2018-Zhang-Penetration Test DEM

DEM Investigation of Particle Crushing Effects
on Static and Dynamic Penetration Tests
Ningning Zhang1(&), Marcos Arroyo1, Matteo Ciantia2,

and Antonio Gens1
1
Department of Geotechnical Engineering and Geosciences,
Polytechnic University of Catalonia (UPC), Barcelona, Spain
[email protected]
2
School of Science and Engineering, University of Dundee, Dundee, UK
Abstract. A 3-dimensional discrete element method model has been developed

to simulate static and dynamic rod penetration test in a calibration chamber. The
chamber has been filled with a scaled analogue of Fontainebleau sand. Crushing
effects on penetration results have been examined. It has been found that particle
crushing reduces penetration resistance for both static and dynamic tests.
Microscale observation of crushed particles has been conducted. It is shown that
dynamic impact causes more crushing events. The crushed particles are dis-
tributed within 2–3 radius from the rod for both tests.
Keywords: Dynamic penetration test Static penetration test Crushing
1 Introduction
In geotechnical engineering, penetration tests have been widely applied to characterize

soils, such as estimating soil strength and stiffness and to evaluate soil liquefaction
potential. Considering different penetration principles, penetrometers can mainly be
divided into two groups: static and dynamic. In static penetration test, a tip is pushed
into soil at a constant speed. In contrast, dynamic penetration involves driving a device
formed by rods or rigid tips into the soil by striking it with a weight. When a rod is
driven into a granular material, grain can be broken due to stress concentration around
the rod. Discrete Element Method (DEM) has been applied as a powerful tool to
address particle crushing in penetration tests (Ciantia et al. 2016a; Lobo-Guerrero and
Vallejo 2007).
It is interesting to investigate the influence of crushing on penetration results under
the two different penetration principles. In this paper, after briefly introducing the
employed methodology for modelling particle crushing, simulations of a rod driven
statically and dynamically into a 3-dimensional sample of Fontainebleau sand are
presented. Particular attention has been paid to the effect of particle crushing on
evolution of rod tip resistance. Microscale observation of crushed particles is also
presented.
© Springer Nature Switzerland AG 2018

W. Wu and H.-S. Yu (Eds.): Proceedings of China-Europe Conference
on Geotechnical Engineering, SSGG, pp. 274–278, 2018.
https://doi.org/10.1007/978-3-319-97112-4_62
DEM Investigation of Particle Crushing Effects 275
2 Methodology
2.1 Particle Crushing Model
The modelling approach for particle crushing developed by Ciantia et al. (2015) was
adopted. For completeness, a general description of the model is given here. Coulomb
friction and the simplified Hertz-Mindlin model were used for contacts. The crushing
limit is based on the formulation proposed by Russell et al. (2009). Breakage of a given
particle is activated if the maximum contact force F reaches the limit condition:
F rlim AF )
8 " 9
< #2=3 =3
d 3=m 3 1 m21 1 m22 1 1 1 ð1Þ
F rlim 0 f ðvarÞ p þ þ
: Nd0 4 E1 E2 r1 r2 ;
Where rlim is the limit strength of the material and AF is the contact area. f(var) is a
function used to incorporate the natural material variability into the model. The limit
strength, rlim, is assumed to be normally distributed for a given sphere size. The
coefficient of variation of the distribution, var, is taken to be a material parameter. The
mean strength value rlim0 depends on the particle diameter (d) where m is a material
constant, d0 is the reference diameter and N is the scaling factor. ri, Ei and vi (i = 1, 2)
are the radii, Young’s Moduli and Poisson’s ratio of the contacting spheres, respec-
tively. Once the limit condition is reached, the spherical particle will split into 14-ball
inscribed tangent spheres. Mass loss issues related to the choice of using such
replacement technique are clearly addressed by Ciantia et al. (2016b). The crushed
fragments assume the velocity and material parameters of the original particle.
2.2 Model Construction and Simulation

The DEM code PFC3D (Itasca 2016) has been used to construct the numerical models.
The construction of a 3-dimensional calibration chamber model was carried out by
adapting the procedure described by Arroyo et al. (2011) (Fig. 1a). The chamber was
filled with an analogue of Fontainebleau dense sand, which was previously tested by
Ciantia et al. (2015, 2018). A scaling factor of 79 was applied in order to achieve a
manageable number of particles. Isotropic compression to 100 kPa was performed by
using wall servo-control.
The flat-ended rod was then created by using frictional rigid walls (Fig. 1a). Par-
ticle rotation was inhibited to roughly mimic the effect of non-spherical particle shapes.
Table 1 shows the model parameters of Fontainebleau sand. A local non-viscous
damping of 0.05 was employed. In static tests, the rod was driven at a rate of 40 cm/s.
As described by Zhang et al. (2018), dynamic penetration was simulated by prescribing
a time-dependent force on the rod. In both tests, particle crushing effects were
examined.
276 N. Zhang et al.
Fig. 1. (a) View of DEM model of calibration chamber and rod. Crushing effects on
penetrograms (b) static test (raw and adjusted curves), (c) dynamic test
Table 1. Model parameters for Fontainebleau sand.

