic and cyclic triaxial tests were performed to explore and determine the liquefaction resistance that will enable a comparison to the empirical correlations based on case-history data.
Queenstown is a known seismically-active area in New Zealand. The nearby mapped active faults – the Alpine, the Nevis and the Cardrona faults – are all capable of producing earthquakes greater than Mw 7. As per the recent seismic hazard study undertaken by the Otago Regional Council for the Queenstown Lakes District (Otago Regional Council, 2015), the primary seismic hazard facing the region is an Alpine Fault earthquake; which is predicted to have a 30% probability of rupturing in the next 50 years.
Given the seismological setting, we sought to evaluate the liquefaction potential of the horizontally interbedded clayey silt / sandy silt laminated Wakatipu Lake sediment soils (lake sediments). To provide a tangible benchmark, an initial liquefaction susceptibility and triggering assessment was carried out following the simplified procedures set out in Modules 1 and 3 of the New Zealand Geotechnical Society (NZGS) “Modules for Earthquake Geotechnical Engineering Practice (2016)”. Additional methods to assess liquefaction triggering were also utilized (eg. dynamic- cyclic triaxial laboratory tests, as well as shear and compressional seismic wave velocity profiling).
Despite the numerous methods used, the data obtained from conventional empirical liquefaction analyses did not converge towards a definitive conclusion and engineering judgement was needed to better interpret the body of evidence.
In this technical note, the results of the cyclic triaxial testing (CTX) on selected specimens from the Frankton area near the existing Kawarau Falls Bridge are presented in continuation of the previous presented work (Awad & Searle, 2016).
2. Geotechnical Properties of Wakatipu Lake Sediment Soils
The X-Ray Diffraction (XRD) method was used in order to test mineralogy of selected samples from the Wakatipu lake sediments. It was found that the laminated lake sediments are rich in mica and chlorite (~ 84% by weight; mica and chlorite are approximately in equal amounts). Quartz makes up only 10% of the mineralogical composition and feldspars make up the remainder.
Several geotechnical investigations have been carried out throughout the Queenstown area, but this paper is dealing mostly with those in the Frankton area where the samples were retrieved from.
Seismic piezocone penetration testing (sCPTu), with down-hole shear-wave velocity (Vs) profiling and boreholes were performed above the sampling location. The lake sediments had a minimum thickness of 25 m in the area of interest and the site was classified as Class D – deep or soft soil sites – in accordance with NZS1170.5:2004.
Table 1: Lake sediment index properties (laboratory testing)
Figure 1: Normalised soil behavior type plot
Apart from the samples obtained for the cyclic triaxial testing, several other soil samples were retrieved from nearby boreholes at depths ranging from approximately 2m to 27m below ground level and tested in the laboratory to determine the soil composition and material properties. The measured lake sediments index properties are presented in Table 1 and the normalized Soil Behavior Type (SBTn, Robertson 1990, 2009) derived from the piezocone penetration testing (CPT) is presented in Figure 1.
Figure 2: Cone resistances (qc, qt), sleeve friction, pore pressure and Soil Behavior Type Index (Ic) records below 8.4m within the lake sediments (from the site-specific sCPT)
Figure 2 illustrates the sCPT profile from 8.4m down to 33.0 m below ground level (lake sediments) and Figure 3 illustrates the shear wave velocity and shear modulus (Vs and Go) profile from the seismic CPT downhole measurements as well as the estimated Vs and Go values (Robertson and Cabal 2014) from the cone resistance.
Figure 3: Shear wave velocity (Vs) and shear modulus (Go) profile with average ± standard deviation values for the lake sediments
The sCPT data indicates that the lake sediments are transitional in behavior (i.e. transitional between sand-like and clay-like soils) as presented in Figure 4, and the transitional nature is confirmed by the index laboratory testing (medium plasticity silts – ML).
Figure 4: Modified Robertson Soil Behavior Type (SBTn)
As shown in Figure 5 (analysis of the sCPT data done with CPeT-IT software by Geologismiki Ltd.), the lake sediments have little microstructure with K*(G) values less than 330 but above 200. The normalized sCPT measurements also suggest that these soils are contractive at large strains.
Figure 5: Normalized rigidity index; soils with K*(G) > 330 can be considered to be with significant microstructure (e.g. age and cementation)
3. LIQUEFACTION ASSESSMENT-cyclic triaxial testing (ctx)
3.1 Previous Work
Bowen et al. (2015) carried out 13 cyclic triaxial tests on low plasticity micaceous lake sediments in Central Otago which have similar properties to the soils we assessed in Queenstown, but with a lower plasticity range (PI = 0 to 7%). The laboratory testing revealed a cyclic mobility behavior (progressive accumulation of limited strains after ru = 100% has temporarily occurred) but no liquefaction flow-type behavior (rapid increase in pore pressure leading to complete loss of shear strength).
