Seepage and consolidation

10.1 Synopsis
In the previous chapters, analysis has been restricted to either drained or undrained soil conditions. While many problems can be solved making one, or a combination, of these two extreme conditions, real soil behaviour is often time related, with the pore water pressure response dependent on soil permeability, the rate of loading and the hydraulic boundary conditions. To account for this behaviour, the seepage equations must be combined with the equilibrium and constitutive equations. This chapter briefly describes the basis behind such a coupled approach and presents the finite element equations. It is then shown how the steady state seepage equations can be obtained from these general consolidation equations. The hydraulic boundary conditions relevant to geotechnical engineering are discussed afterwards. Some nonlinear permeability models are presented, followed by a short discussion on the numerical problems associated with unconfined seepage. The chapter finishes by presenting an example of coupled finite element analysis.

10.2 Introduction
The theory presented so far in this book has been restricted to dealing with either fully drained or undrained soil behaviour. While many geotechnical problems can be solved by adopting such extreme soil conditions, real soil behaviour is usually time related, with the pore water pressure response dependent on soil permeability, the rate of loading and the hydraulic boundary conditions. To account for such behaviour it is necessary to combine the equations governing the flow of pore fluid through the soil skeleton, with the equations governing the deformation of the soil due to loading. Such theory is called coupled, as it essentially couples pore fluid flow and stress strain behaviour together. The chapter begins by presenting the theory behind the coupled finite element approach. This results in both displacement and pore fluid pressure degrees of freedom at element nodes. If the soil skeleton is rigid, the soil cannot deform and the coupled equations reduce to the steady state seepage equations. It is therefore a simple matter to establish the governing finite element equations for this situation from the more general coupled equations. Only pore fluid degrees of freedom at each node are relevant for seepage analyses. As the flow of water within the soil skeleton is now being considered, the

306 / Finite element analysis in geotechnical engineering: Theory hydraulic boundary conditions which control it must be accounted for. These boundary conditions consist of either prescribed flows or changes in pore fluid pressure. Some of the boundary conditions relevant to geotechnical engineering are described in this chapter. In particular sources, sinks, infiltration and precipitation boundary conditions are covered. The latter option accounts for the finite capacity of soil to accommodate the entry of pore fluids from a boundary. Although it is often assumed that the permeability of soil is constant, laboratory and field tests show that this is not so. Fundamentally, one would expect the permeability to depend on the size of the void space between the solid soil grains and therefore depend on void ratio (or specific volume). Three nonlinear permeability models are presented in this chapter. In one of these the permeability varies with void ratio, whereas in the other two it varies with the mean effective stress. Two different types of pore fluid ;ic surface flow can be identified: those which do not involve a phreatic surface (confined flow) and those which do Unconfined flow (unconfined flow), as shown in Figure Sand ^ ^ 10.1. Problems which involve unconfined flow require special V Clay Flow attention in numerical analysis, as it is necessary to determine the position of the phreatic surface. This is not Confined flow straight forward and a brief discussion of how this may be achieved is given. Figure 10.1: Examples of confined The chapter ends by presenting an and unconfined flow example of a coupled analysis.


Finite element formulation for coupled problems

When deriving the finite element equations in Chapter 2, it was assumed, when evaluating the incremental strain energy in Equation (2.19), that the constitutive behaviour could be written in terms of a relationship between increments of total stress and strain: {Acr} = [D]{A£} (10.1)

If the material behaviour is defined in terms of total stress, as for example in the Tresca model, obtaining the constitutive matrix [D] is relatively straight forward. However, if the material behaviour is defined in terms of effective stress, which is the preferred method in soil mechanics and follows from the principle of effective stress, additional complications can arise. It has been shown in Chapter 3 how the [D] matrix can be obtained from the effective matrix [D'] in the special cases of fully drained and undrained soil behaviour. If soil behaviour is somewhere between these two extreme conditions, account must be taken of the time dependency of the

