See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/325999635

METODOLOGY OF ANALYSIS OF 2D RC FRAME MODELS WITH APPLICATION

OF P-Y CURVES AS LINK ELEMENTS ON PILES

Conference Paper · June 2018

CITATIONS READS

0 1,467

3 authors:

Boris Folic Stanko Brcic

University of Belgrade University of Belgrade
31 PUBLICATIONS   53 CITATIONS    41 PUBLICATIONS   83 CITATIONS   

SEE PROFILE SEE PROFILE

Asterios Liolios
Democritus University of Thrace
39 PUBLICATIONS   312 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Stability of Structures View project

WOS indexed journal: Advances in Materials Science View project

All content following this page was uploaded by Boris Folic on 26 June 2018.

The user has requested enhancement of the downloaded file.

 407

UDC: 624.154
Prethodno saopštenje

METODOLOGY OF ANALYSIS OF 2D RC
FRAME MODELS WITH APPLICATION OF P-Y
CURVES AS LINK ELEMENTS ON PILES.

��������������������������**, A. Liolios***
* University of Belgrade, Innovation Center, Faculty of Mechanical
Engineering; [email protected]
** University of Belgrade, Faculty of Civil Engineering; [email protected]
*** Democritus University of Thrace, Dept. Civil Engineering, GR-67100
Xanthi, Greece, (e-mail: [email protected],

ABSTRACT
The paper presents the methodology of analysis of 2D frame model, founded on piles which
represents a skeletal multi-storey building. Numerical analyses related to 2D RC frame are
perfromed, with an accent on the pile-soil interaction. The results are a subject of another
paper, while this paper is related to pile-soil interaction implemented by p-y curves. The P-y
curves are deterimed according to Rees, Cox and Matlock model for sand. Using SAP 2000,
p-y curves are applied on both sides of the pile, as link elements. Properties of link elements
are assumed as the multi-linear plastic type, i.e. as hysteresis envelopes according to Takeda
model. The linear part of link elements corresponds to the initial stiffness of the p-y curve.
Curves are exposed to compression only and to almost negligible tension. Since it is the
analysis of a seismic action, the curve coefficients are selected to correspond to repeated,
i.e. cyclic load, and not for the static one. In such a way, numerical model asymptotically
simulates the real actions and seismic response of structures in dynamic interaction of the
pile-structure-soil system. This model offers a lot of potential for further development and
study of seismic performance.

KEY WORDS: dynamic interaction, pile-structure-soil, p-y curves, link elements, 2D frame

METODOLOGIJA ANALIZE 2D AB RAMA

PRIMENOM P-Y KRIVIH KAO LINK ELEMENATA
NA ŠIPOVIMA
REZIME
U radu je prikazana metodologija analize 2D modela rama fundiranog na šipovima koji
���������������������������������������������������������������������������������������
naglaskom na interakciju šipovi-tlo. Rezultati su predmet posebnog rada, dok se ovaj rad
odnosi na interakciju šip-tlo implementiranu primenom p-y krivih. P-y ������ ��� ���������
prema Rees, Cox i Matlock modelu za pesak. Primenom SAP 2000, p-y krive su nanete na
šip obostrano, kao link elementi. Link elementi su usvojeni kao više-������������������������
408

odnosno kao histerezisne anvelope prema Takeda modelu. Linearni deo link elemenata
odgovara inicijalnoj krutosti p-y krive. Krive su izložene samo pritisku i skoro
zanemarljivom zatezanju�������������������������������������������, izabrani su koeficijenti
krive za ponovljeno, odn. ������������������������������������������������������������������
asimptotski simulira realno dejstvo i se������i �������� ������������� �� �����������
interakciji sistema šip-konstrukcija-tlo. Ovaj model pruža ������������������������������ i
�����������������������������������

