Se.

:emen: insane -me:~ocso'acu

a:inc anc 'ac:ors a "ec:inc

by ERIC E. JORDEN," MSc, PEng

Introduction should be studied for a better under- Schmertmann (1970) calculate settle-
THIS PAPER tabulates methods of cal- stand'ing of the use of the methods. This ment for incremental layers. It seems ex-
culating settlement in sand and reviews is particularly the case for the methods pedient, however, when a relatively in-
the factors affecting the magnitude of this by Alpan (1964) and Parry (1971). compressible stratum is near the surface,
settlement. The point is made that be- Otherwise, the methods can be used to take a proportion of the settlement
cause the methods give different answers without recourse to additional reference calculated by each of the other methods
it is necessary to always calculate material. depending on the thickness of the sand
settlement using several methods. The The first is the original Terzaghi and layer. The proportion might be the
range of values obtained should be re- Peck (1948) method and the next five depth to the incompressible stratum
ported as well as the average value. represent modifications of the original. The divided by 1.5 B. T'e settlement so ob-
Some of these methods have been pre- methods vary in their treatment of the tained cou'Id be compared to th'at obtained
sented recently and others have been in measured N value, the overburden pres- by the more appropriate methods.
use for many years. All 'have been sure, the depth to ground water and the A conservative estimate of maximum
developed by conscientious workers striv- depth of embedment. The correction for settlement may also be made from a
ing to provide as exacting a solution as overburden is the most significant but more recent expression by Meyerhof
possible. These workers have been aware even this correction varies from one (1974) as follows:
of the great reliance placed on penetra- method to another, although Meyerhof pB
tion testing in granular material on the (1976) has suggested on the basis of s
one hand and our imperfect understand- recent studies that the overburden cor- 2q,
ing of the relationships between penetra- rection by Peck (Chart 8) may be close where p is net foundation pressure,
tion resistance and engineering proper- to the correct solution. There appears to qr is the average static cone resistance
ties on the other. be a less conservative approach to ground within a depth approximately equal to the
The Standard Penetration Test and the water as the various methods evolved. footing width B, and all symbols are in
static Dutch cone test on which these The original method by Terzaghi and consistent units.
methods are based have been standardised, Peck is very conservative. It has been A similar expression has been derived
but because of our crude understanding suggested by Sutherland (1974), on the for standard penetration test results as
of the relationship between penetration basis of work by Bazaraa (1967), that follows:
resistance and compressibility (the test the factor 3 in the original formula can s = pVB/2N where B is in inches, p is in
methods are not crude as has been sug- be replaced by 2 and a more realistic tsf. and N is the average within a depth
gested) the various methods give different prediction of the maximum probable B. In silty sand twice the right-hand side
answers for the magnitude of settlement. settlement obtained. It seems possible to is used.
Penetration testing is the most practical go further, however, and replace the 3 by Simons (1975) highlighted the variation
means of assessing settlement in sand one to obtain an average value. The in predicted settlement in a deposit of
and comes closer to a statistical approach range of "average" settlements obtained sand. In this case the water table was
to the problem than more refined and using different methods would -give a near the ground surface. The sand was
expensive methods. A lot of sand and good appreciation of the minimum and generally medium dense for the upper
gravel covers the earth's surface. A proper maximum probable settlement. 11m and dense below this depth.
approach to the utilisation of the various The remaining five methods reflect a It was shown for a 10m square base em-
methods is needed for calculating settle- different approach to the problem of bedded one metre that predicted settle-
ment in these deposits. calculating settlement, in that the com- ment using 10 different methods varied
pressibility or equivalent elastic modu- from 5 to 36mm. The nature of the actual
Methods of calculating settlement lus is calculated for the sand. Correlations values predicted by each method, i.e.,
The methods are tabulated in Table I. are established between penetration test whether high or low, is not significant
Several authors have reviewed and dis- results and the sand modulus, E, or the because these would change as soil con-
cussed these methods (Simons 1975, compression index, C. The methods by ditions and footing geometry changes. It
Sutherland 1974, Schmertmann 1970, D'Appolonia et al (1970) and Parry (1971) is only important to realise the con-
Webb 1959). Most methods tend to give are obvious modifications of the classic siderable range in predicted settlement
a maximum value of settlement, some are elastic equation. These two methods are for a single structure.
more conservative and one or two give based on direct correlations with Standard A revealing comparison was made by
the maximum probable settlement more Penetration Test results. The remaining Simons et al (1974). Eight methods were
often than others. Three methods permit three methods, by De Beer and Martins used to predict the settlements of 6
a calculation of settlement for incremen- (1957), Meyerhof (1965) and Schmert- structures for which settlements have
tal layers. All the methods have been mann (1970), obtain E or C from static been observed. The results are sum-
derived for either the SPT or static Dutch cone test results. marised in Table II.
Dutch cone but all may be used inter- Correlations between E and N (Webb, The authors observed that the methods
changeably with the help of the correla- 1959) might also be utilised in applying by Alpan and Schmertmann gave the best
tions in Chart 18 tempered with en- the method by Schmertmann and possi- agreement with observed settlement on
gineering judgement. bly also in applying the method by De average but that the ranges of calculated
However, it should be noted in using Beer and Martins (1957) and D'Appolonia to observed settlements were very wide.
Chart 18 that Meyerhof (1976) has ob- et al (1970). This should be done with More interesting, perhaps, is the similar-
served no significant change in the ratio caution, however. It must be observed that ity of the nature of the predictions by
qr/N as particle size changes and con- calculated values of E using the correla- the three methods, Terzaghi and Peck.
siders that a ratio of 3 to 5 is appropriate tions by D'Appolonia and by Webb do modified by Meyerhof (1965), Peck and
for both sand gravel sizes. not agree well particularly for low values Bazaraa (1969) and Parry (1971). For
The methods have been tabulated so of N. There appears to be considerable these three methods the averages fall
as to permit a calculation of settlement interest in the approach by Schmertmann. within the narrow range of 0.63 to 0.72
using either the formulae or the accom- The method should be applied where pos- and the range for each method falls
panying charts. The original sources sible and the results carefully analysed within the range of 0.1 to 1.4. The upper
and reported. and lower limits to the ranges exhibit a
aMaritime Testing Ltd., Halifax, Nova Scotia, Only the methods
Canada. Formerly with W. S. Atkins rk Part- by De Beer and very narrow range within themselves
ners, Epsom, UK Martins (1957), Meyerhof (1965) and going from 0.1 to 0.3 and 1.1 to 1.4.
30 Ground Engineering
 CHART 1 CHART 2
40 Ibf /in
80
V
C
0

