Discover millions of ebooks, audiobooks, and so much more with a free trial

Only $9.99/month after trial. Cancel anytime.

Groundwater and Seepage
Groundwater and Seepage
Groundwater and Seepage
Ebook725 pages3 hours

Groundwater and Seepage

Rating: 2.5 out of 5 stars

2.5/5

()

Read preview

About this ebook

The movement of groundwater is a basic part of soil mechanics. It is an important part of almost every area of civil engineering, agronomy, geology, irrigation, and reclamation. Moreover, the logical structure of its theory appeals to engineering scientists and applied mathematicians.
This book aims primarily at providing the engineer with an organized and analytical approach to the solutions of seepage problems and an understanding of the design and analysis of earth structures that impound water. It can be used for advanced courses in civil, hydraulic, agricultural, and foundation engineering, and will prove useful to consulting engineers — or any public or private agency responsible for building or maintaining water storage or control systems.
Among the special features of this book are its coverage of previously unavailable Russian work in the field, an extensive appendix of concepts in advanced engineering mathematics needed to deal with physical flow systems, and numerous completely worked-out and solved examples coupled with over 200 problems of varying difficulty.

LanguageEnglish
Release dateDec 4, 2012
ISBN9780486145198
Groundwater and Seepage

Related to Groundwater and Seepage

Related ebooks

Civil Engineering For You

View More

Related articles

Reviews for Groundwater and Seepage

Rating: 2.5 out of 5 stars
2.5/5

4 ratings0 reviews

What did you think?

Tap to rate

Review must be at least 10 words

    Book preview

    Groundwater and Seepage - Milton E. Harr

    potential

    1

    Fundamentals of Groundwater Flow

    1-1. Scope and Aim of Subject

    The aim of this work is primarily to present to the civil engineer the means of predicting the exigencies arising from the flow of groundwater. The specific problems which are to be dealt with can be divided into three parts:

    Estimation of the quantity of seepage

    Definition of the flow domain

    Stability analysis

    When writing on a subject as broad as groundwater and seepage, it is necessary to presume a minimum level of attainment on the part of the reader. Hence it will be assumed that the reader has a working knowledge of both the calculus and the rudiments of soil mechanics. For example, the problem of the stability of an earthen slope subject to seepage forces will be considered as solved once the upper flow line has been located and the pore pressures can be determined at all points within the flow domain. The actual mechanics of estimating the factor of safety of the slope will be left to the reader. Several texts on the subject of soil mechanics can be found in the references [142, 145].¹ However, such factors as the determination of the uplift pressures under structures, exit gradients, and all pertinent seepage quantities will be considered to be within the scope of this book.

    Although the fundamentals of groundwater flow were established more than a century ago, it is only within recent years that the subject has met with scientific treatment. As a result of the trial-and-error history of groundwater-flow theory, its literature is replete with empirical relationships for which exact solutions can be and have been obtained. Advocates of the empirical approach have long reasoned that the heterogeneous nature of soils is such that rigorous analyses are not practical. This is not so; as will be seen, much of the subject lends itself readily to theoretical analysis.

    Recent developments in the science of soil mechanics coupled with more precise methods of subsurface soil explorations have provided engineers with greater insight into the behavior of earth structures subject to groundwater flow. The mathematics of the theory of functions of complex variables, once the arid theory of imaginary numbers, now allows the engineer to solve problems of otherwise overpowering complexity.

    The engineer can now formulate working solutions which not only reflect the interaction of the various flow factors but also allow him to obtain a measure of the uncertainties of his design.

    1-2. Nature of Soil Body

    In groundwater problems the soil body is considered to be a continuous medium of many interconnected openings which serve as the fluid carrier. The nature of the pore system within the soil can best be visualized by inference from the impermeable boundaries composing the pore skeleton. For simplicity it will be assumed that all soils can be divided into two fractions which will be referred to respectively as sand and clay.