A procedure for achieving this is described below. [Np] is often assumed to be equivalent to [TV]. and Q represents any sources and/or sinks.0. As before the incremental displacements can be expressed in terms of nodal values using Equation (2. The constitutive behaviour. are the components of the superficial velocity of the pore fluid in the coordinate directions. vy and v. expressed in terms of effective stresses: } The equation of continuity.2) where {Aaf} ={Apf. In the finite element approach it is assumed that the nodal displacements and the nodal pore fluid pressures are the primary unknowns.2: (10. coupled with the material constitutive model and the equilibrium equations. Apf.3) where [Ay is the matrix of pore fluid pressure interpolation functions. y and z directions respectively. .1) becomes: } + {Aay} T (10. In addition.6) } where vx.+ -ZL + — s . Using the principle of effective stress Equation (10. similar to [N]. However. Apf.9): {Apf}=[Np]{APf}n (10.5) i^+^+^_2=^ dx dy dz ^ dt v (10 .Seepage and consolidation / 307 changes in pore fluid pressure and effective stress.0} and Apf is the change in pore fluid pressure. The following basic equations have to be satisfied for a soil saturated with an incompressible pore fluid: The equations of equilibrium: dx dx dy dz d<j'v dpf drYV drV7 dy dy ox dz del dpf dr dr — . can be expressed in terms of nodal values using an equation similar to Equation (2.9). it is assumed that the incremental pore fluid pressure. The analysis of time dependent consolidation requires the solution of Biot's (Biot (1941)) consolidation equations. + —y— dz dz dx dy yz = 0 where yx. The choice of [Np] will be discussed subsequently. yy and yz are the components of the bulk unit weight of the soil acting in the x. Apf9 0. see Figure 10.

\ {A£}T{Acr} dVol Vol. but in the opposite direction.10) . ki} are the coefficients of the permeability matrix. If the soil is isotropic with a permeability k. AW is the incremental strain energy and AL is the incremental work done by applied loads. The incremental strain energy term.9) where AE is the incremental total potential energy.4) can be found by considering the principle of minimum potential energy which states that (see Equation (2.8) Vector {iG}={iGx > k]y > hzV is the unit vector parallel.18)): 0 (10.2: Continuity conditions As noted in Chapter 2 a more convenient form of the equations of equilibrium expressed by Equation (10.308 / Finite element analysis in geotechnical engineering: Theory Generalised Darcy's law: K kxz or: Ky K yy kxy k k yz Kyz Kzz dh dx dh dy dh dz (10.7) {v} = -[k]{Vh} where h is the hydraulic head defined as: (10. then kxx= kyy= kz = k and n-xy k = ^xz k = ^yz k =0 yj - Figure 10. [k]. is defined as: = j . (10. to gravity. AW. of the soil.

and approximating dejdt as AeJAt.19) Vol Ut Substituting for {v} using Darcy's law given by Equation (10. gives: ( ) Vol The work done by the incremental applied loads AL can be divided into contributions from body forces and surface tractions.17) = i {ARE}. =±{\ {m}[B]T[Np] G} dVol] (10.13) into Equation (10.14) [B]T[D'][B] dVol) Sj (10.Seepage and consolidation / 309 Using Equation (10.20) can be written in finite element form as: /}«G=[«c] + fi (10. and can therefore be expressed as (see Equation (2. [N]T{AF}dVol\ + j [Nf{AT}dSrf Vol T J j \Srf [m] = { 1 1 1 0 0 0 } (10.Apf]dVol-QApf =0 (10.9) and following a similar procedure to that outlined in Chapter 2 (i.7) gives: f f f voi dt Noting that {Vh} = (1/yj) Vpf+ {iG}.25)).2) this can be written in the following form: = ± J [{As}T[D']{As} + {Aaf}{As}]dVol Vol n0U) Noting that the second term in this equation is equivalent to Apf-Aev.=f.21) to (2.e.6) can be written as: J [{v}1{V(APf)}+^.18) Using the principle of virtual work.21) .15) = Z [LE]. the continuity Equation (10.16) (10. Equations (2.20)): AL= j {Ad}T{AF} dVol + \ {Ad}T{AT} dSrf Vol Srf (l0 l3) ' Substituting Equations (10. Equation (10.12) and (10. gives the following finite element equations associated with equilibrium: [KG]{Ad}nG + [LG]{Apf}nG where: [KG] = l [KE]t =t[\ /=1 /=1 \Vol = {ARG} (10.

5 (Booker and Small (1975)).310 / Finite element analysis in geotechnical engineering: Theory where: (10.3: Approximation of pore curve.26) may now be written in the following incremental matrix form: .14) and (10.5 the area is approximated by 0. if fi=\ the area is fluid integral essentially assumed to be {{pf}tlG)2At. {pf}nG)2 at time t2=tx+At is sought. Substituting Equation (10.e the shape of the curve) is unknown.24) dy To solve Equations (10.^]-In order to ensure stability of the marching process. the ^ integral on the left hand side of Equation (10. but the value of ({pf}nG)\ is known while h h the value of ({pf} tlG)2 is being sought. {pf}tiG)i *s known at time tl9 then the solution ({Ad}nG .26) Equations (10.3 J between /.22) Vol (10.*. If the solution ({Ad}nG .21) a time marching process is adopted. As {pf}nG varies over the time step At.23) (10.3. and t2.25) represents the area under the curve in Figure 10. Alternatively. To proceed it is necessary to assume: J {\-/3){{pf}nG\\At (10.14) and (10. However. < • Equation (10. For example.5 At [({pf}tiG)\+ ({P/}.25) This approximation is shown graphically in Figure 10.25) into (10. it is necessary to choose /?>0.21) gives: [LG]T{Ad}nG -/3At[0G]{Apf}nG = At (10. the manner in which {pf}fiG varies (i. if ft = 0.25) is therefore an approximation of the area under the Figure 10.

