Professional Documents
Culture Documents
ABSTRACT: Virtual Calibration Chambers (VCC) are 3D-DEM models where cone penetration is simulated.
They offer the possibility of substituting for the rather more expensive physical calibration chambers. One
important issue in calibration chamber interpretation is the possibility of chamber size effect under different
boundary conditions. The aim of this study was to explore chamber size effects under BC1 and BC3 conditions
on the VCC. The chamber and cone diameters and boundary conditions were varied to investigate their effects
on cone tip resistance. The scaling laws that are obtained from the VCC results are then compared with those
previously obtained with physical calibration chambers.
Calibration Chamber tests (CC) provide an effective Top & Bottom Boundary Lateral Boundary
way to study Cone Penetration Test (CPT) in sands BC Stress strain stress strain
under strictly controlled conditions (material, density,
BC1 constant – constant –
stress state and boundary). Results from CC tests are
BC2 – 0 – 0
used to establish relationships between the observed BC3 constant – – 0
outcomes like cone resistance (qc ), and the material BC4 – 0 constant –
descriptors like relative density (DR ) and effective
stress state (σ’). Typical boundary conditions applied
in CC differ on whether stresses are kept constant or remarked that also CC CPT’s on loose sands were
displacements are zero on the lateral and top/bottom affected by size effects, but up to a lower limit of
sample boundaries (Table 1). only Rd > 30 − 35. They noted that for the same Rd , qc
depends on the applied boundary conditions, but did
not quantify that effect.
1.1 Chamber size effect Schnaid & Houlsby (1991), using only BC1 tests,
confirmed that for all sand densities the chamber
The use of calibration chamber results to predict field size can affect the results and that the effect is more
performance needs to take into account the size effect significant for dense sand. They also showed that
due to limited chamber dimensions. Size effects are cone resistance and pressuremeter limit pressure were
observed when the test outcomes vary for constant similarly affected. Mayne & Kulhawy (1991) after
sand properties and conditions as a function of equip- examining six data sets from CC CPT’s for differ-
ment dimensions. Size effect is usually explored using ent ranges of Rd proposed a correction factor for size
the parameter Rd , a chamber to cone diameter ratio. effects:
A related aspect is the influence of variable testing
boundary conditions and how these may result in
different size effects.
Since the early 80’s, these issues have been exam-
ined by various researchers. Parkin & Lunne (1982) where qc,corrected = corrected cone tip resistance;
summarized work performed in two different calibra- qc,measured = measured cone tip resistance; Rd = chamber
tion chambers with two differently sized penetrom- – to - cone diameter ratio and DR = relative density.
eters. They did not observe a significant size effect Equation 1 assumes that a ‘free field’ condition is
for loose sand, whereas for dense sand there was a achieved for Rd > 70 and was meant to apply equally
clear influence of chamber size up to Rd > 50. Cone for both BC1 and BC3 conditions.
tip resistance (qc ) increased with Rd for both BC1 A different empirical correction for size effect was
and BC3 conditions, although somewhat faster in the proposed by Tanizawa (1992) while analyzing CC CPT
former case. In later work, Parkin (1988) suggested results on Toyoura sand:
that the Rd value required to eliminate chamber size
effects might be greater then 70. Ghionna et al. (1991)
225
where a & b = f (Rd ) and DR = relative density. The
proposed correction factor does only apply for BC1
conditions (Garizio, 1997).
Salgado et al. (1998) applied a mixture of cavity
expansion and slip line theory to quantify chamber
size effect. The theory predicts that the difference
between ‘free field’ and chamber qc values increases
for decreasing Rd and that qc measured under BC1 or
BC4 conditions is always smaller than the correspond-
ing ‘free field’ values. The difference (or correction
factor) is not only dependent on density, but also on
ambient stress and material parameters. The opposite
held for qc measured under BC3 or BC2 conditions; the
qc value predicted decreased with increasing Rd , and
thus CC results should lie above “free field” values. Figure 1. Grain size distribution of Ticino sand and DEM
That result was contrary to some available experimen- models.
tal evidence, but the discrepancy was attributed to
experimental imperfections in the enforcement of the
no lateral strain condition.
226
Figure 3. View of the DEM model components with indi-
cation of the main relevant dimensions (left) calibration
chamber (right) cone device.
Dcc H dc
Test Series m m mm
227
Figure 5. Tip resistance profiles for different cone tip sizes Figure 7. Tip resistance profiles for different chamber sizes
for p0 = 100 kPa and DR = 75%. for p0 = 100 kPa and DR = 90%.
228
density of 94%, vertical stress of 122 kPa and hori-
zontal stress slightly variable between 42 and 46 kPa.