�������������������������������������������������-šip-tlo, p-y krive, link elementi, 2D

ram

INTRODUCTION

The structure-soil interaction was historically dealt with, directly or indirectly, by many
researchers. Interaction was researched (according to Stevano���, 1999), among the first, by
Zimmerman, 1888, who employed the Winkler hypothesis, published in 1876 for design of
railways. Winkler introduced a model of mutually independent elastic springs, where
settlement at one point depends only upon the force at that point. At a later date, the model
of homogeneous elastic and isotropic semi-space (defined by the soil modulus of elasticity
E 0 and by the Poisson’s coefficient �) was introduced, where the effects of a force in one
point propagates across the entire semi-space. Tajimi, in 1969, developed a method of a
seismic response, based on the elastic theory of wave propagation. Broms used the method
of boundary equilibrium for two types of soil and three ”lengths” of a pile, for the failure
state. The structure-pile-soil system was researched by Hetenyi, Terzaghi, Housner,
Penzien, �������, ��������-�������, �������� �������, Napatvaridze, Vesic, Pender,
Prakash, Trifunac, Tazoh, Stewart, Makris, Badoni, Dobry, Budhu, Mylonakis (Novak,
1974), Nogami, (Poulos and Davis,1980). They formulated the problem using the systems
of ordinary or partial differential equations, with solutions obtained through Bessel’s or
Hankel’s functions, or using the models with springs and dampers. Gazetas used the model
of kinematic and inertial interaction and model of a beam on dynamic Winkler foundation
(BDWF), see (�����, 2017). Wolf used both the frequency domain analysis and the mixed
successive frequency-time analysis, considering the pile elements having the form of the
disc-cone, etc.

Since it is necessary to harmonize the theory with the practice and to determine conditions
and areas of application, there are two approaches. One is based on experiences acquired
during inspection or testing of the buildings and structures after earthquakes, and the other
is based on experiments using the shaking tables or centrifugal tests. The knowledge about
this field, about dynamic properties of soil and soil-structure interaction, considerably
increased during the 20th century (Japanese Society of Civil Engineers, 1996 and 2000).
After a number of strong earthquakes in Japan and USA, detailed researches of the new
phenomena were undertaken. One of them is the Niigata earthquake in 1964 in Japan, when
the liquefaction mechanism was observed and definitely explained. Even though the
mechanism was discovered, the problem is rather complex and there is still no generally
accepted theory which includes all the aspects of this phenomenon.
 409

The model of RC 2D frame with RC piles (Fig. 1, left) is presented in this paper. The
method of calculation and application of p-y curves, derived from the experimental research
(Reese et al. 2001) and the time history (TH) analyses, including dynamic structure-pile-
soil system interaction (DSPSI), is the main subject of the paper. This model of p-y curves
may be used both for the buildings and for the bridges. Piles are imbedded in the soil which
is modeled using horizontal p-y curves, on each side of the pile, at each meter of depth. In
the vertical direction (friction along the surface and pile base) the support is considered to
be linearly-elastic. Piles, generally, may be either clamped or elastically supported at the
base. In this paper clamped support at the pile base is used. P-y curves are determined
(according to Mosher, 2000) based upon the research by Reese, Cox, Koop and Matlock for
the model of sand and for repeated loads.

Figure 1. Left: Model of 2D RC frame founded on RC piles, with joints and FRAME elements. Right:
numbers of link elements (p-y curves)
�������������������������������������������������������������������������grednim elemenatima.
Desno: Numeracija link elemenata (p-y krive)

ABOUT THE MODEL

A 2D RC frame model, which is extracted from the regular symmetrical 3D model of the
building, is used in the paper. At the top and at the bottom of each story, plastic hinges are
introduced in columns. There are two types of beam reinforcement, one above the support
point (intersections with other beams) and the other along the span. For each floor, the
necessary reinforcement of the beams is determined. Considering that it is a regular,
symmetrical building, the reinforcement is the same for all beams within the same floor, but
different for support zones and for spans. The length of the support zones of beam elements
is 0.20 L (the offset zone to be deducted from it), while the zone of the span is 0.60L.
Plastic hinges in the beams are introduced at ends of reinforcement zones.
410

P-y curves are introduced as link elements, on each side of the pile (Figure 1, right), at each
meter of the pile depth. The piles are condensed in the 2D plane, so the group of 3 piles is
represented with 2 piles in plane, one pile being individual (1D60 in figure 2), and the other
having modified properties, because they are multiplied with the coefficient 2 (2D60 in
Figure 2). Analogously, the group of 4 piles is represented with 2 2D60 piles.

