Sunteți pe pagina 1din 13

To cite this paper: Int. J. Rock Mech. & Min. Sci. 34:3-4, paper No. 283.

Copyright 1997 Elsevier Science Ltd

Copyright 1997 Elsevier Science Ltd Int. J. Rock Mech. & Min. Sci. Vol. 34, No. 3-4, 1997 ISSN 0148-9062 To cite this paper: Int. J. RockMech. &Min. Sci. 34:3-4, Paper No. 283

AN E L A S T O - P L A S T I C C O N S T I T U T I V E M O D E L F O R B R I T T L E - D U C T I L E T R A N S I T I O N IN P O R O U S R O C K S
Ketan R. Shah

Cornell Fracture Group and Cornell Theory Center, Cornell University, Ithaca, New York 14853, USA
ABSTRACT

A unified elasto-plastic constitutive model based on the observed behavior of porous rocks is proposed. The model utilizes the concept of critical state (CS), used for soils in the Cam clay models, to separate brittle dilatant deformations from the compactive ones. The CS represents ultimate (residual) state of stress at which large shear deformations occurs with steady level of stresses and fixed porosity. The Cam-clay phenomenology is combined with the poroelasticity theory to develop a comprehensive elasto-plastic model for porous rocks. The yield surface is allowed to evolve with both plastic volumetric and deviatoric strain such that dilatant hardening/softening and compactive yielding can be predicted. The potential function is assumed to have the same form as the yield function, but the constants have different values to predict non-associative dilatancy. The model parameters are evaluated from the experimental data of sandstones and the model capabilities are demonstrated through the calculated mechanical response in triaxial tests under a wide range of confining pressures.
Copyright @ 1997 Elsevier Science Ltd

KEYWORDS C o n s t i t u t i v e R e l a t i o n s E l a s t o - p l a s t i c M o d e l C r i t i c a l State B r i t t l e - d u c t i l e T r a n s i t i o n C a m C l a y M o d e l P o r o e l a s t i c i t y D i l a t a n c y N o n - a s s o c i a t i v i t y Inelastic C o m p a c t i o n P o r o u s R o c k s INTRODUCTION

Stress-strain curve of porous rocks under hydrostatic loading or high pressure has an inflection point which is followed by a large porosity reduction (e.g. Brace, Riley 1972). The failure is through distributed inelastic mechanisms such as pore collapse and grain crushing resulting into cataclastic flow and permanent compaction (Curran, Carroll 1979; Zhang et al. 1990). The required differential stress for the onset of inelastic flow decreases with the pressure and the yield surface is well represented by an elliptical cap in mean-deviatoric stress (p - q) space (Wong et al. 1997). The low-pressure deformations are brittle and dilatant with failure taking place in the form of localization into shear faults. The shear strength has positive pressure-dependency with frictional yield surface such as Mohr-Coulomb one fitting the data well (Jaeger, Cook 1979). At the intermediate range of pressures, the state of deformation is more complex and may be accompanied by multiple conjugate shear bands (kageson-Loe et al. 1993) growing with constant volume or compaction. There are engineering problems in which mean stress varies so much that both brittle and ductile modes of deformation are present. An indenter pressing against the rock surface induces large compressive stresses in its vicinity leading to compaction (Miller, Cheatham 1972; Suarez-Rivera et al. 1990).

ISSN 0148-9062

To cite this paper: Int. J. Rock Mech. & Min. Sci. 34:3-4, paper No. 283.