    In general, sands are composed of macroscopic particles that are rounded (bulky) or angular in shape. They drain readily, do not swell, possess insignificant capillary potential (see Sec. 1-7), and when dry exhibit no shrinkage. Clays, on the other hand, are composed of microscopic particles of platelike shape. They are highly impervious, exhibit considerable swelling, possess a high capillary potential, and demonstrate considerable volume reductions upon drying.

    Sands approach more nearly the ideal porous medium and are representative of the soils primarily to be dealt with in this book. This may appear to introduce serious restrictions; however, in most engineering problems the low permeability of the clays renders them relatively impervious in comparison to the coarser-grained soils.

    Extensive studies have been undertaken by many investigators [13, 45, 75, 84, 136] to calculate the permeability and porosity of natural soils based on their sieve analyses and various packings of uniform spheres. While it is not possible to derive significant permeability estimates from porosity measurements alone [86–88, 99], the pore characteristics of these ideal packings do present some of the salient features of natural soils.

    Let us assume that the soil particles are all of uniform spherical shape. Calling the total volume V and the volume of voids Vv, we have for the porosity

    (1)

    and for the void ratio

    (2)

    For a cubical array of spheres (Fig. 1-1a), V = d³, Vv = d³ − πd³/6, and

    FIG. 1-1

    For a rhombohedral packing (Fig. 1-1b), which represents the most compact assemblage of uniform spheres, the porosity is

    Values for the porosity of some natural soils are given in Table 1-1.

    Table 1-1. Porosity of Some Natural Soils²

    Figures 1-2a and b show the pore volume available for flow for the cubic and rhombohedral array, respectively [45]. It should be noted from these figures that even in the ideal porous medium the pore space is not regular but consists of cavernous cells interconnected by narrower channels. Natural soils contain particles that can deviate considerably from the idealized spherical shape (as in the case of clay) and, in addition, are far from uniform in size. The true nature of the pore channels in a soil mass defies rational description. It is for this reason that groundwater flow was not amenable to a scientific treatment until the advent of Darcy’s law (Sec. 1-4). Fortunately, in groundwater problems we need not concern ourselves with the flow through individual channels. We are primarily interested in macroscopic flow wherein the flow across a section of many pore channels may be considered uniform as contrasted to the near-parabolic distribution of the flow through a single pore.

    FIG. 1-2. (After Graton and Fraser [45].)

    1-3. Discharge Velocity and Seepage Velocity

    The discharge velocity is defined as the quantity of fluid that percolates through a unit of total area of the porous medium in a unit time.

    FIG. 1-3

    FIG. 1-4

    As flow can occur only through the interconnected pores of saturated soils (Fig. 1-3), the velocity across any section must be thought of in a statistical sense. If m is the effective ratio of the area of pores Ap to the total area A, then m = Ap/A. The quantity of flow (also called the quantity of seepage, discharge quantity, or dischargeis called the seepage velocity.

    Let us investigate the nature of m. Designating Ap(z) as the area of pores at any elevation z (Fig. 1-4), we have

    (1)

    The average value of m over the cylinder of height h is

    (2)

    Then

    (3)

    where V is the total volume and the integral is the volume of voids. Hence the average value of m is the volume porosity n and will be so designated. In our work we shall deal mainly with discharge velocities, i.e., superficial velocities, and unless otherwise stated all velocities will be so inferred.

    1-4. Darcy’s Law

    As is well known from fluid mechanics, for steady³ flow of nonviscous incompressible fluids, Bernoulli’s equation⁴ [80]

    where p = pressure, psf

    γw = unit weight of fluid, pcf

    = seepage velocity, ft/sec

    g = gravitational constant, 32.2 ft/sec²

    h = total head, ft

    demonstrates that the sum of the pressure head p/γw elevation head z, and velocity head at any point within the region of flow is a constant. In groundwater flow, to account for the loss of energy due to the viscous resistance within the individual pores, Bernoulli’s equation is taken as (Fig. 1-5)

    (1)

    where Δh represents the total head loss (energy per unit weight of fluid) of the fluid over the distance Δs. The ratio

    (2)

    is called the hydraulic gradient and represents the space rate of energy dissipation per unit weight of fluid (a pure number).

