See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/356290219

Effect of rotation rate on shear vane results in a silty tailings

Conference Paper · October 2016

CITATIONS READS
0 112

1 author:

David Reid
University of Western Australia
62 PUBLICATIONS   250 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Slope Stress Prediction Benchmarking exercise View project

All content following this page was uploaded by David Reid on 17 November 2021.

The user has requested enhancement of the downloaded file.

Effect of rotation rate on shear vane results in a silty tailings
D. Reid
Golder Associates Pty Ltd, Perth, Australia
University of Western Australia, Perth, Australia

ABSTRACT: Undrained strength is a key parameter in assessing the stability of a tailings facility. Howev-
er, the available methods to assess undrained strength directly from piezocone penetration test (CPTu) results
suffer inherent uncertainty. Therefore, when practical, shear vane testing is often used to enable site-specific
correlations between CPTu results and undrained strength to be developed. While the shear vane is recog-
nised to provide excellent estimates of undrained shear strength, it can be effected by drainage during shear
for intermediate soils, including many silt tailings. To assess the effects of drainage conditions on shear vane
results, a series of tests at different rotation rates were undertaken within a silty tailings. Corresponding dissi-
pation tests were undertaken to assess consolidation behaviour, along with CPTu, piston sample collection for
laboratory testing, and ball penetrometer tests. The results indicated that for the silty tailings investigated,
“standard” shear vane rotation rates overestimated undrained shear strength owing to drainage effects during
vane rotation. The rotation rates required to produce undrained shearing were consistent with some of the typ-
ically-applied methods for assessing the effect of drainage, when based on estimates of consolidation behav-
iour from dissipation tests.
1 INTRODUCTION
1.2 Purpose of this paper
1.1 General The purpose of this paper is to present results of
The importance of achieving undrained condi- in situ and laboratory testing of a moderate plasticity
tions for vane shear testing in silts and “intermedi- tailings in WA, including vane shear testing under-
ate” soils has been long understood (Blight 1968, taken at different rotation rates at the same test loca-
Chandler 1988, Castro and Troncoso 1989, Morris tions and depth. These vane shear results, refer-
and Williams 2000, Fell et al. 2014, ASTM 2015). enced to other measures of undrained shear strength
However, despite this recognition, the author is and horizontal coefficient of consolidation (ch) ena-
aware of frequent uses of the vane shear test in tail- ble an assessment of the performance of available
ings investigations in Western Australia (WA), and methods to assess drainage around the vane
elsewhere, where insufficient recognition of partial
drainage on inferred undrained strengths has been 2 IN SITU AND LABORATORY TESTING
applied. Such results have then led to difficulties in
2.1 General
developing site-specific correlations of undrained
shear strength in general, and specifically with re- The tailings facility investigated is upstream-
spect to Piezocone Penetration Test (CPTu) results. raised and hence the undrained strength of the tail-
As tailings often consist of a large proportion of silt- ings are of critical importance to stability of the pe-
sized particles and frequently exhibit significant lay- rimeter embankments. The tailings are of moderate
ering and hence hydraulic anisotropy, they are a ma- plasticity, with plasticity indices (PI) ranging from
terial that is particularly prone to the effects of 10 to 33% for recovered samples from the beach and
drainage during vane shear. As the peak undrained at depth. The material is predominately silt, ranging
strength is possibly the single-most important from sandy silts with up to 40% sand-sized particles
