You are on page 1of 28

# CHAPTER 7

Engineers are often interested in knowing the direction and quantity of flow through porous media. The pore pressure variations resulting from the flow process are also of interest. This information is required in predicting the volume change and shear strength change associated with the flow of water or air. Chapter 5 outlined the driving potentials and the flow laws that govern the behavior of water flow, air flow, and air diffision through water. The hydraulic properties of the soil with respect to each fluid phase were given in terms of coefficients of permeability. These properties are required in the application of the flow laws to engineering problems. In Chapter 6, various methods commonly used to obtain coefficients of permeability were described. This chapter presents the application of the flow laws and the associated coefficients of permeability to the analysis of practical seepage problems. The analysis is performed for air and water flow under isothermal conditions. The effect of air diffision through water, air dissolving into water, and the movement of water vapor are not given consideration. Seepage problems are usually categorized as steady-state or unsteady-state flow analyses. For steady-state analyses, the hydraulic head and the coefficient of permeability at any point in the soil mass =main constant with respect to time. For unsteady-state flow analyses, the hydraulic head (and possibly the coefficient of permeability) change with respect to time. Changes m usually in response to a change in the boundary conditions with respect to time. Steadystate flow analyses are considered in this chapter. The quantity of flow of an incompressible fluid such as water IS expressed in terms of a flux, q. Flux is equal to a flow rate, o, multiplied by a cross-sectional area, A. On the other hand, the quantity of flow of a compressible fluid such as air is usually expressed in terms of a mass rate. The governing partial differential seepage equations are derived in a manner consistent with the conservation of mass. The conservation of mass for steady-state seepage of an incompressible fluid dictates that the flux into an element must equal the flux out of an element. In other words, the
150

Co py rig hte dM ate ria l

net flux must be zero at any point in the soil mass. For a compressible fluid, the net mass rate through an element must be zero in order to satisfy the conservation of mass for steady-state seepage conditions.

7.1 STEADY-STATE WATER F'LOW

The slow movement of water through soil is commonly referred to as seepage or percolation. Seepage analyses may form an important part of studies related to slope stability, groundwater contamination control, and earth dam design. Seepage analyses involve the computation of the rate and direction of water flow and the pore-water pressure distributions within the flow regime. The flow of water in the saturated zone has been the primary concern in conventional seepage analyses. However, water flow in the unsaturated zone is of increasing interest to engineers. For example, the seepage through a dam has commonly been analyzed by considering only the zone below an empirically computed line of seepage (Casagrande, 1937). Recent studies have illustrated that there is a continuous flow of water between the saturated and unsaturated zones, as shown in Fig. 7.l(a) (Freeze, 1971; Papagiannakis and Fredlund, 1984). Another example is shown in Fig. 7.1(b), which illustrates the effect of infiltration and evaporation on the phreatic surface within a slope. A constant water flux across the surface boundary may develop a steady-state water flux through the unsaturated zone of the slope. Water flow through unsaturated soils is governed by the same law as flow through saturated soils (Le., Darcy's law). The main difference is that the water coefficient of permeability is assumed to be a constant for saturated soils, while it must be assumed to be a function of suction, water content, or some other variable for unsaturated soils. Also, the pore-water pressure generally has a positive gauge value in a saturated soil and a negative gauge value in an unsaturated soil. In spite of these differences, the formulation of the partial differential flow equation is similar in

151

largest permeability, and is called the minor coefficient of permeability.
Saturated zone Horizontal drain

l l l l i
Rainfall

\ \
Saturated zone

Figure 7.1 Examples involving flow through unsaturated soils. (a) Water flow in the saturated and unsaturated zones of an earth dam; (b) water flow across the boundary of a slope.

both cases. There is also a smooth transition when going from the unsaturated to the saturated case (Fredlund, 1981).

7.1.1 Variation of Coefecient of Permeability with Space for an Unsaturated S o i l For steady-state seepage analyses, the coefficient of perme-

ability is a constant with respect to time at each point in a soil. However, the coefficient of permeability usually varies from one point to another in an unsaturated soil. A spatial variation in permeability in a saturated soil can be attributed to a heterogeneous distribution of the soil solids. For unsaturated soils, it is more appropriate to consider the heterogeneous volume distribution of the pore-fluid (Le., pore-water). This is the main reason for a spatial variation in the coefficient of permeability. Although the soil solid distribution may be homogeneous, the pore-fluid volume distribution can be heterogeneous due to spatial variations in matric suction. A point with a high matric suction (or a low water content) has a lower water coefficient of permeability than a point having a low matric suction. Several functional relationships between the water coefficient of permeability and matric suction [Le., k,(u, uw)]or volumetric water content [i.e., kw(Ow)]have been described in Chapter 5. Coefficientsof permeability for different points in a soil are obtained from the permeability function. The magnitude of the coefficient of permeability depends on the matric suction (or water content). In addition, the coefficient of permeability at a point may vary with respect to direction. This condition is referred to as anisotropy. The largest coefficient of permeability is called the major coefficient of permeability. The smallest coefficient of permeability is in a direction perpendicular to the