¹50 60
/
C 20 ibflin
0
0
40 10 lbf /in
0
C
Cl N =30 a
C
i0 6000 I-
20
N=20 0
o
Cl
N=20 C
e 4000 0
0 Ill Q
0 N=10 20 40 60 80 100
N=10
o 2000 Relative density 'lo
N=5 N=5

0
0 5 10 15 20
IGIBBS l HOLTZ 1957)
Breadth of footing 8 (ft)
S P T- NUMBERS OF BLOWS- N
Terza ghi and Peck (1948) Ill
20 30 40 50 60
—— -- Peck Hanson and Thornburn(1974)
a
LU
p
p 10 70 dlC

IL'
TERZ AGHI d, PECK
Vl I
1Q
LU
IL:
a.
z 20
IU
O
cc
C HART 30
dl

LU

O 40
LU

I- 50
V
LU
1000 CHART 4
LU (ALPAN 1964)

140
O
2000
0'1
z 20
Cl
I

VI
la
III
4l
I
z
o. vi 100
Ik
0
I-
e 3000 V BQ
I

I
e I-
0 60

> 4000 V
0 40
X
LU

20
a)
5000
1 2 3 4 0
0 10 20 30 40 50
Correction factor - Corrected N value
SPT NUM BER OF BLOWS-
Measure d N value N

CHART 5
(TOMLINSON 1969) (PECK ET AL 1974)
January, 1977 31
 TABLE 1. METHODS OF CALCULATING SETTLEMENT
TERZAGHI 8( PECK GIBBS 8E HOLTZ ALPAN MODIFICATION MEYERHOF MODIFICATION PECK & BAZARAA PECK MODIFICATION
(1948) MODIFICATION (1957) (1964) (1965) MODIFICATION (1969) (1974)
D
Formula See charts See charts
s =C„, C» s =KC»
m
SPT (as measured) for sand (as measured) for sand (as measured) for sand (as measured) for sand (as measured) for sand (as measured) for sand
= 15+0.5 (N —
<o N N N N N N
N,, 15) for very N,, = 15+0.5 (N' 15) for very at foundation level.
fine or silty sand below fine or silty sand below N,. =15+0.5.(N' 15) for very
the water table. the water table. N's N fine or silty sand below
corrected for overburden. the water table. N's N
corrected for overburden.
Overburden No correction to N See charts See charts No correction to N
N,. = — p'(
4N

1+2p'500
for Cx —0.77 log»>-
pi
1bf/ft'-'N
or use Chart (8) parti-
cularly for p'
0.25 tsf.
N„. =
3.25+ 0.5p'500
1bf/ft'-'0
for p'

Water C„, =1.0 for D„)~

—2.0 foi'„=
28 C„. (as for previous method) Increase ws by a factor No correction p't depth 0.58 assuming D„.
C,T 0 2.0 for small ratios of D/B. no ground water level C„, = 0.5+0.5
Increase by 1.5 for D/8~1.0. K= D+8
p't depth 0.58 for existing
ground water level

Embedment C„=1.0 for D/8 = 0 C„(as for previous method) No correction C„(as for T & P method) See Charts (9, 10 or 11)
C„=0.75 for D/8 = 1 C„=1.0—0.4