Copyright 1997 Elsevier Science Ltd

Stresses outside of this compaction zone are low enough so that the brittle fracturing may be induced. Dynamic loading as in blasting and perforating oil wells induces a wide range of stresses and different deformation modes at a single location. Simulation of the above problems requires a unified model encompassing both positive and negative pressure-dependent inelastic deformation. A conventional approach has been to augment a brittle yield surface with an enclosing cap in mean-deviatoric stress space (DiMaggio, Sandler 1971; Lubarda et al. 1996). Such cap models require two different functional forms of the yield surfaces and their separate evolution laws as well as the complex manipulations to track the comer at the junction. The need for a smooth single yield surface has been recognized in the work ofDesai et al. 1988 and Desai 1989. Desai et al. 1988 introduced a family of hierarchical single surface models which require a minimum number of parameters. Cam clay models use a single elliptical or similar yield surface inp - q space and have been successful in predicting the response of different soils (Schofield, Wroth 1968). These models are based on the concept of critical state (CS) which represents a limiting steady state of stress. The critical state for rocks may represent residual state of stress (Georgiannopoulos, Brown 1978). Work hardening/softening in Cam clay is assumed to be a function of the plastic volumetric strain which becomes zero on the critical state line (CSL). The deformations are dilatant and compactive, respectively, to the left and fight of CSL. Cam clay models along with the concept of CS provides a general framework to simulate both brittle and cataclastic deformations for rocks. Cam clay models utilize experimentally observed linear relationship between the void ratio and logarithm of effective mean stress for both loading and unloading. The elasticity becomes nonlinear and pressure-dependent and this is the main way in which Cam clay differs from the classical plasticity formulation which includes single-surface models of Desai 1989. Cam clay models are not directly suitable for the mechanical simulation of porous rocks and will need to be modified to account for their qualitative differences with soils. One of the major weakness of Cam clay is that dilatancy always induces softening. The intercept of a Cam clay yield surface with the p-axis is used as a hardening/softening parameter and it reduces with dilatancy (Schofield, Wroth 1968). The frictional hardening with shear strain may be superimposed such that the dilatant hardening may be predicted. The dilatancy is known to be non-associative which may be accounted for by using a potential function derived from the simple modification of the yield function (Drescher et al. 1995). Cohesion and tensile strength of rocks is generally not negligible and can also be included in the model by simply shifting the yield surface along thep axis. Cam clay disregards the compressibility of solid matrix and pore fluid. This may not be reasonable for porous rocks and the compressible constituents may contribute more than 10% of the volumetric deformations for porous sandstones. Biot's 1941 coupled theory for fluid-saturated medium provides a rational basis to account for solid and fluid compressibility. The constitutive equations of Biot's poroelasticity will be written in incremental form to introduce nonlinearity (Detoumay, Cheng 1993). The Cam clay plasticity will be added to it to formulate a comprehensive constitutive model for rocks. The model developed here will be verified by comparing the results with the experimental data of sandstones given by Wong et al. 1997.

E X P E R I M E N T A L DATA A N D C R I T I C A L STATE F O R R O C K S
Positive pressure-dependent shear strength and dilatancy of rocks in the brittle regime have been well documented (e.g. Jaeger, Cook 1979) and numerous efforts through Mohr-Coulomb or Drucker-Prager

ISSN 0148-9062

To cite this paper: Int. J. Rock Mech. & Min. Sci. 34:3-4, paper No. 283.

