ASSOCIAÇÃO BRASILEIRA DE

MECÂNICA DOS SOLOS E ENGENHARIA

GEOTÉCNICA – ABMS

NÚCLEO REGIONAL DE MINAS GERAIS

CREA - MG

31/05/2017
 CINQUENTA ANOS PROJETANDO
FUNDAÇÕES

UMA RETROSPECTIVA

Luciano Décourt
Engenheiro Civil – Professor Universitário
COMPARAÇÃO ENTRE VALORES
CALCULADOS E MEDIDOS DE
FREQUÊNCIAS NATURAIS
DE FUNDAÇÕES

DÉCOURT, L.

III COBRAMSEF

BELO HORIZONTE, BRASIL, 1966

PARTE 1 PARTE 2

SPT – SPT-T FUNDAÇÕES

DIRETAS

PARTE 3 PARTE 4

FUNDAÇÕES EVENTOS DE
PROFUNDAS PREVISÃO
SPT – SPT-T
 UM CASO ONDE O SPT
FALHOU NA PREVISÃO DO
COMPRIMENTO DE ESTACAS

DÉCOURT, L.

VII COBRAMSEF

PORTO ALEGRE, BRASIL, 1986

STANDARD PENETRATION TEST (SPT)
INTERNATIONAL REFERENCE TEST
PROCEDURE

DÉCOURT, L.; MUROMACHI, T.;

NIXON, I.K.; SCHMERTMANN, J.H.;
THORBURN, S.; ZOLKOV, E.

ISOPT I
ORLANDO, USA, 1988
THE STANDARD PENETRATION TEST
STATE OF THE ART REPORT

DÉCOURT, L.

XII ICSMFE/CIMSTF
RIO DE JANEIRO, BRAZIL, 1989
 “THE MECHANICAL AGING OF SOILS”
25th Terzaghi Lecture
John H. Schmertmann

L. Décourt (1989) observed that the Bazaraa (1967) correlations,

based on field data, if adjusted by an aging factor of 0.4–0.6 would
approximatelly match the SPT vs. relative density correlations
previously obtained by Gibbs and Holtz (1957) and Marcuson and
Bieganouski (1977).
The writer believes that the above aging explanation helps greatly to
explain a big mistery – namely why the Bazaraa correlations did not
fit well with the others.
 BRAZILIAN EXPERIENCE ON SPT

DÉCOURT, L.; BELICANTA, A.;

QUARESMA FILHO, A.R.

XII ICSMFE
SUPLEMENTARY CONTRIBUTIONS BY
ABMS

RIO DE JANEIRO, BRAZIL, 1989

THE SPT-T AN IMPROVED SPT

DÉCOURT, L.
QUARESMA FILHO, A.R.

SEFE II

SÃO PAULO, BRAZIL, 1991

Em 1991, postulei que, para solos fora da
Bacia Sedimentar Terceária de São Paulo,
BSTSP, as correlações estabelecidas para essa
região poderiam ser utilizadas, com sucesso,
para outros locais, desde que os valores de Neq
viessem a ser utilizados no lugar dos valores
medidos de SPT.

Neq foi definido como sendo o valor do torque,

T, medido em kgf.m dividido por 1,2.
 SPT IN NON CLASSICAL MATERIALS

DÉCOURT, L.

U.S. – BRAZIL GEOTECHNICAL

WORKSHOP ON APPLICABILITY OF
CLASSICAL SOIL MECHANICS PRINCIPLES
TO STRUCTURED SOILS

BELO HORIZONTE, BRAZIL, 1992

 PRACTICAL APPLICATIONS OF THE
STANDARD PENETRATION TEST
COMPLEMENTED BY TORQUE
MEASUREMENTS

DÉCOURT, L.; QUARESMA FILHO, A.R.

XIII ICSMFE

NEW DELHI, INDIA , 1994

 A MORE RATIONAL UTILIZATION
OF SOME OLD IN SITU TESTS

DÉCOURT, L.

FIRST INTERNATIONAL CONFERENCE

ON SITE CHARACTERIZATION – ISC

ATLANTA, U.S.A., 1998

 Ei≈45-50% Ei≈72%

Corrections to measured NSPTvalues, in fine Corrections to measured NSPTvalues, in fine

fine sands below the water level (Ei≈45-50%) sands below the water level (Ei≈72%)
In some cases, like the mentioned in Décourt
(1986) these unrealistic low NSPT values were the
most likely explanation for the fact that the
capacities of displacement piles, computed using
Décourt and Quaresma method (1978; 1982)
were much lower than those provided by loading
tests. Once the low NSPT values were corrected as
suggested in slides 21 and 22, the differences
between predicted and measured capacities
become negligible.
 SPT, SPT-T. A PRÁTICA BRASILEIRA:
VANTAGENS, LIMITAÇÕES E CRÍTICAS

DÉCOURT,L.

