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

net/publication/277353701

Correlations between SPT, CPT, DP, PMT, DMT, CH, SP and PLT Tests Results on
Typical Profiles of Saprolitic Soils from Granite

Conference Paper · January 2001

CITATION READS

1 423

3 authors, including:

Nuno Cruz António Viana Da Fonseca

MOTA-ENGIL University of Porto
88 PUBLICATIONS   119 CITATIONS    183 PUBLICATIONS   861 CITATIONS   

SEE PROFILE SEE PROFILE

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

Residual Soil Characterization by in-situ testing View project

In situ testing View project

All content following this page was uploaded by Nuno Cruz on 03 October 2016.

The user has requested enhancement of the downloaded file.

Parametrical Correlations between SPT, CPT, DPL, PMT, DMT, CH and
PLT Tests Results on a Typical Profile of Saprolitic Soils from Granite

António Viana da Fonseca

(PhD, Assistant Professor, Geot.Lab. Director)
José Fernando Vieira de Sousa
(Madeira Reg. Laboratory of Civil Engineering)
Nuno Bravo Cruz
(MSc in SMGE, CICCOPN and FCT University of Coimbra)
Faculdade de Engenharia – Dep. Engenharia Civil – Rua dos Bragas – 4050-123 Porto - Portugal

ABSTRACT: A typical profile of Porto region residual saprolitic soils from granite has been selected for an
extensive in situ testing research, which included: SPT, CPT, DP, PMT, DMT, CH (cross-hole) and PLT.
These plate loading tests were executed over 30, 60 and 90cm diameter plates.

Identification and classification based on such field testing methods and criteria for interpretation of the re-
sults in order to define mechanical properties (stress state, stiffness and shear strength) and geotechnical de-
sign parameters – mainly for shallow foundations purposes are discussed in this paper. Most attention will be
made on the adaptation of current criteria for evaluation of geotechnical properties, developed for transported
soils with similar physical properties, to the particularities of these residual soils, characterised by singular
fabric and structural (cementation) characteristics. Special emphasis will be made on the following levels:
 At rest stress state (K0) and yield locus or virtual preconsolidation stress state;
 Low and medium stiffness parameters, giving attention to non linear constitutive behaviour; relevance will
be also made on the relative position of deformability modulus on stress-strain levels;
 Shear strength, discussed on the ability of each technique to evaluate the dual importance of cohesive and
friction components, both relevant on these residual soils.