the equations can be solved using the procedures described in Section 2. but vary with stress or strain. In the above formulation the permeabilities have been expressed by the matrix [k]. the time steps can be combined with changes in the loading conditions so that the complete time history of construction can be simulated. the matrix [A:] (and therefore [0G] and [nG]) are not constant over an increment of an analysis (and/or a time step). the Authors' experience is that the modified Newton-Raphson scheme. visco-plastic andNewtonRaphson) can be modified to accommodate nonlinear permeability. as demonstrated.g. This problem is similar to that associated with nonlinear stress-strain behaviour where [KG] is not constant over an increment. tangent stiffness. This is necessary even if the constitutive behaviour is linear elastic and the permeabilities are constant.4 Finite element implementation Equation (10. with a substepping stress point algorithm. some of these are more efficient than others. The solution algorithms described in Chapter 9 can therefore be used. As a marching procedure is necessary to solve for the time dependent behaviour. While o Displacement + pore fluid pressure DOF this is theoretically acceptable. and. If the constitutive behaviour is nonlinear. However. Consequently. For example. As noted in Chapter 9. some users prefer to have the same order of variation of both effective stresses Figure 10. All the procedures described in Chapter 9 (e. the strains. the analysis must be performed incrementally. and therefore the effective stresses (at least for a linear material). pore fluid pressures vary across the element in the same fashion as the displacement components. There is therefore an inconsistency between the variation of effective stresses and pore water • Displacement DOF pressures across the element. is the most accurate. [Np] is the same as the matrix of displacement shape functions [N]. if the displacements vary quadratically.4: Degrees of freedom for and pore water pressure. In Equation (10.Seepage and consolidation / 311 10. vary linearly.27). For an eight sn eight noded element .27) provides a set of simultaneous equations in terms of the incremental nodal displacements {Ad}nG and incremental nodal pore fluid pressures {Apf}nG. both the displacements and pore fluid pressures vary quadratically across the element. If these permeabilities are not constant. for an eight noded quadrilateral element.9.3) the incremental pore fluid pressure within an element has been related to the values at the nodes using the matrix of pore fluid shape functions [Np]. Care must therefore be taken when solving Equation (10. If an incremental pore fluid pressure degree of freedom is assumed at each node of every consolidating element. there are several numerical procedures available for dealing with a nonlinear [KG]. However. Once the stiffness matrix and right hand side vector have been assembled.

as this will affect the manner in which the hydraulic boundary conditions are specified. consolidating Cla|y: con iolidatfng elements (i. care has to be taken to ensure the correct hydraulic boundary condition is applied to the nodes at the interface between clay and sand.e.e. For example. if a situation where sand overlies clay is being modelled. It is possible to have some elements within a finite element mesh Sand: nor-consc lidating which are consolidating and some •a • oelei tients which are not. Some software programs insist that the user decides which elements are to consolidate and which are not at the mesh generation stage. or in terms of excess pore fluid pressure. see Figure 10. Similar behaviour can be achieved by only having pore fluid pressure degrees of freedom at the three apex nodes of a six noded triangle. The p/gure IQ. It is important that the user is familiar with the approach adopted by the software being used.14) is therefore not applicable and Equation (10. This will result in the [Np] matrix only having contributions from the corner nodes and therefore differing from [N\. The only degrees of . It is also possible to formulate the equations in terms of hydraulic head. Clearly. Others are more flexible and allow the decision to be made during the analysis stage.21) reduces to: -[<?hl{Pf}nG = [ % ] + 2 (10. In the theory developed above. there can be no soil deformation and only flow of pore fluid through the soil exists. no pore fluid pressure degrees of freedom at the nodes) might be used for the sand. 10. Some software programs allow the user to decide which of these two approaches to use. or at the eight corner nodes of a twenty noded hexahedron.4.28) This is the finite element equation for steady state seepage.312 / Finite element analysis in geotechnical engineering: Theory noded element this can be achieved by only having pore fluid pressure degrees of freedom at the four corner nodes. see Figure 10. see Section 3. Equation (10. whereas ordinary elements (i.5 Steady state seepage If the soil skeleton is assumed to be rigid.4. elements with pore elei nents pressure degrees of freedom at their — nodes) might be used for the clay. In such cases the hydraulic head or excess pore fluid pressure at the nodes will become degrees of freedom.5.5: Choice of elements for sand is then assumed to behave in a consolidating and non-consolidating drained manner by specifying a zero layers value for the bulk compressibility of the pore fluid. the finite element equations have been formulated in terms of pore fluid pressure.