Procedure 1. Use Chart (1) to obtain 1. Use Chart (2) or (3) as 1. Use Chart (4) to cor- 1. Use Chart (1) to obtain 1. Use N,, in Chart (1) or 1. Use smallest average of
q, corresponding to s = 1" shown to obtain N, to use rect
for overburden.
N qa. formula.
in the averages of N values
for largest footing on in formula or in Chart 2. Use Chart (5) to obtain 2. Increase q, by 50% to 2. Use Chart (1) to obtain in each borehole corrected
loosest part of deposit (1) Ny and N, obtain value for s = 1". q,. by C„.. (N corrected to
with no water present. 2. Procedure as for previous 3. Set q, = footing load 3. Apply correction Ca. 3. Increase q„by 50% to N,,).
2. Apply corrections to q,. method. /8'-'=1/F(0.48 y N y + 4. A'Iternatively, use formula. obtain value for s = 1". 2. Use Charts (9, 10 & 11)
3. Alternatively, use formula. y DN,) and solve 4. Divide q, from Step 3 with N, to obtain q,
4. Settlement s increases in for 8 (feet). s (inches), 8 (feet), by Cd and K. corresponding to s = 1".
direct proportion to in- 4. Calculate q, (tsf). q„(tsf) . 5. Alternatively, use 3. Correction C„applied to
crease in q, above value 5. Use Chart (6) to obtain formula. q, (corresponding to
for s = 1" w„and to confirm that s = 1" in dry soil)
q, falls within linear range. obtained if ground water
s (inches), 6. Use Chart (7) to obtain within 8 of base.
8 (feet), q, (tsf). ~s/ w„ for p=unity.
7. Calculate settlement
~s for actual q,
Comments Assumed that buildings can Modification reflects that Original T Ik P correlation Assumes that presence of Original T & P correlation Factor of safety of 2 on q,
sustain a maximum settle- penetration resistance is a between N and relative den- ground water is reflected betwen N and relative from first part of curves.
ment of s = 1". No water function of effective stress sity is used. Method in blow count. Generally density corresponds to the Charts apply for D(B and
correction for o < D„< 28. as well as relative density. acceptable for lower values as original T & P method. 10 psi curve rather than un'iform sand. No account
No account taken of geo- Results in much improved of q,. Original paper provides the 40 psi curve (Assumed). taken of over-consolidation.
logical history of deposit settlement prediction assum- a correction for rectangular Assumption was arbitrary. Values of q„obtained are
(previous loading and over- ing no change made in fac- footings. No apparent fie'Id checks. about 50% higher than
consolidation). Use factor tor 3. F above is assumed factor original T & P method.
2 in place of 3 in formula of safety.
to obtain more correct maxi-
mum settlement. Use 1 in place
of 3 to obtain average value.
 DE BEER & MARTINS (1957)
Ik DE BEER (1965) MEYERHOF (1965) D'APPOLONIA ET AL (1970) SCHMERTMANN (1970) PARRY (1971)
Formula
s =
023
C
/ogrp
p'+ ~p
pl
QH
As previous method
=~o
qB
M
Ui s =C,C,~p — AH s = —
qB

M
C,C„.C,
where p' effective over-
burden pressure
in each layer g„
whee'e M =— E