1 INTRODUCTION
Residual saprolitic soils from granite are widely
dominant in large zones of Portugal. The bonded
structure and fabric of these soils have significant in-
fluence on engineering behaviour, and several par-
ticularities dominate foundation design in these ma-
terials.
Systematic site investigations have been under-
taken for the last years in a different regional weath-
ered profiles. The experimental work that will be de-
scribed here was carried out at sites in which homo-
geneous residual soil develop down to a significant
depth.
2 THE EXPERIMENTAL SITE 1. GENERAL
DESCRIPTION OF THE PROFILE
This site, with a plant of approximately 5030m2,
presented an homogeneous weathered degree sapro-
litic soil 6m thick as previously identified by SPT
and DP exploring campaign.
These residual soils from granite from Porto have
been well characterized (Viana da Fonseca et al.,
1994). Details of geological data as well as of grain
size distribution and Atterberg limits of the saprolitic
soil can be found in Viana da Fonseca et al. (1997).
Generally, results obtained with specimens taken
from the SPT sampler and from blocks reveal a fair-
ly homogeneous ground profile (well graded) classi-
fied as SM (silty sand) or SM-SC (silty clayey sand),
according to the ASTM Classification for Engineer-
ing Purposes. Natural physical properties were eval-
uated from these samples with values of 0.60 to 0.85
for the void ratio and 15 to 27% for the water con-
tent, which correspond to degrees of saturation of 70
to 100%. The natural particle arrangement is charac-
terized by open voids on a cemented structure and
low values of relative density (ID 0.28).
3 DERIVED PARAMETERS FROM IN SITU
TESTS
3.1 Resume of the executed tests
Definite campaign of in situ tests comprised 46 SPT,
3 DPSH, 15 DPL, 9 CPT, 7 Load Tests (3 PLT with
30cm, 3 PLT with 60cm and a real scale footing load
test with 120cm diameter) 5 PMT, 4 SBPT, 32 CH
and 12 DMT. Part from the analysis that have been
made with these results will be briefly described
hereby. Figure 1 presents a trend of the variation of
some of the derived parameters with depth.
3.2 Standard Penetration Testing (SPT)
The variation of (N1)60 reflects a range of values for
the angle of shearing resistance of 36 to 40, with
an average value of 38 (Décourt, 1989). These val-
ues agree with those deducted from back-analsys of
PLT and triaxial tests over undisturbed samples (Vi-
ana da Fonseca, 1996). Nevertheless, cohesion inter-
cept calculated from these late tests is significant
(c’=8-12 kPa), which is a characteristic of residual
soils and, therefore, strength solely defined by  
will be rather conservative.
(a)
(b)
Fig. 1. In situ test: (a) SPT and CPT; (b) CH and DMT.
Values of (N1)60 plotted on Skempton (1986) aba-
cus, Fig. 2, would indicate medium to high densities
while direct evaluation of relative density (Dr0.3)
indicates high porosity. It is believed that analysis
ruled by dilatant theories will be conservative. Be-
sides, the association of OCR values greater than
one, on characterization of these soils from SPT, de-
viates the results towards abnormal low strength
values.
Correlations between NSPT and stiffness parame-
ters are very sensitive to different and crossed factors
which raise different difficulties in getting conclu-
sive models. However, it is reecognised that the rela-
tions between penetration parameters and maximum
shear modulus (G0) best assure some independence
on unavoidable misleading factors, such as scale ef-
fects, non-linearity, etc (Jamiolkowski et al., 1988;
Robertson, 1991).
Fig. 2. (N1)60 versus Dr (Skempton, 1986).
From the experimental data a linear relation could
be obtained:
G0 (MPa)98 0.42  N 60 (1)
The variation of G0 with effective mean stress
(’m0) is very low when compared with other param-
eters, such as N60, with the consequence that correla-
tions between G0 and N60 for the relevant values of
’m0 on shallow foundations strongly underestimate
elastic stiffness of the soil (Stroud, 1988).
3.3 Dynamic Probing (DPSH and DPL)
The two regionally most common energies from
ISSMFE standards were used: DPSH e DPL.
The variation of DPSH parameters with depth is
less pronounced than the NSPT and qc (CPT) varia-
tion. Even an analysis based on the specific energy
of penetration (Nixon, 1988), particularly between
the results of DPSH and SPT, is not conclusive. On
the other hand, the results of DPL tests, executed
aside of CPT and SPT tests, interpreted in terms of
the dynamic cone resistance in the unit of pressure
(Rd – Dutch formula) are very similar to qc values.
This trend has been regionally verified on sandy
soils and on saprolitic soils from granite strongly
weathered with mainly sandy matrix in depths down
to 8–10m. The fact that this trend is undetected on
DPSH test is associated to its very high energy
which leads to a number of blows rather low and
nonsensitive to natural variation of soil characteris-
tics through depth.
3.4 Cone Penetrometer Testing (CPT)
Results of CPT (Fig. 1b) denote an approximately
linear growth of qc with  v0 . It is also observed that
values of sleeve friction grow smoothly with the at
rest effective stress, with ratios of fs/  v0 7, reveal-
ing fair homogeneity of friction over this depth.
Soil classification based on K0 , OCR e f s /  'v0
values, with reference to proposals developed for
transported soils (Masood & Mitchel, 1993), indi-
cates high values of OCR, which are not congruent
with the genesis of these soils and with the low val-
ues of coefficient of at rest earth pressure regionally
observed. Classification by Robertson (1990) chart