Co py rig hte dM ate ria l
Water table

Evaporation

b
(b)

Heterogeneous, Isohpic Steady-State Seepage Permeability conditions in unsaturated soils can be classified into three groups, as illustrated in Fig. 7.2. This classification is based on the pattern of permeability variation. A soil is called heterogeneous, isotropic if the coefficient , , is equal to the coefof permeability in the xdiwtion, k ficient of permeability in the y-direction at any point within the soil mass (Le., k , = ,& , at A and & , = kY at B) [see Fig. 7.2(a)]. However, the magnitude of the coefficient of permeability can vary from point A to point B, depending upon the matric suction in the soil. The variation in the coefficient of permeability with respect to matric suction is often assumed to follow a single-valued functional relationship.

Heterogeneous, Anisohpic Steady-State Seepage Figure 7.2(b) illustrates the heterogeneous, anisotropic case. Here, the ratio of the coefficient of permeability in the xdirection, k , , to the coefficient of permeability in the y-direction, ky,is a constant at any point (i.e., (&,/ky) at A

‘I

(C)

Figure 7.2 Coefficient of pemeability variations in an unsaturated soil. (a) Heterogeneous, isotropic conditions; (b) heterogeneous, anisotropic conditions; (c) continuous variation in permeability with space.

Copyright © 1993 John Wiley & Sons

u. k. In the laboratory measurement of the coefficient of permeability./dy = change in water coefficient of permeability in the y-direction due to a change in matric suction. The steady-state pmedure for measuring the water coefficient of permeability in the laboratory is also a one-dimensional flow example. where k. as illustrated in Chapter 6 . If the saturated soil is heterogeneous(e. dy..4).5) where k. If a saturated soil is homogeneous. u. (7. dy. A comparison of Eqs. As a result.4) and (7. constant coefficient of permeability into Eq.com .5) reveals a similar form.4 One-dimensional water flow through an unsaturated soil element. d Y (7. the hydraulic heads are controlled as boundary conditions at the top and bottom of the soil specimen. However. and & gives the following nonlinear differential equation: dh + dy-. however.. =0 k. dy.) term will be written as k .(u. 7 d Y d Y d Y Substituting Darcy's law into the above equation yields + - (7. the nonlinearity in the unsaturated soil flow equation produces the same form of equation as that required for a heterogeneous.layered soil).5) Copyright © 1993 John Wiley & Sons Retrieved from: www. giving a fixed zero pore-water pressure head. In other words.4) where dk. the flow equation can be written as follows: d'h. In a saturated soil..(u. dz = dimensions in the x-. dk dh. where u. for simplicity. the heterogeneous distribution of the soil solids is the primary factor producing a varying coefficient of permeability. The net flow can be written as follows: 5 dx dy dz = 0. the coefficient of permeability also varies. the water coefficient of permeability.j~.3) and considering the nonzero dimensions for ah.e. Techniques to analyze both head and flux boundary conditions are explained in the following sections. . Since matric suction varies from one location to another. . k .2) The nonlinearity of Eq. = hydraulic head (i. 7. Equation (7.) = water coefficient of permeability as a function of matric suction which varies with location in the ydirection dhw/dy = hydraulic head gradient in the y-direction h. will again vary with respect to location. (7. the coefficient of permeability. The element has infinitesimaldx. k. and z-directions. The flow rate.knovel. (7. and dz dimensions. = saturated coefficient of permeability.7. is assumed to be positive when water flows upward in the y-direction. dz Figure 7.-g d2hw Q%vy The above one-dimensional flow cases involve flux boundary conditions. gravitational head plus pore-water pressure head).3) can be used to solve for the hydraulic head distribution in the y-direction through a soil mass. for the remainder of the formulation.g . the k. the coefficient of permeability is constant for the soil mass. Substituting a nonzero. =0 (7. A steady rate of evaporation or infiltration can be used as the boundary condition at ground surface. Soil thickness.u. (7.1 STEADY-STATE WATER FLOW 153 Formulation for One-Dimensional Plow Consider an unsaturated soil element with one-dimensional water flow in the y-direction (Fig. Rewriting Eq. In an unsaturated soil. Continuity requires that the volume of water flowing in and out of the element must be equal for steadystate conditions: Co py rig hte dM ate ria l k. When the soil becomes saturated. The water table acts as the lower boundary condition.4) is caused by its second term. which accounts for the variation in permeability with respect to space. saturated coefficient of permeability. the variation in the coefficient of permeability is caused by the heterogeneous distribution of the pore-fluid volume occurring as a result of different matric suction values.. y-. can be taken as being equal to a single-valued. dx.. saturated soil. = water flow rate across a unit area of the soil in the y-direction respectively..

