210 SOIL MECHANICS IN FOUNDA TION ENGINEERING SITE INVESTIGATIONS 211

Table 6.9. Abbreviated notation for description of soils of soil on a given site. This information is essential in the design of structural
(to be read in conjunction with Tables 3.8 and 3.9) foundations as well as in the consideration of stability of individual structures.
Soil types and corresponding symbols The problem of determination of soil properties for the purpose of design is
Class of soil
Type of soil Syrnbol Type of soil . Syrnbol Type of soil Symbol
a very complex one because even the most homogeneous strata in the geological
very cohesive sandy clay , S/C e silty clay M/C
sense, as well as from the point of view of petrological and granulometric
clay
cohesive sand-clay S-C sand-clay -sílt S�C-M silt-clay M-C composition, exhibit considerable variations in their physical (water content and
sandy clayey S/C/M clayey silt density) as well as mechanical properties ( compressibility and strength). This is
-�
·� medium clayey sand C/S C/M
cohesive , sílt dueto a number of natural factors.
�(.)
slightly cohesive slightly (C)/S ¡ sandy silt S/M silt M (1) Non-hornogeneity of the parent rock from which the soil was derived.
clayey sand
(2) Variation in the degree of weathering of the mineral constituents of the
clay with silt M//C ! silt with C//M
laminae clay laminae soil.
(3) Variation in the conditions of sedimentation.
cohesionless sand s silty sand M/S
(4) Variation in the conditions of consolidation (at the middle of a stratum
coarse sand Se medium sand Sm fine sand Sf
gravel G sandy gravel S/G sanci-gravei S-G and at its boundaries).
organic organic o organic mud O/C/M slightly (O)/S (5) The effects of deep frost penetration during glaciation periods (formation
organic sand of irregular ice lenses).
peat Pt ( 6) The effects of tectonic movements or folding under the weight of advanc-
ing and retreating glaciers which lead to the formation of structural discontinuities
Several different strata of boulder clays can be encountered at any location;
along the slip planes (slickensides).
varved clays or silts and sands or even peat can be found sandwiched between them.
Soil properties can also be considerably altered by the following.
In certain regions in which the underlying soils were considerably deformed
and folded by the weight of the advancing and retreating glaciers, parent soils (a) Extraction of samples from the bottom of a borehole under conditions of
can often be found in the form of large lenses ( erra tics) within deep strata of seepage.
moraine boulder clays. In places where boulder clays were eroded during subse- (b) Disturbance of soil structure during sampling.
quent formation of river valleys recent alluvial sands can be found directly over (c) Changes in water content during storage of samples: migration of water
the parent rock or soils. from less to more cohesive layers of laminated soils or drying out.
The sequence of strata plotted on longitudinal and transverse sections of any Obviously one must not allow changes to take place in the natural properties
site should correlate with their geological age (Figure 6.18); this should be of soils dueto bad sampling or storage; any affected samples should be completely
carefully checked on ali the sections. Piezometric levels of ground water in elirninated from the considerations leading to the determination of the character-
continuous strata of sand should be fairly uniform. istic properties of soils.
Geotechnical sections should be supplemented with the values of the Particular attention should be given to the sampling of soils containing slip
generalized characteristic properties of soils in the individual strata. planes (slickensides), and finely laminated soils. The samples should be extracted
The accuracy of the interpolation of ground conditions in preparation of in such a manner that the planes of structural discontinuity should be inclined
geotechnical sections depends to a large extent on the spacing between adjacent at the same angle to the planes of shear failure in the laboratory apparatus as
boreholes; the larger the spacing the greater is the likelihood of marked differ- they would be in a given layer in natural conditions beneath a foundation orina