1-v'
where C,= 1—0.5
pa

Ap
where M
~p = increase in stress = 0.25
1-v'-'
t
at each layer
C, =1—0.2!og = 0.25
0.1
P„=overburden pressure at
foundation level.
gp =net bearing pressure.
= 196 + 7.9 N (tsf) E =
2q,, (q,, as measured) E = 50N (kg/cm')
= 1.5 — q,
= 1.9 — or
Compressibility, q„.
Sand Modulus C or C Normally Consoli- For normally consolidated
pl pl datd Sand csturine sediments (Webb 1959). Take N = average value at
E = 1.5 q,. (kg/cm') E = 1.9 q,. (kg/cm') E = 416 + 10.9N (tsf) Submerged SP and SW sands: depth z B.
Take q,, as measured Take q,, as measured Over consolidated E' 5 (N+15) (tsf)
in kg/cm'-'~ tsf) in kg/cm-' (Preloaded) Sand =2.5 (q,.+30) (tsf)
Submerged SP clayey sands:
.Take N as measured. E'=:3.3 (N+5) (tsf)
=1.67 (q,+15) (tsf)
Average E' 2 (p,, + 25) (tsf).
Procedure As previous method. 1. Take average N within 1. Divide soil into layers. Obtain correction C„ from Chart 16.
1. Divide soil into layers the depth VBL. (48) Obtain correction C, from Chart 17.
(6») 2. Obtain influence 2. Obtain E for each layer. Correction C„ for water:
2. For each layer take the factors U„and U, 3. Obtain I, from Chart 15 1. Surface footings, back filled
minimum, the maximum and from Charts 12 Ik 13. for each layer. excavations; reduce N in
the average q,, values and 3. Obtain modulus M from 4. Calculate C, and C, direct proportion to reduction
calculate the corresponding Chart 14 or calculate C„ is a creep correction in p'ue to water level.
C values. using expression above for any time, t, in years 2. Permanent excavations below
3. Calculate the average for E. after construction. water table with a drawdown
settlement and the range 4. Calculate settlement s 5. Calculate settlement of the water level.
of settlement foi each for assumed value of for each layer then sum. D„,
layer and then sum. bearing pressure q. C„, =1+
1,, (dimensionless) D+$ B
s, B (feet) E, E', q,, (tsf) for (0<D„,<D)
q, E, M (tsf) 3. Permanent excavation but water
level below base of excavation,
ie. D„,)D
O„.(2B+D —D )
C„.=1+
2B(D+ 0.75B)
where O<(D„,—
D) <2B
Comments 1. The method is based on the 1. Arbitrary recommendation 1. The influence factors 1. Case studies suggest predicted 1. Method based on limited plate
semi-empirical Terzaghi- that C be increased as U„and U, reflect foot- settlement is about 50% lower bearing tests.
Buisman formula as shown. shown. ing dimensions and relative to the de Beer method 2. Expression for E gives much
This enables the settlement 2. For 17 case studies the embedment as well as layer 2. Laboratory work on sands higher values than normal.
of incremental layers to be increase resulted in a thickness. The factors should (SW) by Schultze Ik 3. Method takes account of stress
calculated thus taking into ratio of predicted to be used with caution. Menzebach (1961) gave changes after the site investi-
account the heterogeneity of observed s ranging from 2. Ground water level ignored E = 71+4.9N (kg/cmz) gation.
the layers. Assesses maximum 0.8 to 2.0 with an on basis that it is reflected = 78+5.4N (tsf) 4. Method applicable to feasibility
differential settlement. average of 1.3. in the measured N. and Schultze Ik Moussa studies. Increases s by 50% for
2. Boussinesq pressure distribution. 3. Method does not differentiate (1961) found that adding desiqn.
3. Carry out at least 3 static between heavily and lightly 4 to 9% silt to clean sands 5. C~, C„, and C, are dimensionless.
cone tests. preloaded soils (preload on (SW) reduced E by a factor D = depth of footing embed-
4. Ratio of predicted to original site was ~ 2.5. Adding 4 to 9% ment.
observed s was 1.9 in approximately 5 tsf). clayey silt reduced E by a D„. = depth of water table.
original case studies. factor ~ 7.5.
TABLE II. CALCULATED SETTLEMENTS BY EIGHT METHODS FOR SIX grains due to interlocking and less riding
STRUCTURES WHERE SETTLEMENTS HAVE BEEN OBSERVED up would both be present. The exhaustive
studies of Schultze and Moussa would
S care. S care. suggest that the former predominates and
S ohs. S obs. that increased angularity results in a net
Method average range increased compressibility. Roundness as
suggested by Jakobson's work may be a
de Beer and Martens 3.22 1.0-4.8 more dominant effect in finer material.
The interlocking phenomenon, however,
Schmertmann 1.48 0.24.0 appears to be relied on near the minimum
Terzaghi and Peck 1.89 0.~.2 void ratio to increase the resistance to
Terzaghi and Peck compression in the vibratory treatment
modified by Meyerhof (1965) 0.70 0.2-1.1 of sands (Greenwood and Tait 1970,
Terzaghi and Peck Schultze and Moussa 1961).
modified by Tomlinson 0.31 0.1-0.6 Mineralogy
Peck and Bazaraa (1969) 0.63 0.3-1.4 Some elastic deformation would occur
Alpan 0.95 0.1-2.4 at inter-particle contact points which
would also contribute
Parry (1971) 0.72 0.1-1.3 bility of a sand deposit tounder
the compressi-
load. Grain
(atter Simona et al 1974) crushing would be a further component
of compressibility. At low confining stress
levels angular, rough grains result in higher
It is also interesting that if the results reflecting different grain size distributions
from the original compressibility due to crushing at the
Terzagh'i and Peck with moisture content varying from dry
method are divided by 3 (effectively re-
contact points. This effect may be less,
to saturated. Correcting N for the effective
placing the factor 3 in the formula by one
however, the finer the material. The
overburden pressure results in a value mineralogy of the sand grains would in-
as suggested above) the average and of N more nearly representative of the fluence compressibility, quartz grains being
range for this method would fall within insitu relative density, the engineering more resistant to crushing and elastic
the narrow limits exhibited by the three property whose change has such a
methods above. deformation than the weaker felspar
markeeI effect on settlement.
There is a significant measure of con- grains, for example. The higher the