ABMS, NÚCLEO REGIONAL

DE SÃO PAULO

SÃO PAULO, BRASIL, 2002

 LEGENDA
I AREIA FOFA N60 = 3,0 + 0,6z

II AREIA MEDIANAMENTE COMPACTA N60 = 8,4 + 1,2z

III AREIA COMPACTA N60 = 21,6 + 1,8z

IV AREIA MUITO COMPACTA N60 = 33 + 3,0z

PARA ARGILAS RIJAS E DURAS
DA BSTSP

Cu (kPa) = 12,5 N72

Cu (kPa) = 10,42 T
 Values of NSPT; Torque, T; Cu (Vane Test) and Cu (Décourt
correlations) for a very soft, quaternary clay of Guarujá, Brazil.
 ÍNDICE de BOSIO
Para solos muito duros/ muito densos
(N60 ≥ 60)
IB = 22 / P50 - P30
P50.......Penetração do amostrador
para 50 golpes

P30.......Penetração do amostrador
para 30 golpes
 CONCLUSÃO

SPT-T+ IB
BOM PARA SOLOS NÃO LATERÍTICOS

S-SPT-T + IB

BOM PARA TODOS OS SOLOS

 O CONCEITO DE N EQUIVALENTE,
Neq, NA PRÁTICA DA ENGENHARIA.
AINDA UM POSTULADO OU JÁ
UMA REALIDADE COMPROVADA?

DÉCOURT, L.

XII COBRAMSEG

SÃO PAULO, BRASIL, 2002

 PARÂMETROS DO SOLO
PARA PROJETOS DE FUNDAÇÕES

DÉCOURT, L.

ENGENHARIA DE FUNDAÇÕES
PASSADO RECENTE E PERSPECTIVAS
UM TRIBUTO AO PROF. NELSON AOKI

SÃO PAULO, BRASIL, 2009

FUNDAÇÕES DIRETAS
 BEARING CAPACITY OF
LONG FOOTINGS, QU

TERZAGHI

qu = cNc + qNq + 1/2γBNγ

 SOME OPINIONS REGARDING
THE VALIDITY OF THE BEARING
CAPACITY THEORIES FOR SANDS
 BRIAUD AND JEANJEAN (1994)
“If is a constant and independent of B,
then N, cannot be a constant and must carry a
scale effect in 1/B.
This major shortcoming plus the difficulty in
obtaining an accurate value of the needed soil
parameter Ø and the documented poor
accuracy of this method (Amar et al. 1984) lead
to the recommendation that:
- The use of this equation should be
discontinued”.
 DÉCOURT (1999)
“What has been done in the past 60 years
was adequate, but, only for that time.
On the threshold of the 21st century,
however, it is mandatory that we clear our
minds of some concepts that helped for a
while, but now prevent any further
development of the art of designing
foundations.
It is being proposed that design approaches
based on these concepts be abandoned.
Emphasis has to be given to settlements.
Foundation designs in the 21st century,
other than for displacement piles, should
rely exclusively on settlement computations,
leaving the ghosts of bearing capacity
theories and failures to the past, duly
exorcised and buried, to show up never
more”.
 FELLENIUS, B.H. (1999)
“It is amazing that the arbitrary nature of the
calculated N-factors has not long ago sent the
formula to the place where it belongs – the
museum of old paradigms whose time has
passed. However, the bearing capacity is so
well entrenched that stating that it is wrong
and should not be used would seem to border
or committing heresy. When I now say so,
should I fear being burned at the stake, or is
the practice ready to accept new views ?”
 POULOS, H.G. (2010)
“Unfortunately, there is now considerable
evidence that demonstrates that this theoretical
conclusion is not confirmed in practice.
Décourt (2008) has re-plotted data from tests
on footings of various sizes and found that,
when normalized with respect to
settlement/diameter (s/b), the load-settlement
curves are unique and not dependent on the
footing size nor on the relative density.
Similar conclusions have been reached
from recent centrifuge test carried out on
model footings by Gavin et al (2009).

Akbas and Kulhawy (2009) have arrived to

conclusions similar to those of Décourt
and Gavin et al”.
FUNDAÇÕES E INTERAÇÃO
SOLO- ESTRUTURA

DÉCOURT, L.

X COBRAMSEF
RELATO GERAL

FOZ DO IGUAÇU, BRASIL , 1994

log (q/qr) = 0,426 + 0,426 log (s/Beq)
 THE BEHAVIOR OF A BUILDING WITH
SHALLOW FOUNDATIONS ON A STIFF
LATERITIC CLAY

DÉCOURT, L.