µ/- E/GPa v/- rlim,0/GPa d0/mm m/- var/-
0.275 9 0.2 1.9 2 10 0.36
3 Outcomes
3.1 Penetration Curves
Penetrograms of static tests for crushable and uncrushable sand exhibit large oscilla-
tions, as shown in Fig. 1b, due to particle size effect. The approach which has been
successfully used by Arroyo et al. (2011) to filter out the noise is adopted. The adjusted
curve of penetration resistance is also shown in Fig. 1b. It is noted that crushing causes
a resistance reduction due to particle rearrangement and finer particle generation.
Additionally, one single impact was conducted at the depth of 15 cm (Fig. 1c). It
shows that during the rising phase of dynamic resistance, particles haven’t reached
limit condition. Afterwards, crushing takes place at the steady-stage of penetration
curve, leading to a reduction of resistance value and an increase of penetration depth.
3.2 Microscale Observations

The rod was driven from 15 cm to 30 cm, then a direct comparison can be made on
crushed particle numbers (Fig. 2a). It is clear that the dynamic impact breaks more
particles. DEM-based model can provide microscale observation at particle or contact
level to provide meaningful insights. It is interesting to visualize the spatial distribution
of crushing events that occur in both static and dynamic penetration tests. Figure 2b, c
show the location of crushed particles in both tests, where it can be clearly observed
that crushing events are distributed within 2–3 radius from the cone.
DEM Investigation of Particle Crushing Effects 277
Fig. 2. Penetration tests from the depth of 15–30 cm (a) crushed particle number, (b) crushed
particle distribution in static test and (c) crushed particle distribution in dynamic test
4 Conclusions
A 3-dimensional DEM model has been created to simulate static and dynamic rod
penetration into calibration chambers filled with a scaled analogue of Fontainebleau
sand. Crushing effects on penetration results has been examined. Particle crushing
causes a reduction of penetration resistance for both static and dynamic tests. Dynamic
impact causes more crushing events. The crushed particles are distributed within 2–3
radius from the rod. In forthcoming research, a more comprehensive study will be
conducted to investigate other penetration test factors, such as sample relative density
and confining pressure.
References
Arroyo, M., Butlanska, J., Gens, A., Calvetti, F., Jamiolkowski, M.: Cone penetration tests in a
virtual calibration chamber. Géotechnique 61(6), 525–531 (2011)
Ciantia, M.O., Arroyo, M., Calvetti, F., Gens, A.: An approach to enhance efficiency of DEM
modelling of soils with crushable grains. Geótechnique 65(2), 91–110 (2015)
Ciantia, M.O., Arroyo, M., O’Sullivan, C., Gens, A., Liu, T.: Grading evolution and critical state
in a discrete numerical model of Fontainebleau sand. Géotechnique (2018)
Ciantia, M.O., Marcos, A., Butlanska, J., Gens, A.: DEM modelling of cone penetration tests in a
double-porosity crushable granular material. Comput. Geotech. 73, 109–127 (2016a)
Ciantia, M.O., Arroyo, M., Calvetti, F., Gens, A.: A numerical investigation of the incremental
behavior of crushable granular soils. Int. J. Numer. Anal. Meth. Geomech. 40(13), 1773–1798
(2016b)
Itasca Consulting Group. PFC3D: Itasca Consulting Group, Inc. (2014) PFC—Particle Flow
Code, Ver. 5.0 Minneapolis, USA (2016)
Lobo-Guerrero, S., Vallejo, L.E.: Influence of pile shape and pile interaction on the crushable
behavior of granular materials around driven piles: DEM analyses. Granular Matter 9(3–4),
241–250 (2007)
278 N. Zhang et al.
Russell, A.R., Muir Wood, D., Kikumoto, M.: Crushing of particles in idealised granular
assemblies. J. Mech. Phys. Solids 57(8), 1293–1313 (2009)
Zhang, N., Arroyo, M., Gens, A., Ciantia, M.: DEM modelling of dynamic penetration in
granular material. In: 9th European Conference on Numerical Methods in Geotechnical
Engineering, Porto (2018)

2018-Zhang-Penetration Test DEM

Uploaded by

Copyright:

Available Formats

2018-Zhang-Penetration Test DEM

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2018-Zhang-Penetration Test DEM

Uploaded by

Copyright:

Available Formats

DEM Investigation of Particle Crushing Effects

on Static and Dynamic Penetration Tests

Ningning Zhang1(&), Marcos Arroyo1, Matteo Ciantia2,

Abstract. A 3-dimensional discrete element method model has been developed

Keywords: Dynamic penetration test Static penetration test Crushing

In geotechnical engineering, penetration tests have been widely applied to characterize

© Springer Nature Switzerland AG 2018

2.2 Model Construction and Simulation

Table 1. Model parameters for Fontainebleau sand.

3.2 Microscale Observations

You might also like