    FIG. 1-5

    In most groundwater problems the velocity heads (kinetic energy) are so small that they can be neglected. For example, a velocity of 1 ft/sec, which is large as compared to typical seepage velocities, produces a velocity head of only 0.015 ft. Hence, Eq. (1) can be written

    and the total head at any point in the flow domain is simply

    (3)

    Prior to 1856, the formidable nature of groundwater flow defied rational analysis. In that year, Henry Darcy [25] published a simple relation based on his experiments on les fontaines publiques de la ville de Dijon, namely,

    (4)

    Equation (4), commonly called Darcy’s law, demonstrates a linear dependency between the hydraulic gradient and the discharge velocity υ. In soil mechanics, the coefficient of proportionality k is called the coefficient of permeability and, as shown in Eq. (4), k has the dimensions of a velocity.

    Although Darcy’s law is presented in differential form in Eq. (4), it must be emphasized that in no way does it describe the state of affairs within an individual pore. Strictly speaking, Darcy’s law represents the statistical macroscopic equivalent of the Navier-Stokes equations of motion [Eqs. (1), Sec. 6-4] for the viscous flow of groundwater. It is precisely this equivalency that permits the subsequent development of groundwater flow within the theoretical framework of potential flow (Sec. 1-10). Stated simply, viscous effects are accounted for completely by Darcy’s law, and hence for all subsequent considerations the flow may be treated as nonviscous or frictionless.

    1-5. Range of Validity of Darcy’s Law

    Visual observations of dyes injected into liquids led Reynolds [121], in 1883, to conclude that the orderliness of the flow was dependent on its velocity. At small velocities the flow appeared orderly, in layers, that is, laminar. With increasing velocities, Reynolds observed a mixing between the dye and water; the pattern of flow became irregular, or turbulent.

    Within the range of laminar flow, Reynolds found a linear proportionality to exist between the hydraulic gradient and the velocity of flow, in keeping with Darcy’s law. With the advent of turbulence, the hydraulic gradient approached the square of the velocity. These observations suggest a representation of the hydraulic gradient as

    (1)

    where a and b are constants and n is between 1 and 2. From studies of the flow of water through columns of shot of uniform size, Lindquist [99] reports n to be exactly 2. Whatever the precise order of n, experiments have shown conclusively that for small velocities (within the laminar range), Darcy’s law gives an accurate representation of the flow within a porous medium.

    There remains now the question of the determination of the laminar range of flow and the extent to which actual flow systems through soils are included. Such a criterion is furnished by Reynolds number R (a pure number relating inertial to viscous force), defined as

    (2)

    where

    v = discharge velocity, cm/sec

    D = average of diameters of soil particles, cm

    ρ = density of fluid, g (mass)/cm³

    µ = coefficient of viscosity, g-sec/cm²

    The critical value of Reynolds number at which the flow in soils changes from laminar to turbulent flow has been found by various investigators [99] to range between 1 and 12. However, it will suffice in all our work to accept the validity of Darcy’s law when Reynolds number is taken as equal to or less than unity, or

    (3)

    Substituting the known values of ρ and µ for water into Eq. (3) and assuming a conservative velocity of ¼cm/sec, we have D equal to 0.4 mm, which is representative of the average particle size of coarse sand. Hence it can be concluded that it is most unlikely that the range of validity of Darcy’s law will be exceeded in natural flow situations. It is interesting to note that the laminar character of flow encountered in soils represents one of few valid examples of such flow in all hydraulic engineering. An excellent discussion of the law of flow and a summary of investigations are to be found in Muskat [99].

    1-6. Coefficient of Permeability

    For laminar flow we found that Darcy’s law could be written as [Eq. (4), Sec. 1-4]

    (1)

    where k is the coefficient of permeability.