identifies this material as cemented, aged or very
stiff natural soil, with a grain size distribution typical
of sands or silt/sand mixtures, although its density
index values are low (Viana da Fonseca et al., 1998).
The evaluation of the angle of shearing resistance
by means of Robertson & Campanella’s (1983) pro-
posal conducted to higher values of ’ ( 44  45º )
than those obtained on triaxial tests, where the cohe-
sive component is non-negligeable. This is conse-
quence of a higher sensitivity of qc to frictional and
cohesive components. Being strictly frictional, that
proposal does not explicitly deduct the cohesive in-
tercept. The value of the angle of shearing resistance
has, however, to be taken as a secant value on the
stress space and, consequently, this late parameter
decreases with the increasing of the vertical effec-
tive stress, fact that is very well defined on the re-
sults plotted in the formerly referred abacus. Never-
theless, care should be taken on the adoption of this
exclusively frictional resistance for the evaluation of
the ultimate load of shallow foundations as it is well
recognised the important influence of the cohesive
intercept on the evaluation of qult, as well as the sig-
nificant non-linearity of the bearing capacity factors
with the variation of ’.
An interpretation of the CPT results was made by
the state parameter concept,  (Been et al., 1986).
Considering the value of the critical state index
(  ss  0, 259) deduced in laboratory and substituting
in the authors proposal becomes:
1 (2)
   ln(0.098  QT )
9.45
being QT the normalised cone resistance. Values of
the state parameter, taken as the difference between
the natural void ratio and the correspondent on the
steady (critical) state line, are plotted in Figure 3 and
compared with the above law.
Fig.3. State parameter () deduced from CPT results
(Been et al. 1986).
From the observation of that figure, it can be con-
cluded that there is a higher reserve of strength
(higher absolute values on the dry side) than those
determined by the natural void ratio. This last index
is singularly high as consequence of the porous
leached matrix, as has been previously referred, and,
when compared with uniquely with the effective
stress state, does not reveal the obvious influence of
fabric and structure of these residual soils.
For stiffness evaluation through conepenetration
results and for the same reasons referred to before,
values of qc were correlated with G0 and the similar
low degree of dependency with penetration re-
sistance was detected. The obtained values where
compared with Robertson’s (1991) proposal for
transported soils with different overconsolidation ra-
tios (Fig. 4). It can be verified that measured values
of G0 are substantially higher than the authors’ corre-
lations, revealing higher ratios of elastic stiffness
properties in relation to penetration parameters. This
is a direct consequence of the non-normalised rela-
tion between stiffness and strength on residual soils,
when referenced to proposals deduced on transport-
ed soils.
Fig.4.G0/qc versus qc1=qc/pa(pa/’vo)0.5 (Robertson, 1991).
Correlations between qc and Young moduli, es-
tablished in different stress-strain levels by triaxial
tests (CID e CAD) with local strain measurement,
confirmed the very strong influence of nonlinearity
on E/qc ratios (Viana da Fonseca, 1996), as well as a
mostly singular pattern of that variation when com-
pared to re-known transport soils proposals
3.5 Pressuremeters Testing – Prebored (Ménard -
PMT) and Selfbored (Cambridge – SBPT)
Pressuremeter tests are progressively considered as
most convenient for soil characterization, as they can
be consistently modeled by complete constitutive
laws. The self-bored (SBPT), allows an insertion of
the cell with a minimum disturbance, making possi-
ble a rigorous theoretical interpretation of the test re-
sults and involves the ideal conditions for the eval-
uation of geostatic total horizontal stress
(Jamiolkowski & Manassero, 1996)
From a systematic analysis of PMT and SBPT re-
sults. Viana da Fonseca concluded that the evalua-
tion of the coefficient of earth pressure at rest, K0,
from PMT suffers from the limitations of the pre-
boring process and its determination has resulted in
excessively high values (0.58). On the other hand,
values of K0 determined from very well controlled
SBPT curves (the “lift-off” presssures were taken
individually from each of the three strain gauges –
see Fig.5), ranged between 0.35 and 0.38, which are
closer to regional experience and were confirmed by
careful K0 triaxial tests with local strain measure-
ment.
Fig. 5. “Lift-off” defined on three strain gauges (120).
For stiffness characterization purposes PMT re-
veals some limitations on the definition of elastic
components, since Epmdr/Epm ratio is rather lower
(2) than it would be expected from the ratio be-
tween E0/Epm (18-20, with E0 taken from G0_CH),
which will reflect the important disturbance resulting
from the preboring process.
Jamiolkowski & Manassero (1996) remarked in
the referred closing address to the Conference in Site
Investigation Practices, that the evaluation of shear
stiffness is specially adequuate by means of self-
bored pressuremeter tests. The usual approach refers
to the unload-reload loops, determining the corre-
sponding modulus, as it is known that Gi - the shear
modulus taken from the initial part of the expansion
curve – is, even in SBPT, strongly influenced by the
insertion process. The value of Gur is also very de-
pendent on factors such as those related to creep
which is directly associated to the adopted range of
the unload-reload cycle. Since the measuring system