ASCE, GSP 40
VERTICAL AND HORIZONTAL DEFORMATIONS
OF FOUNDATIONS AND ENBANKMENTS

TEXAS, U.S.A., 1994

THE STIFFNESS OF LATERITIC SOILS
The stiffness of the clay on which the Rimini building was
founded was much higher than any forecast. The reason for
this extraordinary behaviour of the soil was not immediately
recognized. After some reasoning, the author decided to
search the literature about the behaviour of lateritic soils as
bearing strata for foundations. It was that found practically
nothing had been published on this subject. Even in the
proceedings of the of the First International Conference on
Lateritic and Saprolitic soils, held in Brasilia, in 1985 very little
information was found. An important exception was the
paper by Velloso et al. (1978). Results of plate load tests on
lateritic soils shown very high E values. But no one seemed to
associate these high stiffness with the lateritic behavior of
these soils.
 PREDICTION OF LOAD SETTLEMENT
RELATIONSHIPS FOR FOUNDATIONS ON
THE BASIS OF THE SPT-T

DÉCOURT, L.

CYCLE OF INTERNATIONAL
CONFERENCES LEONARDO ZEEVAERT

MEXICO, 1995
ASSESSMENT OF THE REFERENCE STRESS, qr
ON BASIS OF NSPT, Neq AND T, FOR
COMMON SOILS

Soil type I………………sands

Soils type II…………….intermediate soils
Soils type III……….…..saturated clays (Øu= zero)
 q (kPa) q (kgf/cm2) qr (kPa) qr (kgf/cm2)
SOLO r r

I 95N60/95T 0,95N60/0,95T 114N72/114Neq 1,14 N72/1,14Neq

II 80N60/80T 0,80N60/0,80T 96N72/9,6Neq 0,96 N72/0,96Neq

III 65N60/65T 0,65N60/0,65T 78N72/7,8Neq 0,78 N72/0,78Neq

ASSESSMENT OF THE REFERENCE STRESS,
qr ON BASIS OF NSPT, Neq AND T, FOR
LATERITIC SOILS
For the practicing foundation engineer, what
matters is that lateritic soils (LS) are much stiffer
than non lateritic soils (NLS).
The extraordinary stiffness characteristics of
lateritic clays have been ignored by foundation
engineers. Décourt (1994) showed, for the first
time, that lateritic clays are much stiffer than
non lateritic clays with equal SPT values.
Barros and Pinto (1997) analyzed the
relationships between G0 (MPa) and NSPT, for
lateritic (LS) and non lateritic soils (NLS), showing
that they were quite different.
For lateritic soils (LS): G0 (MPa) = 56 + 20.3N72
For non lateritic soils (NLS):G0 (MPa) = 7/8 N72

In most cases, following Barros and Pinto (1997),

the identification of lateritic clays might be well
established. However, in some other cases it is
not so evident.
The proposed procedure for assessing qr on basis
of NSPT values is not applicable to all soils. As
already pointed out, for saprolites, the method
gives reliable results, provided Neq values are
used instead of NSPT values.

For lateritic clays, however, an increasing factor,

αL, for taking into account the benefits of
laterization, has to be used.
On basis of my previous experience, values of
αL = 2.0 ± 35% are suggested.
Another important parameter to be considered
is the Coefficient of Intrinsic Compressibility,
“C”, Décourt (1999). For most of the soils, the
value C= 0.42, yields to good predictions.
For lateritic soils (LS) “C” values are in the range
0.21 to 0.45.
More precise values, however, are only
obtained analysing loading test results. In the
absence of these tests, values of “C” between
0.28 and 0.42 are being suggested, for
foundation design purposes.
Settlement values for a shallow footing
with B=3.0m as a function q/qr and “C”
BEHAVIOR OF FOUNDATIONS UNDER
WORKING LOAD CONDITIONS

DÉCOURT, L.

XI PAMCSMGE

FOZ DE IGUAÇU, BRAZIL, 1999

 DÉCOURT PROPOSALS (1999)
log (q/qr) = C + C log (s/Beq)
or
log (q/qr) = C (1 + log s/Beq)