    Owing to the failure at the present time to attain more than a moderately valid expression relating the coefficient of permeability to the geometric characteristics of soils,⁵ refinements in k are hardly warranted. However, because of the influence that the density and viscosity of the pore fluid may exert on the resulting velocity, it is of some value to isolate that part of k which is dependent on these properties. To do this, we introduce the physical permeability k0 (square centimeters), which is a constant typifying the structural characteristics of the medium and is independent of the properties of the fluid. The relationship between the permeability and the coefficient of permeability as given by Muskat [99] is

    (2)

    where γw is the unit weight of the fluid and µ is the coefficient of viscosity.

    Substituting Eq. (2) into Darcy’s law, we obtain

    (3)

    which indicates that the discharge velocity is inversely proportional to the viscosity of the fluid. Equation (3) may be used when dealing with more than one fluid or with temperature variations. In the groundwater and seepage problems encountered in civil engineering, where we are primarily interested in the flow of a single relatively incompressible fluid subject to small changes in temperature, it is more convenient to use Darcy’s law with k as in Eq. (1).

    Laboratory and field determinations of k have received such excellent coverage in the soil-mechanics literature [81, 99, 142, 145] that duplication in this book does not appear to be warranted. Some typical values of the coefficient of permeability are given in Table 1-2.

    Table 1-2. Some Typical Values of Coefficient of Permeability⁶

    1-7. Capillarity

    Although the physical cause of capillarity is subject to controversy, the surface-tension concept of capillarity renders it completely amenable to rational analysis.

    If a tube filled with dry sand has its lower end immersed below the level of free water (Fig. 1-6a), water will rise within the sand column to an elevation hc called the height of capillary rise. Consideration of the equilibrium of a column of water in a capillary tube of radius r (Fig. 1-6b) yields, for hc,

    (1)

    Ts is the surface tension of the water (≈0.075 g/cm). It can be thought of as being analogous to the tension in a membrane acting at the air-water interface (meniscus) and supporting a column of water of height hc. α is called the contact angle and is dependent on the chemical properties of both the tube and the water. For clean glass and pure water, α = 0.

    FIG. 1-6

    If atmospheric pressure is designated as pa, then the pressure in the water immediately under the meniscus will be

    (2)

    On the assumption that the rate of capillary rise is governed by Darcy’s law, Terzaghi [144] obtained from the differential equation

    the following expression for the time t required for the capillary water to rise to the height y:

    (3)

    As the capillary water rises, air becomes entrapped in the larger pores and hence variations are induced in both hc and k. From laboratory tests, Taylor [142] found that for y/hc < 20 per cent the degree of saturation is relatively high and Eq. (3) can be considered valid.

    On the basis of Eqs. (1) and (3) we see that although the fine-grained soils exhibit a greater capillary potential than the granular soils their minimal coefficients of permeability greatly reduce their rates of capillary rise. Some typical values for the height of capillary rise are presented in Table 1-3.

    Table 1-3. Typical Values of Height of Capillary Rise⁷

    1-8. General Hydrodynamic Equations, Velocity Potential

    In Fig. 1-7, ūare the components of the seepage velocity at the point in the fluid A (x,y,z) at the time t. Thus they represent functions of the independent variables x, y, z, and t. For a particular value of t, they specify the motion at all points occupied by the fluid; and for any point within the fluid, they are functions of time, giving a history of the variations of velocity at that point. Unless otherwise specified, ū, and their space derivatives (∂ū/∂x) are everywhere bounded.

