th

The 12 International Conference of

International Association for Computer Methods and Advances in Geomechanics (IACMAG)
1-6 October, 2008
Goa, India

Effect on Barrage Raft Floor Due to Intrusion of Rock

Kumar Venkatesh
Dept. of Civil Engineering, Motilal Nehru National Institute of Technology, Allahabad, India

N. K. Samadhiya, A. D. Pandey
Dept. of Civil Engineering, Dept. of Earthquake Engineering, Indian Institute of Technology, Roorkee, India

Keywords: barrage, raft floor, rock, finite element method

ABSTRACT: The analysis for the design of the barrage raft floor, along conventional lines, is done by Hetenyi’s
approach wherein the foundation media is assumed to be uniform. However, field conditions with regard to
foundation media do not present a foundation media having uniform characteristics throughout its extent. It
becomes important to examine the situation when the foundation media is not homogeneous by numerical
techniques such as the finite element method. In order to investigate the effect of non-uniform foundation
condition a typical barrage has been selected, located at the site where the subsoil is partly river bound material
(RBM) and partly rock. Eight noded brick element have been used for modeling the barrage component with
foundation media, whereas the raft floor has been modeled using the plate bending element. Intrusion of the rock
has been considered from the bottom of the models vertically as well as laterally. The analysis has been
performed to find out the variation of moments and deformations at the three sections of the raft floor. The study
has represented that intrusion of rock has affected the moments and deformations behaviour of raft floor under
representative loading condition.

1 Introduction
A barrage is a diversion headwork, which is employed to divert inflows into the canal from a river. In a barrage the
crest is kept at low level and the gates alone affect heading up of water. During the floods, the gates are raised to
pass the high flood flow. When the flood recedes, the gates are lowered and the flow is obstructed, thus
maintaining the required pond level at the upstream of the barrage for feeding the main canal under gravity.
Barrages are usually made of masonry, plain cement concrete or reinforced concrete, depending on the nature of
foundation encountered, availability of construction material, dewatering problems, economy of construction, etc.
A barrage can have gravity or a raft floor. In recent years, the hydraulic and structural engineers have taken up
are seized upon the important task of evolving safe and economic design criteria for barrage raft floor due to
several advantages such as less excavation and dewatering, lesser construction time, superior flexural behavior
etc.
A number of analytical methods are available for the design of raft floors, viz., conventional method (Bowles,
1982), Baker’s method (Baker, 1948), Hetenyi’s method (Hetenyi, 1964) and numerical methods (Desai et al
2000). Out of the above Hetenyi,s method ( Verma, 1981) is widely adopted for analysis and design of barrages
raft floor in India as this method has also been recommended by Indian standard code (IS:11130-1984). The finite
element analyses of barrages have been carried out by Sarkar (2001) and Sasidhar (2002). A comparative
analysis of a barrage raft floor has also been carried by Venkatesh et al (2004) and Pandey et al (2005) on
homogeneous foundation media. However, this paper is an attempt to analyze the behaviour of barrage raft floor
resting on varying foundation media with intruded rock below the raft floor.

2 Overview of the finite element method

The finite element method is a numerical procedure for analyzing structures and continua. It is a powerful tool in
structural analysis of simple to complicated geometries. In the recent years with the advent of compact and
powerful computers, the analyses performed by finite element method have become more acceptable. Finite
element program “ANSYS” has been employed in the present study. The basic steps involved in the finite
element method are as mentioned below.
I. Discretization of the continuum.
II. Calculation of the element stiffness matrices.
III. Assembling the element stiffness matrices.

3224
 IV. Calculation of the element load vectors.
V. Assembling the element load vectors.
VI. Imposition of boundary conditions.
VII. Imposition of external forces.
VIII. Calculation of the displacement vectors.
IX. Calculation of the strains and stress field.

A detailed discussion on linear constitutive relationship for elastic finite element analysis is beyond the scope of
this paper but well documented in standard literature. (Desai and Abel, 2000; Krishnamurthy, 2002; Cook et al.,
1989; Bathe, 1982; Zienkiewicz, 1977).

3 Idealization of barrage bays with raft floor

Typical barrage bays 3-4, have been selected for this study. The barrage raft floor of bays 3-4 is separated by
expansion joints from rest of the bays. The plan of bays 3-4 (Figure 1) with three sections of the barrage raft floor
in transverse direction (across the flow) i.e. upstream section (A-A), section (B-B) and downstream section (C-C)
at different distances from upstream edge have been chosen for the comparison of behaviour. The barrage raft
floor with cut-off of bays 3-4 are completely resting on alluvial soil with single and double pier but suddenly at the
edge of bay 4 towards bay 5 there is discontinuity in foundation soil media due to presence of hard rock as shown
in typical transverse section of the bays 3-4 (Figure 2).