ences between the interpolated soil conditions and 'those revealed by subsequent slope.
excavation. In general one should attempt to determine the mechanical properties of soils
in conditions which as closely as possible resemble the natural conditions in the
6.4.3. DETERMINATION OF GENERALIZED CHARACTERISTIC PROPERTIES OF
SOILS
ground beneath a foundation or in the slope of a cutting.
Taking the above into consideration one finds that even after the elimination
6.4. 3.1. Introduction of the effects of sample disturbance there still is a considerable scatter of the
The fundamental issue in geotechnical engineering is the determination of the results which is partly due to the non-homogeneity of the soil in a given layer
generalized characteristic properties of soils for each individual layer ( or stratum) and partly to the inaccuracies of laboratory testing. Therefore in order to
 212 SOIL MECHANICS IN FOUNDATION ENGINEERING SITE INVESTIGATIONS 213
determine significant characteristic properties of the soil in a given layer it is
necessary to use an appropriate method of analysis of the results.
Table 6.1 O. Student's t values
6.4.3.2. Statistical Methods
Confidence level a
There are a number of statistical methods which can be used in the analysis of Number N
a set of test results with the object of determination of the significant character- 90% 95% 99% 99-9%
istic properties of the whole mass of the soil in a given layer. 1 6-31 12-71 63-66 636-62
A relatively simple method, which is sufficiently accurate for most practica! 2 2-92 4-30 6-92 31-60
purposes is based on the determination of the best estima tes of the arithmetic 3 2-35 3-18 5-84 12-92
mean x and standard deviation aest for the population from a given set of test 4 2-13 2-78 4-60 8-61
results and on the evaluation of confidence limits Xa within which one is o:% 5 2-02 2-57 4-03 6-87
certain that the population mean (i.e. the characteristic property) of the required 6 1-94 2-45 3-70 5-96
soil property is situated. The equations used in this analysis are 7 1-90 2·37 3-50 5-40
N 8 1-88 2-30 3·36 5-04
¿ X¡ 9 1 ·83 2-26 3-25 4·78
- i=l (6.10)
x=-- 10 1-81 2·23 3-17 4.59
N 1-80 2-20 3-11 4.44
11
12 1-78 2-18 3·06 4·32
(6.11) 13 1-77 2-16 3·01 4·22
14 1-76 2·14 2-98 4-14
X =x+- ta Oest 15 1-75 2· 13 2·95 4-07
a -:;Jif (6.12)
16 1-75 2·12 2·92 4·02
where x = arithmetic mean of the given set of test results or the best estima te 17 1-74 2· 11 2-90 3.97
of the arithmetic mean for the population 18 1-73 2-10 2-88 3-92
X¡ = single test result ( e.g. water content or liquid limit, etc.) 19 1-73 2-09 2·86 3-88
N = number of test results in the set 20 1·72 2-09 2-85 3-85
aest = the best estimate of the standard deviation for the population 21 1·72 2·08 2-83 3-82
ta = student's t value obtained from Table 6.1 O for the required level 22 1·72 2-07 2·82 3.79
of confidence o: 23 1-71 2-07 2-81 3.77
o: = level of confidence, i.e. the required degree of certainty that the 24 1-71 2-06 2·80 3.75
population mean lies within the stated confidence lirnits 25 1-71 2-06 2-79 3.72
26 1-71 2-06 2-78 3-71
The above equations refer to a population (in this case the whole mass of soil 27 1-70 2-05 2-77 3-69
within an isolated layer) of normal (Gaussian) distribution. It can be assumed 28 1-70 2-05 2-76 3-67
on the basis of up-to-date practica! experience that geotechnical properties of 29 1-70 2-05 2-76 3-66
soils follow the normal distribution. This can, in each case, be verified by plotting 30 1-70 2-04 2-75 3-65
the results in the form shown in Figure 6.19; values of test results are plotted on 40 1-68 2-02 2-70 3.55
the horizontal axis (linear scale) and the corresponding number of test results 50 1-67 2-00 2-66 3-46
expressed in percentages are plotted cumulatively on the vertical axis (Gaussian 120 1-66 1-98 2-62 3.37
scale ). If the resulting plot approaches a straight line, then the distribution of the 00
1 ·65 1-96 2-58 3-29
results approaches normal distribution; a sudden change in the slope of the line
indicates that samples from two different formations have been included whereas
a general scatter of the points indica tes inadequate size of the set.
 ¡
SITE INVESTIGATIONS 215
214 SOIL MECHANICS IN FOUNDATION ENGINEERING