“C” is approximately constant for a given type of

mineral. For quartz sands, either sedimentary
or residual, the values of “C” are in the range
0.40 ± 15% and for calcareous sands in the
range 0.70 ± 15%.
.
Values of q/quc x s/Beq as a function of “C”
 FELLENIUS PROPOSALS (1999)
q/qr = (s/sr)e
The reference parameters qr and sr are the same
used in Décourt’s method.
The Fellenius “e” value is identical to the Décourt
“C” value. The range of the proposed “e” values
varies between 0.5 and 0.9, which is
approximately, the same range proposed by
Décourt (1999), for calcareous sands.
This is by no means surprising because Fellenius´s
analysis was based on load tests carried out in
calcareous sands.
FUNDAMENTALS OF THE
PROPOSED METHOD AND
SOME EVIDENCES
OF ITS CORRECTNESS
 VESIC (1963)
Vesic (1963) and other authors suggested that
for very dense sands, the stress-settlement
curve displays a peak and next, stresses start
declining with the displacements.
Load tests on model footings. Apud Vesic (1963)
For the densest sand, it seems that a rigid plastic
failure has happened, but, as a matter of fact it
didn’t. Besides, the extrapolation of small scale test
results to prototypes is not straightforward.
Vesic himself (1975), suggested that the results of
the model tests carried out by him could not be
used for real foundations. Unfortunately these
comments have shown up in footnotes, probably
not read by most engineers.
The principles explaining the behavior of the plate
tests and footings are provided by the so called
Critical State Soil Mechanics, Altaee and Fellenius
(1994). It is not the density, or the void ratio, alone
that determines if a soil will dilate or not.
Rather is the density or void ratio in relation to the
steady state line, that is the void ratio distance to the
steady state line and the slope of this line in a void
ratio vs. mean stress diagram. The latter is called the
“upsilon distance”.
According to , Altaee and Fellenius (1994), it is the
density of the model tests governs the density of the
soil for the prototype, which soil must always be
denser than the soil for the model. Therefore, if the
soil for the model is a dense sand, the prototype soil
would have to be a very, very dense sand, most
probably a rock, rather than a soil.
 The slides below show the results of load tests
carried out on a very dense sand and on a soft rock,
not showing the Vesic suggested behaviour

Stress x settlement curve for a Stress x settlement curve or a

footing, on a very dense sand R.J., footing, on a phyllitic soft rock, S.P.,
Brazil. Brazil.
UNICIDADE DA CURVA NORMALIZADA
CARGA (TENSÃO) - RECALQUE
Em 1999, afirmei que a curva tensão-recalque,
normalizada, era única, não dependendo da
densidade da areia, do tamanho da sapata e de
sua profundidade.
As tensões foram normalizadas pela tensão de
referência, qr, que é a tensão que corresponde a
um recalque de 10% do lado de uma sapata
quadrada e o recalque, pela largura da sapata
(ou diâmetro da estaca).
Todas as análises procedidas desde então,
confirmaram essa assertiva.
Stress x settlement curves for footings on sand, Normalized stress x settlement curves. Texas,
Texas, U.S.A. U.S.A.

Stress x settlement curves for footings on a lateritic Normalized stress x settlement curves. Bauru, S.P.,
sand at different depths, Bauru, S.P., Brazil Brazil.
O MÉTODO RDZ
Contrariamente ao procedimento usual para
dimensionamento de fundações que objetiva a
obtenção de uma tensão admissível, o Método
RDZ objetiva a obtenção de recalques diferenciais
zero, entre todas as sapatas. Como consequência
óbvia, as tensões a serem aplicadas ao solo pelas
sapatas terão que ser diferentes.
É muito importante salientar que os “imputs” a
serem utilizados são aqueles fornecidos por
ensaios de rotina, SPT, SPT-T e CPT.
SAPATAS ESTAQUEADAS (ESTACAS-T)
O método Décourt é usado não apenas para fundações
diretas. Ele também é usado para projetos de sapatas
estaqueadas (Estacas–T).
Tipicamente, uma estaca (ou mais) é colocada sob a
sapata no centro da coluna e sua capacidade de carga é
calculada. Para a estaca, sua capacidade é admitida como
sendo totalmente desenvolvida para o deslocamentos do
seu topo, de 15mm.
O procedimento proposto é o de simplesmente subtrair
da carga nominal da coluna o valor de 80% do
correspondente à capacidade calculadada da estaca e
prosseguir com o projeto, como se tratasse de uma
simples sapata, com essa carga reduzida.
 PROVA DE CARGA EM
SAPATA ESTAQUEADA

The Stiffness Plot for the

Piled-Footing

Load-settlement curve for the Piled-Footing

PREDICTED AND MEASURED BEHAVIOR
OF A TALL BUILDING IN A LATERITIC CLAY

DÉCOURT, L.; GROTTA JR., C.;

PENNA, A. S.; CAMPOS, G. C.

5th INTERNATIONAL CONFERENCE ON

SITE CHARACTERIZATION – ISC 5

QUEENSLAND, AUSTRALIA, 2016

SIMULATED LOADING TESTS ON BASIS OF
SETTLEMENT MEASUREMENTS
A couple of years ago, I developed a method for
turning conventional settlement measurements into
full scale loading tests. The program not only
simulates load test results, but also determines the,
eventually unknown, settlement values, at the date
the measurements began. Besides, it makes a
statistical adjustment of the measured values. The
next slide below shows the results of one of
these tests. In this particular case, the settlement
measurements began when the column loads were
about 2/3 of their nominal values.
Corrected load-settlement curves
FUNDAÇÕES
PROFUNDAS
CAPACIDADE DE CARGA DE ESTACAS
A PARTIR DOS VALORES DE SPT

DÉCOURT, L.; QUARESMA , A.R.

