DISCUSSIONS 619
Observations on the undrained strength of a glacial till
M&INLAY, D. G., TOMLINSON, M. J. and ANDERSON, W. F. (1974). G&technique 24,
No. 4,503-516.
A. M&own and A. M. Radwan, University of Strathclyde
The Authors have referred to the fissure pattern studies in tills in the West Central Scotland
area. Fissuring has now been shown to be a characteristic of all drumlinized tills so far investi-
gated in this area from Ayr, in the south, to Stirling in the north (McGown et al., 1974; Radwan,
1974). The finding of these studies therefore place some reservations on the conclusion of the
Authors that conventional site investigation practices of sampling with UlOO open drive
samplers will provide consistent data for determining shear strength of tills, in order to assess
allowable bearing pressures for foundations of structures coming into the class C category.
The principal characteristics of the fissures in the tills are that they are disposed in definite
preferentially orientated sets, usually one set at a low dip angle and two or four sets near verti-
cal. The fissure spacing increases with depth below ground surface and the systematic nature
of the fissure disposition together with differences in the nature of fissure set surface
coatings, induces three-dimensional anisotropy of undrained shear strength (McGown et
al., 1974; McGown and Radwan, 1975). Thus it is important to define the nature of a repre-
sentative test specimen in more than just size. The spatial orientation of a test specimen must
be considered, as the directions of principal strains are critical to the influence of the fissures.
Whether there is planar or three-dimensional strain occurring in the test specimen and whether
this strain pattern is the same as in the prototype must also be considered. To obtain a repre-
sentative test specimen in these soils it is thus necessary to have a sufficiently large specimen,
whose absolute size increases with depth, and to impose stresses and strains during testing
in the pattern and the directions similar to those in the prototype.
In terms of the tests carried out by the Authors, the unconfined compression tests on the
tills would, due to complete stress relief and opening of the fissures, now appear to measure a
particular condition in the soil and not necessarily replicate the confined in situ strength.
Whether or not the 100 or 230 mm test specimens were representative in size would vary with
the fissure spacing and so with depth. Where these sized specimens did not contain a sta-
tistically valid population of fissures the measured strength would have varied somewhere
between intact and fissure strength and from tests in other tills (Radwan, 1974) the ratio of
intact to fissure strength is typically 6 : 1. Large variations in measured strengths of the non-
representative test specimens were thus to be expected and their differences were dependent
on the number and the relative orientation of any fissures that were present in the specimens.
By virtue of the different direction and pattern of strains imposed in the plate bearing tests and
triaxial tests it is considered that strictly they are not directly comparable either in absolute
values or in shear strength-depth relationships.
The fact that these tills have now been shown to be fissured should therefore be taken into
account in the future selection of the size and mode of testing. It is suggested that sampling
and testing requirements in fissured lodgement tills will in fact vary with depth into the soil
and with the direction and pattern of stresses and strains imposed on the soil by the structure
under design.
REFERENCES
McGown A., & Radwan, A. (1975). The presence and influence of fissures in the boulder clays of West
Scotland. Can. Geotech. Jnll3, 1.
Downloaded by [] on [12/09/16]. Copyright © ICE Publishing, all rights reserved.
620 DISCUSSIONS
McGown, A., Sali, A. & Radwan, A. (1974). Fissure patterns and slope failures in boulder clay at Hurl-
ford, Ayrshire. Q. Jnl Engng Geol. 7, l-26.
Radwan, A. (1974). The presence and influence of fissures in the boulder clays of West Central Scotland.
PhD thesis, Strathclyde University.
Authors’ reply to M&own and Radwan
McGown and Radwan have offered valuable comment on our Paper. This Paper was
submitted some two years ago and reports on undrained strengths of tills derived from tests on
vertical samples and plate loading tests. Concurrent studies at Strathclyde University of the
influences of fissuring on the engineering behaviour of tills received mention in the Paper al-
though it was then premature to draw conclusions. Some significant effects of fissures on the
measured behaviour of tills have since been reported and representative sampling is seen to
embrace in quite a complex way the size and the orientation of samples with the latter item
referred in turn to the orientation of the related engineering construction (McKinlay et al.,
1975). We acknowledge the interest and value of this work.
The increase of observed fissure spacing with depth into the tills suggests that 100 mm dia-
meter laboratory specimens from the deeper samples will show properties, e.g. undrained
strength, closer to the value for the intact material. The plate loading tests, encompassing
larger volumes of till, may reflect to a greater degree the influence of fissuring and thus at greater
depths yield the lower strengths observed (Fig. 6 of the Paper). There remains a need for
data from still larger sized plate tests to explore this point and perhaps resources will be made
available for the purpose.
The economics of sampling such stiff stony tills for the assessment of their bearing capacity
are likely to weigh in favour of the use of UlOO samplers, driven vertically preferably coupled
in pairs. The extraction of larger, possibly oriented, samples will be justified on occasions
when loadings are known with some precision and designs are within close limits of accept-
ability. Fissure orientation studies are important, particularly in checking the stability of side
slopes of cuttings and other excavations where the relative orientations of fissure planes to side
slope surfaces should be allowed for. In such excavations, stress relief in the remaining soil
may lead to a softening whose rate would depend in part on the occurrence and silt content, if
any, of fissures.
In the undrained strength test specimens referred to in the Paper, the influence of stress relief
should have been marked in the results of the eight week tests particularly from the samples
taken at the greater depths where stress relief would itself be greater. In the event the eight
week strengths were higher than the strengths measured on the day of sampling at shallower
depths only and at greater depths no difference in strength was detectable. A slight drying
was noted in the samples from shallower depths during the period of storage under conditions
of commercial practice.
Interpretation of the plate loading tests presumes careful field work and isotropic shear
strength behaviour in the tills. The tills are clearly anisotropic but in practical terms the plate
test so interpreted remains an economically acceptable measure of undrained shear strength if
only to yield approximate values in their proper relative magnitudes.
REFERENCE
McKinlay, D. G., McGown, A., Radwan, A. M. & Hossain, D., (1975). Representative sampling and testing
in fissured lodgement tills. Symp. Engng Behaviour of Glacial Materials, Univ. of Birmingham.
Downloaded by [] on [12/09/16]. Copyright © ICE Publishing, all rights reserved.