Figure 2 „Condensation“ principle of a group of 3 piles into the plane group of 2 piles (1D60 –
individual pile, 2D60 – double pile).
Slika 2 Princip „kondenzacije“ grupe od 3 šipa u ravansku grupu od 2 šipa (1D60 samostalni šip,
2D60 dvostruki šip).

P-Y CURVES

Originally, the idea of modeling the soil using p-y curves was created using an analogy with
the soil behavior curves in triaxial test. Meymand stated that it was most likely first
introduced by McClelland and Focht, as early as in 1958. They recommended the procedure
for correlation of data of triaxial stress-strain test with the pile force-displacement curves
for specified depths, via the expected modulus of reaction of subsoil, for each soil layer.
The p-y curve construction is presented in Figure 3. Both designations of the characteristic
points of p-y curves, which can usually be found in literature, are written in Fig. 3.

Reese, on the basis of experiments, observed and displayed the concept of the soil failure in
the shape of a wedge which occurs close to the soil surface. If the pile is sufficiently long
two forms of the soil failure would appear. The first, upper one, has a wedge-like form,
while the other, lower one, has a block form. The depth at which the soil failure changes
form is called the critical depth.

P-y curve in Figure 3, according to the original method (Reese at all, 2001) and to others
consists of four parts, those being:
The first part, where there is a linear dependence of displacements and stresses,
from the point 0 to the point y k /p k (a different designation can also be encountered:
y a /p a ),
The second part, where there is a parabolic dependence of the stress and
displacement from the point y k /p k to the point y m /p m (a different designation is
y b /p b ). Parameters of the parabola in this part are coefficients C and n. It is not the
same n as the n which is applied in formula (10).
The third part of the curve is an inclined straight line from the point y m /p m to the
point y u /p u .
 411

The fourth part is the constant stress: after reaching displacement y u there is no
further increase in stress, i.e., with the increase of displacement after this point, the
stress retains the constant value p u .

Figure 3. Construction of the characteristic shape of p-y ������������������

Slika 3. Konstrukcija ���������������� oblika p-y krive ������������

The points y m and y u depend exclusively on geometry of the pile, and have a fixed value,
b/60 and 3b/80 respectively, while the value of the deformation at the boundary of linearity
y k is determined by calculation, as well as the stresses p m and p u .

where:

b – width of the pile normally to direction of load action.

The calculation of characteristic points of the curve y k /p k , y m /p m and y u /p u , is performed in

a number of steps. First, the coefficients C 1 , C 2 i C 3 are determined (formulae (4), (5) and
(6), or Figure 6, which can be used as a control). The coefficients depend exclusively on the
angle of internal friction and the self weight of the soil, so their employment yields the
values of lateral resistance of the soil, per unit of length of a pile. Then follows
determination of the inclination k 0 of the initial part of 0-k (0-a), according to formula:

k0 � k � z (1)

The coefficient k in Eq. (1) is determined according to the table 1 (limits of k 0 for dry and
submerged sand with respect to the relative density in %, are provided in Figure 7).
412

a) b)

Figure 4. Two types of soil failure a) wedge mode b) block mode, or flow (according to Reese at all,
2001)
Slika 4 Dva tipa sloma tla: a) klinasti oblika, b) blok oblik, ili ������� (prema Risu i dr, 2001)

Figure 5. Coefficients of reduction of lateral resistance for sand - A and B (according to Mosher at all,
2001)
Slika 5. Koeficijenti redukcije horizontalne otpornosti za pesak - A i B (prema Mosher i dr, 2001)
 413

