Seismic downhole, CPT, and DMT correlations in sand

( \ Proceedings: Second International Conference on Recent Advances in Geotechnical Earthquake Engineering and Soil Dynamics,
March 11-15, 1991, St. louis, Missouri, Paper No. 1.37
~
Seismic Downhole, CPT, and DMT Correlations in Sand
Thomas G. Thomann
Roman D. Hryciw
Staff Engineer, Woodward-Clyde Consultants, Wayne, NJ
Assistant Professor of Civil Engineering, University of Michigan,
Ann Arbor, Ml
Results from seismic cone penetration and dilatometer tests performed in a cohesionless
SYNOPSIS:
soil are presented and compared with previously published correlations for estimating the elastic
Poor correlations were obtained between the elastic shear modulus, the dilatometer
shear modulus.
modulus and the cone tip resistance; however, somewhat better estimates of the elastic shear modulus
were obtained using the coefficient of lateral stress and total unit weight empirically determined
from the dilatometer.
with
decreases
Rg,
termed
commonly
Baldi, et.al.
relative density.
(1989) developed a correlation between Gelqc and
qcl (Ovo'lo.s based on CPT calibration chamber
tests and resonant column and cross hole tests.
E0 ,
INTRODUCTION
~ncreasing
The elastic shear modulus <eel of cohesionless
soils is primarily a function of the void ratio
(e) and the effective confining stresses in the
and wave
(CJi 'l
direction of particle motion
variables which can be
(CJ j ')
propagation
and
(CPT)
determined from cone penetration
Therefore, empirical
dilatometer (DMT) tests.
between Ge and penetration tests
~orrelations
costly
to
alternative
viable
a
offer
may
when
tests
seismic
in-situ
or
laboratory
estimates of ce are required.
Empirical correlations for determining Ge of
cohesionless soils from the void ratio and
confining stress have existed for nearly three
The original equation proposed by
decades.
Hardin and Richart (1963), has undergone minor
The most
modifications since its inception.
recent version by Hardin and Blandford (198 9)
takes the form
Correlations between penetration tests and c;e
Jamiolkowski,
have previously been developed.
et.al. (1988) related ce to the DMT modulus, E 0 ,
and Baldi, et .al. (1989) related c;e to cone tip
Both approaches were based
resistance, qc.
sands.
in
tests
laboratory
on
primarily
However, in-situ tests have often revealed the
in
correlations
these
of
shortcomings
1989;
(Sully and Campanella,
predicting Ge
Better predictions may be
1990).
Hryciw,
obtained when the coefficient of lateral stress
(K0 ) and the total unit weight <Ytl from the DMT
are used in an empirical correlation for ce
The present paper compares the
(Hryciw, 1990).
results from seismic cone penetration tests
(SCPT) and DMT tests performed at a site in
northern Michigan with the approaches suggested
by Jamiolkowski, et.al. (1988), Baldi, et.al.
(1989), and Hryciw (1990).
celj
(1)
=
F(e)
2 (l+V)
BACKGROUND
is the
(overconsolidation ratio)
where OCR
ratio of the maximum historical mean effective
stress to the current mean effective principal
stress, k is a function of the plasticity index
(PI) (k = 0 for cohesionless soils), Sjj is the
stiffness coefficient in the ij plan(:!, Pa is
the atmospheric pressure, F(e) is a v•)id ratio
= 0.3 + 0.7(e)2, and v is the
function
In seismic down hole testing,
Poisson's ratio.
cri' is the effective horizontal stress (oh' l
and oj' is the effective vertical stress <crv'l.
Hardin and Blandford (1989) suggest using s 1 j =
The obvious
0.1.
0.5, and v
1400, n
disadvantage of equation (1) is that not only
must the void ratio be known but the vertical
as well as the horizontal stresses must be
evaluated.
DMT calibration chamber tests, resonant column
tests and in-situ cross hole tests performed in
Ticino sand and Po River sand by Jamiolkowski,
et.al. (1988) indicate that the ratio of ce to
Recognizing that the DMT offered the ability to
determine the soil density as well as the
horizontal stress, Hryciw (1990) developed an
empirical equation similar to equation (1).
97
to a depth of 21.3 m.
The water table was
never
encountered
augering
or
during
penetration
testing;
the
however,
moisture
content was approximately 4%.
The
model
is
based on
results
from
nine
different test programs where DMT and in-situ
shear wave measurements were performed.
The
soils
ranged
from
clays
with
dilatometer
indices (ID) of 0.1 to sands with ID's as high
as 8.
The proposed equation takes the form:
Two sets of SCPT and DMT tests were performed
at a distance of 3.35 m apart.
At each
location, the SCPT and DMT tests were performed
approximately 0.61 m away from each other.
The
qc, ED, and DMT horizontal stress index
(KD)
from the two sets of tests are shown in Figure
2.