strength parameter for many tailings facilities, it is of (>75 µm), and finer zones with up to 30% clay-sized
importance that where vane shear testing is used to particles (<2 µm). Surficial and piston tube samples
supplement other assessment methods, the potential from depth indicated an average bulk density of
for overestimation of strengths through partial drain- 1.8 t/m3. The material is cyclically deposited across
age is accounted for and avoided where possible. the tailings cell, meaning that significant solar dry-
ing and densification of the material typically occurs
prior to placement of additional tailings. Additional
information regarding the mine site and location sessment of dissipation test results suggested that at
cannot be provided, in order to maintain anonymity 12º/min, partial drainage was likely. Therefore,
for the site. higher rates of both 90 and 240º/min were attempted.
The site investigation undertaken consisted of Comparison of these results is the primary purpose
CPTu probing with regular dissipation testing, with of this paper.
vane shear testing, ball penetrometer, and piston Electronic data collection was employed during
sampling at key locations. Three test locations are vane shear rotation, with the system as used at the
relevant to this study, as outlined in Table 1. It is time of the investigation only able to record data at
noted that ball penetrometer testing was only possi- 4 second intervals. Owing to this minimum save in-
ble for locations on the tailings beach, as this device terval, the results for tests with higher rotation rates
was unable to penetrate the stiff compacted crest of are somewhat sparse. As a result of the limitations
the embankment. of the system seen during the investigation outlined
here, the contractor subsequently had the data re-
Table 1. Test location summary
______________________________________________ cording system modified to enable smaller data re-
Test Information CPT1 CPT2 CPT3 cording time intervals in future investigations.
______________________________________________
Location Crest Crest Beach 2.4 Laboratory test methods
Dissipation Depths (m) 5 3.5, 7.5 4, 9
Piston Samples Y N Y
Samples extruded and trimmed from piston tubes
Vane Testing Rates (º/min) 12, 240 12, 90 240 were used to undertake direct simple shear (DSS)
Ball Testing N N Y
_____________________________________________
tests. The DSS system used was of the Swedish Ge-
otechnical Institute-type, where the sample was lat-
2.2 Methods – CPTu, ball, piston sampling erally restrained by a membrane and Teflon rings,
CPTu probing was undertaken with a 10 cm2 and sheared under constant volume conditions at 5%
cone, with pore pressure recorded at the u2 shoulder strain per hour. Samples were consolidated in in-
location. Cone resistance was corrected for the une- crements to their estimated in situ vertical effective
qual area effect. The cone had an equal end area stress, to assess sample disturbance prior to selection
friction sleeve. The ball penetrometer used was of a vertical effective stress for testing. Significant
80 mm in diameter, and did not include a piezome- vertical strain was observed in reconsolidated the
ter. Piston samples were taken using a “Gouda- specimens to their in situ vertical effective stress, in-
type” sampling system, in 63.5 mm internal diameter dicating sample disturbance. This is likely a result
tubes with an outside cutting edge angle of 10º and of the long distances required to transport the sam-
with no inside clearance. CPTu probes were con- ples from site to the testing laboratory in Perth (>400
ducted first at each location to provide stratigraphy km). Owing to the likely sample disturbance, some
information to assess suitability and target zones for of the samples were consolidated to vertical stresses
other tests. Other tests were undertaken within a higher than in situ values to attempt to provide more
3 m radius from the initial CPTu. relevant normally consolidated strength ratios.