Table 1. Coefficient of lateral soil reaction for sand. Initial modulus of p-y curve as function of
relative compactness and the level of underground water table (WT)
������������������������������������������������������������������������p-y ����� ��
����������������������������������������������������
� Soil modulus Coefficient k for sand
Relative
Loose Medium dense (Hard) dense
compactness
Submerged sand 5,430 KPa/m 16,300 KPa/m 33,900 KPa/m
Sand above WT 6,790 KPa/m 24,430 KPa/m 61,000 KPa/m

Then a limit value of lateral resistance, as a lower value, is calculated for the wedge mode
of soil failure, close to the ground surface, p st , Eq. (2), or for the block mode of soil failure,
at a greater depth, p sd Eq.(3):

pst � (C1z � C2b) � ' z (2)

psd � C3b � ' z (3)

K tan � Sin� tan 2 � Sin (� / 2)

C1 � � � K tan � (tan � Sin� � tan(� / 2)) (4)
tan( � � � ) cos (� / 2) tan( � � � )

K tan �
C2 � � tan 2 (45 � � / 2) (5)
tan( � � � )

C3 � K tan � tan 4 � � tan 2 (45 � � / 2) (tan8 � � 1) (6)

where:
�- effective unit weight of sand
z – depth with respect to the soil surface
K – coefficient of lateral soil pressure, selected value is 0.4
��– angle of interior friction

� � 45 � � / 2 (7)

z cr - critical depth at which the soil failure from the wedge mode, Eq. (2), becomes the block
mode, Eq. (3)

After that, lateral resistances for the transitional points of the p-y curve are calculated (the
third part of the p-y curve, Fig. 3, y m and y u ; or y c and y b ), according to:

yu = 3 � b / 80 , pu = A � (pst , z � zcr ; psd , z � zcr ) (8)

ym = b / 60 , pm = B � (pst , z � zcr ; psd , z � zcr ) (9)
414

Reduction coefficients A and B depend on the type of load and for dynamic analysis the
cyclic load curves are used. In this calculation, the estimated error of determination of A
and B coefficients for the cyclic load, from the diagram in Fig. 5, is around ±1% and in an
extreme cases it does not exceed ±3%. The A and B coefficients are provided in the
diagram in the interval from 0 to 5 z/b. After the five pile depths, the curves representing
coefficients A and B have a constant value, as given in Fig. 5, so in this calculation they are
completely in agreement with the theory.

Construction of p-y curve according to API recommendations (according to Mosher, it was

proposed by Murchison and O'Neill) is performed using the formula which is more
convenient than the original Reese formula:

P � kH �
� n � A tanh �� y �� (10)
pu � n � A � pu �
where:
p u – is ultimate strength at the depth H (the same as p c ),
k – initial modulus of soil reaction (according to the table 1, for sand)
y – lateral displacement of the pile

The part n·A in Eq. (10) refers to the type of load and shape of the pile, so for the cyclic
load and the constant cross-section of the pile (prismatic), the formula is:

� k ( z) �
P � 1.0 pu tanh �� 0 y �� (11)
� 0.9 pu �

where k 0 (z) is determined according to the formula (1), and p u from the set of previous
formulae, conclusive with (8). Therefore, when using the formula (11), it is necessary to
know only p u and k 0 , at each meter of pile depth.
 415

Figure 6 Factors for calculation of ultimate soil resistance for laterally loaded pile in sand (according
to Mosher at all, 2001)
��������������������������������������������������������������������������������������������� (prema
Mosher i dr, 2001)

Table 2 presents parameters of p-y curve, which consists of 4 parts, depth of the pile - z,
pile diameter 60cm, and soil parameters as given in the table caption. Both usual
designations for characteristic points of p-y curve: p a , p b , p c , i.e., p k , p m , p u are given in
the table.

