Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

Stability Analysis of Rock Slope by Using RFPA-SRM Method

Cheng Jia
School of Civil Engineering, Yancheng Institute of Technology, Yancheng, China e-mail: jctonm@163.com

Huihui Chen*
School of Civil Engineering, Yancheng Institute of Technology, Yancheng, China e-mail: chenhuihui@ycit.edu.cn * Corresponding Author

ABSTRACT
To intuitively exhibit the failure process of engineer structure as strength reduces, slope stability is analyzed by using RFPA-SRM method based on theories of FEM and statistical damage. The RFPA-SRM can be used to model the observed evolution of damage or crack initiation, propagation and coalescence in brittle materials by allowing the linear elastic elements to fail in a brittle manner. In this paper, the stability of simple rock slope and No.3 dam section of the Three-Gorges Project were analyzed. It is shown that the RFPA-SRM is a valid and superior method and could be an effective tool for analyzing complex slope.

KEYWORDS: rock slope; stability analysis; strength reduction method; failure surface

INTRODUCTION
Slope stability is a subject that classic soil mechanic attempts to solve the earliest but not solve satisfactorily. For many years, a lot of researchers proposed many methods for this area, such as limit equilibrium method, limit analysis, finite element method and so on, promoting the development of slope stability analysis. Finite element method based on shear strength reduction as a tool for analyzing slope stability started from 1970's. Zienkiewicz[1] first put forward shear strength reduction technique in 1975. Ugai[2] assumed the soil as ideal elastic-plastic material, analyzed stability of vertical slope, inclined slope and heterogeneous slope using shear strength reduction technique. And other researchers, such as D. J. Naylor, E. M. Dawson, I. B. Donald and J. M. Duncan also studied shear strength reduction technique, analyzed stability of many practical slopes using this technique. Finite element method based on shear strength reduction has some unique advantages, of which most important one is the slip surface could be acquired automatically. Although conventional finite element method could do well in simulating nonlinear behavior in rock - 1189 -

Vol. 19 [2014], Bund. E

1190

deformation, the nonlinear behavior has been assigned by introducing a nonlinear constitutive law to elements which are considered to be homogeneous material. Though this method has gained a sufficient degree of functionality, it may not always help in our understanding of why rock material demonstrates nonlinear behavior. In this paper, slope stability was analyzed by using RFPA-SRM method which can be used to model the observed evolution of damage or crack initiation, propagation and coalescence in brittle materials.

BASIC PRINCIPLE OF RFPA


RFPA(Rock Failure Process Analysis Code), based on finite element method, developed by Tang, Northeastern University, P.R. China, is a comprehensive two-dimensional rock failure package. The code is based on the idea that the heterogeneity is the source of nonlinearity [3], using continuum mechanics and the linear method for numerically processing nonlinear and discontinuum mechanics problems in rock failure. Details of the RFPA code are in published literature [4], and so only a brief outline will be given here. Statistical damage constitutive model is adopted for reflecting heterogeneity of rock material and randomness of defects distribution. The statistical distribution hypothesis for stone-like material properties is combined with finite element method. And broken treatment for elements is accomplished during the simulation according to a given strength criterion. In the present investigation, a Coulomb criterion envelope with a tensile cut-off [5] is used so that the elements may fail either in shear or in tension. Thus, the failure progress of heterogeneity material could be simulated. For heterogeneity, the material properties (failure-strength c and elastic modulus Ec) for elements are randomly distributed throughout the specimen by following a Weibull distribution [4]:

( ) =

m m 1 ( ) exp ( ) m 0 0 0

(1)

where is the element strength and 0 is the mean strength of the elements for the specimen. For an elastic modulus, E, the same distribution is used. The variable m is defined as the homogeneity index of the rock[4]. According to the definition, a larger m implies a more heterogeneous material and vice versa.

RFPA WITH STRENGTH REDUCTION METHOD (RFPASRM)


The fundamental principles of the strength reduction method (SRM) are incorporated into the RFPA code to produce an RFPASRM method for analyzing the failure process and stability of rock and soil slopes. The RFPASRM method not only satisfies the global equilibrium, strain-