Results from this simulation series are presented in the
next section.
229
despite the major simplifications in particle size dis-
tribution, particle behavior and model construction
required to obtain practical results using only desktop
computers.
REFERENCES
Arroyo, M. Butlanska, J., Gens, A., Calvetti, F. &
Jamiolkowski, M. 2009. Cone penetration tests in a virtual
calibration chamber (under review).
Bellotti, R. (1985) Chamber size effects and boundary con-
ditions effects, Seminar of cone penetration testing in the
laboratory, University of Southampton, pp 26–29.
Butlanska, J., Arroyo, M. & Gens, A. 2009. Homogeneity and
Figure 12. Experimental and DEM results for CPT in a CC symmetry in DEM models of cone penetration. Powders &
with no radial strain boundary condition. All samples with Grains, vol 1145: 425–428.
DR 92–95% and initial horizontal stress 42–50 kPa. Calvetti, F. & Nova, R. 2005. Micro – macro relationships
from DEM simulated element and in – situ test. Powders &
Grains, vol. II: 245–250.
Cundall, P.A. & Strack, O.D.L. 1979. A discrete numeri-
cal model for granula assemblies. Geotechnique 29(1):
47–65.
Garizio, G.M. (1997) Determinazione dei parametri geotec-
nicie in particolare di K0 da prove penetrometriche, Tesi
di Laurea, Politecnico di Torino.
Ghionna, V.N. & Jamiolkowski, M. 1991. Acritical appraisal
of calibration chamber testing of sands, Calibration cham-
ber testing, A. – B. Huang Editor.
Jamiolkowski, M., Lo Presti, D.C.F. & Manassero, M. 2003.
Evaluation of relative density and shear strength of sands
from CPT and DMT, in Germaine, Sheahan & Whit-
man, Soil behavior and soft ground construction, ASCE
Geotechnical Special Publication 119: 201–238
Jiang, M.J., Yu, H.-S. & Harris, D. 2006. Discrete element
modeling of deep penetration in granular soils, Interna-
tional Journal for Numerical and Analytical Methods in
Figure 13. Horizontal section of normalized mean stress Geomechanics, 30, 335–336.
intensity at the chamber top after anisotropic consolidation Ma, M.Y. 2004. A numerical study of cone penetration test
of a VCC model with Rd = 33. in granular assemblies. PhD Thesis, Clarkson University.
Mayne, P.W. & Kulhawy, H. 1991. Calibration chamber
The numerical tests are all in Rd range below that database and boundary effects correction for CPT data.
of the physical tests. Increasing Rd in the numerical Calibration chamber testing, A. – B. Huang Editor.
tests to overlap the physical range it is both numer- Parkin, A.K. & Lunne, T. 1982. Boundary effects in labora-
ically costly and prone to mistakes. For instance we tory calibration of a cone penetrometer for sand, Proceed-
ings of the Second European Symposium on Penetration
have noticed that the usual procedure of numerical cal-
Testing, Amsterdam.
ibration chamber preparation results in strong stress Parkin, A.K. 1998, Calibration of cone penetrometers, Pro-
inhomogeneity during 1D compression in large cham- ceedings of the First International Symposium on Pene-
bers. Figure 13 illustrates this issue for a chamber tration Testing, Orlando.
with Rd = 33. Circumferential arching between the Peterson, R.W. 1988. Laboratory investigation of the pene-
two radial walls leaves understressed the near – axis tration resistance of fine cohesionless materials. Proceed-
zone, where the CPT penetration takes place. ings of the First International Symposium on Penetration
Testing, Orlando: 895–901.
Salgado, R., Mitchell, J.K. & Jamiolkowski, M. 1998.
Calibration chamber size effecs on penetration resistance
5 CONCLUSIONS in sand. Journal of Geotechnical and Geoenvironmental-
Engineering, ASCE, 124(9): 878–888.
The use of 3D-DEM models offers an interesting Schnaid, F. & Houlsby, G.T. 1991. An assessment of cham-
avenue to explore large-deformation problems in gran- ber size effects in the calibration od in situ tests in sand.
ular materials, such as CPT in virtual calibration Geotechnique 41(3): 437–445.
chambers. Due to necessary scaling of mean particle Tanizawa, F. 1992. Correlations between cone resistance and
size another size effect, due to high particle size to mechanical properties of uniform clean sand. Internal
cone diameter ratio, appears in the virtual calibration Report. ENEL-CRIS, Milano.
chamber. This effect is shown to be easily dealt with by
filtering the penetration resistance curve. Remarkable
quantitative agreement with physical tests is obtained
230