A numerical solution can be used as an alternative to a closed-form solution.13) and (7. in the xdirection. i.1. can be rearranged after assuming equal Equation (7. as shown in Fig. The finite difference method will be used to illustrate the solution to the flow equation for an unsaturated soil. varies in space.14) Equations (7. 4 )can be written in a finite difference form for point (i): ..(i + 1 ) I Yi. dy3 i water pressure head. Similar approximations can be derived for a function.e.12)and again neglecting the higher order derivatives gives the second derivative of the function at point (0: (\$)i = hi+l + hi-1 . (i . One-dirnensional.14). A central difference approximation is then applied to the hydraulic head and coefficient of permeability derivatives in Eq. 7.7. at point i.2Ay ' (7. Eq. h. h( y).av'av. 7 . The use of an iterative finite difference technique in solving flow problems is illustrated in the following sections. 4 ) . Consider the situation where a function. 7 ) .15) where kwy(r3. while another illustrates the use of a flux boundary condition. (7.12) the higher order derivatives result in the first derivative of the function at point (i): (\$)i Co py rig hte dM ate ria l (s).l ) .A tensiometer is installed near the ground surface to measure the negative pore-water pressure.15) Copyright © 1993 John Wiley & Sons Retrieved from: www... (7.. h(x). kV(i- kv(i+ = water coefficients of perme- h. + Figure 7. ( 7 . +e (e) 2! -&ye) .hi+l . (7. One example involves the use of a head boundary condition. Ay. h( y). (7.l ) . h.11)and neglecting Subtracting Eq. (7.6 Function h ( y ) shown in a finite difference form.1 STEADY-STATE WATER FLOW 155 M i t e Dztfcrence Method The seepage differential equation can be written in a finite difference form. and (i l). from Eq.6. hp.13) (7. steady-state flow is assumed when the tensiometer reading remains constant with respect to time.I). and (i + l ) .12) 3! ' Ay dy2 + * .knovel.11)and (7. respectively = hydraulic heads at points (i).11) hi (t).For example. The hydraulic head distribution along the length of the column is given by Q. . 7.Values of the function at points along the curve can be computed using Taylor series: h i + l = hi + AY hi- 1 where i .7.The column length is first discretized into (n) equally spaced nodal points at a distance Ay apart (Fig. ( 7 . These approximations can be used to solve the differential equation.(. the water table) is equal to zero. The linearity in the hydraulic head and the pore-water pressure head distributions is the result of the conswt water coefficients of permeability.2hi AY (7. Yi Yet1 Y ability in the ydirection at points (i). ( 7 .+. . 4 ) . respectively. Head Boundary Condition Steady-state evapomtion from a column of unsaturated soil is illustrated in Fig. Summing Eqs. (i .14)are called the central difference approximations for the first and second derivatives of the function.13)and (7.This equation can be solved using the finite difference approximations in Eqs. 4 ) ]requires a more complex solution than that for a saturated soil.hi-1 . ( 7 .1 . i + 1 = three consecutive points spaced at increments. The pore-water pressure at the base of the column (i.e. distribution is linear under steadystate seepageconditions. Eq. (7.com .. The equation for one-dimensional steady-state flow through an unsaturated soil [i. + Ay2 d2h (\$)i = +\$(\$)>.

the flux is assumed to be positive in an upward direction and negative in a downward direction A = cmss-sectional area of the soil column.1 STEADY-STATE WATER vw FLOW 157 infiltration may be established as a result of sprinkling irrigation. is known.4). The hydraulic head boundary condition at the ground surface is assumed to be unknown.knovel.. The hydraulic head distribution can be determined by solving the finite difference form of the steady-state flow equation [Le. i i Water flux (4w) (n . point (i) can be expressed in terms of the hydraulic heads 1) and (i 1) using Darcy's law: at points (i + - where qW = water flux through the soil column during the steady-state flow. Discretization "(Q . I i . Ay (Fig.7.8 Steady-state evaporation through an unsaturated soil column. The water flux at Steady-state infiltration O Co py rig hte dM ate ria l m u r e 7. The soil column is first discretized into (n) nodal points with an equal spacing. (7.9).16)].9 One-dimensional steady-state water flow through an unsaturated soil with a flux boundary condition.com .1 ) ~ @ 1 qw Unsaturated soil L Water table 't Iqw Datum T 1@ -hwl=O Nodal points I t Figure 7. 7. and is constant throughout the soil column for steady-state conditions. Eq. However. qW. Copyright © 1993 John Wiley & Sons Retrieved from: www. (7. Steady-state flow can be described using EQ. the water flux. Boundary conditions L I I i i n i . Let us assume a constant downward water flux of qW.