Boundary conditions can also be prescribed at internal nodes of the finite element mesh. Once steady state conditions have been reached. If the permeabilties are constant and the flow is confined.28). {Apf}nG. As only pore fluid pressures are calculated.27). Such boundary conditions will affect the whole structure of the system equations. They are dealt with in a similar fashion to prescribed displacements.3.Seepage and consolidation / 313 freedom are the nodal pore fluid pressures. soil deformations are zero and the solution is therefore equivalent to that given by Equation (10. . it is only possible to have permeabilities varying with pore fluid pressure. Prescribed values of incremental nodal pore fluid pressure affect only the lefthand side (i. This is achieved by giving the soil fictitious linear elastic properties and applying sufficient displacement constraints to prevent rigid body motion. it is still possible to use it to obtain a steady state seepage solution.1 Introduction With either coupled or steady state seepage analysis there are pore fluid pressure degrees of freedom at the nodes. It is also possible to tie nodal pore fluid pressures in a similar manner to that described for displacements in Chapter 3.7. infiltration and precipitation boundary conditions.28). {Apf}f]G) of the system equations. Equation (10.e. the user must be fully aware of the default condition that the software assumes. 10. as described in Section 3. If a particular piece of finite element software can deal with the coupled formulation given by Equation (10. The various hydraulic boundary condition options that are useful for g^Qtechnical engineering are discussed below.2 Prescribed pore fluid pressures This option allows the user to specify a prescribed incremental change in nodal pore fluid pressure. but not the steady state formulation given by Equation (10. If this is the case and/or if the flow is unconfmed. Prescribed nodal flow values affect the right hand side vector (i. an iterative approach must be used to solve Equation (10. As pore fluid pressure is a scalar quantity.6. 10. sinks.6.e. Clearly. for one or several of the boundary nodes. An analysis is then performed applying the correct hydraulic boundary conditions and sufficient time steps for steady state conditions to be achieved. They are treated in a similar way to prescribed nodal forces.28) can be solved by a single inversion of the matrix [<PG]. and account for this when specifying the boundary conditions for an analysis. They can be specified in the form of sources. and for each node on the boundary of the mesh (or of that part of the mesh consisting of consolidating elements) it is necessary to specify either a prescribed pore fluid pressure or a prescribed nodal flow. Q) of the system equations.6 Hydraulic boundary conditions 10. This usually takes the form of zero nodal flow. most software packages will assume a default condition. If a condition is not specified by the user. Prescribed changes in pore fluid pressures are dealt with in a similar way to prescribed displacements. local axes are irrelevant.28).

6. for all the nodes along the boundary AB. for the beginning of the increment. once excavation is completed. as the nodal degree of freedom.7. Therefore. given by the user for the end of the increment. the excavated soil surface is assumed to be permeable.3 Tied degrees of freedom This boundary condition option allows a condition of equal incremental nodal pore fluid pressure to be imposed at two or more nodes. it can evaluate Apf. it is often more convenient for the user to specify the accumulated value at the end of a particular increment. no pore fluid pressure boundary condition is prescribed along this surface and a default condition of zero nodal flow is imposed. It is noted that not all software packages have this facility. and the value stored internally in the computer. It should also be noted that some software packages may use change in head or excess pore fluid pressure instead of pore fluid pressure. thefinalaccumulated value (i. the boundary conditions will have to be consistent. there is only one tying option. This concept is explained in detail for displacements in Section 3.4 and therefore is not repeated here.6: Prescribed pore fluid increment of the analysis. The first boundary conditions increments of the analysis simulate excavation in front of the wall and it is assumed that the excavated surface is impermeable. Throughout the analysis it is assumed that on the right hand side of the mesh the pore fluid pressures remain unchanged from their initial values. pf= 0) is specified along CD. separated by interface elements shown in Figure . It is then left to the software to work out the incremental change from the prescribed value.20. consider the excavation problem shown in Figure 10.e. whilst the magnitude of the incremental nodal pore fluid pressure remains unknown. Apf = 0) is specified for every Figure 10. compared to the several that are available for displacement which is a vector. 10. Because pore fluid pressure is a scalar quantity. Consequently. Consequently. As an example of the use of tied pore fluid pressure. consider the example of two consolidating layers of soil. As an example of the use of prescribed pore fluid pressures. for the increment after excavation has been completed. see Figure 3.314 / Finite element analysis in geotechnical engineering: Theory Although it is the incremental change in pore fluid pressure that is the required quantity when solving Equation (10. As the program knows the accumulated pore fluid pressure at the nodes along this boundary at the end of excavation.e.6. with a zero pore fluid pressure.6.27). a zero incremental pore fluid pressure (i. For subsequent increments the pore fluid pressure remains at zero along CD and consequently Apf = 0 is applied. However. Consequently. as shown in Figure 10.