The evaluation of x and Oest on the basis of Equations (6.10) and (6.11) is Example. The test results are grouped in 1-0% water content intervals and are
tedious ( unless computer facilities are available) and therefore the following tabulated as shown in Table 6.11 ( columns 1, 2, and 3), a convenient arithmetic
approximate, but still sufficiently accurate, simpler and quicker method can be mean Xa is assumed on the basis of Figure 6.19 and calculations in columns 4, 5,
used. and 6 are completed.
The method is based on the technique of grouping of the results in equally The arithmetic mean of the water contents can be determined according to
spaced intervals and on the use of the following equations: Equation (6.13):
_ Lxn x _Lxn - 1468-5 - 34-86 ""'34.9
X=-¡;¡- (6.13) N 45
or
or according to Equation ( 6.14)
x=xa +.!!.n
N a (6.14)
and x = 34.5 +.!..:2. x 16 = 34.5 + 0-36 ""'34.9
45
( 6.15)
The standard deviation is evaluated from Equation ( 6.15):
where x = central value for a group (1·0 )2)
J{
a est =+- l-02
-x218- -x16 =± 14-72=±2-17
n = number of test results in a given group 45 45 V
Xa = convenient assumed arithmetic mean of the complete set
a = group interval When the standard deviation is expressed as a percentage of the mean, one
D0 = difference of sums S 1 and S2 (see example in Table 6.11) obtains the coefficient of variation:
S = Sa + 2S� (see example in Table 6.11)
V= ª.:5t x 100 (6.16)
The results of water content tests given in Figure 6.19 will now be analysed X
to illustrate the above method.
For the considered example V= 6-2%. In general, values of Vbelow 10
99· 9 99-9 indicate very good uniformity within a given set whereas values between 10-13,
.,
99· 5 V
99·5 13-16, 16-20, and above 20 indicate, respectively, good, average, low, and
99 99
l.,, 98
./
98 very poor uniformity.
gl r 97
96
95
The percentage error in evaluation of the population mean for the various
95 11,,<
degrees of confidence can be taken as equal to
::,
� 90 90
�,!9
0 80 80
� 70 ., � 70
t
p = a Oest 100 (6.17)
§
e:
60
50
11'
1
1
60
50
x x.JN x
�
0
40
30
., -
• .I' 1
1
40
30 For the considered example
Á
o
v;
1
20 20
o. 1 2-01 X 2-17
Px = 34_9 45 100= 1-9%
"'>
15
10
/ "''
1 1
::1gl
1
10
5
(1) for a= 95% V X

:5
E
::,
5

,V t, 1
1 3-50 X 2-17
u 1 ::i (2) for a = 99-9% Px = 34-9y45 100 = 3-3%
1
0·5
� 1
�
1
0·5
X
i1
O· 1
T1 1 �
46
i.e. there is a 5% (1 in 20) chance of the error being 1-9% and only 0-1 %
=34-9
29 30 31 32 33 1 34 36J 37 38 39 40 41 42 43 44 45
33-4 -
Water content
w
36-4
(%)
(1 in 1000) chance of there being a 3-3% error.
1- 1213141619181514131 i 1-1-1-1-1- 1-1 This indica tes considerable accuracy in determination of the population
Number of results in o group
Figure 6.19. Distribution of results of water content tests plotted on Gaussian scale.
arithmetic means of the considered property (water content).
216 SOIL MECHANICS IN FOUNDATION ENGINEERING SITE INVESTIGATIONS 217
Table 6.11. Determination of mean water content by tabular method homogeneous layers, one has to obtain a large number of test results in arder
(example) to determine the generalized characteristic properties with a reasonable degree
Central Assumed mean x0 = 34·5 of accuracy. At the same time, however, it is possible to use simpler methods
Group limits Numbe
value Product of analysis in determination of these properties.
(Figure ma Sum of numbers in groups x.n
6.19)
for a
groupn A method based on determination of the guaranteed average minimum and
groupx First Second maximum values according to Equations ( 6.18) and ( 6.19) can be used for
1 2 3 4 5 6 approximate estimation of the confidence limits. After 10% of the lowest or