sistency in these results although more stresses at the contact points the more
Grain shape crushing would occur.
comparisons must be made before their Void ratio and relative density change The separate and varying effects of
reliability can be accepted. Sutherland as a result of (a) particle rearrangement, roundness, angularity and grain crushing
(1974), for example, is of the opinion that (b) elastic deformation and (c) grain would seem to exhibit a combined effect
the modification by Meyerhof (1965) crushing. of generally increased compressibility
gives a reasonably good estimate of the With respect to particle rearrangement with increased angularity.
maximum probable settlement rather than the degree of roundness influences the
the something less than maximum sug- tendency towards rearrangement and the Grain size distribution
gested by the above comparison. closer packing together of the particles. The greater the number of different
However, it may tentatively be con- Particles ride up on one another during sizes of particles
cluded that if the average of the results represented (the
rearrangement so that the more rounded broader the grain size distribution), the
of these four methods of calculating set- the grains are, the more likely this riding greater is the likelihood that voids formed
tlement is increased by one half a up will occur. With increased angularity
reasonable prediction of settlement would by larger particles are infilled with smaller
a greater amount of particle interlocking particles. The result is a decreased void
result. occurs inhibiting particle rearrangement ratio (increased relative density) and thus
and thus decreasing compressibility. decreased compressibility.
Factors affecting settlement (This effect is modified by grain crush-
A proper evaluation of settlement in ing, however — see below under Miner-
The predominant size of particle also
seems to have an effect on compressi-
sand can only come about through an alogy). bility. Limited data (Bazaraa, 1967) sug-
understanding of the factors affecting If the angle of internal friction of sand gests that submerged silty sands are
settlement or compressibility. T'his under- reflects in part the roundness or angularity 50'/, less compressible than a corres-
standing is the essence of engineering of the particles, which is well established, ponding clean sand at the same relative
judgement. It will enable the engineer, for then work by Jakobson (1957) illustrates density. This is inconclusive, however.
example, to arbitrarily increase predicted the effect of roundness on compressibility. The increased resistance to penetration
settlement, which is obviously too low, Two sands studied by Jakobson were in silty sands, suggesting a decreased
when he is aware of the well developed identical to the naked eye exhibiting the compressibility when in fact it may be
angularity of the sand particles, or to same grain size distribution curves and the same, may be due to negative pore
decrease settlement when he suspects the degree of roundness. Sand A exhibited pressures associated with dilatancy.
sand deposit has been preloaded. considerably greater compressibility than A larger percentage of gravel size
Compressibility, or the tendency of a Sand B and also had a lower angle of particles may also suggest the same situ-
sand to settle or compress under load, is internal friction in triaxial tests. The ation but for a different reason. The ap-
influenced by (i) relative density, (ii) difference was attributed to the fact that parent increased resistance to penetration
grain shape, (iii) mineralogy, (iv) grain the grains in Sand A were more polished in this case, however, would result from
size distribution, (v) overburden pressure, than Sand B as observed by stereo micro- large pieces of gravel giving a misleading
(vi) water, (vii) precompression or insitu scope. A polished surface accompanies indication of penetration resistance.
stress state, and (viii) cohesive admix- a high degree of roundness.
tures. Of all the factors relative density The work by Schultze and Moussa Overburden pressure
and overburden pressure have the most (1961) would seem to suggest that in- An increased overburden pressure has
marked influence. In fact, the mechanism creased compressibility would be asso- the effect of increasing the confining
of compressibility is fundamentally one ciated with increased angularity, the pressure on an element of sand and reduc-
of changing relative density. opposite of that which has been suggested ing the lateral strain. The effect is
above. Increased angularity would perhaps similar to an increased triaxial cell pres-
Relative density result in a more card-house, open struc- sure. The greater the restriction on lateral
As the initial relative density of a sand ture to the fabric of the sand deposit strain the more dominant is vertical strain
deposit decreases the degree of compres- which would be reflected in a greater (axial compression) relative to volumetric
sion or settlement of the sand increases difference between the maximum and compression. Axial compression is less
under a given load. This has been mirrimum void ratio. This opportunity for than volumetric compression (Skempton
observed by many investigators. Schultze a greater change in void ratio would result and Bishop, 1954) so that at increased
and Moussa (1961) for example reported in an increased compressibility. overburden pressure compressibility is
that as the initial void ratio increased The two effects of (1) a more open decreased.
the strain (settlement) experienced at a structure wth angular grains and thus in- This effect of overburden pressure in
given pressure also increased. This was creased compressibility and (2) a decreas- decreasing compressibility occurs even
the case for a variety of sands tested, ed tendency towards the rearrangement of though relative density may decrease. The
34 Ground Engineering
 CHART 6 CHART 7
10 10
J'
9 5
in ft/t 8
08— 4 — — ——
8