Table 7 . steady-state flow can be expressed as follows: .) = water coefficients of permeability as a function of matric suction. Eq. ak. respectively.28) ay ay where k. isotropic case. is positive when water flows in the positive x-direction. = water coefficient of permeability in the x.26)becomes more obvious after The nonlinearity of E an expansion of the equation: (7. = 0 (7.2(b)J.7.25)results in a nonlinear partial differential equation: =0 (7.. and the y-direction.. (7. Water flow is in the cross-sectional plane of the dam. and dz is considered. (7. (kwJkv). .. Until recently. ax ax ak. k. for simplicity. (7.27)produces nonlinearity in the governing flow equation./ax = change in water coefficient of permeability in the x-direction.11 Two-dimensional water flow through an unsaturated soil element.u. A soil element with infinitesimal dimensions of dr.(u. k. The flow rate.and y-directions. = k .. 7. v.27)can be written as follows: k.com . is assumed to be constant at any point within the soil mass. 1 summarizes the relevant equations for two-dimensional steady-state flow through unsaturated soils. (3 +T ) + -. . Copyright © 1993 John Wiley & Sons Retrieved from: www. k. . unsaturated soil [Fig.11).26)describes the hydraulic head distribution in the x-y plane for steady-state water flow.. anisotropic. The flow rate.u..1 STEADY-STATE WATER FLOW 159 Co py rig hte dM ate ria l where k. The following two-dimensional formulation is an expanded form of the previous onedimensional flow equation. The coefficients of permeability in the x-direction. k. 7. but flow is assumed to be two-dimensional (Fig.) are written as k . are assumed to be related to the matric suction by the same permeability function.and y-directions ace equal (i.24) vw+7dy dz The spatial variation of the coefficient of permeability given in the thirrl and fourth terms in Eq.) and k. The formulation is called an uncoupled solution since it only satisfies continuity. For the remainder of the formulations. = water flow rate across a unit area of the soil in the x-direction.25) Substituting Darcy’s laws into Eq. Figure 7.. The analysis presented herein assumes that water flows in both the saturated and unsaturated zones in response to a hydraulic head driving potential. Continuity for two-dimensional. Therefore.and y-directions. (7. = k. and k. ah.(u. the coefficients of permeability in the x.3 Two-Dimensional Flow Seepage through an earth dam is a classical example of two-dimensional flow. vWx. the net flux in the x..u.).+ -a2h. For a rigorous formulation of two-dimensional flow. while flow perpendicular to the plane is assumed to be negligible. where v ./ax = hydraulic head gradient in the x-direction.knovel. is positive for flow in the positive y-direction. ah. & (7.26) Formulation for Two-Dimensional Flow The following derivation is for the general case of a hererogeneous.1.27) where ak. dy. the permeability can vary with location in the xdirection ah. For the heterogeneous. v. kwx(u.and ydirections is.). The ratio of the coefficients of permeability in the x.) dr dz = 0 (7. continuity should be coupled with the force equilibrium equations. Therefore. Equation (7.e.(u.u.. it has been conventional practice to neglect the flow of water in the unsaturated zone of the dam. 7.