e. If the interface is to be treated as a permeable boundary. there is no Figure 10. zero incremental nodal flow). the solution is to tie the incremental pore fluid pressures of adjacent nodes across the interface elements. Because the interface elements do not account for consolidation.7. where it is assumed that rainfall provides a flow rate qn on the soil surface adjacent to the excavation. most software programs will treat each row as an impermeable boundary (i. To apply such a boundary condition in finite element analysis. CD.7.4 Infiltration When it is necessary to prescribe pore fluid flows across a boundary of the finite element mesh for a particular increment of the analysis.. Many finite element Fjgure W. ^ooo^ trtrxrc 10.6. In the situation shown Interface in Figure 10. These flows are treated in a similar fashion to boundary stresses as described in Section 3.7.8: Example of infiltration programs will do this automatically boundary conditions for generally distributed boundary flows and for arbitrary shaped boundaries.7: Tied pore fiuid seepage link between these two rows pressures of nodes and. the flow over the boundary must be converted into equivalent IIIUIII I nodal flows.Seepage and consolidation / 31 5 10. there is a set of nodes elements along the underside of soil layer 1 and Layer 2 another set on the upper surface of soil layer 2. infiltration boundary conditions are used. the flow rate may vary along the boundary over which it is active. corresponding to the upper and lower side of the row of Pf ~ Pf=Pi interface elements respectively. For example.. In general. they are not usually Layer 1 formulated to account for consolidation. The nodal flows equivalent to the infiltration boundary condition are determined from the following equation: .6. Because interface elements have zero thickness. tie the incremental pore fluid pressures for nodes AB.8. etc. An example of an infiltration Rainfall intensity = qn boundary condition is shown in Figure 10.. unless a boundary condition is specified for these nodes. .

As with boundary stresses. if the pore fluid pressure is more tensile than/?^.6. the magnitude of which gives an accumulated pore fluid pressure equal to pfh at the end of the increment.6. Alternatively. involving a row /I of extraction wells (sinks) within an excavation and. are specified. or if the current flow rate at the node exceeds the value equivalent to qn.29) where Srfis the element side over which the infiltration flow is prescribed. If it is. Both an infiltration flow rate. to limit excessive // settlements behind the retaining wall. this integral can be evaluated numerically for each element side on the specified boundary range.5 Sources and sinks A further option for applying flow boundary conditions is to apply sources (inflow) or sinks (outflow) at discrete nodes.316 / Finite element analysis in geotechnical engineering: Theory Qinfi. see Section 3. and the effect of the injection wells could be modelled by applying Figure 10. . The effect of the extraction wells could be modelled by applying a flow rate equivalent to the pumping rate at node A. At the start of an increment.9: Example of sources and sinks boundary conditions a flow rate equivalent to the injection rate at node B. An example of a source and sink Injection well boundary condition is shown in Extraction well / / Figure 10. in the form of prescribed nodal flows.7. pfh. The following two examples show how this boundary condition may be used. 10.6. ft Z 10. the boundary condition for that node is taken as a prescribed incremental pore fluid pressure Apf. the boundary condition is taken as a prescribed infiltration with the nodal flow rate determined from qn. each node on the boundary is checked to see if the pore fluid pressure is more compressive than pfh.9 in the form of a simple dewatering scheme. qm and a pore fluid pressure. For plane strain and axi-symmetric analyses these are essentially line flows acting perpendicularly to the plane of the finite element mesh.6 Precipitation This boundary condition option allows the user to essentially prescribe a dual boundary condition to part of the mesh boundary.} = gn dSrf Srf ( 1 0 . a row of injection wells (sources).