is able to overpass limitations of compliance of
strain measurement and enables to implement short
but significant looping stresses, effects of creep are
minimized (hysteresis is very low) and the value of
Gur represents the drained stiffness at intermediate
strain level (110-4    110-3), being relatively in-
sensitive to soil disturbance caused by the probe in-
sertion.
This matter was studied on the pressuremeter test
results from PMT and SBPT. It was concluded that
even on implementing unload-reload cycles, PMT do
not assure enough precision to overpass limitations
of strain measurement, which reveals that PMT is
strongly limited for stiffness characterization pur-
poses. Their limitations were also very well defined
on the pseudo-elastic versus virgin modulus ratios,
since Epmdr/Epm is rather lower (2) than it would be
expected from the ratio between E0/Epm (18-20,
with E0 taken from G0_CH), proving that important
disturbance resulting from the preboring process also
remains.
SBPT turn to be the opposite as they have proved
to give a good insight of stiffness properties. Figure
6 represents one of the executed tests, which have
included careful controlled unload-reload cycles.
Fig.6. SBPT with a Cankometer equipment with inclu-
sion of unload-reload cycle (Viana da Fonseca,
1996).
Taking into account the results of SBPT for eval-
uation of the deformation modulus, particularly of
the unload-reload modulus, the ratio between G0 and
Gur for the same stress levels involved in the several
tests were found to range between 2.6 and 3.0. These
values are substantially lower than the ratio (10),
reported by Tatsuoka & Shibuya (1992) on Japanese
residual soils from granite.
Taking the non-linearity hypothesis of Akino –
cited by the previous authors - developed for a high
range of soil types, including residual soils from
granite, and expressed simply by:
Esec  E0 ,  104 (2)
Esec  E0  ( /104 ) 0,55 ,  104 (3)
the SBPT unload-reload modulus correspond to se-
cant values for shear strain of 5.7-7.410-4 (610-4),
which agrees very well with the above indicated
trends. One of these values is implanted in the rela-
tion of Gs versus c on Figure 7, using Jardine
(1992) model.
Fig.7. Non-linearity - Gdr over relation Gsec versus c
(Jardine, 1992; Akino, 1990).
The evaluation of strength parameters from PMT
by means of yield or limit pressures methods, lead to
a rough underestimation of strength, even taken as
exclusively frictional. On the other hand, results
from Cambridge SBPT interpreted by the Hughes-
Wroth-Windle method (for purely frictional materi-
als), has conducted to very similar values of peak
angle of shearing resistance to the ones obtained by
CPT interpretation, but still remains the indefinition
of an explicitly cohesion evaluation. A reinterpreta-
tion of such results by recent cavity expansion theo-
ries for cohesive-frictional materials can solve this
indefinition. Besides, a similar analysis on the large
deformation levels (=0) – as close as possible to
critical condition - of the pressuremeter curves has
enabled to define a value of ’cv=31.6 which is
equivalent to the value obtained from specific triaxi-
al tests (Viana da Fonseca, 1996).
3.6 Cross-Hole Seismic Test (CH)
The results of cross-hole tests are very homoge-
neous and reveal a very smooth growth of G0 with
depth. In Fig. 7 these variations are compared with
the lines corresponding to Ishihara’s (1986) pro-
posals, for natural alluvial sands, aged and cement-
ed, corresponding to:
G0 (MPa)  (7.9 to14.3)  0.4
2.17  e2  p  103 0.4 (4)
1 e
0
while for the present results the following is found:
G0 (MPa)  65 
2.17  e
2
  p 103 
0.07
(5)
1 e
0
It can be seen that the value of the constant for the
maximum shear modulus expression is much higher
than the corresponding for sandy transported soils
while the factor (m) of dependency on the octahedral
at rest effective stresses is far lower.
Fig.7. Comparison of observed trends of G0 versus p0
and Ishihara (1986) proposals.
Fig.7. Comparison of observed trends of G0 versus p0
and Ishihara (1986) proposals.
From the reinterpretation of the observed trends –
see Fig.8 – with reference to the most common cor-
relations between N60 and Vs(CH) for transported
aged soils (Seed et al., 1986), new values for the
constants were found: fixing =0.84 and fb =1.3, one
gets fb =1.51. Reinterpreting the same values with
the direct correlations between G0 and N60 (Stroud,
1988), the following was obtained:
G0 (MPa)  57  N 600,2 (6)
concluding that it is the distinct dependency of both
G0 and N60 on ’v0 that explains such a reformula-
tion of the correlation between them.
3.7 Dilatometer Test (Marchetti - DMT)
Material classification in grain size terms based on
Marchetti ID index is fairly consistent (SM) opposing
the results based on CPT parameters, which identify
finer groups from f s / qc , probably due to high per-
centage of laminar particles (mica) which induce
higher sleeve resistances on these soils.
Campanella & Robertson (1991) proposal for
strength evaluation gives very accurate values, only
if the relation between qc/’v0 and KD is changed to
the new trends found on the experimental campaign,
i.e.: qc /  v0 =8.4 K D against qc /  v0 =33 K D pro-
posed by the authors for transported soils. With the
same relation the evaluation of the coefficient of
earth pressure at rest, K0, by means of K DDMT corre-
lations, will be adapted to the following law:
K 0  0.736  0.024  K D  0.00172 qc /  v0 (7)
Recent trends for association of K0 with KD_DMT
by using the state parameter,  (Bellotti et al., 1994)
were reinterpreted, resulting on a more adjusted law:

K0  K D / 46.8 103,82  (8)
Stiffness characterization for shallow foundations
settlement assessment, having recourse to the corre-
lations between the moduli ED(DMT) and G0 or Es10%
(Baldi et al., 1989), resulted in new correlations with
much higher ratio between those parameters, than
the referenced proposals:
G0 / E D  16.7  16.3  log10 ( p0 N ) (9)
Es10% / ED  2.35  2.21 log10 ( p0 N ) (10)
This last correlation, however, is situated between
the laws that define – for those authors – the behav-
iour of NC and OC transported soils.
3.8 Loading tests (PLT and footing load test)
Strength parameters in residual soils are character-
ised commonly by friction and cohesive compo-
nents. These two components are usually determined
directly, in an homogeneous space, by laboratory
tests over undisturbed samples, or indirectly by in
situ tests interpretation based on curve-fitting analy-
sis, such as those apllied on pressuremeter tests, such
as those described by Schnaid & Mantaras (1998), or
by back-analysis of two or more loading tests in dif-
ferent sizes, considering a specific failure model.
The need for the execution of more than one load-
ing test to define failure patterns is complemented by
advantage of being able to elaborate an integrated
analysis of the different pressure-displacement
curves enabling to study the importance of the pair
of factors “stiffness-depth of influence”. This subject
is discussed elsewhere in this Symposium (Viana da
Fonseca & Sousa, 2001).
On this experimental site a set of load test under
different loading areas was executed: a large scale
footing loading test (D=120cm) and two PLT with
30 and 60cm diameter. The incremental loading
steps have followed the same stress levels. General
curves can be seen in the other paper to this Sympo-
sium (Viana da Fonseca & Sousa, 2001).
It shall be noted that all these tests, similarly to
the footing tests, do not appear to develop a classical
generalised failure surface but present a punching
shear failure mode. This pattern of behaviour results
in monotonously increasing pressure-settlement
curves, with smooth curvatures, giving rise to very
little definition of the inflexion zone of that curve. A
log-log plotting can be a better way to detect it.
This has been used to preliminarily interpret the
results of the loading tests (Viana da Fonseca, 1996)
and the following was obtained:
 B = 0.30 m  qrot = 700 kPa
B = 0.60 m  qrot = 821 kPa
B = 1.20 m  qrot = 950 kPa
qrot means “punching failure” which can be distinct
from the ultimate equilibrium defined by limit state
analysis (Terzaghi-Meyerhof-Vésic).
Note that the last value, respecting the footing, cor-
roborates the developing signs inferred from inflec-
tions in vertical and horizontal displacements -
synthomatic of the appearance of shear bands, lead-
ing to failure (Viana da Fonseca et al., 1997).
If we use the values of those failure loads, in the
bearing capacity formulation (Eurocode 7, 1994),
taking account the water level position, we obtain
three equations. These can be optimised to get the
two strength parameters range. The values obtained
were: c' = 6.3–6.9kPa and ' = 36.5-37.0, revealing
a fair agreement with the results obtained in exten-
sive laboratory testing over undisturbed samples
(Viana da Fonseca, 1998) and in situ testing (Viana
da Fonseca et al., 1998).
4 CORRELATIOS BETWEEN IN SITU TESTS
PARAMETERS
4.1 CPT – SPT correlations
Danzinger et al. (1998) collected data from tests
sites of different parent rocks of residual soils, to
conclude that correlations between CPT ad SPT pre-
sent a large scatter attributing it to intrinsic hetero-
geneity.
Fig.7. Ranges of the relation of qc/N versus D50 on re-
sidual soils – Danzinger et al. (1997)
Attributing it to intrinsic heterogeneity, they have
concluded that different parent rocks generally pro-
duce different correlations for the same particle size
distribution (a pattern of soil type). It is assumed that
from brazilian data, there is a general trend to of de-
creasing values of qc/N with D50 and generally lower
values than the expressed by Robertson et al. (1983)
average line. Our results, however show the opposite
as it is expressed in Figure 7 and more clearly in Fu-