�k-
­ - e­­+­ highest results in a given set have been rejected

�
-k-­2} 2}
30·0­31·0 30·5 2 ­ ­ 61·0
31·0­32­0 31·5 3 ­
��
: -➔- ­­5
S2
- l---� ­
. 1-➔? - 1 s; 94.5
Xavmin -
_X+ Xmm
2
(6.18)
32·0­33·0 32·5 4 ­ :: -➔- ,...9 -: �➔- ­16 130·0
33·0­34·0 33.5 6 ­ : ---➔- ­15 ­ 201·0 _X+Xmax (6.19)
Xavmax - 2
34·0­35·0 34.5 9 Assumed mean 310·5
35.5 In the considered example one obtains, from Figure 6.19, xmin = 31­9 and

­21) s, - : ,­2­
35·0­36·0 8 ... : ­ 284·0

�----
r­,.�
H­­
Xmax = 37.7, and therefore the average minimum and maximum values

·. ­ ­26}
�­
36·0­37·0 36·5 5 =::: ,__➔.:::: :=­ l3 -= : 182·5
37·0­38·0 37.5 4 .. � �r:;::,,,. =­13 S' 150·0
-
8 ➔:::
,_� 1 34.9 + 31­9
38·0­39·0 38·5 3 ... : ,_ ➔::: =-- ·4 ::::.... 5
: 115·5 Xavmin
2
33­4%
­­
,­..�­
39­0­40·0 39.5 1 �­ 1-➔-:::::::,, 1 ➔-- -- 1 39.5
34.9 + 37.7 36­3%
N = 45 S1=21+13+8+4+1=47 s; = 26 + 13 + 5 + 1 = 45 b.468·5 Xavmax= 2
S2=2+5+9+15=31 S�=2+7+16=25 There is also a possibility of considering the lower and upper quartiles as
Sa = 4 7 + 31 = 7 8 s� =45+25=70 approximate limits of the confidence interval; with sufficient accuracy for
D0=47 ­ 31 = 16 D� = 45 ­ 25 = 20
practica! purposes these values can be determined from Figure 6.19, by rejecting
S =78+2x70=218
25% of the test results from each end of the plot; the two ends of the shortened
cumulative plot correspond to the quartiles. The lower quartile in Figure 6.19
Note. Figures in column 4 are obtained by successive summation of figures in column 3,
downwards and upwards to the assumed mean; similarly figures in column 5 are obtained is 33­4% and the upper 36­4%; the upper significant characteristic value of the
by successive summation of figures in column 4. water content would then be taken as Wrep = 36­4%.
As can be seen from the considered example the values of water content
The confidence limits Xa are evaluated from Equation ( 6.12). For a given
corresponding to the confidence limits obtained by different methods fall
number N of test results in a set, ta values corresponding to the required level
within a fairly narrow band, the intermediate values being those obtained by the
of confidence (a= 90, 95, 99, and 99­9%) are taken from Table 6.10 and on
method of quartiles. Considering the simplicity of application of this method and
substitution in the equation the upper and lower confidence limits Xa are
of its clear physical interpretation it is recommended as the most effective method
obtained (Table 6.12).
for general use.
Table 6.12. Lower and upper confidence limits Xa In the case of investigations of special importance and when V is greater than
Values of Xa 16 the authors recornmend the use of the 'confidence limits' method; otherwise
a the simple 'quartiles' method is considered to be sufficiently accurate (Table 6.12).
Lower Upper
The above­described methods of evaluation of the confidence limits for a
90 1­68 0­5 34.4 35.4
95 2­02 0­75 34·2 35­6 given population can be applied in determination of any geotechnical property,
99 2·69 0­9 34­0 35­8 provided a sufficiently large number of results is available from tests carried out
99.9 3.53 ¡.¡ 33­8 36­0 on samples which are representa tive of the soil in a given layer.
This condition can be satisfied in the case of water content tests which are
Taking into consideration the low level of accuracy associated with geo­ relatively simple and are usually carried out in sufficiently large numbers. In
technical testing and the considerable natural variation of soils, even in apparently the case of the more time­ and work­consuming investigations (e.g. determination
 SOIL MECHANICS IN FOUNDATION ENGINEERING SITE INVESTIGATIONS 219
218
of cornpressibility or shear strength) execution of a sufficiently large number of 11
.? .,, NE � 1, Generolized chorocleristic
tests is very difficult and an analysis of results of a small number of tests may -o � 3
7C :z �. i--
1
""t�....,_,1 11 �"1
JJI
1
1
properlies
���� 6i: " 2
be subject to considerable accidental errors because the samples may not be �:§º ,E ¡':": scot-ttt-ttH-1+-iJIUUlit;_jj_ Emax =66 MN/m
� •· ..., 1 ,..... E�;n = 58 MN/m2
sufficiently representative of the soil in a given layer. o, Ñ
401 H -tt-tttittH1ttt-ttft-itfu•tttt11t-tltt-H-tl++-.+--jf--=�--=-:....:..:.::...::...::�-
111 1� 1 1
E 40 L +-- 1
:z = 36 MN/m2
.!:: "' Emox
In connection with the above the authors recommend the use of the -o ;-
,A 1 1