. flow of water from the tunnel into the soil would result. consequently a flow boundary condition with qf = 0 (i.Seepage and consolidation / 317 Tunnel problem flow from tunnel Because pf more tensile than/7^. When this occurs. because there is unlikely to be a sufficient supply of water in the tunnel. At any one increment some nodes can have a prescribed pore fluid stress boundary condition.e. see Volume 2 of this book. This problem can be dealt with using the precipitation boundary option with qn = 0 and p/h = 0. zero flow boundary conditions adopted a) Prescribed pore fluid pressure as boundary condition in short term b) Precipitation boundary conditions in short term /^more compressive than/?^: pore fluid boundary conditions adopted PjmoxQ tensile than/?^: flow boundary conditions adopted pf at all nodes more compressive than j ^ : pore fluid boundary conditions adopted at all nodes c) Precipitation boundary conditions at intermediate stage d) Precipitation boundary conditions in long term Figure 10. no flow) is adopted. If for subsequent increments of the analysis (tunnel boundary now permeable) a prescribed zero accumulated pore fluid pressure boundary condition is applied to the nodes on the tunnel boundary. with a magnitude set to give an accumulated pore fluid pressure at the end of the increment equal topfh. the pore fluid pressures at the nodes on the tunnel boundary are more tensile thanpjh. This implies that the boundary condition can change at individual nodes at different increments of the analysis. see Figure 10.10: Precipitation boundary conditions in tunnel problem After excavation for a tunnel.10a. With time the tensile pore fluid pressures reduce due to swelling and eventually become more compressive thanpjh. the pore fluid stress boundary condition is applied. The pore fluid pressure checks are made on a nodal basis for all nodes on the tunnel boundary for each increment. the pore fluid pressure in the soil adjacent to the tunnel could be tensile.lOd. while others can have a flow condition. see Figure 10. This is unrealistic.10b. see Figure lO. Initially (after excavation). assuming the tunnel boundary to be impermeable.10c. see Figure 10.

• compared with permeability of soil. see Figure 10.11a.1 Introduction When performing coupled (or steady seepage) analysis it is necessary to input permeability values for the soils undergoing seepage. see Figure 10. 10. There High rain fall intensity is a finite depth to such ponding. For coupled analysis it is also necessary to input the constitutive behaviour.e. Some of the models that the Authors find useful are briefly described below. with qn set equal to the rainfall intensity set to have a value more compressive (i. 10.7. However. at any point. the permeabilities can vary spatially. or they can vary nonlinearly as a function of void ratio or mean effective stress. rainfall intensity is small. which will pond on the surface.7. the soil will not be able to absorb the water. For example.318 / Finite element analysis in geotechnical engineering: Theory Rainfall infiltration In this case the problem relates to a boundary which is subject to rainfall of a set intensity. If the soil is of sufficient permeability and/or the Low rain fall intensity compared with permeability of soil. and b) consequently a pore fluid pressure boundary condition would be applicable.11: Rainfall infiltration possible to decide which boundary boundary conditions condition is relevant before an analysis is undertaken. defined by a single value of k. if the soil is less permeable and/or the rainfall intensity is high. If during the analysis the pore fluid pressure becomes more compressive than/?/A. Because pf] is more tensile than/?^. permeability and geometry. the soil can be assumed to be isotropic or anisotropic. Several options exist for specifying permeabilities.1 lb. j \ j \ / \ / uuuu 11111111. the boundary condition will switch to that of a prescribed pore fluid pressure. in most soils the permeability varies with void ratio . equivalent to the ponding level the initial value of the pore fluid stress at the soil boundary. the soil can no surface ponding absorb the water and a flow boundary condition is appropriate.7 Permeability models 10.2 Linear isotropic permeability This model assumes permeability to be isotropic and. a flow boundary condition will be assumed initially. surface ponding which is problem specific. However. because the behaviour will depend on soil stratification. The dilemma can be overcome by using the precipitation boundary condition. However. it is not always Figure 10.

12: Permeability profile In this model the permeability is for London clay assumed to be direction dependent. e. 10. From a fundamental point of view. -10 101 101 L io. ym . Again. it is often convenient to adopt an expression in which k varies with mean effective stress. Consequently. It is therefore convenient to Weathered London clay i have the option to vary k spatially. This implies that the initial value. at the beginning of the analysis.8 Figure 10. but to vary as a function of the void ratio. With this model it is necessary to know the void ratio at any stage of the analysis. see Figure 10. The following relationship between the coefficient of permeability.0 simulate a permeability varying with j> depth. For example. This enables the permeability matrix in Equation (10. This is usually performed automatically by the software. it is useful to have the option of varying the permeability values spatially.4 Nonlinear permeability related to void ratio In this model the permeability is assumed to be isotropic. the permeability coefficients must be transformed. It should be noted that this matrix is associated with the global coordinate axes and therefore. must be input into the finite element program. With this model. or 30 Thames gravel depth. zm) and values for the coefficient of permeability in each direction specified (kxm.7. k (m/s) 10.Seepage and consolidation / 319 and therefore mean effective stress. the permeability at an integration point Woolwicji & Reading bed cla[y remains constant throughout an Woolwich & Reading bed sand analysis. and void ratio. kym. However. allowing the permeability to vary with void ratio makes sense. if the material and global axes differ. k.12 (note: 3 coefficient of permeability is plotted on a log scale).7) to be obtained. is assumed: k= +be) (10. This option can be used to .7. A set of permeability axes are defined (xm . kzm).3 Linear anisotropic permeability Coefficient of permeability. .30) where a and b are material parameters. there is often little laboratory or field data available to determine the parameters a and b. options are often 20 available for varying A : in a piece wise ^ Unweathered London clay / linear fashion across a finite element w mesh. Two such models are presented below.