There- Table 7. k. uppermost boundary when constructing the flownet. steady-state seepage beneath a sheet pile wall has the boundary conditions shown in Fig. For the saturated portion. +-ax ax Eq. Isotropic fore.12(a). Flow through the saturated soil can be considered as a special case of flow through an unsaturated soil. homoge- Co py rig hte dM ate ria l Copyright © 1993 John Wiley & Sons Retrieved from: www. 1979). isotropic saturated soil [Le.160 7 STEADY-STATE FLOW Table 7. k.32) Homogeneous Eq. analog methods. (7. or numerical methods. but also an uppermost flow line. the water coefficient of permeability becomes equal to the saturated coefficient of permeability.12(b)].. ah.29) Eq. + ksy aY2 ak. The flownet solution has been used extensively to analyze problems involving seepage through saturated soils.2. (7.34)] is called the Laplacian equation. The saturated coefficients of permeability may vary with respect to location due to heterogeneity. However. (7. A summary of steady-state equations for saturated soils under different conditions is presented in Table 7. The flownet solution results in two families of curves. The saturated coefficients of permeability in the x. and is explained in most soil mechanics textbooks. The solution can be obtained using closed-form analytical methods. 7. 7. A head boundary condition or an impermeable boundary condition can readily be imposed for most saturated soils problems.31) Eq. In the past. The solution of this equation describes the head at all points in a soil mass.and y-directions. steady-state seepage through saturated-unsaturated soils can be analyzed simultaneously using the same governing equation. may not be equal due to anisotropy. the conditions are more difficult to assign when dealing with unsaturated soils.. Eq. It is a linear. Eq. (7. Either the head or the flux is prescribed along the boundary. (7.34) are specialized forms that can be derived from the steady-state flow equation for unsaturated soils [i. The assumption that the free surface is a phreatic line requires that the pore-water pressures be zero along this line. 1937)... The flownet technique has been developed primarily to analyze steady-state seepage through isotropic. it is assumed that there is no flow across the free surface.27)]. ah. This uppermost boundary [i. Often. A network of flow lines and equipotential lines is sketched by trial and e m r in order to satisfy the boundary conditions and the requirement of right-angled. Equipotential lines must intersect the free surface at right angles since it is also an uppermost flow line. (7. The position of the free surface is usually determined using an empirical procedure (Casagrande. 7.33) Eq. referred to as flow lines and equipotential lines.respectively.com . partial differential equation. The flownet can then be constructed.12(b)] is not only considered to be a phreatic line.e.1 Two-Dimensional Steady-State Equations for Unsaturated Soils Heterogeneous. Anisotropic Heterogeneous.30) Seepage through a dam involves flow through the unsaturated and saturated zones. In other words. The uppennost boundary is referred to as a free surface under these special conditions (Freeze and Cherry. Eq.31)-(7. Let us consider steady-state seepage through an e a r t h dam [Fig. and ksy.. the position of the free surface is unknown. ax ak.2 Two-Dimensional Steady-State Equations for Saturated Soils Anisotropic Isotropic ksx a’h. 1937).34) Solutionsfor Two-Dimensional Flow The differential equation describing two-dimensional steady-state flow through a homogeneous. Boundary conditions for the soil domain must be known prior to the constructionof the flownet. and it must be approximated prior to constructing the flownet. the assumption has generally been made that the flow of water through the unsaturated zone is negligible due to its low permeability. In other words. a graphical method referred to as drawing a “flownet” has been used to solve the Laplacian equation (Casagrande.e. However. line BC in Fig.knovel. +-ax Heterogeneous a’h. For example. (7. (7. the phreatic line is assumed to behave as an impervious. A boundary condition exception is the case of a free surface. equidimensional elements. Equations (7.

the boundary conditions that are satisfied on the free surface specify that the pressure head must be atmospheric and the surface must be a streamline. Triangular and quadrilateral shapes of elements are commonly used for Copyright © 1993 John Wiley & Sons Retrieved from: www.. 7. Figure 7.13 compares two solutions of a saturated-unsaturated soil system. . There is an inherent problem associated with applying the flownet technique to saturated-unsaturated flow. saturated soils.1 STEADY-STATE WATER FLOW 161 neous. The flownet in Fig.varies with respect to the matric suction in the unsaturated zone. Steady-state flow in the saturated and unsaturated zones can be analyzed simultaneously using the same governing equation [Le. The water coefficient of permeability in the saturated zone is equal to k . The flownet technique is no longer applicableto saturated-unsaturated flow modeling when the governing flow equation is not of the Laplacian form. heterogeneous soil systems. isotropic earth dam. (b) steady-state seepage throughout a homogeneous. isotropic saturated soil. The general flow equation can be solved using a numerical technique such as the finite difference or the finite element method. The flownet shown in Fig. The incorrect assumption regarding the uppermost boundary condition can be avoided by realizing that there is flow between the saturated and unsaturated zones (Freeze.12 Flownet constmctions to solve the Laplacian equation. Papagiannakis and Fnxllund. Eq. The free surface is a close approximation of the phreatic line from the saturated-unsaturated flow modeling..knovel.13(b) was constructed using an empirically defined free surface. Seepage Analysis Using the lcpnite Element Method The application of the finite element method requires the discretizationof the soil mass into elements. thereby neglecting flow in the unsaturated zone. The flownet technique becomes complex and difficult to use when analyzing anisotropic. (a) Steady-state seepage throughout a homogeneous. (7. Both zones are treated as a single domain. Co py rig hte dM ate ria l I Boundary conditions: AB: hw=H1 BC and DE: qw= 0 EF: hw = HZ AH and FG: qw = 0 HG: qw=O Assumed impervious impervious Equjpotential'lines Y Horizontal drain Boundary conditions: AB: hw=H1 BC: free surface.13(a) was drawn based on a numerical method solution for a saturated-unsaturated flow system. 7. . it's location is unkown CD: hw=O DA qw=O (b) Figure 7.. 1971.com . Whereas the first of these conditions is true." Figure 7. the second is not. ". Freeze (1971) stated that. The water coefficient of permeability. . 1984).26)].14 shows several typical solutions by Freeze (1971) involving saturated-unsaturated flow modeling.7.. The following section briefly describes the fonnulation of the finite element method in analyzing steady-state seepage through saturated-unsaturated soils. k.