3.2 Soil sampling and laboratory testing procedure
Twelve (12) undrained cyclic triaxial tests were performed on undisturbed samples by the Soil Mechanics, Foundation and Geotechnical Earthquake Engineering Laboratory of the Aristotle University of Thessaloniki (AUTh-SMFGEE) on behalf of Miyamoto International NZ Ltd., and the lead author. The objective was to estimate the liquefaction resistance (cyclic triaxial tests) of the Wakatipu varved lake sediments and the post-liquefaction undrained shear strength (monotonic triaxial tests).
The physical properties of the tested soils are given in Table 2, with the grading curves of two representative samples, after they were loaded cyclically, to be illustrated in Figure 6.
The samples were block samples retrieved by constant pressure via plastic tubes with sharp cutting edges and typical steel samplers above the river level (GWT – saturation level), within the unsaturated zone to avoid disturbance while travelling for New Zealand to Greece. The samples were then carefully extruded, and saturated to represent the site conditions. It should be noted that the comparison between the void ratio, plasticity indices, unit weights etc. with those from the New Zealand tested laboratory samples are similar without providing clear evidence of significant sample disturbance during sampling, travelling, and handling.
Table 2: Summary of physical properties of tested soils
Figure 6: Grain size distribution of tested soils which lie within the suggested bound gradation curves for potentially liquefiable soils (after Tsuchida 1971)
All triaxial tests were performed using a closed-loop automatic cyclic triaxial apparatus (M.T.S. Systems Corporation).
Saturation of samples was achieved by applying a series of consecutive cell and back pressure increments, while maintaining an effective confining stress of 10kPa, until a Skempton B = Δu/Δσ value equal or greater than 95% was obtained. Following saturation, samples were isotropically consolidated under an effective isotropic stress, p’o, of 50kPa, 150kPa, and 300kPa. During consolidation, the volume change and the axial displacement deformation of samples were recorded to calculate the post-consolidation void ratio. A period of time equal to double the consolidation time (approximately 50 minutes) was allowed before cyclic loading. The change of void ratio of the tested samples with consolidation stress is shown in Figure 7.
Figure 7: Variation of void ratio, e, with mean effective stress, p’o, for the tested samples
Cyclic loading was performed by applying a sinusoidally varying deviatoric axial stress (σd) at a frequency of f = 0.1Hz, under undrained conditions. In this work, the occurrence of double amplitude axial strain, εDA = 5%, was used as a reference point for the onset of liquefaction of the tested samples. For this reason, a series of cyclic triaxial tests with different cyclic stress ratios, CSR = σd / 2p’o, were carried out in order to determine the number of loading cycles (N) required for the development εDA of 5%. It is noted that the development of Δu practically equal to p’o according to the relationship: Δu/p’o ≥ 95%, was also examined as a reference point for the onset of liquefaction and compared to the aforementioned criterion of εDA = 5%.
In view of the equivalent number of loading cycles of actual earthquakes (10 to 20 for an earthquake of Mw = 7.5), in this work the onset of liquefaction and thus the cyclic resistance ratio, CRR15, was considered as the CSR, required to produce εDA = 5% in 15 load cycles.
After cyclic loading, which resulted in liquefaction, samples were consolidated once more at the mean effective stress level prior to loading, in order to determine the volumetric deformation, εv, after liquefaction.
Following the second consolidation stage, samples were loaded monotonically under undrained conditions to determine their post-cyclic undrained shear strength (Sumax). Samples were subjected to undrained compression at a constant rate of axial displacement of 0.1mm/min. The undrained shear strength, Sumax, is equal to half the maximum deviatoric stress, qmax, during compression.
1 I-X-Y: I (distinctive number), X (A=short PVC tube / B=long steel tube), Y (mean effective stress, p’o)
2 Total number of loading cycles
3 Volumetric deformation during consolidation after liquefaction
4 Post-cyclic undrained monotonic strength during loading performed after liquefaction and consolidation
5 Contained plant roots
Table 3: Summary of cyclic triaxial test (CTX) conditions, cyclic loading and post-cyclic monotonic strength
3.3 CTX results and discussion
Table 3: presents the cyclic test conditions and results for tests on undisturbed samples.