The friction ratios from the CPT tests were
generally less than 1%.
ID ranged from 2 to 4.
Complete results of the SCPT and DMT tests are
given by Thomann (1990).
530
(2)
where Yt = DMT-based total unit weight and Yw
unit weight of water.
Hryciw (1990) also showed that when a range of
soil
types
is
considered,
very
little
correlation
exists
between
Rg
and
the
horizontal
stress
index
(KD)
of
the
DMT.
Hryciw postulated that the poor correlations
may be due to differences in the strain level
when measuring ~ and ED.
DETERMINATION OF K0 AND DR
Comparisons between the results of this study
and
the
previously
mentioned
correlations
require determination of the relative density,
dry unit weight, and confining stresses.
These
values
were
determined
by
empirical
correlations developed for the SCPT and DMT.
SITE CHARACTERISTICS
Marchetti and Crapps
(1981)
found that the
density
increased
as
ED
increased.
They
developed an empirical chart relating ED and ID
to the soi 1 type and density.
This chart was
used for determining YD and av' at the Douglas
Lake site for use in equation (2)
The
in-situ
tests
were
performed
at
the
University of Michigan Biological Station in
Pellston,
Michigan.
The
soil
conditions
consisted of a medium to fine,
subangular,
light brown sand.
The grain size distribution
is shown in Figure 1.
The Unified Soil
Classification was SP
poorly graded sand.
The specific gravity
(G8 )
was 2.65 and the
maximum and minimum void ratios (emax and eminl
were 0.82 and 0.54, respectively.
SAND
Several relationships exist for determining K
0
of cohesionless soils from results of the DMT
test.
Schmertmann (1983) developed a semiempirical equation for estimating K0 based on KD
and the thrust necessary to drive the DMT
blade.
Extensive calibration chamber tests
performed by Baldi, et.al. (1986) found that K
0
for natural, predominantly quartz, uncemented
sand was best predicted by using :
SILT
Fine
100
-1-l
..c:
0'>
-~ 80
3::
0.376 + 0.095 KD- 0.00461 qclav'
(3)
>.
..0 60
The K 0 values at the Douglas Lake site,
as
determined by equation (3), are shown in Figure
3.
'c:"'
(!)
-.-l 40
rx..
-1-l
coet!. of Uniformity -C 11
c:
(!)
u
20
coeff. of Gradation- Ce
Uni!ied Classification -
-2.42
Results
from
228 calibration chamber tests
performed by Jamiolkowski,
et.al.
(1988)
on
both normally consolidated and overconsolidated
Ticino
Sand
samples
revealed
a
strong
relationship
between
Dr,
qc,
and the
mean
effective confining stress (CJ 0 ' ) :
- 1.04
SP
'"'
(!)
p..
1
0.5
0.2
Grain Size
Figure 1.
0.1
0-05
0.02
(mm)
Grain Size Distribution of Douglas
Lake Sand.
1
-----2.93
The grain size distribution of the soil, as
observed
from
several
augered
holes,
was
consistent to a depth of 9. 1 m.
One CPT test
outside the study area indicated similar soil
ln
(4)
The relative density at the Douglas Lake site
is shown in Figure 3.
The results of equations
(3)
and (4)
were used to estimate Ge via
98
(bars)
(bars)
60 0
40 0
0
8
12
16
20
0
Douglas Lake
Sand
1
2
8
3
..C4
'
-
.j...)
0..
())5
0
6
7
8
Figure 2.
Ev, and Kv from SCPT and DMT tests in
Douglas Lake Sand.
Dr
Ka
0 0.2 0.4 0.6 0.8
qc,
1.2 1.4
0
0.2
0. 4
0. 6
The empirical total unit weights
equation (1).
from the DMT were not used for determining Dr
because they were found to be unreasonable .
0.8
OrT~~~~~-r~~~r~~~~~~~~~
..................
Test il
~
1
Essentially no correlation was found between Rg
This figure also
and Dr, as shown in Figure 4.
from Sully and Campanella
results
contains
(1989) which also do not coincide with results
Very little
of Jamiolkowski , et.al. (1988).
relationship was also found between the Douglas
Lake results and the correlation proposed by
Baldi, et.al. (1989) (Figure 5) .
Test t2
2
8'3
...
.C4
the
to
be due
correlations may
The poor
difference in strain level between seismic and
The strains associated with
penetration tests.
G€ are much lower than those imposed when
determining E 0 or measuring qc.
.1-l
p.
QJ
0
5
6
(1) and (2) yield somewhat better
Equations
Although both
estimates of Ge (Figure 6) .
G€,
for
values
similar
predict
equations
equation (2) does not require CPT, emax and emin
Therefore, the DMTtests to be performed,.
based approach for determining Ge is more
desirable.
Douglas Lake
Sand
Figure 3.
CORRELATIONS FROM DMT AND SCPT RESULTS
K0
and Dr from equations
in Douglas Lake Sand.