SOIL

4.5 m Upstream Section

2.25 m
2.25 m

11 m 11 m
A 4 A
13.8 m
49.526 m

Bay 3 Bay 4
SOIL Ogee Section ROCK
Double Pier
Double Pier

Single Pier

B B

24.6 m

Downstream Section

C C

30.5 m
SOIL
Figure 1. Plan of barrage bays 3-4.
Double Pier
Double Pier

Single Pier

Raft Floor
Side Cut-off
Rock
Expansion Soil Expansion
joint joint

Figure 2. Transverse section of barrage bays 3-4.

4 Condition of analysis
The investigations are based on linear elastic model for representative load cases. The self-weight of

3225
the soil and rock media has not been considered as it has been assumed that entire soil and rock
media is already settled by its own weight. It has also been assumed that within the entire soil & rock
media, elastic modulus and Poisson’s ratio remain the same as well as soil and rock junction has
been assumed to be in contact with each other. The intrusion of the rock has been taken up to study
the effect of intruded rock on barrage raft floor bays 3-4, if it extends toward the barrage raft floor
bay 4 and bay 3 from bottom and side. The intrusion of rock is taken in to consideration from the
bottom towards the raft and from the edge of bay 4 towards bay 3. The vertical extent of rock is
studied in 3 cases and the lateral extent of rock is studied in 2 cases thus making a total of 6 cases as
given in Table 1.

Table 1. Vertical and lateral intrusion of rock below bays 3-4 raft floor.
Vertical extent of Lateral extent of rock
Barrage Bays 3-4 rock (From edge of the bay4
(From bottom in m) double pier side in m)
Case 1
15.25
(Figure 3) 20
Case 2 30.5
Case 3 15.25
40
Case 4 30.5
Case 5 15.25
Case 6 60
30.5
(Figure 4)

Bay 3

Bay 4
Bay 3

Bay 4

30.5 m

Rock
Soil Soil Rock

Intruded
60 m

15.25 m Rock

Intruded
20 m

Rock

Figure 3. Intrusion of rock (Case1) in bays 3-4. Figure 4. Intrusion of rock (Case 6) in bays 3-4.

5 Finite element modelling and meshing

Three-dimensional eight noded isoparametric brick elements have been used for the modeling of soil and rock
media (King, 1977). The cut-off, pier, abutment wall and beam have also been modeled using eight noded
isoparametric brick elements. The element is defined by eight nodes having three degrees of freedom at each
node, translations in the nodal x, y, and z directions. The four noded three-dimensional isoparametric shell
elements have been used for barrage raft floor modeling to simulate the behaviour of barrage raft floor as plate
bending element (King, 1977), having six degrees of freedom per node capable of taking loads normal to the
plane. In this model the depth of the soil and rock media considered is 80m from the crest level. The extent of
surrounding soil and rock up to 35m on both sides of the transverse section of the raft and 50m on both in
upstream and downstream side equivalent to the length of the raft floor along the flow has been considered.
Several iterations were made for refining the mesh of the models from coarser to finer till the values of moments
at the same section under study in the two consecutive models converged under gravity load. The material
properties of various components of barrage as well as for soil and rock media are as given in Table2. The
adopted model with finite element mesh consisting of the pier and beam structure with the supporting ogee
shaped raft floor is shown in Figure 5. The finite element mesh for the entire structure-raft-foundation soil and

3226
rock system has been presented in Figure 6. The dark grey portion in the figure resembles the rock portion. The
total number of elements used for the adopted finite element model is 18744, which resulted in 21204 nodes in
the model. The boundary condition imposed on the finite element models consist of restraining the limiting
boundary of the foundation soil and rock in such manner that displacement normal to the boundary surface are
restrained i.e. the ends along and across the direction of flow, foundation media is restrained against the
horizontal displacement but the base of the foundation media at the depth of 80 m have been restrained against
vertical and horizontal displacement.

Table 2. Material properties used in bays 3-4 model.

Components Modulus of Unit Poisson’s Ratio
Elasticity (E) Weight (μ)
2
(kN/m ) (γ)
3
(kN/m )
7
Pier/Abutment 2.5 x 10 25 0.15
Raft Floor 2.5 x 107 25 0.16
7
Cut-off 2.4 x 10 25 0.18
5
Foundation soil 1 x 10 20 0.3
Foundation rock 1 x 107 26 0.25

Single pier
Double
pier Double
pier

Raft
floor

Upstream
Cut-off Side Cut-off

Downstream Cut-off

Figure 5. 3D-finite element discretization of the pier and raft floor with cut-off of the bays 3-4.

Rock

Soil Soil

Figure 6. 3D-Finite element discretization of the pier, raft, soil and rock system of bays 3-4.