Figure 8 presents the cyclic triaxial test of sample 2-A-50 (as per Table 3) showing:
– the variation of (a) single amplitude cyclic deviatoric axial stress,σd, (b) double amplitude axial deformation, εDA and (c) excess pore water pressure, Δu with versus time, t;
– the variation of (d) deviatoric stress, q, and (e) Δu versus εDA;
– (f) their stress paths, in terms of q and p’, and
– (g) the stress – strain behavior during cyclic loading.
The plots demonstrate a characteristic cyclic mobility bahavior with limited strain potential and the typical inverted s-shaped loops on the CSR vs shear strain plots. Due to change of soil stiffness the deviator stress is decreasing with cycles for the specific testing conditions of sample 2-A-50 and low confining pressure.
Figure 8: CTX test results on sample 2-A-50
Similarly, Figure 9 illustrates the cyclic test results for sample 4-A-150 demonstrating an ‘elastic’ response during cyclic loading.
Figure 11 presents the variation of CSR with the number of cycles, N, required to reach εDA = 1, 2.5 and 5%, as well as Δu/p’o ≥ 95% for tested samples at p’o = 50kPa, 150kPa and 300kPa. It is shown that regardless of the p’o level, at relatively low levels of CSR the required N to reach the aforementioned characteristic deformations practically coincides, whereas at higher CSR the N value at each εDA is distinctive from another.
Furthermore, it is shown that at the smallest p’o level the two criteria for the onset of liquefaction practically coincide. With increasing p’o, however, liquefaction is triggered first based on the εDA = 5% criterion and afterwards according to the Δu criterion.
Figure 9: CTX test results on sample 4-A-150Save
The results for each triaxial compression test performed after cyclic loading, are presented in Figure 10 for the above samples, showing:
– the variation of (a) deviatoric stress, q, and (b) excess pore water pressure, Δu, with axial deformation, εσ and
– (c) their stress paths, in terms of q and p’
Figure 10: Post-cyclic monotonic triaxial test results on samples 2-A-50 and 4-A-150
Figure 11: Variation of CSR with number of cycles, N, for tested samples at (a) p’o = 50kPa, (b) 150kPa and (c) 300kPa
Figure 12 presents the variation of CRR15 with void ratio, e for samples tested under p’o = 50kPa, 150kPa and 300kPa. It is shown that for the studied void ratio range, CRR15 is practically unaffected by void ratio (density). Furthermore, with increasing p’o, liquefaction resistance of the tested samples decreases, as shown also in Figure 14, since soil specimens becomes more contractive at higher confinement pressures.
Figure 12: Variation of CRR15 with void ratio, e, for tested samples at (a) p’o = 50kPa, (b) 150kPa and (c) 300kPa and mean effective stress, p’o
Figure 13 presents the effect of liquefaction on the undrained shear strength, Sumax of the tested samples. For the studied density range (void ratio), it is shown that the normalized post-cyclic undrained monotonic strength over p’o is not significantly influenced by density – void ratio. The variation of Sumax/p’o with p’o is also presented in Figure 13 and it can be seen that the normalized post-cyclic Sumax/p’o decreases with increasing p’o
Figure 13: Variation of the normalised post-cyclic Sumax/p’o with void ratio and mean effective stress p’o
Figure 14: Post-cyclic stress paths of the tested samples
Figure 14 presents the post-cyclic stress paths for the tested samples. The qpcy-PT and p’pcy-PT values corresponding to the Phase Transformation State (point of transition from contractive to dilative, minimum p’) were calculated according to the equation:
qpcy-PT = Mpcy-PT·p’pcy-PT (r2=0.998) 
According to the above equation, the post cyclic stress ratio at the Phase Transformation state, Mpcy-PT was derived equal to 1.209 and the friction angle at that state, φ’pcy-PT=30.2o.
In-situ and laboratory investigations were carried out to evaluate determine the liquefaction resistance of the micaceous clayey silt / sandy silt laminated lake sediments in Queenstown. This includes undertaking cyclic triaxial testing (CTX) on selected specimens of typical lake Wakatipu soils from the Frankton area near the existing Kawarau Falls Bridge.
The results from the conventional laboratory tests to determine liquefaction susceptibility were inconclusive and the CPT-based methods of assessing liquefaction triggering produced significantly varied results.
Based on the current assessment, we consider that these soils do not exhibit flow liquefaction but rather cyclic mobility with the following important characteristics:
- It was shown that regardless of the p’o level, at relatively low levels of CSR the required N to reach the liquefaction criteria practically coincides, whereas at higher CSR the N value at each εDA is distinctive from another;
- For the studied void ratio range (tested samples), CRR15 is practically unaffected by void ratio;
- With increasing the mean effective stress p’o, liquefaction resistance of the tested samples decreases;
- For the studied void ratio range, it is shown that the normalized post-cyclic undrained monotonic strength over p’o is not significantly influenced by void ratio;
- The normalized post-cyclic Sumax/p’o decreases with increasing p’o.