    FIG. 1-7

    A particle of fluid originally at point A (x,y,z) at time t will move during the time δt ). The time rate of change of any velocity component, say δū/δt, will be

    Now if δt , the total acceleration in the x direction is

    (1a)

    Similarly, in the y and z directions,

    (1b)

    (1c)

    Let the pressure of the center point A (x,y,z) of Fig. 1-7 be p and the density of the fluid be p and let the components of the body force per unit mass be, respectively, X, Y, and Z in the x, y, and z directions at the time t. In groundwater flow the common body force is that due to the force of gravity. With a pressure p at point A (x,y,z), the force on the yz face of the element nearest the origin is

    and that on the opposite face is

    By Newton’s second law of motion, the product of the mass and acceleration in the x direction must equal the sum of the forces in that direction; hence

    Simplifying and substituting the value of the total acceleration from Eq. (1a), we have

    (2a)

    Similarly, in the y and z directions,

    (2b)

    (2c)

    Equations (2) are Euler’s equations [80] of motion for a nonviscous fluid.

    , etc., can be neglected and Eqs. (2) become (velocity component without bar is discharge velocity)

    (3a)

    (3b)

    (3c)

    When the state of flow is independent of the time (steady state), the left sides of Eqs. (3) vanish. For example, Eq. (3a) becomes

    (4)

    Hence, substituting for p from Eq. (3), Sec. 1-4, Eq. (4) takes the form (ρ is constant)

    (5)

    where i is the hydraulic gradient. Applying Darcy’s law, we have finally

    (6a)

    Similarly, for Y and Z we obtain

    (6b)

    (6c)

    Equations (6) indicate that for steady-state and laminar flow the body forces are linear functions of the velocity.

    Assuming that the coefficient of permeability k is independent of the state of flow and substituting Eqs. (6) into Eqs. (3), we obtain the following dynamical equations of flow:

    (7)

    For steady flow Eqs. (7) reduce to the vectorial generalization of Darcy’s law.

    (8a)

    Although the steady-state assumption is generally made to set aside the inertia terms in Eqs. (7) [88, 99], it can be shown that these terms are of negligible order in a wide range of situations, even when velocity changes do occur with time. To demonstrate this, we note that each of Eqs. (7) can be expressed in the form (i is hydraulic gradient)

    which, with the change of variable

    (8b)

    reduces to the expression

    Substituting conservative values for n, g, and k [n = ½, g ≈ 1000 cm/sec² and k = 0.1 cm/sec (coarse sand)], we obtain

    (8c)

    Comparing the orders of magnitude of the two terms on the right side of this expression, we see that if the ratio (∂i/∂t)/υ∗ is not exceedingly large, the first right-hand term is a small quantity and can be neglected. Thus, the remaining equation is

    which, after integration, yields

    The right part of this equation very rapidly tends to zero and hence within a fraction of a second we can consider υ∗ = 0 and Eq. (8b) will reduce to Eq. (8a). Analytical studies conducted by Pavlovsky [110] and Davison [26] indicate that unless reservoir water elevations are subject to excessive variations in time the contribution of ∂i/∂t in Eq. (8c) can be neglected. Physically, this implies that the speed of flow through natural soils for laminar flow is so slow that changes in momentum are negligible in comparison with the viscous resistance to flow. In some measure, the nominal contribution of the inertia terms was implied previously in the critical value of Reynolds number which assured laminar flow and the validity of Darcy’s law. If one recalls (Sec. 1-5) that Reynolds number is the ratio of inertia force to viscous force and that for laminar flow in soils R ≦ 1, it is evident that the viscous forces are at least of the order of magnitude of the inertia forces. This is in contradistinction to considerations of laminar flow through pipes (Rcritical = 2,000) where the forces due to momentum changes may be very much greater than the viscous forces resisting the flow.

    Equation (8a) contains the four unknowns u, v, w, and h. Hence one more equation must be added to make the system complete. This is the equation of continuity which assumes that the fluid is continuous in space and time.

    The quantity of fluid through the yz face of the element of Fig. 1-7 nearest the origin is

    nū dy dz

    and that through the opposite face is

    The net gain in the quantity of fluid per unit time in the x direction is

    Similarly, the gains in the y and z directions are, respectively,

    If the fluid and flow medium are both incompressible, the total gain of fluid

    Enjoying the preview?
    Page 1 of 1