Professional Documents
Culture Documents
https://doi.org/10.1007/s40091-019-0213-9
ORIGINAL RESEARCH
Received: 15 March 2018 / Accepted: 25 January 2019 / Published online: 2 February 2019
© The Author(s) 2019
Abstract
The work presents a finite element modeling of beam resting on a two-parameter layered soil. The behavior of the soil con-
tinuum and the beam are assumed to be linear and homogeneous isotropic. Using the strain energy expressions, the shear
strain of the beam element and the soil foundation are taken together in the analysis. In this approach, the stiffness matrix of
each component is elaborated and integrated in the finite element analysis. First, various examples are elaborated to show
the effectiveness of the proposed approach and the ability of the numerical program developed for this concern. Second, the
analysis is widened to study the influence of the soil properties on the interface continuum and on the beam responses. Third,
a parametric study is carried out to highlight the effect of the position of springs at the interface continuum, the properties
of soil, the deepest of the soil foundation and the ballast layer on the response of the interface and the beam itself. Moreover,
shear deformations are presented to show the crucial influence on the beam, on the structure and on the interface behaviors.
Obtained results show pertinent results corresponding to the interface continuum and the beam responses.
Keywords Static soil-beam interaction · Layered soil · Two-parameter elastic foundation model · Finite element method ·
Elastic foundation · Ballasted layer · Soil properties
13
Vol.:(0123456789)
22 International Journal of Advanced Structural Engineering (2019) 11:21–30
Due to the complexity of the interaction problems, analyt- elastic foundation, a finite element procedure was developed
ical solutions are rarely used and desired alternatives would (Gan and Kien 2014). Finally, the effect of material non-
be a numerical approach (Dinev 2012; Hassan and Doha homogeneity and the two-parameter elastic foundation were
2015; Ai and Cai 2016). The FEM is still popular method in used to quantify the response of simply supported beams.
this domain (Bourgeois et al. 2012; Su and Li 2013) but it The foundation medium behavior is assumed to be linear
has disadvantages compared to BEM and spectral element elastic, homogeneous and isotropic with two parameters
method (SEM) (Omolofe 2013; Mokhatari et al. 2016) using describing the reaction of the elastic foundation on the beam
absorbent frontier modeling. The FEM requires the discre- (Avcar 2016).
tization of the domain to high number of finite elements but In this analysis, the soil-structure interaction problem has
the problem can be solved with BEM where only the bound- been studied using effectively two-parameter model of the
ary of the domains involved can be discretized (Padron et al. layered soil. Both the beam and the substrate are described
2011; Ai and Cheng 2013; Ribeiro and Paiva 2014). by means of FEM integrating shear deformations. However,
Beam using Winkler foundation model (1867) is used in to ensure vanishing displacement at the frontier of the sub-
various practical problems. In this approach, the founda- strate, mesh has to be extended far away from the loaded
tion flexibility is considered as a set of continuous springs region. To improve the computational efficiency, two-
and has employed to model soil-structure interaction (Kim dimensional finite elements are refined in the loaded area.
and Yang 2010; Chore et al. 2010; Sapountzakis and Kan- The modeling uses plane stress state for the shear beam in
pitsis 2011; Raychowdhury 2011; Limkatanyu et al. 2012). adhesive contact with plane strain state of the soil founda-
Springs introduced provide a resistance in the vertical direc- tion. Adding, a parametric study is elaborated to show the
tion only confining the deformation of the soil foundation. influence of (1) the horizontal behavior of the soil founda-
Evidently, the one-parameter soil foundation model suffers a tion, (2) mechanical properties of the soil foundation and
handicap due to the discontinuity in the supporting medium. (3) ballasted layer.