Cc.5 Nonlinear permeability related to mean effective stress using a logarithmic relationship Again the permeability is assumed to be isotropic. The relationship between permeability and mean effective stress takes the form: (10.13.8 Unconfined seepage flow Situations which involve unconfined flow.13: Variation of above the phreatic surface and the permeability with pore fluid pressure *5fc . Such methods are not applicable to coupled consolidation analysis. The derivation of the logarithmic law therefore incorporates the assumption that mv is constant (Vaughan (1989)).g. A typical variation of permeability logk with pore fluid pressure is shown in Figure 10. 10. if the accumulated pore fluid pressure becomes more tensile than pj2 . incorporating the initial void ratio at zero mean effective stress and the coefficient of volume compressibility. 10.7. given by one of the models described above) is adopted. There are several different algorithms available. remains constant (Vaughan (1989)). where it is necessary for the software to determine a phreatic surface.31) where ko is the coefficient of permeability at zero mean effective stress and a is a constant. the soil is assumed to be Figure 10. If the accumulated pore Log k = log kn \ogR/— fluid pressure is more compressive than pfu the soil's normal KJR permeability (e. because there does not appear to be a robust algorithm for finding and accommodating the phreatic surface. (+ve) Pn Pore pressure However.320 / Finite element analysis in geotechnical engineering: Theory 10. can be problematic. A more general approach is to reduce the permeability when the soil sustains a tensile pore fluid pressure associated with a position above the phreatic surface.32) where again ko and a are material parameters. according to the following relationship: (10./?'.7. mv. Some of these involve adjustments to the finite element mesh so that the phreatic surface follows a mesh boundary. but to vary as a function of mean effective stress.6 Nonlinear permeability related to mean effective stress using a power law relationship This model has been derived assuming that the compression index.

It is therefore well suited to a nonlinear solution strategy of the modified NewtonRaphson type. R. The 9 boundary conditions employed are consolidation problem -1 noted on this figure.Seepage and consolidation / 321 normal permeability is reduced by a large factor. Some of the results from b = 6a these analyses are now presented. The Authors' experience is that numerical instability can sometimes arise with this approach and that more research is required to obtain a robust algorithm. Exact solutions have been found only to problems involving linear elastic materials. The water table was assumed to be at the ground surface. under conditions of plane strain. 10. subjected to a load of intensity q over a width 2a. " " Originally. If the accumulated pore fluid pressure is between/fy and/?/2. c: -_ 2Gk (10. This B problem was used by the Authors as one of a series of validation exercises Fixed. T: Smooth. so that boundary nodes can automatically switch from a prescribed pore fluid pressure to a prescribed flow. Usually R takes a value somewhere between 100 and 1000.34) . with simple geometry and subjected to simple boundary I I conditions. impermeableb ound ary consolidation problems are not easy X to obtain. Analysis involving a phreatic surface also require the use of the precipitation boundary option. One such problem is that of a porous elastic half-space. permeable boundary f rp _ Ct (10. Clearly.33) and the adjusted coefficient of consolidation.14 was used. the finite element mesh Fi ure 10 14: FE mesh for shown in Figure 10. impermeable boundary when coupled analysis was first coded into ICFEP.9 Validation example a Closed form solutions for Free. the permeability is found using a linear interpolation between the two extreme values. The results are expressed in terms of the adjusted time factor. This is especially true when dealing with elasto-plastic materials. this approach requires an iterative algorithm.