In the upper part of table 3, parameters of the p-y curves for the first 4 meters of depth are
given. In the lower part there are the pairs of columns of y-p values, ready for input into
SAP2000, as link elements. Angle of �=35° is the boundary value between the medium
dense and dense sand, Figure 7.
416

Table 2. p-y curve: �= 35°; D=0,60 m; �= 10 kN/m3; k= 33900 kPa/m.

Tabela 2. p-y kriva: �= 35°; D=0,60 m; �= 10 kN/m3; k= 33900 kPa/m.
i; z; k0; ya; p a =p k ; p b =p m ; p c =p u
1 1 33900 8,15E-04 27,64 42,86 52,23
2 2 67800 7,00E-04 47,47 105,34 144,80
3 3 101700 3,53E-04 35,90 178,72 285,95
4 4 135600 5,63E-04 76,29 303,64 485,82
5 5 169500 8,18E-04 138,73 461,23 737,98
6 6 203400 1,12E-03 227,78 651,51 1042,41
7 7 237300 1,47E-03 347,96 874,45 1399,13
8 8 271200 1,86E-03 503,70 1130,07 1808,12
9 9 305100 2,29E-03 699,40 1418,37 2269,39
10 10 339000 2,77E-03 939,43 1739,34 2782,95
11 11 372900 2,88E-03 1074,71 1952,70 3124,33

Table 3. p-y curve: �= 35°; D=0,60 m; �= 10 kN/m3; k= 33900 kPa/m.

Tabela 3 p-y kriva: �= 35°; D=0,60 m; �= 10 kN/m3; k= 33900 kPa/m.
ko= 33900 ko= 67800 ko= 101700 ko= 135600
z= 1 Z= 2 Z= 3 z= 4
pc= 52,23 pc= 144,80 pc= 285,95 pc= 485,82
y (m) p(kPa/m) y (m) p(kPa/m) y (m) p(kPa/m) y (m) p(kPa/m)
1,0E-06 1,0E-05 1,0E-06 1,0E-05 1,0E-06 1,0E-05 1,0E-06 1,0E-05
0 0 0 0 0 0 0 0
-0,001 -29,82 -0,001 -63,24 -0,001 -97,62 -0,001 -132,19
-0,002 -44,98 -0,002 -106,22 -0,002 -174,86 -0,002 -246,15
-0,003 -50,15 -0,003 -128,35 -0,003 -225,42 -0,003 -332,50
-0,0045 -51,93 -0,0045 -140,58 -0,0045 -263,57 -0,0045 -412,93
-0,005 -52,07 -0,005 -142,14 -0,005 -270,08 -0,005 -429,66
-0,007 -52,22 -0,007 -144,39 -0,007 -282,04 -0,007 -466,69
-0,009 -52,23 -0,009 -144,74 -0,009 -285,00 -0,009 -479,47
-0,01 -52,23 -0,01 -144,78 -0,01 -285,48 -0,01 -482,18
-0,015 -52,23 -0,015 -144,80 -0,015 -285,94 -0,015 -485,60
-0,02 -52,23 -0,02 -144,80 -0,02 -285,95 -0,02 -485,81
-0,0229 -52,23 -0,0229 -144,80 -0,0229 -285,95 -0,0229 -485,82
-0,09 -52,23 -0,09 -144,80 -0,09 -285,95 -0,09 -485,82
-0,18 -52,23 -0,18 -144,80 -0,18 -285,95 -0,18 -485,82

Using SAP 2000, the p-y curves are applied on both sides of the pile, as link elements. The
link elements are assumed as a multi-plastic type, i.e. as hysteresis envelope according to
the Takeda model. The link elements in SAP 2000 are applied as the two-node elements, so
for each link element it is necessary to introduce another node, away from the pile, which,
together with the link element, represents nonlinear soil. The first node is part of the pile in
which two link elements are joined (Fig. 1, right). The link element itself has a linear and
non-linear part, Figure 8. The linear part is used for linear analysis, and for determination of
the first natural mode. In the nonlinear analysis both linear and nonlinear parts are used, in
ratio which depends on the level of realized strains.
 417