To improve the Winkler model, two-parameter and three-
parameter soil foundation models have been developed tak-
ing into consideration shear deformations of the soil. To Physical modeling
overcome the one-parameter foundation model deficiencies,
many researchers proposed various foundation models to Transverse deflections of a thin elastic beam can be pro-
describe rigorously soil response insuring the interconnec- duced due to the reciprocal effect of the loading and the
tion between vertical springs. foundation reactions. Really, the reaction forces of layered
In the first approach, the interaction between springs soils are proportional to the deflection at each point at the
has been established by an elastic tensioned membrane interface medium. Among deficiencies of the Winkler
(Filonenko 1968) and the second one uses beam elements or model is that the displacement discontinuity which appears
a plate to interact between them (Hetényi 1966). In this case, between the loaded and unloaded regions of the foundation
the tensioned membrane is quantified by a shear parameter. surface. The displacement field must be a continuous func-
Third, to model the mutually effect between springs, Kerr tion along the all surface (Fig. 1).
(1964) integrated a shear layer dividing the soil medium of In the Winkler model, vertical soil sections of the foun-
foundation to two different layers. dation are modeled by identical, discrete elastic springs
The two-parameter soil foundation has shown consider- characterized by their mechanical property, Ks (Fig. 1b).
able developments during the last decade. A new finite ele- To improve prediction of the entire system behavior, the
ment formulation was developed to study the shallow and the contribution of the interface medium between the founda-
raft foundation response eliminating the limits of Winkler tion and the beam must be integrated. The literature shows
model (Mullapudi 2010). In the way to study large deflec- many methodologies interpreting the soil-beam interac-
tions of functionally graded beam resting on two-parameter tion response. This interaction was established by means
q(x)
W=0 Ks
w
(a) (b)
13
International Journal of Advanced Structural Engineering (2019) 11:21–30 23
Ks w(x) ( )2 ( )3
x2 x3
N2 (x) = x − 2 L + L2 , N3 (x) = 3 Lx − 2 Lx and
x2 x3 { }
N4 (x) = − L + L2 , and qe is the degree of freedom vector.
Fig. 5 Shear effect model
The strain energy for the beam element, U, is combined of:
U = UB + UF , (4)
Lateral springs
UB and UF are the strain energies for the bending beam and
Ks KG
the soil foundation.
In this case, the strain energy of the soil foundation can
be divided in two components as:
Fig. 6 Shear spring elements
UF = Us + UG . (5)
For the two-parameter foundation model, Shen (2012) gave
of flexural element (beam or plate) (Fig. 2), shear layers the strain energy expression, UF, as:
(Fig. 3) or pre-tensioned membranes (Fig. 4).
On the other hand, to take into account the lateral inter- ( )2
2∫ 2∫
1 1 𝜕w(x)
action due to the soil (cohesion, Ks w2 (x)dx + (6)
) a plate is used introducing
UF = KG x.
d2 w(x)
𝜕x
the shear effect ( KG dx2 ) (Fig. 5). Further, a set of L L
springs are introduced in the longitudinal direction of the Substituting the relation (3) into the Eq. (6), we obtain:
soil foundation guaranteeing the interaction between lon-
gitudinal and transverse behaviors (Fig. 6).
13
24 International Journal of Advanced Structural Engineering (2019) 11:21–30
∫
Ks [N(x)]t [N(x)]dx = 1 ⎢ numerical solution.
156 −22L2 ⎥⎥
KMs =
420 ⎢
⎣ Sym
Figure 9 depicted horizontal displacements of interface
L
4L2 ⎦
continuum nodes. The figure shows important displacements
(9)
in the left region of the applied force but a slight variation
⎡ 36 3L −36 3L ⎤
accompanied by weak displacements at far nodes.
� �t � �
� � K2 ⎢ 4L2 −3L −L2 ⎥ Figure 10 shows the horizontal interaction model. In this
∫
dN(x) dN(x)
KMG = KG
dx dx
dx = ⎢
30 ⎢ 36 −3L2 ⎥⎥
,
case, springs are introduced in vertical and horizontal direc-
L ⎣ Sym 4L ⎦
2
tions, simultaneously.
(10) Horizontal springs can be justified to avoid the handicap
KG b
with K1 = Ks bL and K2 = L raised in the Winkler model. When the lateral interaction is
In the same manner, the strain energy of the beam bend- considered in the analysis, horizontal displacements are dif-
ing is ferent from those obtained by Winkler approach. In this case,
an improvement is highly observed varying from 45.13% at
x = 0 m to 100% for x ≥ 5 m along the soil length (Fig. 11).