Vol. 19 [2014], Bund. E

1191

consistent, and nonlinear constitutive relationship of rock and soil materials but also takes into account the heterogeneous characteristics of materials on the micro- and macro-scales. When the proposed method is used for slope stability analysis, both the critical failure surface and the safety factor can be obtained directly without any assumptions regarding the shape and location of the failure surface. The key problem of slope stability analysis is how to judge the critical instability state based on the results. Usually, slope failure is accompanied by large deformation. And large deformation will necessarily leads to element broken. In RFPA-SRM method, the reduction step when the largest number of broken element occurs in the process of computation is used as critical step of slope instability. Essentially, this approach is consistent with two criterions[6-7]. RFPA-SRM code could automatically record the number of broken elements during computation process. Accordingly, judging slope instability by this approach is very simple and effective. Safety factor, also called stability factor, is an important conception in analysis of slope stability. Base on the strength reduction method, combined with the element constitutive model of RFPA, initial strength f0 in RFPA-SRM is linearly degraded as follows

f 0trial =

f0 Fstrial

(2)

where f0 is unified element compressive strength and shear strength, Fstrial is the trial safety factor and f 0trial is the trial strength of the element. The trial strength f 0trial is used in RFPASRM to investigate the strength of the geological medium (in this case, the rock masses). Slope stability simulation in RFPA-SRM is run with the trial strength f 0trial until the critical failure surface in slope is determined. Each f 0trial corresponds to a Fstrial during slope stability simulation. Provided that the number of broken element reaches the maximum value, a macro failure surface forms and slope failure occurs. Simultaneously, the corresponding Fstrial is the safety factor Fs of the slope.

VALIDATION AND APPLICATION OF RFPA-SRM


Simple rock slope
The simple rock slope is adopted as numerical example to verify the validity of the RFPASRM model in the analysis of rock slope stability. Figure 1 shows the computational model which is discretized into 14080 (11200) mesh elements. The element parameters include Youngs modulus E, the uniaxial compression f0, and Poissons ratio. The elements provide resistance against compressive or tensile deformations which are governed by the constitutive law. The input parameters, i.e. Youngs modulus E, compressive strength f0, Poissons ratio, and density

Vol. 19 [2014], Bund. E

1192

of the material, are 1200 MPa, 1 MPa, 0.32 and 24 kN/m3, respectively. The trial safety factor Fstrial is gradually increased until the critical failure surface is determined.
Figure 2 and Figure 3 show the critical failure surface and displacement vector respectively. It can be seen that larger displacement occurs and maximum displacement reaches 91 cm when the critical failure surface is completely formed. Figure 4 shows the failure surface obtained using conventional limiting equilibrium analysis. And it agrees well with results from RFPA-SRM. The number of broken elements recorded during computation, reaches peak at the reduction step of 353 which indicates that the critical failure surface has formed. Thus the safety factor of the slope is 1.54, while that is 1.42 obtained from limiting equilibrium method. Their close agreement verifies that RFPA-SRM is reasonably effective in slope stability analysis.

20 m 10 m

40 m 600

70 m
Figure 1: Geometry and boundary conditions for rock slope

Figure 2: Critical failure surface obtained by RFPA-SRM

Vol. 19 [2014], Bund. E

1193

Figure 3: Displacement vector obtained by RFPA-SRM

Figure 4: Critical failure surface obtained by limiting equilibrium analysis

No.3 dam section of Three-Gorges Project


Further analyses were performed to study complex slope. The Three-Gorges water conservancy complex located about half way along the Yangtze River is the largest multipurpose water conservancy project ever built in China. No.3-No.5 dam foundations of left powerhouse of Three-Gorges Project locate in relative developed area of gently dip angle structural surfaces where long and large gentle dip angle fracture and rock bridge distributed together. The main engineering geologic problem of left powerhouse of Three-Gorges Project is deep-sliding stability of No.3-No.5 dam foundations along the potential slip surface constituted by these structural surfaces. And No.3 dam section plays a dominate role in stability analysis because No.3 dam section is the weakest part in whole left powerhouse dam foundations. The crest elevation is 185 m, dam foundation plane elevation is about 85.0 to 90.0m and powerhouse foundation plane elevation is 22.2m. Thus a high and steep slope arises behind dam toe.

Vol. 19 [2014], Bund. E

1194

As mentioned, the stability of No.3 dam section is analyzed in this paper. Figure 5 shows the simplified sliding model of No.3 dam section and the values of material parameters are list in table 1.The dimensions of the dam foundation to be modeled are 170m in depth from foundation surface of dam downward, 180m in length directing upstream from the dam heel and 150m in length directing downstream from the powerhouse site. Three main slip surfaces, JI ABCDEabc, are taken into account in model. Horizontal freedoms at left and right boundary are restrained, and bottom is clamped boundary. Design upper water level is 175 meters and downstream water level is 62 meters. MohrCoulomb strength criterion is adopted in simulation.