8 �� 35 � 1
E' 1
if"� l;;:;- 25
::�=>�30-
graphical correlation method. ::,_ 1 E�in = 32 MN/m2
� 201H++-+-!!1+-++-+-WI-IH!lfHHH!!IIHI-Hll-l-tl++l++-H-+t+-l---...j:....jf-------_J
a1 10 - +- 1 ll,
1

�=
·¡¡¡z 60 11
111r-,.,..._,_.,,1•111r:·
- 1
6.4.3.3. Graphical Correlation Method u <;, = 47 kN/m2
50 1 � ¡..:'

=--
1
The method is based on a graphical summary of results of tests carried out to u,.J' 40 1, 1 1 '"t--,-
��K+t-t-111-Htll-HH!-+Hl!-HHltl:+HllHHH+i!+-rt++I+-++¡..+-+.-+------
11 �
�
determine various soil properties, with reference to a 'leading' property 0 .1.1 ;
O> 1 1 1 :
��� 16 - 1 ¡,, 1 1 � u = 13 º
(Figure 6.20); it is suggested that for cohesive soils the water content is taken .[-g� 14 11
as the leading property, and for cohesionless soils the density index. When all � 12 1
IOt-++t--Htt-Ht-l-Hlil'+Hff+-ll �
the results are plotted correlation curves cari be drawn. Frequently the plotted � � 60 w. Wt = 58%
-%l � 5ok..L.il-.UU.H+-I-Httlfflltt1i
11 111
1

�o. 11
results are not evenly distributed along the reference line (horizontal axis); it is ·- E 40 1 1
3=i���=t1fft::tm:!lfflttflffi!ljlj
wl �1 tt:jfjttit:tlt:t±:J..._�__
1 P2_6_
� __ º _j
therefore suggested that the reference line is divided into equal length sections
(groups) so that within each group at least two results are present; weighted
1
1 I' 1
i
i f'mox = 1·91 g/ml

l
k; l'il� 1
average values are then evaluated for each group and the new points are used
for plotting correlation curves ( or straight lines). �-- 185,1-++!-+-!!1+4+-I-+ le, on
1: ]: -
,.
f min = 1·88 g/ml

The ordinate corresponding to the point of intersection of the correlation

l� 00 -1-lllLlW+li��'
I 111�1111!11
roclion
:Tiiñí+tt:trrN
1 � 1 P S. D.:
1 1rn
curve of the considered property with a vertical line drawn through the appro-
. s:i] ;e
60
�� � 40
&i:i 2c
si 1, tro�MI I�
111111111
klv,ltracllon
? 1 ¡
'
I
l Sand
Sil!
Clay
18 %
62%
20%
priate confidence limit of the leading property is taken as the significant Water 030 ) ....31-..u.u
L.J. 3 ��34;._,�¡¡¡¡¡,;,�36�;w�·7.u..w...
I 23 �.¡.3... 3 8W...3-l-
3 �9 40¡____:w{'l1/c:: .: .!)._
o _..=::�___J
characteristic value of the property.
Toe recommended method has the following basic attributes. :==:rT=
a=
e:/%;
i oN �=:
t I=
ip=
1
. 51�=::6=
1 1
1 �.......,...;;19....,
1-(�I 1
�1 lf � 1 ! 1
.,J.5 I1 i: IA
(1) It forms a synthesis of the complete range of investigations in a given Borehole �- 12 i 3"- 374
�
� i:; 13 1 ,,1,. 4