VI COBRAMSEF

RIO DE JANEIRO, BRASIL , 1978

PREDICTION OF THE BEARING CAPACITY
OF PILES BASED EXCLUSIVELY ON N
VALUES OF THE SPT

DÉCOURT, L.

ESOPT II

AMSTERDAM, THE NETHERLANDS , 1982

ULTIMATE BEARING CAPACITY OF LARGE
BORED PILES IN A HARD SÃO PAULO CLAY
DÉCOURT, L.

DE MELLO VOLUME. A TRIBUTE

TO PROF. VICTOR F. B. DE MELLO

SÃO PAULO, BRAZIL, 1989

 PREDICTION OF THE BEARING
CAPACITY OF PILES BASED
EXCLUSIVELY ON N VALUES OF THE SPT

DÉCOURT, L.

2nd INTERNATIONAL GEOTECHNICAL

SEMINAR ON DEEP FOUNDATIONS
ON BORED AND AUGER PILES

GHENT, BELGIUM , 1993

ON THE LOAD-SETTLEMENT
BEHAVIOR OF PILES

DÉCOURT, L.

SOILS AND ROCKS

3rd ODAIR GRILLO LECTURE

SÃO PAULO, BRAZIL, 1995

 PROJETO DE ESTACAS

DÉCOURT, L.

CAPÍTULO 8.1 - FUNDAÇÕES, TEORIA

E PRÁTICA – ABMS, ABEF

SÃO PAULO, BRASIL , (1996/2016)

PREVISÃO DE CAPACIDADE DE CARGA
DÉCOURT & QUARESMA

Qu (kN)= α qp Ap + β qs As

Qu (kN)= α K Np Ap + 10β (Ns/3+1)

 Valores do coeficiente K para a estaca de
referência em função do tipo de solo
qP (kPa) = KN60

Tipo de solo Valores de K (kPa)

Areias 333

Solos Intermediários 185

Argilas Saturadas 100

Valores do coeficiente α em função do tipo de
estaca e solo
* Esses valores são não mais que indicativos, devido a escassez de dados disponíveis.

Injetada
Escavada Escavada FDP
CFA Raíz sob altas
em geral (bentonita) (Ômega)
pressões

Areias ≤ ,30 ≤ 0,30 ≤ 0,30 0,45 0,50* 1,0*

Solos
Intermediários
≤ 0,40 ≤ 0,40 ≤ 0,40 0,45 0,50* 1,0*
Argilas
≤ 0,50 ≤ 0,50 ≤ 0,50 0,45 0,50* 1,0*
Saturadas
Valores do coeficiente β em função do tipo de estaca
e solo
•Esses valores são não mais que indicativos, devido a escassez de dados disponíveis.

Injetada
Escavada Escavada FDP
CFA Raíz sob altas
em geral (bentonita) (Ômega)
pressões

Areias 0,50 0,60 1,0 1,5* 1,50* 3,0*

Solos
Intermediários
0,65 0,75 1,0 1,5* 1,50* 3,0*

Argilas
0,80 0,90 1,0 1,5* 1,50* 3,0*
Saturadas
 CAPACIDADE DE CARGA DE
ESTACAS EXECUTADAS NO CAMPO
EXPERIMENTAL DA UNIVERSIDADE DE
LONDRINA. ALGUMAS PONDERAÇÕES
.
DÉCOURT, L.

XII COBRAMSEG

SÃO PAULO, BRASIL , 2002

 AS RELAÇÕES ENTRE Qs (PC) E
Qs (Décourt & Quaresma) FORAM:

1,71; 1,84; 2,33

NA MÉDIA

Qs (PC) / Qs (DQ) = 1,96 ≈ 2,0

DESENVOLVIMENTO DO ATRITO LATERAL
ENTRE ESTACA E SOLO COM BASE EM
PROVAS DE CARGA USANDO CÉLULAS
EXPANSIVAS HIDRODINÂMICAS

DÉCOURT, L.