cr
or 3 —— .,c,
-'4'
~N> I
r

I
0
I
O
e
,r' 1 2 4 5 6 7 8 9
0 Bin ft
(TERZAGHI & PECK 1948)

0 I

0 10 20 30 40 50
SPT- NUMBER OF BLOWS N

(ALPAN 1964)

Np =1
e Correction factor
I CN=
4 4 f.
N field
III 040608 10 12 141.6 1B 2.0
II!
e 0
I
0. 0 5 (c) D I 8= 0.25
C (a) Df /8=1 (b) Df I 8=0. 5
e 50
4350 43 50 N=
I
Jl 5 I
I cr Ng 40 ( NI 40
0
0
I/
III

III
2.5 /
III
C
0
e 3
4
/', N,'o 4) 30

V
C
0 3.0 / I
Ne 20 NI 20
Q

" /.'l~~
III
3.S N- 15
0

V
0
,,,, II I
a
0
1
//i Yr
Y '~S
/lEM~-:
445
NI 10
4= 5

LU
,,I 0
0 1 2 3 4 0 1

Width of footing;
2 3 4
8, ft
0 1 2 3 4 5 6

CHART 8 CHART 9 CHART fo CHART 11

(PECK ET AL 1974)

1.0 ~~
0.9
a O.a
0
I- 07 IA
ci L'Ei.---I 1 ~2 5'K)
~
X 50 N
<<Chil&
0.6 ~is.
CHART 12 0.5 ITS%~~~ ~ ~ ~ ~ l