gure 8.
Figure 8. Correlation between qc(MPa) and N60 - refer-
ence to Burland & Burbidge (1985)
Figure illustrates the comparison of the range of
qc/N60 for the experimental site with the chart of
Burland & Burbidge (1985) developed for sandy
transported soils. The results, corresponding to
qc/N60 = 0.7 - 0.8 (qc, MPa), are above the mean
trends of the charts, which is probably a conse-
quence of the greater sensitivity of CPT cone re-
sistance to the cohesive strength component of the
residual soil. This, however, corroborated other bra-
zilian authors (Rocha Filho, 1986) findings in resid-
ual soils. These results demand for more data and
regional development of specific proposals.
4.2 PMT versus CPT/SPT correlations
Ratios between PMT and SPT or CPT parameters
(see Table 1) do not have always correspondence, in
grain size terms, with transported soils. Relevance is
made on the similarity between N60/Epm values and
those of Martin (1987), for Piedmont saprolitic soils.
Table 1 - Ratios between SPT, CPT and PMT parameters
* * *
qc / p L f s / p L N 60 / p L N 60 / E pm E pm / pl* Epmd / Epm
(MPa) (MPa)
14.3 0.390 14.6 1.4 10.6 1.41.9
5.EXPERIMENTAL SITE 2. GENERAL PRO-
FILE
Located in Maia, surroundings of Porto, in
CICCOPN facilities, experimental site 2 has is part
of the same granitic complex of experimental site 1.
The site, that has been used for research for quite a
while, was characterised by 4 DPSH, 8 CPT, 12
screw plate (SP) tests performed in 3 main align-
ments, 3 DMT, 2 PMT, plus the laboratory testing.
Identification and physical properties determined by
laboratory testing fall within the same ranges of site
1 (Cruz et al, 1997).
DMT and CPT results identify these soils as silty
sands and sandy silts, sometimes clayey siltes, which
fits in the ASTM Classification. DMT, on other
hand, showed surprising accuracy evaluating the unit
weight (+1kN/m3)
5.1. Stress history
Since OCR has no meaning in this type of soils,
coefficient of earth pressure at rest (K0) is the stress
history parameter that was evaluated.
Following the same path of experimental site 1,
the ratio between qc/’V0 and KD is 12,5. This ratio
leads to values of K0 of 0,4 – 0,5 that are recognized
as typical of the complex.
PMT results, by their side, show higher values
(>1) which confirms the general idea of the small ef-
ficiency of the test to evaluate K0.
On the other hand, the ratio P0DMT/P0PMT falls in
the range 2 – 3.
5.2. Strength Parameters
Results of DPSH tests show values of dynamic
point resistance, qd, varying between 5 and 15 MPa,
often between 5 and 6 MPa. This range finds good
correspondence with CPT results of qc (4 - 8 MPa).
The other parameter of the test, f(s), ranges between E’i =  qc
0,1 and 0,3. The ratio N20 / qc, vary from 0,6 to 0,8. (11)
The compairison between results obtained from
PMT and CPT show values of qc/PL* and fs/PL* where  was taken equal to three since this value
falling in the ranges 4 – 6 and 0,10 – 0,25, respec- was obtained from comparison with SP tests (ASCE,
tively, which are lower than those obtained in site 1. 1994). In this case, E’i ranges from 12 to 24 MPa.
Concerning to friction angle, CPT results ob- The evaluation of E from the results of SP tests
tained by using Robertson and Campanella chart can be based on three interpretation models: Norwe-
(1993) vary in the range 38º - 42º (Fig. 9), while gian, International (elasticity) and Elasto-plastic in-
DMT’s, determined by Marchetti’s latest correlation terpretations. More detailed data can be found in
(1997) established for sandy soils, assume values Strout (1998). The results of E’i obtained from Elas-
from 37 to 39º. It should be noted that this correla- to-plastic interpretation fall in the interval 20 – 50
tion is assumed to underpredict ’, since it results MPa. It is also relevant to stress the comparison be-
from the lower limit of 3 curves based on the as- tween SP and PLT obtained results in a different ex-
sumptions of K equal to K0nc, to 1 and to the square perimental site, that showed surprising agreement
root of coefficient of passive earth pressure, Kp. (Cruz et al., 2000). In Table 2, results from CPT and
SP tests are presented. DMT data is also included.
Friction Angle