XIII COBRAMSEG

CURITIBA, BRASIL , 2006

Procedi a um programa de pesquisa,
analisando os resultados de cerca de 200
ensaios bi direcionais. Parte dessa pesquisa foi
apresentada em Décourt (2006). Uma prova de
carga muito importante será a seguir
apresentada para ilustrar o método. Trata-se
de prova executada na Coréia do Sul pela
empresa norte americana Load Test, que na
época se constituiu em recorde mundial.
α
Um ponto importante a ser enfatizado é de que o
valor de qs não atinge um valor limite, para
deslocamentos de cerca de 10,0mm, como
admitido por muitos autores. Ao contrário, o atrito
lateral unitário, qs continua crescendo com as
deformações. Com a utilização do gráfico de
rigidez, fica fácil a avaliação da carga de ruptura
convencional, por atrito lateral (Qs)c assim
como, da carga correspondente à ruptura
física (Qs)u, mesmo nos casos onde o
carregamento foi interrompido prematuramente.
(Qs)10/(Qs)c em função de tgα
PROVAS DE CARGA EM ESTACAS PODEM
DIZER MUITO MAIS DO QUE TÊM DITO

DÉCOURT, L.

SEFE VI

SÃO PAULO, BRASIL , 2008

 LOADING TESTS: INTERPRETATION AND
PREDICTION OF THEIR RESULTS

DÉCOURT, L.

FROM RESEARCH TO PRACTICE IN GEOTECHNICAL

ENGINEERING.
ASCE, GSP 180

SYMPOSIUM HONORING TO
DR. JOHN H. SCHMERTMANN

NEW ORLEANS, U.S.A., 2008

EXTRAPOLATION OF LOAD
TESTS RESULTS
 THE DÉCOURT STIFFNESS METHOD
In 1996 I proposed a method for a better interpretation of
load test results. One of the most important features of
this method is the recognition that a few foundations may
fail, but most of them will never fail.

This procedure makes use of the so-called Stiffness Plot

and is known in Brazil as the Stiffness Method. Using load
test results, the stiffness is computed (stiffness is the load
or stress divided by the corresponding movement) and
plotted against the applied load. The physical failure (Qu)u
is defined as the load corresponding to zero stiffness.
Zero stiffness corresponds to infinite deformation, so it
cannot be reached in practice.
Two possibilities exist:

I - The stiffness decreases linearly with increasing loads.

The ultimate pile capacity (Qu)u is defined as the load for
which the extrapolated straight line intersects the
horizontal axis (abscissa).

In this paper, for both sands and clays the conventional

bearing capacity (Qu)c is considered as being the load
corresponding to a pile top movement, "s", of 10% of the
pile diameter, "d".

II - The stiffness decreases non-linearly with increasing

loads and only conventional capacity is defined (Qu)c.
Stiffness plot for a displacement pile
(After Décourt - 1999)
Stiffness plot for a shallow foundation
(after Décourt - 1999)
 PREDICTION OF THE
LOAD-SETTLEMENT CURVE

Points for predicting the load-settlement curve

COBRAMSEG 2010
PREDICTION EVENT
βL = 2
PROGRAMA PARA INTERPRETAÇÃO
DE PROVA DE CARGA BI-DIRECIONAL

ESTACA TIPO ESCAVADA

d = 1200mm
L = 17,3m + 3,0m
CASO A
CASO B
CASO C
 EVENTOS DE
PREVISÃO DE CAPACIDADE
DE CARGA E DA CURVA
CARGA-RECALQUE DE
FUNDAÇÕES DIRETAS
E PROFUNDAS
Entre os anos de 1982 e 2012, o autor
participou de 5 concursos de previsão de
capacidade de carga, a saber:

1º. Caso........................ESOPT II – Amsterdam.…………1982

2º. Caso…………...Texas A&M University – Texas…………1994
3º. Caso..............................SEFE VI – São Paulo.....……2004
4º. Caso.................XV COBRAMSEG – Gramado....….…2010
5º. Caso.......XVI COBRAMSEG – Porto de Galinhas...…2012
CLASS A PREDICTIONS AND BENEFITS
DERIVED FROM THEIR ANALYSES

DÉCOURT, L.

ISC - 4

PORTO DE GALINHAS, BRAZIL, 2012

THE ESOPT II PREDICTION EVENT

A prediction event was held in Amsterdam in

1982 during the Second European Symposium
on Penetration Tests ESOPT II.
The information here reported have been
supplied by the Delft Soil Mechanics
Laboratory, Heijnen and Janse (1985). The pile
had a square 0.25m cross section and length
of 15.0m.
Results of SPT– Amsterdam
LOAD TEST
The tested pile was a pre cast concrete pile with
0.25m square cross section and length of 15.0m.

During the 10th load increment the settlement rate

reached the value of 10mm per minute. At that
moment the load was approximately 1200 kN.

In accordance with the information supplied to the

participants this load was taken was the ultimate
bearing capacity QU of the pile.
Thus: QU = 1200 kN
Quoting Heijnen and Janse (1985).

“The best prediction of ultimate pile bearing capacity

QU out of the predictions obtained from countries
outside Holland, was presented by Prof. L. Décourt.
Prof. Décourt precasted QU = 1157 kN with basis on
the N30 - Values of the SPT. The method is described in
detail in the paper by Prof. Décourt for ESOPT II.”
During the 10th load increment the settlement rate reached the
value of 10mm per minute. At that moment the load was
approximately 1200kN.