2.3 Vane shear testing 3 ASSESSING VANE SHEAR DRAINAGE

The system used was a GeoMil FFL 100 vane CONDITIONS
shear, with a rectangular vane 100 mm in length and
50 mm in diameter. The vane shear rods were in- Literature review undertaken as part of this study
serted within larger-diameter casing, and a slip cou- suggests there currently exists two different ap-
pling was used to estimate, and correct for, rod fric- proaches to assessing whether a vane shear test has
tion on each test prior to commencing shear of the occurred under undrained conditions:
tailings. Shaft rotation commenced approximately  The method of Blight (1968), as updated by
60 seconds following insertion of the vane into a Chandler (1988). This method assumes the pri-
new depth of soil. This was the fastest practical time mary excess pore pressure generation relevant to
in which the torque application and measurement drainage during shear occurs during vane rotation.
system could be fitted to the rods and activated after Further, this method assumes that drainage path
advancing the vane to a new depth. The rods typi- length is equal to the diameter of the vane.
cally required rotation of between 10 and 20º before Therefore, time factor T can be directly related to
rotation of the vane and hence shearing of the soil degree of drainage U.
commenced. Tests were undertaken generally at 0.5  The method of Morris and Williams (2000). This
or 1.0 m depth intervals. method assumes that the primary excess pore
Tests were initially conducted at a rotation rate of pressure generation relevant to drainage during
12º/min. However, initial assessment of the strength shear occurs during vane insertion. Further, the
results obtained from these tests on-site indicated drainage path length relevant to application of
they were likely erroneously high. Partial drainage this method is not generally known a priori, and
was suspected as the cause. Initial screening as- hence directly relating T and U is more difficult.
 4.3 Undrained strength synthesis
The two methods outlined approach the vane In situ peak undrained strengths were assessed
shear assessment problem in considerably different through CPTu, ball penetrometer, vane shear, and
frameworks. Both have been used to interpret the comparison to DSS tests previously outlined. A
data presented in this paper, to assess their respec- comparison of results at CPTu 1 (crest location) is
tive performance. made in Figure 1. The compacted crest material in
this area was estimated to be between 3 – 4 m in
4 RESULTS AND INTERPRETATION depth from the surface. Peak undrained strengths are
4.1 Laboratory undrained strength inferred from the CPTu based on an Nkt of 10 and an
Nu of 5, based on peak undrained strength ratios ob-
The results of four DSS tests undertaken are pre- tained from the DSS across the sampled depth inter-
sented in Table 2. The results indicate peak un- val. The vane results are shown for rotation rates of
drained strengths ranging from 0.20 to 0.27, which 12 and 240 degrees per minute. The vane shear re-
are reasonable values for a normally consolidated, sults indicate that at 12º/min, the results appear to
moderate plasticity material loaded in the simple significantly overestimate undrained strengths com-
shear direction (for example, Ladd 1991). Rigidity pared to other techniques. When the rate is in-
index (Ir) calculated for each test are included, as creased, the resulting undrained strengths align more
these values were used to assist in inferring ch from
favourably to those from other tests.
dissipation tests. Ir was calculated based on stresses
and strains recorded at 50% of the failure shear
stress, i.e. G50. As discussed previously, the vertical
effective stresses selected for the tests are higher
than the estimated preconsolidation pressures that
are likely to have occurred owing to air drying on the
tailings beach.
Table 2. DSS test summary
______________________________________________
CPTu DSS Sampled su / σ’v Ir
No. Test No. Depth
(m) (-) (-)
______________________________________________
1 1 6.5-7.0 0.27 70
3 2 7.5-8.0 0.24 100
3 3 7.5-8.0 0.20 135
3_____________________________________________
4 7.5-8.0 0.22 98

4.2 Phreatic conditions and dissipation tests

Horizontal coefficient of consolidation ch was in-
ferred based on Teh and Houlsby’s (1991) solution,
assuming an average Ir as indicated from the DSS
tests undertaken. Dissipations tests undertaken at
locations relevant to this study are summarized in
Table 3. The values are seen to range from approx-
imately 100 to 500 m2/year. These values are in ex-
cess of that suggested to provide undrained shearing
at “typical” vane rotation rates (Chandler 1988).
Table 3. Dissipation test summary
_______________________________ Figure 1. CPTu 1 summary
CPTu Depth ch
No. (m) (m2/year)
_______________________________ The results obtained from CPTu 3, a beach loca-
1 5.0 132 tion, are presented in Figure 2. At this beach loca-
2 3.5 348 tion ball penetrometer testing was undertaken adja-
2 7.5 505 cent to the initial CPTu, to provide another
4 4.0 334
4_______________________________
9.0 311 assessment of peak undrained shear strengths. The
ball penetrometer was inferred based on the methods
of Randolph (2004) and DeJong et al. (2011), with
Dissipations tests indicated approximately hydro- an assumed Nball of 11. The results suggest that for
static pore pressure profiles with depth. Evidence of this test location, solar drying processes following
small amounts of base drainage were observed at test deposition result in overconsolidated tailings to a
locations other than those assessed in this study.
depth of approximately 5 m. This suggests a pre- number of tests with T<0.05 still indicate peak un-
consolidation pressure imparted by solar drying of drained strength ratios exceeding 0.3 this may be a
approximately 70 kPa, which seems reasonable in result of sandy seams of material in these areas. Im-
the context of the typically observed beach condition portantly, the linked tests for CPT 1 undertaken at
in fallow areas. rates giving time factors higher and lower than
Chandler’s theorized drained/undrained threshold
appear to support that proposed drainage criterion.