Table 2 – Results of E
qc (bars)

0 50 100 150 200

0
0.1
0.2
CPT SP DMT
0.3
0.4
' = 48º E’i (MPa) 12 - 24 20 – 50 (1) 15 – 40
' v, (bars)

0.5
E’10 (MPa) 13 – 14 (2)


0.6 46º
0.7
0.8 E’r (MPa) 10 – 16 (2)
0.9
44º (1) Elasto-plastic interpretation
1 30 32 34 36º 38º 40º 42º
CPT1.1 CPT1.2 CPT2.1 CPT2.2 CPT3.1 CPT3.2 CPT4.1. CPT4.2
(Robertson e
(2) International interpretation

Figure 9. Friction angle from CPT (Robertson & Cam-

panella, 1993) DMT results were determined following Mar-
chetti’s correlation (1980). Data reduction in DMT
CID and CAD triaxial testing sustains values leads to constrained modulus, M, which was con-
ranging from 37º to 41º. In fact, those results were verted to E assuming Poison’s coefficient, ’ = 0.3.
expected since sampling disturbance has no im- As it can be seen in Table 2, DMT and SP (Elasto-
portant effects on ’. On the other hand, the cohesion plastic interpretation) results fall within the same
intercept give results within 2 – 3 kPa. However, it is range, while CPT values are quite lower and seem to
well known the effect of sampling disturbance in fit the International interpretation results. If one pre-
these results. tends to compare data from CPT and SP (elasto-
Another way of determining c’ is proposed by plastic interpretation) it is find an interval of 5 to 6,
Cruz (1995) following a simple path. In fact, KD re- e.g., double of the  assumed. On the other hand the
flects the resistence of soil, so it is logical to expect ratios between ED/EPMT, E’SP/EPMT and E’SP/qc are
that either c’ and ’ have influence on this parame- 1.5, 2 – 3 and 5 – 6, respectively.
ter. Then, if ’ from triaxial testing is assumed, Finally, results from triaxial testing sustain much
correspondent KD can be back-calculated. The dif- lower values (4 – 6 / 15 - 20 MPa) reflecting the
ference between both KD (measured and back- level of sampling disturbance in this type of soils.
calculated) will reflect effective cohesion. More de-
tailed information can be found in Cruz et al. (1997). 5.4. Correlations between in situ tests parameters
From available data which is not yet considered suf-
ficient to establish reliable correlations, the results In table 3 comparative parameters between in situ
for both experimental sites 1 and 2 are 9 - 12 kPa tests are presented.
and 2 – 4 kPa, respectively. Notice the convergence
of DMT and triaxial values. Table 3 – Ratios obtained from in situ tests
qd/qc 0.75 – 1.25 P0DMT/P0PMT 2–3
5.3. Deformation parameters
N20/qc (MPa-1) 0.6 – 0.8 P1DMT/PfPMT 1
Derived Young modulus, E’i, were taken from qc/PL* 4–6 ED/EPMT  1.5
CPT, DMT, SP and triaxial testing. E’i (CPT) results
fs/PL* 0.10 – 0.25 E’SP/ EPMT 2–3
were based on the general formula oftenly used:
EPMT/ PL* 12 E’SP/qc 5–6

CONCLUSIONS
REFERENCES
ASCE. Settlement Analysis. American Society of Civil Engi-
neers. ASCE Press, 1994.
Cruz, N. (1995). Avaliação de Parâmetros Geotécnicos pelo
Dilatómetro de Marchetti. MSc thesis (University of Coim-
bra).
Cruz, N., Almeida e Sousa, J., Aguiar, A. (2000). Ensaios com
Screw Plate. Uma Experiência em Solos Residuais. VII
Congresso Nacional de Geotecnia, pp 123-132.
Cruz, N., Viana, A., Coelho, P., Lemos, J. (1997). Evaluation
of Geotechnical Parameters by DMT in Portuguese Soils.
XIV Int. Conf. on Soil Mechanics and Foundation Engi-
neering, pp 77-80.
Marchetti, S. (1980). In Situ Tests by Flat Dilatometer. ASCE
Jnl GE, Vol 106, No.3, 229-321.
Marchetti, S. (1997). The Flat Dilatometer Design Applica-
tions. III Geotechnical Engineering Conference, Cairo
University.
Strout, J. (1998). Evaluation of the Field Compressometer Test
in Sand. PhD thesis (NTNU).
5 REFERENCES
5.1 CPT – SPT correlations
Danzinger et al. (1998) collected data from tests
5.2 References
In the text, place the authors’ last names (without in-
itials) and the date of publication in parentheses (see
examples in Section 5). At the end of the paper, list
all references in alphabetical order underneath the
heading REFERENCES (Reference heading tag).
The references should be typed in small text (10 pt
on 11 pt) and second and further lines should be in-
dented 5.0 mm (0.2") (Reference text tag). If several
works by the same author are cited, entries should be
chronological:
Larch, A.A. 1996a. Development ...
Larch, A.A. 1996b. Facilities ...
Larch, A.A. 1997. Computer ...
Larch, A.A. & Jensen, M.C. 1996. Effects of ...
Larch, A.A. & Smith, B.P. 1993. Alpine ...
5.2.1 Typography for references
Last name, First name or Initials (ed.) year. Book title. City:
Publisher.
Last name, First name or Initials year. Title of article. Title of
Journal (series number if necessary) volume number (issue
number if necessary): page numbers.
5.2.2 Examples
Grove, A.T. 1980. Geomorphic evolution of the Sahara and the
Nile. In M.A.J. Williams & H. Faure (eds), The Sahara and
the Nile: 21-35. Rotterdam: Balkema.
Jappelli, R. & Marconi, N. 1997. Recommendations and preju-
dices in the realm of foundation engineering in Italy: A his-
torical review. In Carlo Viggiani (ed.), Geotechnical engi-
neering for the preservation of monuments and historical
sites; Proc. intern. symp., Napoli, 3-4 October 1996. Rot-
terdam: Balkema.
Johnson, H.L. 1965. Artistic development in autistic children.
Child Development 65(1): 13-16.
Polhill, R.M. 1982. Crotalaria in Africa and Madagascar. Rot-
terdam: Balkema.
5.3 Notes
These should be avoided. Insert the information in
the text. In tables the following reference marks
should be used: *, **, etc. and the actual footnotes
set directly underneath the table.
5.4 Conclusions
Conclusions should state concisely the most im-
portant propositions of the paper as well as the au-
thor’s views of the practical implications of the re-
sults.
6 PHOTOGRAPHS AND FIGURES
Number figures consecutively in the order in which
reference is made to them in the text, making no dis-
tinction between diagrams and photographs. Figures
should fit within the column width of 90 mm (3.54")
or within the type area width of 187 mm (7.36").
Figures, photographs, etc. should be in black on-
ly. Paste copies of the same size onto the typescript
where you want them to appear in the text. Do not
place them sideways on a page; however if this can-
not be avoided, no other text (except the figure cap-
tion) should appear on that page. Figures, etc. should
not be centered, but placed against the left margin.
Leave about two lines of space between the actual
text and figure (including caption). Never