Copyright 1997 Elsevier Science Ltd

plasticity have been made to model them. The onset of dilatancy and plasticity is followed by hardening, peak strength, and softening. Localization into shear faults accompanies the softening and ultimately the level of stresses reaches the limiting residual strength. This steady state also has positive pressure-dependency and is primarily determined by friction along the faults. Brace, Riley 1972 compressed different rocks under uniaxial strain condition up to the pressure of 30 kilobars (3 GPa) and observed that appreciable permanent compaction takes place. Most of the high-porosity rocks showed an inflection point in their stress-strain response which was followed by rapid inelastic compaction and hardening. Their microstructural observations for Indiana limestone indicated that pores are eliminated through the plastic deformation of calcite grains. Cataclastic flow in porous silicate rocks is due to pervasive intragranular cracking (Edmond, Paterson 1972; Hirth, Tullis 1989; Zhang et al. 1990). The onset of compaction yielding has negative pressure-dependency (Hirth, Tullis 1989; Wong et al. 1997). The yield is followed by the hardening at a decreasing rate and the stresses approach the limiting residual values. At large strains, dilatancy may initiate and cause the faulting (Hirth, Tullis 1989; Wong et al. 1992). The residual strength in such a case is governed by the friction similar to the brittle deformations. It can be concluded from these experimental data that in both brittle and ductile regimes, a limiting, positive pressure-dependent residual state of stress can be defined. The concept of critical state (CS) has been based on the observed critical void ratio for soils at which large shear deformations initiate at peak loads without further change in the volume (Schofield, Wroth 1968). The level of stresses do not alter and so the critical state represents the steady residual state of stress. Since the experimental data for porous rocks show an existence of residual strength, they may be used to define the critical state. Although the experimental data are not conclusive, it would be required that pore volumetric strains are zero in the residual state to define it as CS. The volume change in the cataclastic regime is negative whereas for the brittle deformation it is positive. There generally exists an intermediate state of stress in which large shear deformations persist at constant volume (Wong et al. 1997). The CS separates dilatant deformations from the compactive ones and so represents the brittle-ductile transition. Wong et al. 1997 provided a comprehensive database for the mechanical behavior of a variety of sandstones. Their data for the onset of dilatancy and peak strength in brittle regime and for the initiation of inelastic compaction in ductile regime for Adamswiller sandstone are shown in Figure 1. They showed that the differential stress at the onset of cataclastic flow has a negative pressure-dependency (shown by filled squares in Figure 1) and an elliptical cap fits the data well. The onset of dilatancy (shown by ) takes place at less than 60% of the failure stress ( in Figure 1) and considerable hardening is present.

C A M CLAY M O D E L F O R R O C K S
Consider a representative element of volume V in a porous rock with mean stress tensor (yO.and pore fluid pressure P as state variables. Biot's 1941 formulation uncouples the bulk shear deformation from the fluid flow for an isotropic material and so deviatoric strain is dependent on the deviatoric stress only. Writing this in an incremental form:

where sij = (Yij -PSij is the deviatoric stress tensor withp being the mean stress (yk#/3and 8~. being the

ISSN 0148-9062

To cite this paper: Int. J. Rock Mech. & Min. Sci. 34:3-4, paper No. 283.

Copyright 1997 Elsevier Science Ltd

Kronecker delta tensor, eeij is the elastic part of deviatoric strain tensor defined as Deij - Dek/flij/3 where Dij is the strain tensor and the superscript e is used to denote the elastic part. Compression will be assumed to be positive for stresses, pressure and strains. The bulk shear modulus ~t is in general a function of state variables (y~).and P. The bulk volumetric strain and the change in specific volume v (bulk volume per unit volume of solids) are linearly related to the effective pressure p ' = p - P and pore pressure P (Biot 1941; Rice, Cleary 1976; Detournay, Cheng 1993):

de~k = dt dP K -t K~

(2) 1

(3)

--

where K i s the drained bulk modulus, = (v - 1)/v is the porosity, and K s is the solid bulk modulus. These equations are simplified from that of Biot 1941 relationships under the assumption of isotropic and inert solid and continuous pore space (Rice, Cleary 1976). Although not mentioned explicitly, bulk moduli K and K s are in general the function of stress tensor and pore pressure. It would be assumed here that the solid is elastic and bulk modulus K s remains a constant. It can be seen from Eqn. 3 that elastic incremental change in specific volume is directly proportional to the incremental effective pressure. The void ratio e and so the specific volume (v = 1 + e) are assumed to be an explicit function of effective pressure for soils in Cam clay models. Specifically, v for elastic loading or unloading is linear in the lnp'. This relationship breaks down w h e n p ' is zero or negative and since rocks may have finite tensile strength, we assume v to be a linear function of In (p' + Po) withpo being the isotropic tensile strength.
=