4.0 Ol 0.2 0.5 1 2 5 10 20 50 100

I I

D/B RATIO
L=LEN GTH CHART 13 ( D'APPOLONIA ET AL 1970)
P
&II/ ~/ Lca~ --<.~" 50 qi'.0

II
D

)8 .~" 20
~ia
~

- ~~~
0 CHART 14
/// Wy//WyPSV//I.~y//
0 2.0 1600
LI S=U~ U,I PB04 dl preloaded'r
E 12 00 ~compacted
2 (II ~ IInIe I

!
0 1 800 w 0 I

'' rr
.!! nor m ally'
gelr ~ loaded sand
~
~

~r 4oo
fI
o.o 0
0.1 0.2 0.5 1 2 5 10 20 50 1CO
0 0
HI B RATIO 0 20 40 60 80
( D'APPOLONIA ET AL 1970) AVE RAGE MEASURED SPT
January, 1977 35
investigations of Gibbs and Holtz (1957) The effect of precompression or pre- test it would seem reasonable to expect
show an increased resistance to penetra- loading of a sand deposit is to decrease a correlation between the two tests.
tion with increasing overburden pressure the compressibility. The preloading may Many workers have suggested corre-
even though relative density is decreas- result from previously imposed founda- lations between N and q,. Billam (1974)
ing. Similar studies by Schultze and Men- tion loads, construction loads, water, ice has provided the compilation shown in
zebach (1961) and Schultze and Melzer or eroded overburden. Higher penetration Chart 18 on which the correlation sug-
(1965) confirmed these results. resistance resul'ts after removal of the gested by Thorburn (1970) has been
The relationship between compressi- preload because of "locked-in" residual superimposed. The average curvilinear
bility and overburden pressure p can be stresses. The locked-in stresses reduce correlation shown by the author may be
seen in the following expression for the lateral strain in a similar way to an in- more appropriate than the suggested
modulus of sand, creased overburden pressure. The resi- linear relationship. The linear correlation
dual stress affecting penetration remains is conservative when going from N to
E=up"'here

because resistance appears to be affected qr and liberal for the opposite case.
u is a coefficient. more by the mean stress than by the Rodin (1974) has reported that most
Work by Schultze and Moussa (1961) effective overburden preload (de Mello, workers agree that particle size distri-
and Chaplin (1961) suggests w 0.5 = 1971). The preloading has the effect of bution is an important factor. In particular,
for a well graded sand with a moisture interlocking and prestressing the sand care should be taken to distinguish be-
content from zero to 6%. The value of particles with prestressing probably more
w varies from 0.3 to 0.8 for the soils
tween differences in particle size distri-
dominant. This action is more pronounced bution and difference in particle size.
studied. in statically preloaded deposits than in The suggested relationship shown in the
The value of u has been defined by dynamically preloaded deposits, particular- correlation is not understood partly be-
Schultze and Melzer (1965) as follows:
u = 246.2 log N — 263.4p + 375.6 ~ 57.6
ly at low relative
and Moussa,
densities (Schultze
1961). Dynamic loading
cause the effects of dynamic loading on
soil are not well understood.
OI'
For this
= 301.1 log qr —382.3p + 60.3 ~ 50.3
disturbs the sand particles which then
fall under the influence of gravity to form
reason the correlation should be used
with caution, particularly as the average
The coefficient can be seen to be a a more compact interlocked arrangement
function of overburden
size of particle increases.
pressure p and but one which may be characterised by a
penetration resistance N or q,. Both p card-house structure with minimal pre-
and q, are in kg/cmz while u is dimen- Conclusion
tressing. The net effect (reduced com- The magnitude of settlement in sand
sionless. pressibility) of interlocking and prestress- as predicted by the available methods
ing is, therefore, less than in static varies considerably. The individual meth-
Water preloading.
There is no apparent change in the ods are not consistent in being either
It would seem possible to assess the above or below average for varying soil
compressibility of a sand deposit upon effect of precompression on the basis
saturation but at moisture contents less conditions and foundation geometry. This
of the resistance to penetration. De Mello variability is due to the many factors
than full saturation the compressibility is (1971) has reported that the the pene- affecting settlement and our lack of
decreased (Schultze and Moussa, 1961). tration resistance depth profile may
The capillary forces acting in partially understanding of their influence and the
appear to bulge some distance below general relationship between penetration
saturated sands are responsible for the de- the surface when precompression has
crease. Their effect is to retard the re- resistance and compressibility.
occurred. It is interesting that Schmert- Settlement should always be calculated
arrangement of grains. An accurate as- mann (1970) has observed that strain
sessment of the influence is only pos- by several methods and a range of values
is a maximum some distance below the reported as well as an average value. To
sible in fine sand (Schultze and Menze- level at which load is applied. D'Appolo-
bach, 1961) where relatively undisturbed facilitate this approach the various
nia (1970) has suggested a correlation methods have been tabulated and the
samples can be obtained for correlation. between an equivalent elastic modulus
The resistance to penetration calculation procedure outlined for each.
N was and N for preloaded dune sand. The cor-
observed in the same studies to decrease Engineering judgement should be used
relation is based on an investigation of in deriving the final settlement for each
by an average of 15% relative to values over 300 footings on a construction site.
above the ground water level. Schultze method, keeping in mind the factors
The correlation remains to be tested more affecting settlement and the published
and Melzer (1961), also observed a notice-
thoroughly on other sites, however. E comparisons between predicted and
able change in penetration resistance can at best be only appropriate to the observed settlement for various methods.
(N and q,). unloaded soil so that the value applying Recent work suggests that a reasonable
These investigations would seem to sug- to a foundation loading condition is prediction of settlement would be obtained
gest that no consideration need be given under-estimated (Greenwood 1974).
to the presence of ground water when by averaging the results from the methods
assessing compressibility by Schmertmann, Meyerhof, Peck and
and only a
Cohesive admixtures
nominal decrease 'in bearing capacity need
Bazaraa, Alpan and Parry where this
The compressibility of clean sands is possible.
Vol.
be made. The factor of safety against
reduced by the addition of fine grained
bearing capacity failure is not normally
materials, Th compressibility is reduced References
critical for footing widths greater than
1.3m. The spread in the change of the by approximately 2 —,'imes with the ad- 1. Alpen, I., (1964): "Estimating settlements of
dition of 4% to 9% of silt. Similar foundations on sands". Civil Eng'g. and
values of penetration resistance was not Public Works Rev., Vol. 59, p. 1415.
quantities of clayey silt reduced the com- 2. Bazaraa, A.R.S.S., (1967): "Use of the