The published empirical triggering curves based on the routine penetration-based procedures used throughout New Zealand to assess liquefaction triggering may lead to overestimating the liquefaction risk for sites in the Queenstown area. However, further investigation, sampling and testing would be required to confirm these findings beyond the tested Frankton area and its applicability for the majority of the Wakatipu lake sediments.
The experimental work on the lake sediments performed by the Soil Mechanics, Foundation and Geotechnical Earthquake Engineering Laboratory of the Aristotle University of Thessaloniki (AUTh-SMFGEE) under the direction of Prof. Theodora Tika.
This research wouldn’t be possible without the support and funding of Miyamoto International NZ Ltd.
The writer is grateful for the many discussions, feedback and valuable comments provided by Ass. Prof. Katerina Ziotopoulou, Dr. Peter Robertson (Professor Emeritus and Technical Advisor to Gregg Drilling & Testing), Ioannis Antonopoulos (Principal Geotechnical Engineer at Coffey Services, NZ) and Nick Barounis (Senior Geotechnical Engineer at Cook Costello).
Awad, A., and Searle, J. (2016). Liquefaction potential of the micaceous lake sediments in the Queenstown area”, 11th Young Geotechnical Professional Conference, Queenstown, New Zealand.
Boulanger, R.W., and Idriss, I.M. (2006) Liquefaction susceptibility criteria for silts and clays. Journal of Geotechnical and Geoenvironmental Engineering, 132(11), 1413-1426pp.
Boulanger, R.W., Idriss, I.M. (2007). Evaluation of cyclic softening in silts and clays. Journal of Geotechnical and Geoenvironmental Engineering, 133(6), 641-462pp.
Boulanger, R.W., and Idriss, I.M. (2014) CPT and SPT liquefaction triggering procedures. Report No. UCD/CGM-14/01, April 2014, Center for Geotechnical Modeling, Department of Civil and Environmental Engineering at the University of California, Davis, California.
Bowen, H.J., Ashby, GL, and Stringer, M.E. (2015). Case study – Undisturbed sampling, cyclic testing and numerical modelling of a low plasticity silt. 6th International Conference on Earthquake Engineering, Christchurch, New Zealand.
New Zealand Geotechnical Society (NZGS) and Ministry of Business, Innovation and Employment (MBIE) (2016). Earthquake geotechnical engineering practice Module 1: Overview of the guidelines, March 2016.
New Zealand Geotechnical Society (NZGS) and Ministry of Business, Innovation and Employment (MBIE) (2016). Earthquake geotechnical engineering practice Module 3: Identification, assessment and mitigation of liquefaction hazards, May 2016.
Otago Regional Council (2015). Seismic Hazard in Queenstown Lakes District: Study Report, April.
Robertson, P.K. (1990). Soil classification using the cone penetration test. Canadian Geotechnical Journal, 27(1): 151-158
Robertson, P.K. (2009). CPT interpretation – a unified approach, Canadian Geotechnical Journal, 46: 1-19
Robertson, P.K., and Wride, C.E. (1998). Cyclic Liquefaction and its Evaluation based on the CPT Canadian Geotechnical Journal, 1998, Vol. 35, August.
Robertson, P.K, and Cabal, K.L. (2014) Guide to Cone Penetration Testing for Geotechnical Engineering (6th Ed), Gregg Drilling & Testing, Inc., California, USA
Tsuchida, H. (1971) Estimation of liquefaction potential of sandy soils. In Proceedings of the 3rd Joint Meeting, US-Japan Panel on Wind and Seismic Effects.
Youd, T.L., Idriss, I.M., Andrus, R.D., Arango, I., Castro, G., Christian, J.T., Dobry, R., Finn, W.D.L., Harder, L.F. Jr., Hynes, M. E., Ishihara, K., Koester, J., Liao, S.S.C., Marcuson, W.F. III,
Martin, G.R., Mitchell, J.K., Moriwaki, Y., Power, M. S., Robertson, P.K., Seed, R. B., and Stokoe, K. H. II (2001). Liquefaction resistance of soils: Summary report from the 1996 NCEER and 1998 NCEER/NSF workshops on evaluation of liquefaction resistance of soils, Journal of Geotechnical and Geoenvironmental Engineering, ASCE,127(4): 297-313