layer. 1-0 1·0

2-0 l+ll+-+t+H+lf--Ht--+-l2·0
(2) It enables one to reject accidental errors or mistakes. 3-0 ! 3-0
(3) It gives a good insight into relationships between individual properties ] 40
5-0
I-Htt-tt-++++lf-+++--+-l4 · 0
5·0
and enables one to carry out a comprehensive analysis of the considered geo- "'
.e
6·0 I-Htt-+t++++-lf-+++--+--l6-0

technical problem.
The graphical correlation method is considered to be sufficiently accurate
for most practical purposes. If in certain cases the use of numerical correlation
methods is considered necessary then, again, it is suggested that as the first step
in the analysis the results are divided into groups and then the correlation is
carried out using the weighted group averages. This is justified in the physical
sense, because it enables one to eliminate the effects of accidental accumulation
of results in several groups which otherwise might have influenced the results of
the correlation analysis by giving incorrect coefficients of regression; the group-
ing also simplifies and reduces numerical work.
In cases when tests of different accuracy are used in determination of a
particular soil property ( e.g. determination of compressibility of soil from
results of oedometer and plate-loading tests) the correlation curves should be
drawn separately for each method of testing (Figure 6.20); the lower correlation
�
7-0 7-0
o 8-0 1-Htt-+t++++-lf-H+--+--10-0
90 ULL u..J.U...u..J.J..J.J...L-J...1..J..._.J........J 9·0
□ Representative value of given property
• Samples token from tnal pits for plate-lood1ng tesis
• Samples token from boreholes
s Stiffness moduli from plate-looding tests
Figure 6.20. Graphical summary of geotechnical properties of soil.
curve for the less accurate oedometer method is obtained on the basis of a large
number of results and it can be used for compara tive reference purposes in
plotting the higher correlation curve for the smaller number of the more
accurate results of the plate-loading tests. The latter correlation curve should
be used in determination of the significant characteristic properties.
Toe correlation curve relating stiffness moduli to water content is used for
determination of the upper and lower values of the characteristic stiffness
220 SOIL MECHANICS IN FOUNDATION ENGINEERING
SITE INVESTIGATIONS 221
moduli �x and «:
for a given layer; these are obtained by plotting vertical
As can be seen, the strength parameters obtained from Figure 6.20 are
lines through the lower and upper water content confidence limits (see
<l>umin =13° and Cumin = 47 kN/m2 and, from Figure 6.21, <l>umin = 13° and
Figure 6.20).
Cu min = 50 kN/m2•
The obtained characteristic stiffness moduli E:nax and � can be used in
evaluation of the expected total and differential settlements of foundations of a 6.4. 3.4. Generalized Characteristic Soil Properties and Factors of Safety
given building. In connection with this, one should verify whether the most The values of factors of safety at present used in geotechnical designs have been
compressible zones of soil (having E'< E:nin) are not, for example, situated established on an empirical basis. For any specific problem wide ranges of values
under the central part of the building while the less compressible (having are frequently quoted, as, for example, in the case of determination of
E'> E:Uax) under the end walls, or vice versa. allowable bearing stresses where factors of safety between 2 and 3 are used,
Shear strength correlation curves (<Pu, w) and (cu, w)curves, are usually used when the ultimate bearing capacity is considered, and wide ranges of allowable
only for the determination of the lower significant characteristic values <Pu min settlements or angular distortions ( e.g. 1 : 200-1 : 400) are used, when settle-
L\11 = 118-73 = 45 ments are taken into consideration; no reference is made with regards to the
:. Ion �u= � =0·225
z accuracy of determination of the geotechnical properties of the soil.
and �u= 13'1 Undoubtedly the large values of the factors of safety used possess a
cu= 73-100 tan 13° sufficiently large margin of safety to cover for the inaccuracies associated with
=73-23=50 kN/m2 the methods of design and soil parameters.
There is therefore a possibility of linking the values of factors of safety with
the accuracy in determination of soil properties.
"'E A general principle can be adopted that, in cases where more accurate
......
z methods of testing and determination of generalized characteristic soil proper-
""" 1201-----¡------t---+--+----+-- ties are used, lower factors of safety are acceptable.
.e In such cases a comprehensive analysis of the ground conditions and test
O'
� I00t--"-...::t-----,f-----+---+----+----.+---+--+----+---+---'--+-� results is essential; a graphical summary of all tests results of the type shown in
vi
o Figure 6.20, at the bottom of which borehole and trial pit reference numbers
� 80t---+----+-----. and depths of samples are plotted, can be very useful.
This enables one to obtain a clear three-dimensional picture of the ground
conditions which will indicate, for example, whether the weaker soils in a given
31 32 33 34 35 36 37 39
38 40
layer are associated with a particular area on the site or whether the lower
Water conlent w (%) values are part of the natural scatter of results in a given set.
Figure 6.21. Graphical determination of significant characteristic values of Cu and <f>u In the case of design of important structures the graphical summary of test
using shear strength rt- results should forman integral part of the site investigation report.
and cu min, í.e. the values corresponding to the upper water content confidence 6.4.4. DESCRIPTION OF SITE, COMMENTS, AND RECOMMENDA TIONS
limit. The description of the site, comments, and recommendatíons relevant to the
If the scatter of <Pu and Cu values is greater than that in Figure 6.20, then a
proposed works should cover the following.
more accurate method can be used which is based on the consideration of the
(a) Topography of the site.
shear strength Tf =Cut an tan <Pu- By assurning two values of an (one, for
(b) Geological structure, i.e. indication of the origin of the soils and their
example, 100 kN/m2 smaller and the second 100 kN/m2 greater than the
geological characteristics.
expected value of a0) the corresponding values of Tf are calculated and are
(c) Ground water conditions with indication of possible fluctuations of the
plotted as shown in Figure 6.21. Correlation curves are drawn and the upper
ground water table and present and probable differences in water levels in
and lower significant values 7f max and Tf min, corresponding to the lower and
adjoining surface reservoirs.
upper water content confidence limits, are used in evaluation of the character-
( d) Description of soil strata: the type and consistency of soils, their
istic values of <Pu min and Cu min·
physical and mechanical properties and indication of the range of probable