reported, however, and it is significant Standard Penetration Test for estimating
that effective shear strength is decreased pressibility of the order of 7— ', times settlements of shallow foundations on sand".
(Schultze and Moussa, 1961, as reported Ph. D. Thesis, Univ, of illinois, Urbane. (Not
by approximately one half with submer- seen),
gence. The discrepancy arises in that it by Greenwood and Tait 1970). The fine 3. de Beer, E., (1965): "Bearing capacity and
grain materials fill the void space and settlement of shallow foundations on sand".
is not well understood to what extent Proc. Sym. Bearing Capacity and Settlement
effectively prevent the sand particles
the resistance to penetration is an index of Foundations, Duke University, 15-33.
of the physical properties of a soil deposit. moving closer together. 4. Billam, J., (1974):Personal communication.
There would be a limit to this effect, 5. Chaplin, T. K., (1961): "Compressibility of
This uncertainty is reflected in the way sands and settlements of model footings".
however. With greater quantities of clay 5th ICSMFE, Paris, Vol. II, p. 33.
in which each method of predicting set- 6, D'Appolonia, D. J., et al., (1968): "Settle-
and silt the sand particles would come to
tlement considers the presence of ground ment of spread footings on sand", ASCE
water. It is well established, be generally separated and the mixture 94, SM 3, p. 735: Vol. 96, SM 2, p. 754.
however, Discussions: Vol. 95 SM 3, p. 901, Holtz,
would behave more like a cohesive soil.
that penetration resistance is a reliable W. G. and Gibbs, H. J i Peck, RB., and
The compressibility of the mixture would Bazaraa, A.R.S.: Bolognesi, A.J.L.
approach to assessing compressibility 7. D'Appolonia, D. J., et al., (1970): Discussion
then be influenced by the consolidation
when pore water pressure build-up does on Settlement of Spread Footinqs on Sand,
not occur below the penetrometer. These characteristics of the clay. J. Soil Mech. 8i Found. Div. A.S.C.E., SM 2
March 754-761.
conditions exist in coarse non-cohesive 8. Fletcher, G. F. A., (1965): "Standard Pene-
soils and lightly cohesive soils when the Correlation of SPT and Dutch cone tration Test: its uses and abuses", ASCE,
Vol. 93, SM 4, Pt. 1, p. 67. Discussions: Lo
degree of saturation is less than 85% test results Pinto, V. S. and Mohr. H. A., 1966, Vol. 92,
SM 1, p. 195. Schnabal, J. J. et al., 1966,
(Schultze and Menzebach 1961). Compressibility is affected by many Vol. 92, SM 2, p. 184, Parsons, J. D., 1966,
factors and so are the methods used to Vol. 92, SM 3, p. 105. Schmertmann, J. H.,
Precompression 1966, Vol. 92, SM 5. p. 130.
assess this quality. The objective is to 9. Gibbs, H. J. and Holrz, W. H., ( 1957):
Most natural sands are at least slightly measure relative density and whether the "Research on determining the density of
over-consolidated sands by spoon penetration testing", 4th
(Greenwood, 1974). approach used is a dynamic test or static ISCMFE, London, Vol. 1, p. 35.
36 Ground Engineering
 5 0

4 0
~

4l I d I
O
0 C
L
0 CD
J tl
a 0
Cl
n" I
"~-", t
cave pl/t$ QB Lfa
0
Ink Cp tv P 1+$ $ t 6 L I
0 0.2 Ovt 06 08
Vertical str aln Influence factor I> 0 2 4 6 8 10
Dg

CHART 15 CHART 16

(SCHMERTMANN 1970) (PARRY 1971)

CHART l8
CH ART 17 20
1.0 BOILING
NOTED
IO
H~K
P 'o
0 5 B 6
/I
A

15 qc
0 05 1 0 2 0 N
Kq
E
T J
>e ( PARRY 1971) BURN (I 97 0)

F.I,
HJ
10. Greenwood, D. A., and Tait, J. B., (1970):
"Prediction of Foundations on Sands". Sym-
posium on Foundations on Interbedded Sands,
CSIRO Perth.
11. Greenwood, D. A., (1974): Discussion, Proc.
Conference on the Settlement of Structures,
Cambridge. .002 .006 .02 .06 .2 .6 2 6 20 60 mm
12. Jakobson, B., (1957): "Some fundamental
properties of sand", Proc., 4th ICSMFEC, CLAYI.FJM
LT
Ic I F IM
SAND
I c I FI
GRAVE
M I C IcowES
London, Vol. I, p. 167. S I I I
L
13. de Me/lo, I/. F. B., (1971): The Standard Pene-
tration Test, Proc. 4th Penamerican Con-
ference on Soil Mechanics & Foundation GRAIN SIZ E mm
Engineering Brazil.
14. Meyarhol, G. G., (1965): "Shallow founda- of p, and
tions", ASCE, SM2, Vol. 91, March, p. 21. Correfation N (sfter J. BILLAM, 1974)
15. Meyerhol, G. G. (1974): State-of-the-art of
penetration testing in countries outside A. Schmertmann, 1970 G. Kantey, 1965
Europe." Proc. European Symp. Penetration B. Meigh & Nixon, 1961 H. da Costa Nunes, 1961
Testing, Stockholm, Vol. 2.1. C. Sutherland, 1963 I. Narahari & Aggarwal, 1967
16. Meyerhol, G. G. (1976): Personal communi- D. Rodin, 1961 J. FrankipileLtd. Data I — L from
cation. E. Meyerhof, 1956 K. de Alencar Velloso, -1959 Sang lerat, 1972
17. Parry, R. H. G„(1971): "A direct method F. Schultze & Knausenberger,
1957
L. Spanish tests
of estimating settlements in sands from SPT
values", Proc. Symp. Interaction of Structure
and Foundation, Midland Soil Mech. and
Foundation Eng'g. Soc., Birmingham, pp. 29-
37.
18. Parry, R. H. G., (1974): Personal communica-
tion.
19. Peck, R. B., and Bazaraa, A.R.S.S., (1969J:
Discussion of Paper by D'Appolonia et al,
Journal Soil Mech. & Found. Div. Proc. ASCE,
Vo I . 95, S M 3. 26. Schulrze, E. and Melzer, K. J., (1965): "The (Review Paper), Proc. Conference on the Settle-
20. Peck, R. B., Hanson, W. E., and Thornburn, determination of the density and the modulus ment of Structures, Cambridge.
T. H., (1974): Foundation Engineering, 2 ed. of compressibility of non-cohesive soil by 32. Terzaghi, K. and Peck, R. B., (1948J: "Soil
John Wiley & Sons, Inc. soundings", 6th ICSMFE, Montreal, 1965, Vol. mechanics in engineering practice", Wiley;
21. Perrijohn, F. J., Porter, P. E., and Slaver, I, p. 354. Also 2nd Edition, 1967.
R., (1973): Sand and Sandstone, Springer- 27. Schulrze, E. and Mousse, A., 1961: "Factors 33. Thorburn, S., (1963): "Tentative correction
Verlag, New York. affecting the compressibility of sand', 5th chart for the Standard Penetration Test in
22. Rodin, S., er al, ( 1974): Penetration Testing ICSMFE, Paris, Vol. 1, p. 335. non-cohesive soils", Civil Eng'g. and Public
in the Ijnited Kingdom, European Symposium 28. Simons, N. E. er al (1974): Discussion, Proc. Works Review., June, Vol. 58, p. 752. (Not
on Penetration Testing, Stockholm. Conference on The Settlement of Structures, seen).
23. Sang/erat, A., (1972): The Penetrometer and Cambridge. 34. Thorburn, S., ( 1970): Discussion on the
Soil Exploration, Elsevier Publishing Company, 29. Simons, N. E., and Menzies, B. K. ( 1975): Paper by Thorburn and MacVicar, Proc. Inst.
Amsterdam. A Short Course in Foundation Engineering, Civil Engrs. Conf, on Behaviour of Piles,
24. Schmerrmann, J. H., (1970): "Static cone to IPC Science and Technology Press Limited, London.
35. Webb, D. L., (1959): "Settlement of struc-
compute static settlement over sand", ASCE England.
Vol. 96, SM 3, p. 1011. 30. Skempron, A. W., and Bishop, A. W., (1954): tures on deep alluvial sandy sediments in
25. Schulrza, E. and Menzebach, E. (1961): "Soils", Chap. 10 of Building Materials, Durban, South Africa," Conf. In situ
"Standard Penetration Test and compressi- Their Elasticity and Inelasticity, Amsterdam, Investigations in Soils and Rocks, British
bilitv of soils", 5th ICSMFE, Paris, Vol. 1, North Holland Pub. Co. Geotech. Soc., London, p. 181. Discussion:
p. 527. 31. Sutherland, H. B., (1974): Grandular Materials Meigh, p. 199, Rowe, p. 202.

January, 1977 37