The linear part of link element corresponds to the initial stiffness of p-y curve. The curves
are applied on both sides (Figure 1, right) because they are exposed to pressure only, and
negligible tension, which simulates the property of the soil to convey only pressure.
Therefore, asymmetry of the seismic response may be determined (as presented in another
paper for this Conference). Since the seismic action is analyzed, the selected curve
coefficients correspond to repeated (cyclic) load, rather than for a static one. It was
attempted that this model asymptotically simulates the real actions and seismic responses of
structures in dynamic interaction of the pile-structure-soil system. This model offers a great
potential for further development and study of seismic performance, such as dissipation of
energy along the depth of the pile etc. Another paper is presenting the change of the first
natural mode and states of plastic hinges depending on PGA.

In the case of pushover analysis it is also necessary to use cyclic p-y curves, because the
loads in the pushover analysis are quasi-static and not static. This quasi-static load attempts
to cause the effects upon the structure which would be as similar as possible to the ones
caused by an earthquake. Quasi-static load is used for determination of seismic
(performances) properties of the structure and also to determine the sequence of formation
of plastic hinges, when the structure-pile-soil system eventually becomes a mechanism. It
should be mentioned that the initial plastic hinges, in pushover analysis and time history
analysis can be significantly different. However, as the number of plastic hinges increases,
the differences usually decrease.

y = 12,748x2 - 203,44x + 2787,1

80000
R2 = 0,9999
70000

60000
k0 (KN/m2/m)

50000
y = 6,5674x2 - 65,882x + 2142,6
40000 R2 = 0,9993
30000
20000

10000
0
0 20 40 60 80 100
Dr (%)

Figure 7. Initial modulus of lateral reaction of soil as function of the water table level, density and
��������������������������������������������������������������������������������������������������Right
curves k 0 as function of compactness, with formulae for aproximate calculation.
���������������������������������������������������������������������������������������������������
���������������������������������������������������������������������������������������������������
krive zavisnosti k 0 od zbijenosti, sa formulama za približ������������
418
�

The soil in this model is represented using coupled p-y springs (Figure 4 left), which,
regarding that they are subjected to compression only, are placed on both sides of the pile.
These link elements are of multi-linear plastic curve type, i.e. non-linear hysteresis Takeda
model. The modeled multi-linear link elements also enable the onset of the gap effect of the
hysteresis curve.

If instead of p-y curves the secant elastic stiffness is used, their upper boundary of
inclination represents the initial stiffness of p-y curves (Figure 7 and table 1). The initial
linear lateral stiffness of the p-y curves for sand increases proportionally with the depth z,
(in this case) according to the formula: kx=33900·z, but only up to the value pa (pk in Fig.3).
Further, the p-y curve behaves as a hyperbole with a slant asymptote, while the maximum
value of pressure pc, according to the formula (11), represents a horizontal asymptote.
Lateral stiffness of elastic springs is ostensibly smaller in comparison to the p-y curves, but
only in relation to the initial stiffness of the p-y curves, because such elastic springs provide
considerably higher soil resistance forces at higher values of displacements and do not have
hysteretic loss of capacity, as is the case of p-y curves. It is the larger displacements which
occur in the upper layers of soil. If the secant stiffness was used (Pando, 2013) for elastic
springs, the result would be different, at least in the part referring to the displacements
smaller than the intersection point of the p-y curve and the secant stiffness line. In (Foli�,
2017), it was shown that the usage of secant stiffness can have a negligible effect, but also
it can considerably change the form of failure of the structure and arrangement of plastic
hinges, for the middle bridge frame. The middle bridge frame for different soil types,
migration of plastic hinges and damping was discussed by (Suarez, 2005). The structural
response largely depends on the used accelerogram, intensity of PGA, inclination of secant
curve, change of natural mode of the structure-pile-soil system and response spectrum.
Therefore, changes of natural modes of the SPS system, using various secant stiffnesses
should be considered in relation to the response spectrum for a chosen accelerogram. �
�

Figure 8. Soil model as a link element in SAP 2000. P-y multilinear plastic, according to Foli� at all
(2013).
Slika 8. Model tla kao link element u SAP 2000. P-y višelinearno plasti�no, prema Foli� i dr. (2013).
 419

The p-y curves from table 3 should be entered into the properties of link elements, in the
lower left part of the dialog form, as seen in Fig. 8.