In (p' + po)

(4)

where ~: is a slope o f v versus In (p' +Po) line and v o is the specific volume a t p ' = -Po + 1 in elastic regime. Differentiating the above equation with respect to p ' and comparing with Eqn. 3 gives

1 - -

K(p')

1
(1 - ) K ~

~;

(1:/+po)v

(5)

It would be assumed that the shear modulus remains constant and hence the Poisson's ratio v is a function of effective pressure given by

3 K ( p ' ) - 2t, "(P') = 2(3K(p')+ u)

(6)

The Poisson's ratio is pressure-dependent and may become negative for low values of confining pressure which is consistent with the experimental data of Wong et al. 1997. The alternative formulation in which v is assumed constant may be used to alleviate the negative v. It can be shown from the elementary relationship V = v V s where Vs is the solid volume that the bulk

ISSN 0148-9062

To cite this paper: Int. J. Rock Mech. & Min. Sci. 34:3-4, paper No. 283.

Copyright 1997 Elsevier Science Ltd

volumetric strain decomposes into the following form:

AV V

Av -- kk = v

AV, V,

(7)

Since it is assumed that solid matrix behaves elastically, bulk plastic volumetric strain is negative of the AvP/v, i.e.,

gk =

Av/'
v

(s)

where the superscriptp denotes the plastic part. This relationship (Eqn. 8) is used subsequently in the development of the elastoplastic model.

Cam clay plasticity


The modified Cam clay model uses an elliptical yield surface in mean-deviatoric stress space such that it intersects the mean stress axis at 0 andpc. The consolidation pressurepc is the past maximum effective pressure under hydrostatic loading and is utilized to define hardening and softening. During an isotropic elastoplastic loading, specific volume is linearly related to In (p' + Po) similar to the elastic part: v = vl - A In (p' +

Po)

(9)

where )~ is a slope ofv versus In (p' +Po) line and Yl is the specific volume atp' = -Po + 1. The plastic volumetric strain from Eqns. 8, 4, and 9 can be written as:

d~kk = _ (dVv

dve)v = (~ - + Po)

(10)

For hydrostatic loading, p ' = Pc and so the evolution law forpc becomes:

dv~ = v(p' + po)d~k


Although Cam clay models have been developed with only logarithmic (original) and elliptical (modified) yield surfaces (Schofield, Wroth 1968), a possible family of surfaces may be developed by requiring that each surface intersects the effective pressure axis exactly twice. These two intersections must be -Po andp~, withp~. > -Po. This general yield surface F depends upon the effective pressurep' and a parameter related to the second invariant of deviatoric stress tensor sij denoted as q. It is defined to be

~/3sqsq /2 such that it equals the differential stress for triaxial tests. Associativity is generally assumed
with Cam clay models, but this requirement is relaxed here and a different potential function G(p', q) = 0 similar to Drescher et al. 1995 will be defined. The plastic strains are given by

ISSN 0148-9062

To cite this paper: Int. J. Rock Mech. & Min. Sci. 34:3-4, paper No. 283.

Copyright 1997 Elsevier Science Ltd

OG
The critical state is that limiting state in which incessant deformations take place with infinitesimal change in the stresses. Mathematically, within the context of elastoplastic modeling, this requires do~).= 0 = P; dF = 0 and dD~). 0. Ifp~. is the only hardening parameter then this implies dG. = 0 = dDpkk (Eqn. 11). The necessary condition for this from the flow rule (Eqn. 12) is that

OC/Op'=O
Unified model f o r porous rocks

(13)

The above formulation improved the Cam clay models by accounting for cohesion (through the parameterpo ) and solid matrix compressibility. The failure surface for a variety of rocks may be well approximated by an ellipse (Figure 1). The first yielding in the brittle regime or the dilatancy initiates at a significant lower level of stresses. A general yield function which may be skewed is needed to better fit the experimental data. A function which allows to adjust the value of effective pressure at the peak of the function and as a special case is reduced to the ellipse is given by:

F= q 2 M~pc(P' +---P~"(Pc-P')= 0

\Pc

(14)

where M 1 and n are the parameters. For n = 1, F reduces to be an ellipse and M 1 becomes the slope of critical state line for the associated flow rule. The skewed surface that was used by Drescher et al. 1995 t is obtained by substituting n = 2. The value of effective pressure p peak at which q is maximum is

Ppeak - -

npc-Po
n + 1

(15)

Since the equation for the slope of the line connecting the point (-Po, 0) to the peak is not trivial, a slightly different form of the function is written such that the parameter M equals that slope.

F = q2_ M2(n + l)n-1 (Pt +Po~r' nn_2 (Pc+Po) \-PcT~oo,I (Pc-P')=0

(16)

Drescher et al. 1995 formulated a non-associative Cam clay model to predict the peak and post-peak response of a normally consolidated soil in undrained condition. The non-associativity was introduced by using another parameter N replacing the M in the yield function to define the potential function. The parameter Nwas required to be higher than M. The same idea is used here to obtain G from Eqn. 16

G=q2_N2(n+l)"-I (g +P) (P'+Po) '~

(t7)

Since G is defined as a surface passing through a given (p',q) point following the above Eqn. 17, it does

ISSN 0148-9062

To cite this paper: Int. J. Rock Mech. & Min. Sci. 34:3-4, paper No. 283.

Copyright 1997 Elsevier Science Ltd

not intersect the horizontal axis atp~., but atpg c. It can also be shown that when N > M, I 3G/3pt <
I cPF/3pt and so the magnitude of plastic volumetric strain is less than that for the associated flow rule.

The above form of potential surface effectively models the experimentally observed nonassociated dilatancy. The dilatancy in the brittle regime forces Pc. to reduce and shrinks the yield surface along the horizontal axis leading to softening. The onset of dilatancy and the peak load are coincident and Cam clay models in its present form do not predict dilatant hardening. One way to account for the hardening is to allow parameters M and n to evolve simultaneously with Pc in such a way that it expands the yield surface in vertical direction. Since M is related to the slope of the yield surface in brittle regime and so to the internal friction angle, the evolution of M is equivalent to the frictional hardening or to the mobilization of friction. M and n are assumed to harden with the plastic deviatoric strain similar to the hardening/softening Mohr-Coulomb plasticity. The following form of functions are chosen for M and n such that the their initial values are M o and n o, respectively, and they approach the ultimate values M~ and n~.
M = (n= Mo)e no)e-Z'

n =

(19)

The constants c~ and 13are to be chosen so as to fit the experimental stress-strain response. The plastic deviatoric strain Dpq = ~ = ~/2COC/j/3 is a work conjugate to q. The parameter X would be assumed to be a constant times M for the lack of any better suitable representation with that constant being greater than 1.

COMPARISON WITH EXPERIMENTAL DATA


The proposed model for rocks needs eight additional constants besides that needed for the modified Cam clay model. Bulk modulus of solids K s, isotropic tensile strengthpo, parameters @, ]3, n o, n~, and M~ which define the shear hardening of the yield surface, and the constant ratio of N with M measuring the non-associativity are to be fitted to the experimental data. Cam clay models require 3, ~c, and v 1 defining the functional form for the specific volume v, bulk shear modulus p, and the slope made by the peak of the yield surface M ( M o here). The only parameter that records the history in Cam clay models is the past maximum consolidation pressurepc. The additional history parameter that is required here is deviatoric plastic strain or any of the constants M, N and n. The above set of parameters will be evaluated based on the experimental results of Wong et al. 1997 for Adamswiller sandstone. The results of specific volume versus the hydrostatic pressure are shown on a semi-log plot in Figure 2. The bilinear curve is a good fit to the data with the intersection representing the first yielding and the consolidation pressure Pc. = 190 MPa. The slope of the lines to the left and right of o f p ' = Pc. defines ~c(=0.0159) and 3, (=0.1567), respectively. The intersection of the right line with p ' = -Po + 1 determines the constant v 1 (=2.065). If the specimen volumetric strain measurements are available, then solid bulk modulus K s may be evaluated using Eqn. 5. K s may also be found from the known mineralogy & t h e solids and published mineral properties and was taken to be 38.4 GPa (Wong