x3Y2) + (Y2 .15 shows the cross section of a dam that has been discretized using triangular elements.162 I STEADY-STATE FLOW - ----__ = Equipotential line r n =Impervious Flow line a) Homogeneous boundary bl Cutoff E E I . - - {L} = matrix of the element area coordi- Copyright © 1993 John Wiley & Sons Retrieved from: www. &. (7. .e. 1984): ]s{L}%wds = 0 where nates (i. The finite element formulation for steady-state seepage in two dimensions has been derived using the Galerkin principle of weighted residuals (Papagiannakis and Fredlund. 5 = area coordinates of points in the element that are related to the Co py rig hte dM ate ria l c) Internal.13 Steady-state seepage in a saturated-unsaturated soil system.. The hydraulic head at each nodal point is obtained by solving the governing flow equation and applying the boundary conditions. y = Cartesian coordinates of a point within the element A = area of the element [* :*-J = matrix of the water coefficients of permeability (Le.X3)Y) r. { L . 3) = Cartesian coordinates of the three nodal points of an element x. 1979).com . yi(i = 1. that is.X2)Y) k? = 1 / u { ( X 3 Y l XlY3) + (Y3 .)X + (x1 . 1 6 ) : "1 = 1 / u ( ( x 2 Y 3 . The lines separating the elements intersect at nodal points. & &}) L .]) {h..knovel. = 1/244{(XlY2 .Y2)X + (x2 X l ) Y ) x i .Y.Y3)X + (x3 . 1971). basal Freesurface I _---- = Flow line line = Equipotential (b) e) Sloping core and Figure 7. 2. (b) flownet constxuction by considering flow in the saturated zone (after Freeze and Cherry.14 Typical solutions for saturated-unsaturated flow modeling of various dam sections (fromFneeze. \ \ ' ' I : : ! D Figure 7. 7 . (a) Flownet constructed from \$arurare~-u~\$a~ura~e~ flow modeling.35) Cartesian coordinates of nodal points as follows (Fig.X2Y1) + (Y1 . [k.} = matrix of hydraulic heads at the nodal points. Figure 7. two-dimensional problems.

The second term in Eq.com . (7.36).15 Discretized cross section of a dam for finite element analysis.35) yields a simplified form for the governing flow equation: 1“ [BIT[k. respectively. (7. the coefficient of permeability within an element is set to a value depending upon the average matric suction at the three nodal points. in the vertical direction must be converted to a normal flow rate. a spec- yl Figure 7. The normal flow rate is in turn converted to a nodal flow.” A negative nodal flow indicates evaporation or evapotranspiration at the node and that the node acts as a “sink.XI) 3- * = Nodal point i =1 yl) -W (1.Y2) - x2) (XI . In this way. the second term in Eq. (7.” When the flow rate acmss a boundary is zero (e. Q. impervious boundary).37) Copyright © 1993 John Wiley & Sons Retrieved from: www. Specified hydraulic heads at the boundary nodes are called Dirichlet boundary conditions.0. Equation (7. I.36) accounts for the specified flow rate measured in a dimtion normal to the boundary. 7.7.and y-dimtions can be computed for an element by taking the derivative of the element hydraulic heads with respect to n and y. This is performed while satisfying nodal compatibility (Desai and Abel.36) is nonlinear because the coefficients of permeability are a function of matric suction. ow = external water flow rate in a direc- tion perpendicular to the boundary of the element S = perimeter of the element..knovel. QWi and Q W j . as illustrated in Fig.l[BldA {h. which can be written as -[ 24 1 Either the hydraulic head or the flow rate must be specified at boundary nodal points.17 shows the computation of the nodal flows.Yl) (Y1 . A positive nodal flow signifies that there is infiltration at the node or that the node acts as a “source.g. k. at the boundary nodes (i) and ( j ) .x3) (x2 . The hydraulic head gradients in the x.(u. the global flow equations are linearized and can be solved simultaneously using a Gaussian elimination technique. {h. =0 (7. Co py rig hte dM ate ria l Figure 7.0)(XI. For each iteration. The finite element equation [Eq. u.36)] can be written for each element and assembled to form a set of global flow equations.(7. The hydraulic heads are the unknown variables in Eq.1 STRADY-STATE WATER FLOW Node number 1 Element number 1 163 8 195 Rearranging Eq. Z. (7. Equation (7. For example.16 Area coordinates in relation to the Caltesian coordinates for a triangular element.17.36) is solved by using an iterative method.) - where [B] = matrix of the derivatives of the area coordinates.. 1 S tLITZ.) . Figure 7. Nodal compatibility requires that a particular node sharect by the sumunding elements must have the same hydraulic head in all of the elements (Zienkiewicz. Desai 1975a). Cartesian coordinates Area coordinates ified flow rate. which is related to the hydraulic head at the nodal points.36) disappears. The above steps are repeated until the hydraulic heads and the coefficients of permeability no longer change by a significant amount. (Segerlind 1984).). A specified flow rate across the boundary is ref e d to as a Neuman boundary condition.36) (Yz’(x3 Y3) (Y3 . respectively: - (7. The computed hydraulic head at each nodal point is again averaged to determine a new coefficient of permeability from the permeability function. 1972).