8 solution. was derived for an infinite Figure 10. pexcess Iq Figure 10.2 0. in the code 8 analysis there was a quadratic variation of pore fluid pressure across each finite element. b=X8a pore fluid pressure. 1969 squares) over predicts the excess FE prediction. with adjusted time factor. see Figure 10. and t denotes time. both using the mesh shown in Figure 10. particularly at FE prediction. In the other analysis. typically five per 'log' cycle. is given for two specific points in the half-space. were used thereafter. . there is very little difference between the predictions from the two finite element analysis. Variation of half-space. In addition.1 andPoisson's ratio ju=0.6 0. it was suspected that this normalised excess pore pressure discrepancy was due to the close proximity of the bottom and/or lateral boundaries in the finite element mesh. all eight nodes had a pore fluid pressure degree of freedom. pexcess/q. b-6a greater depths beneath the loaded area. Consequently.15 (Schiffman et al (1969)). Also shown in Figure 10.15 for the specific case of T= 0.. Two predictions were made. pexcess /q.7 0. are shown in Figure 10. AT". it was found that the position of the lateral boundary has the greatest influence. It can be seen that the predictions from both finite element analyses are in excellent agreement with the closed form solution.16. Logarithmic time increments.322 / Finite element analysis in geotechnical engineering: Theory where G is the elastic shear modulus. The variation of normalised excess pore fluid pressure. Because the closed form 0 0.15xlO"5).]. ° beneath the centre of the loaded area is given in Figure 10.17 the variation of normalised excess pore fluid pressure. it was 1.1 0. resulting in an undrained response.15. Eight noded isoparametric elements were employed and in one analysis. but three times the width.16. T. The load intensity q was applied in a very small time step (expressed in terms of the adjusted time factor.5 0. labelled code 4. labelled as code 8. except for a slight discrepancy at the base of the mesh. with different forms of the element pore fluid pressure interpolation matrix. only the four corner nodes had pore fluid pressure degrees of freedom.14. [7V/.15 as solid squares. which has a similar depth to that given in Figure 10. The effect of the position of both boundaries on the predictions was investigated and. surprisingly.3 0. presented by the full line in Normalized excess pore pressure.16. and the results for this iog' cycle are not presented. whereas in the code 4 analysis the distribution was linear. yf\s the bulk unit weight of the pore fluid. In Figure 10. The response during the first iog' cycle was essentially undrained. Results from the mesh shown in Figure 10. These results are in excellent agreement with the closed form solution.4 0. It can be seen that the finite element prediction (open Schiffinanefa/. k is the coefficient of permeability.17 is the closed form solution.

5 x/a =1. to enable a solution to be obtained.II yla = 0. code 4 A.5 B.17: Variation of normalised excess pore pressure 10.E. a time marching algorithm is necessary.Schiffinan£tfa/. This involves combining the equations governing the flow of pore fluid through the soil skeleton. Schiffinanefa/. T Figure 10.0 > A: y/a = 0. This chapter has considered the modifications to the finite element theory that are necessary to enable time dependent soil behaviour to be simulated. The resulting finite element equations involve both displacement and pore fluid pressure degrees of freedom at element nodes.0001 0. This involves .16: F.1 Adjusted time factor. with the equations governing the deformation of the soil due to loading. FE prediction. code 8 A. code 4 B.001 0.01 0. impermeable boundary Z>=18a Figure 10.Seepage and consolidation / 323 B:y/a = x/a = 0. 2. mesh for consolidation problem . FE prediction... code 8 0. In addition. FE prediction. 1969 B. FE prediction. 1969 A.0 ' Free.10 Summary 1. permeable boundary Fixed.

to accommodate the time marching process. at each node. However. in which the analysis has to determine the position of a phreatic surface.324 / Finite element analysis in geotechnical engineering: Theory an assumption of the magnitude of the average pore fluid pressures over each time step. 3. . 7. The permeability can take several different forms: it can be isotropic or anisotropic. 4. 6.5. it can vary spatially. the average pore fluid pressure during the time step is linearly related to the values at the beginning and end of the time step. infiltration. It is also necessary to input the coefficients of soil permeability. It is assumed that. can be accommodated. This results in a considerable reduction in the complexity of the governing finite element equations. 5. there can be no deformation and only flow of pore fluid through the soil occurs. Even for linear material behaviour analysis must be performed incrementally.e a zero nodal flow condition). This involves the parameter/?. 8. For numerical stability /?> 0. These consist of prescribed nodal pore fluid pressures. Problems involving unconfined seepage. In the latter case the permeability at each integration point will vary during the analysis and the software must have the appropriate algorithms to cope with this. In coupled analysis hydraulic boundary conditions must be considered. 10. For all nodes on the boundary of the finite element mesh either a prescribed pore fluid pressure or a flow boundary condition must be specified. tied pore fluid freedoms. If a boundary condition is not set for a boundary node. Only pore fluid pressure degrees of freedom need to be considered. 9. Further complications arise if the soil is partly saturated. at present the algorithms that are available are not robust. If the soil skeleton is rigid. which can lead to numerical instability. or it can vary as a function of void ratio or mean effective stress. most software packages will implicitly assume that the node represents an impermeable boundary (i. sinks and sources and precipitation. The coupled theory presented in this chapter assumes the soil to be fully saturated. It is possible to have some elements within a finite element mesh which are consolidating and some which are not. Further research is required.

Sign up to vote on this title
UsefulNot useful