l
222 SOIL MECHANICS IN FOUNDATION ENGINEERING

bearing stresses, their frost susceptibility (mainly for road construction), their
7
suitability for construction of embankments and indication of the local sources
of road materials.
(e) A discussion of the possibility of landslides in natural slopes, new cuttings,
and excavations.
(f) Toe type of construction and state of existing buildings and roads.
(g) A discussion of any special problems associated with the type of the pro-
posed works. Nomograms for Determination
of Physical Properties of Soils

7.1. Introduction
'
The knowledge of a wide range of physical properties of soils is necessary in
the investigation of their shear strength and compressibility characteristics in
the laboratories as well as in the actual calculations leading to the determination
of safe bearing stresses, thrust on retaining walls, or the stability of slopes. The
following properties belong to this range.
(1) Pd, dry density of soils.
(2) Psat, density of fully saturated soil.
(3) n, porosity of soil.
(4) e, voids ratio.
(5) Wsat, saturation water content.
(6) Sr, degree of saturation.
All the above properties can be computed from the knowledge of the
following basic properties.
(a) Gs, specific gravity of solid particles.
(b) p, natural density of soil.
(c) w, natural water content.
The computations are, however, tedious and time consuming and the use of
appropriate nomograms can be of considerable assistance.

7 .2. Description of Nomograms

Five nomograms have been constructed for the following individual values of
the specific gravity of the solid particles: Gs = 2-65, 2-67, 2-69, 2·72, and 2-75.
For soils having particles of specific gravity between these figures the nomogram
with G¿ closest to the actual value is used. Water contents w in percentages are
plotted on the horizontal axes of the nomograms while the natural densities p
are plotted on the left-hand vertical axis .
•