Figure 1. Caption of a typical figure. Photographs will be
scanned by the printer. Always supply original photographs.
place any text next to a figure. Leave this space
blank. The most convenient place for placing figures
is at the top or bottom of the page. Avoid placing
text between figures as readers might not notice the
text. Line drawings (as well as photographic repro-
ductions of these) should be in black (not grey) on
white. Keep in mind that everything will be reduced
to 75%. Therefore, 9 point should be the minimum
size of the lettering. Lines should preferably be 0.2
mm (0.1") thick. Keep figures as simple as possible.
Avoid excessive notes and designations.
Photographs should be black and white, with
good contrast and on glossy paper. Photographic re-
productions cut from books or journals, photocopies
of photographs and screened photographs are unac-
ceptable. The proceedings will be printed in black
only. For this reason avoid the use of colour in fig-
ures and photographs. Colour is also nearly always
unnecessary for scientific work.
7 PREFERENCES, SYMBOLS AND UNITS
Consistency of style is very important. Note the
spacing, punctuation and caps in all the examples
below.
 References in the text: Figure 1, Figures 2-4, 6,
8a, b (not abbreviated)
 References between parentheses: (Fig. 1), (Figs
2-4, 6, 8a, b) (abbreviated)
 USA / UK / Netherlands / the Netherlands instead
of U.S.A. / U.K. / The Netherlands
 Author & Author (1989) instead of Author and
Author (1989)
 (Author 1989a, b, Author & Author 1987) instead
of (Author, 1989a,b; Author and Author, 1987)
 (Author et al. 1989) instead of (Author, Author &
Author 1989)
 Use the following style: (Author, in press); (Au-
thor, in prep.); (Author, unpubl.); (Author, pers.
comm.)
Danzinger, F. A. B., Politano, C.F. & Danzinger,
B. R. (1998)
Schnaid & Mantaras (1998)
Jamiolkowski, M. & Manassero, N. (1996). The
role of in situ testing in geotechnical - thoughts
about the future. Closing address. Advances in
Site Investigation Practice. Pp. 929-951. Thomas
Telford, London.
Been, K., Crooks, J.H.A., Becker, D.E.; Jefferies,
M.G. (1986). The cone penetration test in sands:
part I. State parameter interpretation. Géotech-
nique, Vol.36, Nº 2, pp.239-249
Always use the official SI notations:
 kg / m / kJ / mm instead of kg. (Kg) / m. / kJ.
(KJ) / mm.;
 20°1632SW instead of 20° 16 32 SW
 0.50 instead of 0,50 (used in French text); 9000
instead of 9,000 but if more than 10,000: 10,000
instead of 10000
 14C instead of C14 / C-14 and BP / BC / AD in-
stead of B.P. / B.C. / A.D.
 × 20 instead of 20 / X20 / x 20; 4 + 5 > 7 in-
stead of 4+5>7 but –8 / +8 instead of – 8 / + 8
 e.g. / i.e. instead of e.g., / i.e.,
8 SUBMISSION OF MATERIAL TO THE
EDITOR
The camera-ready copy of the complete paper print-
ed on a high resolution printer on one side of the pa-
per as well as two copies of the paper should be sent
to the editor. Include the original photographs.
Check whether the paper looks the same as this
sample: Title at top of first page in 18 points, authors
in 14 points and all other text in 12 points on 13
points line space, except for the small text (10 point
on 11 point line space) used in tables, captions and
references. Also check if the type width is 187 mm
(7.36"), the column width 90 mm (3.54"), the page
length is 272 mm (10.71") and that the space above
the Abstact is exactly as in the sample. Write your
name and the shortened title of the paper in pencil in
the bottom margin of each page and number the pag-
es correctly.
 9 DEADLINE

The above material should be with the editor before

the deadline for submission. Any material received
too late will not be published. Send the material by
airmail or by courier well packed and in time. Be
sure that all pages are included in the parcel.

View publication stats