� �t � �
� �t ⎡ d2 N(x) d2 N(x) ⎤� �
⎢∫
1
UB = Eb Ib qe ⎢ dx⎥ qe . (11) The lateral interaction appears with no effect on vertical dis-
2 dx2 dx2 ⎥
⎣L ⎦ placements (Fig. 12).
Really, the calibration of horizontal stiffness plays a pri-
Substituting the relation (3) into the Eq. (11), the stiffness mordial contribution in this research theme. In this study, the
matrix can be deduced as: same stiffness values of horizontal and vertical springs are
considered due to the continuity of the medium.
⎡ 12 6L −12 6L ⎤
� � Eb Ib ⎢ 4L2 −6L 2L2 ⎥
Kb = 3 ⎢
12 −6L ⎥⎥
. (12)
L ⎢ Analysis of the beam on a rigid base
⎣ Sym 4L ⎦
2
{ }
The loading vector Fe depends on the distributed forces A beam of length L = 5 m, a width b = 0.60 m, and height
along the element side. For a uniform load q(x) = q0 , the h = 0.80 m, supported on a rigid soil foundation is consid-
corresponding loading vector can be formulated as: ered. The load on the beam is 500 kN applied at the middle.
For the symmetry reason, only one half of the beam is con-
{ } sidered in the analysis (Fig. 13).
∫ ∫
Fe = [N(x)]t q(x)dx = q0 [N(x)]t dx. (13)
L L
2m P =100 kN
Numerical study
0.25 m
Validation of the program
14@1m Ks=104 kN/m3
First, the beam on Winkler foundation model is used to
show the ability of the numerical program conceived for
Fig. 7 Beam on elastic soil foundation
13
International Journal of Advanced Structural Engineering (2019) 11:21–30 25
2.00E-03
-4.00E-03
-6.00E-03
Distance along of the beam (m)
(a) (b)
Fig. 8 Vertical deflection along the beam. a Binesh (2012). b The present approach
0 2.5 5 7.5 10 12.5 15 length L = 54 m, Young’s modulus Es variable and Poisson
-2.00E-04
ratio 𝜈s = 0.30.
-4.00E-04 Figures 17 and 18 show the influence of mechanical
properties of soils on the continuum interface responses.
-6.00E-04 When the mechanical properties of the soil are feeble,
-8.00E-04 horizontal and vertical displacements are very impor-
tant. The enhancement of soil properties of 35.33% is
-1.00E-03 observed when E s increases from 15 MPa to 45 MPa.
Distance along the beam (m)
The maximum horizontal displacement decreases from
1.15 × 10−3 to 0.357 × 10−3 m (Fig. 17) and from 11 × 10−3
Fig. 9 Horizontal displacement along the beam nodes
to 3.89 × 10−3 m for vertical displacement (Fig. 18). In
this context, an improvement of soil behavior is highly
The beam reposing on a rigid base shows important hori- observed, the horizontal displacement passes from
zontal displacements of upper nodes in the vicinity of the 0.357 × 10 −3 to 0.154 × 10 −3 m (43.13%) (Fig. 17) and
applied load area (Fig. 14) but vertical displacements are from 3.89 × 10 −3 to 1.85 × 10 −3 m (47.55%) for vertical
very significant at the middle of the beam and decrease at displacement (Fig. 18) when soil properties change from
far points (Fig. 15). Es = 45–100 MPa.