ISSN 0148-9062

To cite this paper: Int. J. Rock Mech. & Min. Sci. 34:3-4, paper No. 283.

Copyright 1997 Elsevier Science Ltd

e t a l . 1997). The bulk shear modulus ~t was found to be 3.75 GPa from the initial elastic modulus at 5 MPa confining pressure.

The parameterpo is the isotropic tensile strength and may be estimated from the uniaxial tensile strength u t by substituting p ' = ut/3 and q = u t in Eqn. 16. It may also be evaluated by extending the failure surface for the negative values of mean stress as shown in Figure 1 and was found to be 10 MPa. The Figure 1 shows two surfaces which correspond to the first yielding and the failure. These two surfaces coincide for the ductile regime, but the failure stress is significantly larger in the brittle regime. The evolution from the first yielding to the failure determines the functional form of M and n. The parameter n is a measure of the skewness of the surface, whereas the M is related to the vertical extent of the surface. For the experimental data in Figure 1, n o = 2, n~ = 1.22, M o = 0.89, and M~ = 1.14 were a good fit. The parameters cz and 13were taken high enough so that they cancel the softening effect ofpo and allow the dilatant hardening and were selected to be equal (400.0). The parameter N was adjusted to reproduce the observed dilatancy angle and residual strength and was taken to be 1.3M. Figure 3 shows the predicted axial strain versus differential stress for several values of confining pressures. Corresponding change in porosity are shown in Figure 4. The qualitative nature of both the sets of curves closely resemble the experimental data in Figure 1 of Wong e t a l . 1997. The results of pressuredependency of peak strength, inflection point and residual state match very well with their data for Adamswiller sandstone. At low values of confining pressures (5 and 20 MPa), peak load is followed by softening and steady state. Porosity decreases (porosity change increases) first and then increases due to the dilatancy. The initiation of dilatancy is prior to peak loads in agreement with the experiments. The slope of dilatancy versus axial strain curve which is related to the dilatancy angle match the data for 5 MPa pressure. The axial strain at the peak loads is almost constant (2%) contrary to the experiments in which it increases with the confining pressure. The elastic modulus has the correct trend, but the model over-estimates the increase in the modulus with the pressure. These observations suggest that the elasticity in brittle regime is not simple and Eqns. 5 and 6 with constant shear modulus may not be sufficient. For larger confining pressures, there is no softening and the inflection point in the stress-strain response is followed by compactive hardening. The axial strain at the inflection point has the correct trend as well as the value that matches the experiments. The elasticity of the model captures the behavior in cataclastic regime well. Compaction rate as defined by the slope of a curve in Figure 4 for the intermediate range of confining pressures (20, 40, and 60 MPa) are more negative (dilatant) than the experimental measures. Although a more careful parameter fitting would reduce these errors, the differences may be indicative of the model limitations. The experimental data for 40 MPa has positive compaction rate and mild softening. The present model can predict softening only if their is dilatancy. CONCLUSIONS A comprehensive Cam-clay type model for the mechanical response of porous rocks was developed. The major contributions and model limitations are summarized below: The concept of critical state (CS) and linear relationship between the void ratio and the logarithm of effective pressure, which differentiate Cam clay from other elastoplastic models, were shown to be applicable to porous rocks. The critical state represents the residual state of stress and separates brittle and cataclastic deformation modes in stress space. Although the assumption of constant

ISSN 0148-9062

To cite this paper: Int. J. Rock Mech. & Min. Sci. 34:3-4, paper No. 283.