The values of the conventional bearing capacity, Quc and of the

ultimate bearing capacity, Quu, were, respectively, 1050.3kN and
1246.5kN.
THE TEXAS A&M PREDICTION EVENT
A very important prediction event was held in
1994, at the campus of Texas A&M University.
The main organizers of this event were the
present President of the ISSMGE, Professor Jean
Louis Briaud and Mr. Robert Gibbens.
Load tests on five footings were carried out. All
footings were square, with widths of 1.0m;
1.5m; 2.5m and 3.0m (2 footings).
QUESTIONS
Two questions were asked by the organizers.
I – What would the stresses for a settlement of 25mm
be?
II – What would the stresses for a settlement of
150.0mm be?
The first question was related to the concept of
allowable stress.
The second question was related to the concept of
bearing capacity.
The organizers considered predictions to be correct in
the range of plus or minus 20% of the measured
values.
PREDICTIONS
A total of 31 predictions were received from 8
different countries, half from consultants and half
from academics. The main conclusion reached from
comparing the predictions and the measurements
was that nobody gave a complete set of answers
consistently falling within ± 20% of the measured
values. At the time, the author had not yet developed
the method of settlement predictions that he broadly
uses nowadays.
The author’s predictions were based on the values of
the maximum shear modulus G0 and their
correlations with results of cone penetration tests,
CPT and other in situ tests. This procedure was quite
laborious.
Anyway the ratios of the predicted stresses to the
measures ones were 0.916; 0.863; 0.761; 1.035
and 0.825 with an average of 0.88.
Therefore, four out of five predictions were in the
range of ± 20% of the measured values.
The question “II”, related to the bearing capacity
demonstrated clearly the incapacity of the author
to correctly predict the bearing capacity of footings
on sand.
 Average values of
Qpred/Qmeas

Footing s= 25mm s = 150mm

width (m)
1.0 0.71 0.65
1.5 0.84 0.81
2.5 0.68 0.99
3.0 (s) 0.70 1.12 (S)
3.0 (n) 0.69 1.08 (N)

Overall Average 0.724 ---------------

For 25mm, these ratios didn’t show any tendency of
variation with the foundation width. An average value
of 0.724 was computed. For 150mm settlement these
ratios increased linearly with the width of the
footings.

A statistical correlation can be established.

Qpred/Qmeas = 0.457 + 0.215B

R2 = 0.986

Therefore, the errors increase in direct proportion to

the width of the footings.
 Variation of Qpred/Qmeas with B

This means that averages (for the 31 predictors) of the

ratio Qpred/Qmeas of the footing increased directly with the
increase in the footing width, figure above.
It seems that all predictors believed that, for sands, the
bearing capacity increased linearly with the footing width.
The load test results present a clear evidence that
the stresses corresponding to the bearing capacity
does not increase with the footing size.
Therefore the reality was quite different from that
the predictors assumed.
Some months later, Décourt, (1994) proposed a
new method for predicting the full stress-
settlement curve. This new method, Décourt
(1994-1999) was developed after a deep analysis
of these loading test results.
The method uses a normalized stress-
settlement curve.
THE SEFE VI PREDICTION EVENT
As part of SEFE VI a prediction event was
held using data from the Foundation
Experimental Field of Unicamp.
A total of 11 predictions were received.
The goal was the prediction of the
ultimate capacities of four root piles and
their displacements for loads half of their
ultimate ones.
Informations regarding this event are
given in Carvalho et al (2004).
Assuming that the bonds are little affected or
even, not affected by the CPT, the ratio of
qc/NSPT could be useful for the assessment of βL.
Considering that for depths over of 10m the
ratio qc/NSPT is about 221 and for the upper
layer, it is 4.93, the ratio 4.93/2.21 = 2.23 could
be considered to be a lower bound limit for βL.
Considering that the depth of the lateritic layer
was about 8.0 to 10.0m and that βL should be
equal or larger than 2.23, it was decided that,
for prediction purposes, this depth was
assumed to be 8.0m, and that βL was equal to 3.
PILE CHARACTERISTICS
Four root piles have been tested, two under
compression and two under tension.
Their lengths were 12.0m and 23.0m and their
diameters, 410.0mm and 310.0mm.

Questions

The organizers asked the predictors to determine.