7. can be calculated from the hydraulic head gradients and the coefficients of permeability in accordance with Darcy’s law: = where v. The weighted average is computed in proportion to the element areas. 7.3*) mospheric. The first example is an isotropic earth dam with a hori0 m height of zontal drain. = hydraulic head gradient within an element in the x. The cross section and discretization of the probA 1 0 m height of water is lem are illustrated in Fig. The hydraulic head gradient and the flow rate at nodal points are computed by averaging the corresponding quantities from all elements surrounding the node. Examples o f fro-Dimensional Problems The following examples m presented to demonstrate the application of the finite element method t o steady-state seepage through saturated-unsaturated soils...knovel. 0 X lo-’ m/s. [kwl[Bl{h. applied to the upstream of the dam. vw. 7.19. A zero hy- Ngure 7. The effects of anisotropy.The saturated coefficient Of permeability. respectively. The element flow rates.17 Applied flow rate across the boundary expressed as nodal flows. Copyright © 1993 John Wiley & Sons Retrieved from: www.The 1 water on the upstream of the dam gives a 1 0 m hydraulic head at each node along the upstream face. The base of the dam is chosen as the datum.and y-directions. Lam (1984) has solved several classical problems of seepage through a dam using a saturated-unsaturated finite element seepage analysis. infiltration. i . and the use of a core and a horizontal drain on seepage through the dam are illustrated later using additional examples.1 6 4 7 STEADY-STATEFLOW where i.15. Therefore. 7.18 are numerically equal to the pore-water pressures. The pore-air pressure is assumed to be at- Co py rig hte dM ate ria l Figure 7... k. tlw = water flow rates within an element in the x- and y-directions.18.com . is 1 ..18 Specified permeability functions for analyzing steady-state seepage through a dam. as shown in Fig.) (7. the matric suction values in Fig. and can be expressed as a pore-water pressure head. h. respectively.The permeability function used in the analysis is shown as function A in Fig.

= Equipotential line (m) - = Nodal flow ratevector.----. (a) Equipotential lines and nodal flow xate vectors through the dam. (m) .00 Figure 7.7 x 10. Vw (m/s)lwith thf scale = 4.56 Free surface from flownet construction m F H o r i z o n t a l drain4 -2 m -4 m .knovel. \ Scale for geometry: -= 1. Copyright © 1993 John Wiley & Sons Retrieved from: www.----- = Isobar.19 Seepage through an isotropic earth dam with a w n t a l dmh. (b) contours of porn-water pressurn head ( i bars) through the dam.m/s - -T Phreatic line or Co py rig hte dM ate ria l Reservoir level 20 Phreatic line from finite element m o d e .com .

Copyright © 1993 John Wiley & Sons Retrieved from: www. 4 111 - = 1.---.00 c 9 Figure 7.56 Horizontal drain 4 (a) ___-__ = Isobar (m) Phreatic line or zero isobar .com .20 Seepage through an anisotropic earth dam with a horizontal drain. (a) Equipotential lines and nodal flow rate vectors throughout the dam. (b) contours of pore-water pressure head (isobars) throughout the dam.Equipotential line (m) Phreatic line Reservoir level = Nodal flow rate vector. - vw (m/s) with the scale = 4-0 m/s Scale for geometry: 9m Co py rig hte dM ate ria l ..knovel.