Copyright 1997 Elsevier Science Ltd

volume during the residual state needs to be verified, CS provides a firm basis to develop a unified model for porous rocks. The Cam-clay utilizes a single surface to model yielding in both brittle and ductile modes and predicts the corresponding dilatant and compactive deformations. It alleviates the need for two different functions used in Cap models and so is simpler to implement. A generalized skewed yield surface in mean-deviatoric stress space was introduced which allows to represent complex failure surfaces. A suitable hardening with respect to plastic shear strain was added to model the observed dilatant hardening. The non-associative dilatancy was accounted for by having a potential surface which has the same form as the yield surface but uses different values for certain constants. The model was formulated through the framework of the poroelasticity theory and so it accounts for solids and fluid compressibility. The model can be easily extended to couple the fluid flow with the mechanical response. The model response of the initiation of inelastic deformation, volume change, and peak and residual strength matched well with the experimental observations. The elastic response in the brittle regime and the volume change in the transition regime need to be improved.

Acknowledgments:
Partial support was provided by National Science Foundation award number 9503981 and the Cornell Theory Center. B. Birgisson and T.-F. Wong are thanked for reviewing this manuscript and for useful discussions. W. Zhu and T.-F. Wong provided with the experimental data. A. Drescher introduced the author to the field of critical state soil mechanics.

FIGURES

ISSN 0148-9062

To cite this paper: Int. J. Rock Mech. & Min. Sci. 34:3-4, paper No. 283.

Copyright 1997 Elsevier Science Ltd

Paper 283, Figure 1.


.,-%

L".4-L~ " - r -

...........................................................................

,', "-, ,', ",

, , , " , , " , , ", , - ~

120
l O0

~0 60 40
qi

a Peak brittle strength Onset of dilatancy

20 0

il

Inliation o[ cataclastic flow


0 5 l) l I')0 150 200

. . . . . . . . . . . . . .

-50

Mean stress (MPa)

Figure 1. The experimental data for the onset of dilatancy, peak strength, and for the initiation of cataclastic flow in Adamswiller sandstone are shown (Wang et al. 1997). Two surfaces are fitted to represent first yielding and failure.

Paper 283, Figure 2.


l.g8
:.-_ ,1~_ _ ----!~--_ _ --0--___ ,

1.2 ~. ~. 1.16 ~3
L .L'J'

,\

Exl~erimental data
,
[

\
\*
',

1,01~

]0

1 O0

1000

EffecU~'e pressure 10 (Ml'a)

Figure 2. The experimental data for the variation of specific volume under hydrostatic loading in Adamswiller sandstone (Wang et al. 1997). The bilinear curve fits the data well and defines the parameters ~: and X.

ISSN 0148-9062

To cite this paper: Int. J. Rock Mech. & Mm. Sci. 34:3-4, paper No. 283.

Copyright 1997 Elsevier Science Ltd

Paper 283, Figure 3.


3OO 250 T 2O0-

150.

~,.....:..:::............::.,.. ......:Z

........ ,o~,~.~.

' !y
OI

5
Axial strain (%)

10

15

Figure 3. The model prediction of differential stress versus axial strain for various values of confining pressure.

Paper 283, Figure 4.


_.-1 ..

/ 5 MPa
/ /

" i

-2-

_.._=.~_yi,;,

.......

6 -

L0

......

[0

15

Figure 4. The model prediction of porosity change versus axial strain for various values of confining pressure.

ISSN 0148-9062

To cite this paper: Int. J. Rock Mech. & Min. Sci. 34:3-4, paper No. 283.