Figure 3. Vane summary, based on Chandler

(1988)

4.5 Vane shear interpretation (Morris and

Williams 2000)
Figure 2. CPTu 3 summary
The peak undrained strength ratios results are
4.4 Vane shear interpretation (Chandler 1988) plotted again on Figure 4, where T is now based on
The peak undrained strength ratio results of all of the methods suggested by Morris and Williams
the vane shear tests undertaken at CPTu 1, 2 and 3 (2000), where time to failure is taken from insertion
locations are presented in Figure 3, with reference to of the vane. The relevant T below which undrained
the inferred time factor T, based on the start of conditions would prevail cannot be definitely esti-
shearing. Time factor was calculated from the start mated for the data in this study based on the method
of rotation of the vane (not the rods, owing to slip of Morris and Williams, as their method requires ad-
coupling), assuming drainage path length is equiva- ditional information to directly link T to degree of
lent to the diameter of the vane (Chandler 1988), and drainage U, which varies for different materials.
applying a ch value from the nearest dissipation test. However, based on the data they present, the lowest
A time factor of 0.05 is also indicated on Figure 3, required T value relevant to their method to define
which represents the approximate threshold suggest- undrained conditions would be approximately 1.3.
ed by Chandler between drained and undrained This value is indicated on Figure 4. The test results
shearing. For results at CPT 1, where vane shear undertaken at CPT 1 at the same depths, but with
tests were undertaken at the same depths at 12 and different rotation rates, are again linked for clarity.
240 degree/min., the paired results for each depth are The summary of the results as shown in Figure 4
connected for clarity. does not, in general, support the interpretation meth-
The results indicate a weak trend between peak od of Williams and Morris. Where tests were under-
undrained strength ratio and time factor. However, taken at the same depth, at different rates, the result-
the results where partial drainage is likely are all in ing calculated change in T are such that either rate is
excess of the expected normally consolidation un- suggested to be essentially undrained. This appears
drained strength ratio range of 0.20 – 0.30 for this to be contradictory to the strength results, which in-
material. This suggests that the effects of partial dicate significantly lower strengths in many case at
drainage are providing overestimates of undrained the higher shear rate. The reasons for this lack of
strength when low rotation rates are used. While a agreement are unclear. However, it is noted that ap-
plication of the Morris and Williams method is type index (SBT) inferred from the CPTu did not
based, by current necessity, on the assumption that T vary significantly. The highest ch value, at CPT 2,
of 1.3 be taken as potential limit on undrained be- was undertaken in a location that may have included
havior, from a limited number of data points. Addi- sandier material, as there is a local drop in SBT
tional data to refine this criterion may be useful (as around this location.
was pointed out by the authors of this method in It is interesting to note the effect of rotation rate
their publication). on the time to commencement of shearing when us-
ing a slip coupling. For example, when using a rota-
tion rate of 12º/min, it can take up to two minutes of
shaft rotation prior to the commencement of vane ro-
tation. This is an important consideration if as-
sessing the potential for partially drained conditions
in the context of Morris and Williams’ (2000) meth-
od, where time to failure commences from vane in-
sertion.
It is noted that in parallel to the effect of rotation
rate on potential drainage around the vane, rotation
rate also effects shear rate. This can itself affect un-
drained shear strengths, with faster shear rates lead-
ing to higher measures of shear strength. Indeed,
much of the typically-applied vane correction factors
are suggested to be a result of such rate effects, and
their increase with increasing plasticity (for example,
Ladd and DeGroot 2003). Fortuitously, the index
properties of the material considered in this study
suggest that rheological rate-effects may be low,
when viewed in the context of Bjerrum’s (1972)
Figure 4. Vane summary, based on Morris and vane correction method. Alternatively, laboratory-
Williams (2000) based studies have suggested that variation in shear
rate is likely to have significant effect on resulting