- The bearing capacity of the piles.
- The displacement of the piles under loads half of the
ultimate ones.
CAPACITY AND DISPLACEMENT PREDICTIONS
The author’s capacity predictions were made using the
Décourt and Quaresma (1978) formula, that usually
gives reasonable results, provided correct average Nspt
values are used.
As postulated by the author, Décourt (1996) the Nspt
values have to be increased by increasing factors, for
taking into account the lateritic characteristics of the
soil.
In terms of the lateral friction capacity, the increasing
factor βL was assumed to be 3.
Probably, the only predictor that recognized the
importance of identifying lateritic soils was the author,
(predictor nº 6).
Although 8 out 11 of the predictors used the Décourt and
Quaresma (1978) method, the non consideration of βL
was one of the major reasons for their, not so successful
predictions.
In the next figure, the ratios of measured to predicted
displacement values ,δpred/δmeas made by the 11
participants are presented.
The author, predictor nº 6, wrote a single paragraph and
predicted a value 1.45 times the measured one. Predictor
nº 8, probably an academic, wrote 36 pages, used
practically all available methods, including the finite
elements method, and predicted a value 150 times the
measured one.
Values given by 11 predictors
This may seem strange. But, as matter of fact, it is not
necessarily a surprise.
Poulos (1999) pointed out that “It is axiomatic that an
essential ingredient for successful settlement
prediction is the selection of appropriate geotechnical
parameters. It is the author’s experience that
settlement predictions are far more sensitive to the
geotechnical parameters and site characterization
than to the method of analysis (e.g. Poulos, 1989).”

Besides, as already mentioned – “No prediction can

be more reliable than the test data used in the
computations.”
THE XV COMBRAMSEG PREDICTION EVENT
The fourth case is a prediction event held during the
XV Brazilian Conference on Soil Mechanics and
Geotechnical Engineering, Gramado, 2010.
Only a few weeks before the event the organizers
announced that a prize would be offered to the
predictor that provided, not only the correct value of
the ultimate bearing capacity of a pile, but also the
correct load-settlement curve.
The results of a SPT-T boring carried out close to the
load test site were provided. The SPT-T is the
traditional SPT complemented with torque
measurements. No other information was given, not
even the place where the load test was carried out.

The SPT-T test was carefully analyzed by the author,

focusing on both Nspt values and torque
measurements, T.

The agreement between these two types of data

(Nspt and T) convinced the author that the SPT-T had
been properly performed.
However, a few minutes before the presentation of
the "winner" of the contest, the author was advised
that the loading test had been carried out in Londrina,
Paraná State, a region where the soils are lateritic,
Cesar (2010)a.
Décourt (1994) had demonstrated that for lateritic
soils, the SPT/SPT-T fails to recognized the stiffness of
such soils. The NSPT and T values resulted too low.
The actual bearing capacity and the stiffness were
many folds higher than, the SPT based, predicted
ones. In another words, both the lateral friction and
the end bearing, computed with basis on SPT/SPT-T
have to be multiplied by increasing coefficients ∝L and
βL, Décourt (2002).
For Londrina soils, the values of βL required to make
Nspt based predictions to match load test results
varies from 1.8 to 2.2, with an average value of about
two, Décourt (2002).
The reason why for these soils SPT yields to
unrealistic values of soil resistance, was suggested by
Décourt (1994).
Since nobody knew were the pile load test was
carried out and the organizers of this event didn’t
provide any information regarding their lateritic
characteristics, it became obvious that, all predictions
made, were deemed to be big failures.
βL = 2
5º. CASO - XVI COBRAMSEG
Tema: previsão da capacidade de carga de duas
estacas, uma do tipo Hélice Contínua (CFA) e outra do
tipo Ômega (FDP), L=15,0m e d= 500,0mm

Estaca Hélice Contínua (CFA) Estaca Omega (FDP)

Qs=81,64tf (D&Q) Qs=107,25tf (D&Q)

Qs=61,55tf(monitoramento) Qs=122,0tf (monitoramento)
Qs(tf)=71,5 ± 10,1 ou ± 14% Qs(tf)=114,6 ± 7,4 ou ± 6,5%
Qs≤73,1tf (PC) Qs≤116,6tf (PC)
Qp≥84,59tf (PC) Qs≥94,2tf (PC)
CONCLUSÕES RELATIVAS AOS CINCO CASOS

1 – As previsões de capacidade de carga das 8 estacas, efetuadas com

base no método Décourt & Quaresma e dados de solo adequados,
conduziram a resultados bastante satisfatórios (casos 1; 3; 4; 5).

2 – As previsões de carga, para recalque de 25,0mm, deram

resultados satisfatórios para as 5 sapatas.

3 – As previsões de carga, para recalque de 150,0mm foram

totalmente insatisfatórias, ,demonstrando, claramente, que os
procedimentos de cálculo utilizados eram inadequados.

4 – Deve se enfatizar a imperiosa necessidade de se reconhecer que

solos lateríticos não seguem o que foi estabelecido para solos não
lateríticos.
MUITO OBRIGADO