Copyright 1997 Elsevier Science Ltd

References

References
Blot M.A. 1941. General theory of three-dimensional consolidation. J. Appl. Phys., 12, 155-164. Brace W.F., Riley D.K. 1972. Static uniaxial deformations of 15 rocks to 30 kb. Int. J. RockMech. Min. Sci., 9, 271-288. Curran J.H., Carroll M.M. 1979. Shear stress enhancement &void compaction. J. Geophys. Res., 84(B3), 1105-1112. Desai C.S. 1989. Single surface yield and potential function plasticity models: A review. Comput. Geotech., 7, 319-335. Desai C.S., Somasundaram S., Frantziskonis G. 1988. A hierarchical approach for constitutive modeling of geologic materials. Int. J. Num. Analyt. Meth. in Geomech., 10, 225-257. Detournay E., Cheng A.H.-D. 1993. Fundamentals ofporoelasticity. In Comprehensive Rock Engineering, volume 2, chapter 5, pages 113-171, Pergamon. DiMaggio F.L., Sandler I. 1971. Material model for granular soils. J. Engn Mech. Div., ASCE, 97, 935-950. Drescher A., Birgisson B., Shah K. 1995. A model for water-saturated loose sand. In G.N. Pande and S. Pietruszczak, editors, Proc. o f the 5th International Conference on Numerical Models' in Geomechanics, pages 109-112, A.A. Balkema, Rotterdam. Edmond J.M., Paterson M.S. 1972. Volume change during the deformation of rocks at high pressure. Int. J. Rock Mech. Min. Sci., 9, 161-182. Georgiannopoulos N.G., Brown E.T. 1978. The critical state concept applied to rock. Int. J. RockMech. Min. Sci. & Geomech. Abst~, 15, 1-10. Hirth G., Tullis J. 1989. The effect of pressure and porosity on the micromechanics of the brittle-ductile transition. J. Geophys. Res., 94, 17825-17838. Jaeger J.C., Cook N.G.W. 1979. Fundamentals' o f Rock Mechanics. Chapman and Hill, London, Third edition. Kageson-Loe N.M., Jones M.E., Petley D.N., Leddra M.J. 1993. Fabric evolution during deformation of chalk. Int. J. Rock Mech. Min. Sci. & Geomech. Abst~, 30, 739-745. Lubarda V.A., Mastilovic S., Knap J. 1996. Brittle-ductile transition in porous rocks by cap model. J. Engrg. Mech., 122(7), 633-642. Miller T.W., Cheatham J.B. 1972. Analysis of the indentation of a compacting material by a perfectly rough wedge. Int. J. Rock Mech. Min. Sci., 9, 475-492.
Rice J.R.,

Cleary M.R 1976. Some basic diffusion solutions for fluid-saturated elastic porous media with compressible constituents. Rev. Geophys. Space Phys., 14(2), 227-241. Schofield A.N., Wroth C.R 1968. Critical State Soil Mechanics. McGraw-Hill, London. Suarez-Rivera F.R., Cook N.G.W., Cooper G.A., Zheng Z. 1990. Indentation by pore collapse in porous rocks. In W.A. Hustrulid and G.A. Johnson, editors, Rock Mechanics Contributions and Challenges, pages 653-660, A.A. Balkema, Rotterdam.

Wong T.-F., David C., Zhu W. 1997. The transition from brittle faulting to cataclastic flow in porous sandstones: Mechanical deformations. J. Geophys. Res., In Press.

ISSN 0148-9062

To cite this paper: Int. J. Rock Mech. & Mm. Sci. 34:3-4, paper No. 283.

Copyright 1997 Elsevier Science Ltd

Wong T.-F., Szeto H., Zhang J. 1992. Effect of loading path and porosity on the failure mode of porous rocks. AppI. Mech. Rev., 45(8), 281-293. Zhang J., Wong T.-F., Davis D.M. 1990. Micromechanics of pressure-induced grain crushing in porous rocks. J. Geophys. Res., 95(B1), 341-352.

ISSN 0148-9062

S-ar putea să vă placă și