In the same concept, the displacements of upper nodes
Influence of soil properties on the interaction of the beam are described on Figs. 19 and 20 where hori-
response zontal displacements appear with considerable soil prop-
erties (Fig. 19) but with a partial independency between
Mechanical properties of the beam and the soil foundation soil properties and vertical displacements of beam nodes
are mentioned on Fig. 16. A fine meshing is selected in the (Fig. 20). Important displacements using Es = 15 MPa
horizontal direction but a uniform meshing is assumed in the are obtained followed with a considerable bound of
Kh
Kv
14@1 m
13
26 International Journal of Advanced Structural Engineering (2019) 11:21–30
6.00E-04 1.00E-05
4.00E-04 0.00E+00
0 0.5 1 1.5 2 2.5 3
2.00E-04 -1.00E-05
0.00E+00 -2.00E-05
0 2 4 6 8 10 12 14 16
-2.00E-04 -3.00E-05
-4.00E-04 -4.00E-05
-6.00E-04 -5.00E-05
-8.00E-04 Vertical springs
13
International Journal of Advanced Structural Engineering (2019) 11:21–30 27
0.8 m
Beam-cross section
Ballast layer
12@2.5 m
Sub-grade
E s, s
5@0.5 1 1.5 2 2 3 6 9m
4.00E-04 0.00E+00
0 0.5 1 1.5 2 2.5 3
2.00E-04 -5.00E-05
0.00E+00
-1.00E-04
-2.00E-04 0 5 10 15 20 25 30
-4.00E-04 -1.50E-04
Es=15 MPa
-6.00E-04 -2.00E-04
Es=45 MPa
Es=15 MPa Es=65 MPa
-8.00E-04 Es=100 MPa
Es=45 MPa -2.50E-04
-1.00E-03 Es=65 MPa
-1.20E-03 Es=100 MPA
Fig. 19 Horizontal displacement of the beam
-1.40E-03
-6.00E-03 -1.50E-02
Es=15 MPa
-8.00E-03 Es=45 MPa Fig. 20 Vertical displacement of the beam
Es=65 MPa
-1.00E-02 Es=100 MPa
-1.20E-02 5.00E-04
13
28 International Journal of Advanced Structural Engineering (2019) 11:21–30
2.00E-03 0.00E+00
0.00E+00 0 0.5 1 1.5 2 2.5 3
-2.00E-03 0 5 10 15 20 25 30 -5.00E-05
Ballast 30 cm
-4.00E-03 Without Ballast
-1.00E-04 Ballast 60 cm
-6.00E-03
-8.00E-03 -1.50E-04
-1.00E-02 H=30 m
H=40 m
-1.20E-02 H=20 m -2.00E-04
-1.40E-02
-2.50E-04
Fig. 22 Vertical displacements of the interface nodes
Fig. 24 Horizontal displacements of beam nodes
Ballast layer
H=30 m
L=27 m
13
International Journal of Advanced Structural Engineering (2019) 11:21–30 29
13
30 International Journal of Advanced Structural Engineering (2019) 11:21–30
Reissner E (1967) Note on the formulation of the problem of the plate Su D, Li JH (2013) Three-dimensional finite element study of a single
on an elastic foundation. Acta Mech 4(1):88–91 pile response to multidirectional lateral loadings incorporating
Ribeiro DB, Paiva JB (2014) Mixed FEM-BEM formulations applied to the simplified state-dependent dilatancy model. Comput Geotech
soil-structure interaction problems. In: Proceedings of the World 50:129–142
Congress on Engineering, London, 2–4 July 2014. Lecture Notes Tezzon E, Tullini N, Minghini F (2015) Static analysis of shear flexible
in Engineering and Computer Science, 2014, pp 1178–1183 beams and frames in adhesive contact with an isotropic elastic
Sapountzakis EJ, Kanpitsis AE (2011) Nonlinear analysis of shear half-plane using a coupled FE-BIE model. Eng Struct 104:32–50
deformable beam-columns partially supported on tensionless Zienkiewicz OC, Taylor RL, Zhu JZ (2005) The finite element method:
three-parameter foundation. Arch Appl Mech 81(12):1833–1851 Its basis and fundamentals. Elsevier, Boston, p 733
Sayeed M, Shahin M (2018) Design of ballast railway track foundation
using numerical modeling. Part II: applications. Can Geotech J Publisher’s Note Springer Nature remains neutral with regard to
55(3):369–396 jurisdictional claims in published maps and institutional affiliations.
Shen HS (2012) A novel technique for nonlinear analysis of beams on
two-parameter elastic foundation. International Journal of Struc-
tural Stability Dynamics 11(6):999–1014
13