-0.00 (b) Figure 7. ------.---. !\s Permeabilitv - Core with permeability function B Horizontal drain 4 Phreatic line or zero isobar -.96 6. @) contours of pore-water pressure head (isobars) throughout the dam.94 5.95 9 3 d.knovel.Equipotential line (m) Reservoir level = Nodal flow rate vector.83 F 1 . vw(m/s) with the scale =1..99 .56 .s-g28fi 8.82 +-0.I .s2 8. 9m -7m Scale for geometry: - = 1.93 8. /'.com . 8 6 r 0 .99 8. 1 2 + .86 U .97 2.94 8.00 8.2 0 ~ 8.99 2.l .-O.82 299 9. (a) Equipotential lies and nodal flow rate vectors throughout the dam.= Isobar (m) 6-99 5.59 *-0.06 x lo-*m/s - Co py rig hte dM ate ria l Phreatic line -.U8 r 0 .98 6.21 Seepage through an isotmpic earth dam with a core and a horizontal drain. W 3 * . Copyright © 1993 John Wiley & Sons Retrieved from: www.90 L828.90 5. 9 6 .86 5.

(b) contours of pore-water pressure head (isobars) throughout the dam.56 I t.com - = 1.1 I __---.ontal drain under steady-state infiltration. vw(m/s) with the scale = 4. Copyright © 1993 John Wiley & Sons Retrieved from: www. (a) Equipotential lines and nodal flow rate vectors throughout the dam.knovel. m2/s Scale for geometry: Co py rig hte dM ate ria l Phreatic line or zero isobar \ (b) Fagore 7.= = Nodal flow rate vector.= Isobar (m) .2 x lO-&m/s - = tp'.22 Seepage through an isotropic earth dam with a horiz.

com . vw(m/s) with the scale = 2.----_= Equipotential line (m) = Nodal flow rate vector. (a) Quipotential lines and nodal flow rate vectors throughout the dam.5 x m/s - Scale for geometry. Copyright © 1993 John Wiley & Sons Retrieved from: www.56 Co py rig hte dM ate ria l Phreatic line \ 9 m -8 m Free surface from flownet construction ------ = Isobar (m) Phreatic line or zero isobar -00 Figure 7.knovel. (b) contours of pore-water pressure head (isobars) thmghout the dam.23 Seepage through an isotropic earth dam with an impervious lower boundary. - = 1.

as long as the soil is isotropic. Let us consider the case where steady-state water flow is established within the slope and the phreatic line is parallel to the ground surface. increasingly larger flow channels are required in order to maintain the same quantity of water flow. A negative pore-water pressure head at this point indicates that the assumed exit point is correct. As a result. The hydraulic heads at or below the assumed exit point are specified as being equal to their gravitational heads. Isobars are parallel to the phreatic line.. After convergence. During the analysis. The water flow in each channel is one-dimensional.com forming the analysis. the condition in the central section of a homogeneous dam. The coefficient of permeability is essentially independent of the pore-water pressure in the saturated zone. Therefore.7. Lines separating the flow channels are r e f e d to as flow lines.24. This is similar to Co py rig hte dM ate ria l Figure 7. The direction of the water flow indicates that there is no flow perpendicular to the phreatic line. However. In other words. The above pmcedure is repeated until the correct exit point is obtained. as ground surface is approached. the hydraulic head gradient is equal to zero in a direction perpendicular to the phreatic line. the nodal flows above the assumed exit point are set at zero. and is parallel to the phreatic line.24 Steady-state water flow through an infinite slope. as explained in Section 7. however. the same type of finite element seepage analysis can be applied to other problems involving saturated-unsaturated flow. The exit point can then be revised after each iteration by reevaluating the seepage face boundary condition. Similarly. the location of the exit point must be assumed in order to commence the analysis. Water flows through both the saturated and unsaturated zones. the lines drawn normal to theqhreatic line are equipotential lines. the pore-water pressure head at the node directly above the assumed exit point is examined. The water coefficient of permeability depends on the negative pore-water pressure or the matric suction in the unsaturated zone. In the case of the vertical column. The pore-water pressure decreases from zero at the phreatic line to some negative value at ground surface. 0therwise. the saturated zone can be subdivided into several flow channels of equal size.1. the seepage face boundary is revised by assuming a higher exit point for the phreatic line. water flux. In this case. 7. The above examples illustrate that equipotential lines and flow lines intersect at right angles for unsaturated flow problems.1 STEADY-STATE WATER FLOW 171 Znfinite Slope A slope of infinite length is illustrated in Fig. the coefficient of permeability varied in the flow direction. Therefore.2. the permeability decrcases f r o m the phreatic line to ground surface. The coefficient of permeability varies in the direction perpendicular to flow. qw)flows through each channel. Heterogeneity with respect to the coefficient of permeability results in varying distances between either the flow lines or the equipotential lines. as shown earlier. in a direction parallel to the phreatic line. these lines cross at 9 0 ' . This condition can be compared to the previous case of water flow through a vertical column. Copyright © 1993 John Wiley & Sons Retrieved from: www. and the equipotential lines were not equally distributed throughout the soil column. .knovel. An equal amount of water (Le. The above examples deal with seepage through earth dams. qw.