3227
6 Influence of intrusion of rock on raft floor
Analyses have been carried out for intruded and non-intruded cases. The effects of intrusion of rock on the
moments of raft floor for all the six cases have been shown in Figures 7 to 9 for upstream, ogee and downstream
sections respectively for gravity load case. The compared moments for all the cases indicates that there is a
significant difference in the moments when rock is intruded beyond 40 m from the bottom of the foundation
media. The moments obtained from first four cases resemble the non-intruded condition whereas the last two
cases represent the change in the magnitude of the moments at upstream, ogee and downstream sections of
barrage raft floor. The quantitative variations at the centre of raft between first four and last two cases at
upstream and ogee section are around 16% whereas for downstream section the variation is about 25% which
specify that intrusion is affecting the moments largely at downstream section compared to upstream and ogee
section.
Figures 10 to 12 represents deformations of raft floor at different sections considering intruded and non intruded
cases. Differential settlement of raft floor can be seen as barrage raft floor is resting on varying foundation media.
Significant reductions in the deformations have been seen with the increase in the intrusion of the rock.
Deformations of the first four cases resemble the non-intruded condition, whereas the last two cases which are at
the height of 40 to 60 m indicate the large quantitative difference at all the section compared to the other cases.
The moments and deformations of raft floor have reduced as the overall stiffness of the foundation media has
also increased.

Figure 7. Moments 'Mz' at upstream section (A-A) considering gravity load with and without rock intrusion.

Figure 8. Moments 'Mz' at ogee section (B-B) considering gravity load with and without rock intrusion.

3228
Figure 9. Moments 'Mz' at downstream section (C-C) considering gravity load with and without rock intrusion.

Figure 10. Deformation at upstream section (A-A) considering gravity load with and without rock intrusion

Figure 11. Deformation at ogee section (B-B) considering gravity load with and without rock intrusion

3229
 Figure 12. Deformation at downstream section (C-C) considering gravity load with and without rock intrusion

7 Conclusions
Based on the comparison of moments and deformations at a various sections of the barrage raft floor for intrution
and non-intrution of rock, the following conclusions may be drawn:

i) The moments and deformations reduces with the intrusion of rock and it has been attributed to the
increase in the stiffness of foundation media which influenced the behaviour of raft floor at every
section.
ii) The intrusion of rock from 40 m to 60 m was most critical from the foundation base which signifies
that intrusion of rock within the 40 m of ground level is going to change the behaviour of raft floor
significantly.
iii) The three-dimensional system as represented by finite element method is able to account for the
geometrical disposition of the raft floor and spatial variation of stiffness for foundation media.

8 References
Baker A.L.L. 1948. Raft foundations, the soil line method of design, Concrete Publications Limited, London (UK).
Bathe K.J. 1982. Finite Element Procedures in Engineering Analysis, Prentice-Hall, Englewood Cliffs, New Jersey (USA).
Bowles J.E. 1982. Foundation analysis and design, McGraw Hill, New York (USA).
Desai C.S., Abel J.F. 2000. Introduction to the finite element method, CBS Publisher and Distributors, New Delhi (India).
Verma C.V.J. 1981. Design and Construction features of selected Barrages in India, Publication No. 149, Central Board of
Irrigation and Power, New Delhi (India).
Hetenyi M. 1964. Beams on elastic foundations, Seventh printing, The University of Michigan Press (USA).
IS: 11130: 1984. Criteria for structural design of barrages and weirs, Bureau of Indian Standards, New Delhi (India).
King G.J.W. 1977. An Introduction to Superstructure/Raft/Soil Interaction, Proc. Int. Symposium on Soil-Structure Interaction,
University of Roorkee, Roorkee (India), 453-466.
Krishnamoorthy C. S. 2002. Finite element analysis, Theory and Programming, Ninth reprint, Tata McGraw Hill Publishing
Company Limited, New Delhi (India).
Pandey A.D., Sharma N., Venkatesh K., Kulkarni M.D. 2005. Comparative study on analysis of barrage raft by Hetenyi’s
method and FEM, Water and Energy International Journal, 62 (1), 40-47.
Sarkar S. 2001. FEM analysis of barrage under varying subsoils condition, M. Tech. Dissertation, Department of Earthquake
Engineering, IIT Roorkee (India).
Sasidhar T. 2002. 3-D finite element analysis of a barrage, M. Tech. Dissertation, Department of Earthquake Engineering, IIT
Roorkee (India).
Venkatesh K., Pandey A. D., Samadhiya N. K. 2004. Comparative analysis of raft foundation for a barrage in India. Proc. Int.
Conf. on Geotechnical Engineering, Sharjah ( UAE), 468-473.
Zienkiewicz O.C. 1977. The Finite element method, McGraw-Hill, London (UK).

3230