It is important to emphasize at the end of this section, that unless sand is clean, and has a
considerable cohesion, the mentioned procedures cannot be implemented. Instead,
specially constructed curves, which include these effects as well, should be used (Meyer at
all, 1979, Reese at all 2001, Foli� B. 2017).�

CONCLUSION�

The presented model of 2D RC frame founded on RC piles, with the soil modeled with
horizontal p-y curves, offers a range of options for numerical research of seismic
performances of the structure-pile-soil system. This model of p-y curves is applicable for
buildings and bridges, for small pile diameters, and also for the bridges with pile diameters
of 150-180cm. The algorithm of calculation of the p-y curves for sand is provided in detail,
as well as the method for application on the piles of 2D models of RC frames. �

Acknowledgment
The research described in this paper was financially supported by the Ministry of Education
and Sciences Republic of Serbia within the Project TR 36043. This support is gratefully
acknowledged.�

LITERATURA:�

Foli� B., Sedmak S., �osi� M., La�inovi� �. (2013): The forming of the bridge structure founded on
RC piles and solution seismics strenthening middle frame. 5th International Scientific and
Expert Conference of the International TEAM Society.�
Foli� B. (2017): Seizmi�ka analiza betonskih konstrukcija fundiranih na šipovima. Doktorska
disertacija. FTN. Univerzitet u Novom Sadu. Novi Sad.�
Foli�, B., Foli�, R. (2018): Komparativna nelinearna analiza interakcije šip-tlo AB 2D okvira.
Gra�evinski materijali i konstrukcije. DIMK Srbije. Beograd. 2018.
Japanese Society of Civil Engineers. Dynamic analysis and earthquake resistant design. Rotterdam,
Brookfield. Balkema, A. Vol.1, 1997. and Vol 2, 2000.
Meyer B. J., Reese L. C. (1979): Analysis of single piles under lateral loading. Center for
Transportation Research, University of Texas at Austin. FHWA/TX-79/38+244-l.�
Meymand P. J. (1998): Shaking Table Scale Model Tests of Nonlinear Soil-Pile-Superstructure
Interaction In Soft Clay. Thesis (Ph. D.) in Engineering-Civil Engineering. University of
California, Berkeley.�
Mosher R., Dawkins W. (2000): Theoretical Manual for Pile Foundations, U.S. Army Corps of
Engineers, Report ERDC/ITL TR-00-5, Washington, USA.�
 420
�

Nogami T. Ed. (1987): Dynamic Response of Pile Fundations. Experimental, Analysis and
Observation. ASCE. Geotechnical Special Publication. Geotechnical Engineering Division.
No. 11.�
Novak M. (1974): Dynamics stiffness and damping of piles. Canadian Geot. Journal, No 11/1974. pp
574-598.�
Pando M. (2013): Analyses of Lateral Loaded Piles with p-y Curves - Observations on the Effect of
Pile Flexural Stiffness and Cyclic Loading. NCDOT 7th Geo3T2, Raleigh, NC, Thursday,
April 04, 2013. Paper: 3B-1_A49�
Reese L., Van Impe W. (2001): Single pile and pile groups under lateral loading, Balkema,
Rotterdam, 2001.�
Poulos H.G., Davis. E. H.(1980): Pile foundation analysis and design. New York.
Suarez V. (2005): Implementation of Direct Displacement Based Design for Pile and Drilled Shaft
Bents. NCSU. North Caroliona State Univesity. October.
Stevanovi� S. (1999): Fundiranje gra�evinskih objekata. Gra�evinski fakultet Univerzitet u Beogradu
i �asopis Izgradnja. Beograd.

View publication stats