Figure 5: The simplified sliding model of No.3 dam section

Table 1: The values of material parameters


Type Dam concrete Rock Structural surfaces Fault Foundation surface of dam Foundation surface of powerhouse Powerhouse Youngs modulus (GPa) 26 35 1 10 22 22 22 Poissons ratio 0.167 .023 0.3 0.3 0.167 0.167 0.167 24.5 Density (kN/m3) 24.5 27.0 26.0 26.0 Friction coefficient 1.1 1.7 0.7 0.9 1.1 1.25 1.1 Cohesion (MPa) 3.0 2.0 0.2 0.8 1.3 1.5 3.0

Vol. 19 [2014], Bund. E

1195

(a) failure process at step 1-1

(b) failure process at step 653-16

(c) failure process at step 653-17

(d) failure process at step 653-18

(e) failure process at step 654-1

(f) acoustic emission graph at step 654-1

Figure 6: Failure processes and acoustic emission


The potential failure processes of the slope are shown in figure 6(a)-(e). It can be seen that the first failure appears at the bottom of slip surfaces JI and ABCDE. Then several broken elements arose at top of slip surfaces JI, ABCDE and abc. Subsequently, with the reduction in strength, two major cracks are first initiated in the middle of JI and ABCDE. In step 653-18, two major cracks grow in two ends and a major crack is formed at bottom of slip surfaces abc. Finally, the crack at slip surfaces JI coalesces and a critical failure surface is formed in the slope.

Vol. 19 [2014], Bund. E

1196

As pointed out by Griffiths[6], failure is triggered at the weakest elements because the strength profile of the rock material is randomly distributed with specified mean and variance. Figure 6(f) is the acoustic emission graph when failure occurs in slope. The red and white areas stand for tensile failure and shear failure respectively. As can be seen in figure 6(f), distribution of acoustic emission agrees well with the failure model of slope. The major failure mode of the slope is a combination of shear and tensile fracturing. Thus the final safety factor is 2.62 according the reduction step when the slope failure occurs. This is consistent with safety factor 2.4 obtained from FEM[8]. This case study indicates that the heterogeneity of the model has a notable influence on the determination of the fracture paths and the final failure patterns. The RFPA-SRM model taking heterogeneity into consideration provides a realistic approach to the study of slope failure mechanisms.

CONCLUSIONS
In this paper, the RFPA method incorporated with the fundamental principle of strength reduction(RFPA-SRM) is briefly introduced. The RFPA-SRM model could satisfy the global equilibrium, strain-consistent, and non-linear constitutive relationship of rock and soil materials. Specially, heterogeneity of the material is introduced into the model. By using RFPA-SRM, both the critical failure surface of the slope and the safety factor can be determined directly without any assumptions regarding the location and shape of the critical failure surface. Compared to the conventional FEM, RFPA-SRM could visually show the slope failure process as the strength is linearly degraded while conventional FEM can only give the final results corresponding to a given trial safety factor, with no intermediate results reflecting the failure process. The stability analysis of simple rock slope and No.3 dam section of Three-Gorges Project are conducted using RFPA-SRM. The results show that RFPA-SRM is valid and superior and could be a convenient and effective method for analysis of complex slope.

ACKNOWLEDGMENT
This study was supported by Scientific Research Project for Introduced Talents of Yancheng Institute of Technology (project no. XKR2011016).

REFERENCES
1. Zienkiewicz O C, Humpheson C, Lewis R W(1975) Associated and nonassociated viscoplasticity and plasticity in soil mechanics,Geotechnique, Vol. 25, No. 4, pp 671-689.

Vol. 19 [2014], Bund. E

1197

2. Ugai K(1989) A method of calculation of total factor of safety of slope by elastoplastic FEM, Soils and Foundations, Vol. 29, No. 2, pp 190-195. 3. Tang C A(1995) Numerical simulation of rock failure process, Proceedings of the 2nd Youth Symposium on Rock Mechanics, Chengdu. 4. Tang C A(1997) Numerical simulation on progressive failure leading to collapse and associated seismicity, International Journal of Rock Mechanics and Mining Sciences, Vol. 34, No. 2, pp 249-261. 5. Brady BHG, Brown ET(1993) Rock Mechanics. 2nd ed, London:Chapman & Hall. 6. Giffiths D V, Lane P A(1999) Slope stability analysis by finite element, Geotechnique, Vol. 49, No. 3, pp 387-403. 7. LIU Zuo-qiu ZHOU Cui-ying DONG Li-guo(2005) Slope stability and strengthening analysis by strength reduction FEM, Rock and Soil Mechanics, Vol. 26, No. 4, pp 558-561. 8. Chen Guorong(1997) A research report on stability analysis of NO 3 dam section of Three- Gorges,Nanjing: engineering mechanics institute of hohai university, pp 510.

2014, EJGE

You might also like