(3) and (4,)
99
I
5
Ge (bars)
0
500
1000
1500
Ge (bars)
2000
0
4.5
2500
0
200
400
600
800
1000 1200
4
fSD
3.5
Ge
1 ..
.
3
ED
2.5
.
2
1
I
•
. .
.
.
1.5
!
-
20
60
40
I
I
I•
(Jamiolkowski ,
DMT in Laing Bridge South Sand
(Sully and Campenella, 1989)
•
DMT Test #1 in Douglas Lake Sand
A
DMT Test #2 in Douglas Lake Sand
Figure 4.
r~sults-Test
t2 ]
Estimated G€ by equations (1) and (2).
CONCLUSIONS
in CC with Ticino Sand (Jamiolkowski , et.al., 1988)
~
SCPT
(%)
Mean and Standard Deviation of DMT
~ DMT in Po River Sand
(2)
100
80
Relative Density
(1)
I
Figure 6.
0
Equation
--- Equar_lon
et.al.,
The results from SCPT and DMT tests performed
in a
cohesionle ss soil were compared with
previously
published
correlatio ns
for
estimating G€.
Poor correlatio ns were obtained
when equations relating G€ to Ev and qc were
used.
The poor correlatio ns may be due to
difference s in strain magnitude between seismic
and penetratio n tests.
The most accurate and
efficient method for empiricall y determinin g G€
from penetratio n tests appears to be one that
utilizes the DMT-based Yt and K0 •
1988)
Correlatio n between Rg and Dr
ACKNOWLEDGMENTS
Funding for this research was provided through
NSF Grants CES-880713 4 and CES-870874 7.
The
authors
wish
to
thank
the
University
of
Michigan Biological Departmen t for use of their
facilities at the Biologica l Station.
Thanks
to Mr. Ing-Tsang Hoe for helping perform the
in-situ tests.
24
22
20
.
Correlation proposed by
Baldi, et.al., 1989 .
~~ lo
18
.
. .
16
Ge
14
12
qc
10
•
..
REFERENCES
••
Baldi, G., Bellotti, R., Ghionna, v., Jamiolkows ki,
M., Marchetti, S. and Pasqualini, E. (1986).
"Flat
Dilatomete r
Tests
in
Calibration
Chambers",
In Proceeding s of In-Situ
'86,
pages 431-446,
Geotech.
Eng.
Div.,
ASCE,
Blacksburg, VA.
8
6
Douglas Lake Sand
•Test fl
2
.&Test 12
0
200
400
300
Figure 5.
soo
600
1000
800
Baldi, G., Bellotti, R., Ghionna, V., Jamiolkows ki,
M., Lo Presti, D.C.F.
(1989). "Modulus of
Sands from CPT' s and DMT • s", In Proc. of the
12th Int.
Conf. on Soil Mech. and Found.
Engin.,
Vol. 1, pages
165-170, Rio de
Janeiro, Brazil.
1400 1800
1200
1600 2000
Correlation between G€ and qc.
100
G.E.,
Blandford,
and
B.O.
Hardin,
"Elasticity of Particulate Materials",
of
Geotechnical
Engineering
(1989) .
Marchetti,
and
S.
Dilatometer
Journal
Division, ASCE,
Flat
( 1981) .
D.K.
Crapps,
GPE,
Gainesville,
Inc.,
Manual,
FL.
115 ( 6) :·788-805.
B.O.
Hardin,
"Elastic
and
Wave
F.E.,
Jr.,
Richart,
Velocities in Granular
Journal
ASCE.
DMT
Digest
411B,
GPE
Foundation
(1989).
R.G.
Campanella,
and
J.P.
Sully,
"Correlation of Maximum Shear Modulus with DMT
Test Results in Sand", In Proceedings of the
"Small Strain Shear Modulus of
to be published in November
Soil
on
Conference
International
12th
Mechanics and Foundation Engineering, Volume
and
Mechanics
of Soil
Journal
Engineering, ASCE, 89(SM1) :33-65.
Hryciw, R.D. (1990).
Soil by DMT",
(1983).
J.H.,
Schmertmann,
Inc., Gainesville, FL.
(1963).
Soils",
of Geotechnical Engineering Division,
I,
Thomann,
pages 339-343, Rio de Janeiro, Brazil.
and Strength
Stiffness
(1990).
in Cohesionless Soils due to Stress
Ph.D.
Disturbance,
Dynamic
and
T.G.,
Changes
History
Lancellotta, R.
Ghionna, V.N.,
Jamiolkowski, M.,
and Pasqualini, E. (1988). "New Correlations
of Penetration Tests for Design Practice", In
Dissertation, Uni~ersity of Michigan.
International
First
the
of
Proceedings
ISOPT-1,
Symposium on Penetration Testing,
pages 263-295, Orlando, FL.
101

Download Report