5 DISCUSSION AND LIMITATIONS undrained shear strengths for a material with the in-
The results obtained in this study provide greater dex properties relevant in this study (for example,
support to the methods of Blight (1968) and Chan- Sheahan et al. 1996). This is an area of ongoing un-
dler (1988) compared to that of Morris and Williams certainty and research in geotechnical practice.
(2000). However, it must be acknowledged that the
site procedures, while undertaken in a manner that 6 CONCLUSIONS
controlled for the effects of vane rotation rate, were
not adequately planned to fully assess the effect of A series of shear vane tests were undertaken adja-
time from vane insertion. Therefore, the available cent to CPTu and ball penetrometer tests within a
data is insufficient to directly relate time factor to silty, moderate plasticity tailings. Dissipation tests
drainage coefficient as proposed by Morris and Wil- within the tailing suggested that vane shear testing
liams (2000). Further, a relatively small number of may not achieved undrained conditions without con-
tests are available at similar depths to enable direct sideration of the time to failure and/or rate of shear.
comparison. Vane shear tests were therefore conducted at differ-
It is noted that the results of the current study, in ent shear rates to assess their effect on the results ob-
the context of the above comparison, suffer a num- tained. In general, rapid shear rates resulted in clos-
ber of relevant limitations, including (i) reliance in er agreement with other available estimates of
some cases upon dissipation test data from different undrained shear strengths. This result enabled an as-
depths to that of the vane shear tests interpreted, and, sessment of two current methods for assessing the
(ii) the potential that “paired” vane tests at the same drainage behaviour of vane shear tests.
depth, yet 1 to 3 m apart laterally, may not test iden-
tical material owing to the inherent variability within REFERENCES
a typical tailings deposit. Despite these uncertain-
ties, it is observed that application of the vast ma- ASTM D2573 / D2573M-15. 2015. Standard Test Method for
jority of the range of ch values typical of the results Field Vane Shear Test in Saturated Fine-Grained Soils.
West Conshohocken , ASTM International.
obtained in the investigation would not materially Bjerrum, L. 1972. Embankments on soft ground: SOA Report.
affect the results of the comparisons made. Further, Proceedings Specialty Conference on Performance of
where ch values from adjacent depths were applied Earth and Earth-Supported Structures: 1 – 54. Purdue,
to the interpreted vane shear results, soil behaviour ASCE.
Blight, G.E. 1968. A note on field vane testing of silty soils.
Canadian Geotechnical Journal 5: 142-149.
Castro, G. and Troncoso, J.H. 1989. Effects of 1985 Chilean
earthquake on three tailings dams. Fifth Chilean Confer-
ence on Seismology and Earthquake Engineering. Santiago.
Chandler, R. 1988. The in situ measurement of the undrained
shear strength of clays using the field vane. In A. Richards
(ed) Vane shear testing in soils: field and laboratory stud-
ies, Tampa, Florida, January 1987 .Baltimore: ASTM.
DeJong, J.T., Yafrate, N..J, and DeGroot, D.J. 2011. Evalua-
tion of undrained shear strength using full-flow penetrome-
ters. Journal of Geotechnical and Geoenvironmental Engi-
neering 137(1): 14 – 26.
Fell, R., MacGregor, P., Stapledon, D., Bell, G. and Foster, M.
2014. Geotechnical Engineering of Dams. Boca Raton:
CRC Press.
Ladd, C.C. 1991. Stability evaluation during staged construc-
tion (22nd Terzaghi Lecture). Journal of Geotechnical En-
gineering 117(4): 540-615.
Ladd, C.C. and DeGroot, D.J. 2003. Recommended practice
for soft ground characterisation – The Arthur Casagrande
Lecture. Proceedings of the 12th Panamerican Conference
on Soil Mechanics and Geotechnical Engineering: 3-57.
Morris, P.H., and Williams, D.J. 2000. A revision of Blight’s
model of field vane testing. Canadian Geotechnical Jour-
nal 37(5): 1089-1098.
Randolph, M.F. 2004. Characterisation of soft sediments for
offshore applications. In V. da Foneseca and P. Mayne
(eds.), Proceedings ICS-2, 2nd International conference on
site characterization: 209-232. Rotterdam: Millpress.
Sheahan, T., Ladd, C., Germaine, J. 1996. Rate-dependent un-
drained shear behaviour of saturated clay. Journal of Ge-
otechnical Engineering 122(2): 99-108.
Teh, C.I. and Houlsby, G.T. 1991. An analytical study of the
cone penetration test in clay. Géotechnique 41(1): 17-34.

View publication stats