Classification of cut slopes in weathered metasedimentary bedrocks

comprise quartz mica schist, graphitic schist and quartzite layers. Field investigations revealed that these meta-sedimentary rocks have gradational weathering profile based on differences particularly in textures, hardness, lateral changes in colour, and consistency of material extension. The results from uniaxial compressive strength tests confirmed field observations whereby failure occurred mostly on outcrops having joints almost perpendicular to foliation. From the kinematic analyses, the investigated cut slopes are unstable with possibilities of wedge and planar failures. Application of rock mass classification schemes including Rock Quality Designation (RQD) and Rock Mass Rating (RMR) yielded almost similar poor to good quality ranges for each investigated rock mass. While Slope Mass Rating (SMR) classified the cut slopes from stable to unstable slopes, this study categorized them into one actively unstable, four marginally stable and five stable slopes.


Introduction
Information gathered from field investigations by geologists and engineers is insufficient to predict geotechnical behaviour of rocks and rock masses, which calls for laboratory investigations to ascertain the response of rocks under a wide variety of disturbances.The uniform definition of geotechnical engineering by Murthy (2002), Venkatramaiah (2006), and Das and Sobhan (2013) is that it deals with the application of the principles of soil and rock mechanics to the design of foundations, retaining structures, and earth structures.In this investigated area, previous studies by Mohd Asbi & Associates (2005) and Andrew Malone Ltd (2007) classified the stability of the cut slopes using kinematic analyses.To ensure further understanding of the characterization of the slopes, this study carried out field observations, laboratory and computational analyses of the geotechnical properties of the rocks and then utilized the recorded and derived data for rock mass and slope mass classifications.

Geological setting of investigated area
The study area is along Pos Selim Highway in Perak State Malaysia.It traverses through the Titiwangsa Main Range mountainous terrain reaching highest altitude of 1587m above sea level at Gunung Pass and terminating at 1420m above sea level at Perak/Pahang states boarder.This study begins at latitude 4° 33' 44" and longitude 101° 18' 86" extending to 17.7km and terminating at latitude 4° 35' 95" and longitude 101° 20' 80".Granite and schist are the two major Formations covering 60% and 40% of the area respectively (Fig. 1).This investigation focused on the schist, which according to Tajul (2003)

Description of weathering profile over quartz mica schist
Since the failures in quartz mica schist occurred in the weathered zones, the characteristics of these weathered earth materials were studied in order to interpret the features of the morphological zones.The photograph and schematic diagram of weathering profile of location 010 shown in Fig. 3 is typical of quartz mica schist units.Employing BSI (1981) code of practice for site investigations, the weathering profile was graded.By identifying the lateral changes in colour, texture, hardness, and consistency of material extension, the weathering profiles over the quartz mica schist are broadly differentiated into three zones; pedological soil zone, intermediate zone and bedrock zone corresponding to zones I, II and III respectively.Based on differences particularly in textures and structures of original bedrock as well as degree of preservation of the constituent minerals, these three broad zones are further differentiated into thinner characteristic horizons.Zone I with vertical thickness of less than 8m which corresponds to grade 6 is further sub-divided into IA, IB, and IC .Horizons IA and IB are brownish to reddish brown pedological soil profile, whereby IA which is less than 1m is friable sandy clay, whereas IB which is less than 2m is firm sandy clay.Horizon IC is completely weathered, stiff yellowish brown sandy clay devoid of any distinct discontinuity plane.Greater attention is focused on zone II because of its extensive thickness, complexity, and easy accessibility.With vertical thickness measuring up to 72m, this zone II comprises about 24% vertical thickness of higher horizon IIA (corresponding to grade 5) of highly weathered brownish unit devoid of distinct discontinuity plane, but with intercalations of relicts of moderately weathered units.The lower horizon IIB is about 55% vertical thickness of the entire zone II, comprising moderately weathered grayish coloured quartz mica schist.It corresponds to grade 4, exhibits conspicuous quartzite veins, and distinct joint and foliation planes, but indistinct fault plane.Horizon IIC is the lowest horizon of zone II, and it corresponds to grade 3.With vertical thickness of about 21% of the entire zone II, this slightly weathered dark gray unit comprises distinct relict discontinuity planes and unweathered core-boulders which are very prominent.Zone III which is schematically represented only by morphological zone IIIA is unweathered bedrock which experiences the effect of weathering only along and between structural discontinuity planes.

Measurement of rock mass jointing
To ensure adequate representation of the joint parameters, the joint orientations were recorded using scanline and random survey techniques.Joint spacing and persistence were measured with ruler and tape while the aperture was measured with caliper.According to Palmström (2005), joint frequency is defined as the number of joints per meter length, and this was calculated as the inverse of joint spacing.The variations of joint spacing and frequency of joint sets are shown in Table 2.The average volumetric joint count Jv was calculated as the sum of the average frequency of the joint sets while field recording of the conditions of the joints include their hydraulic conditions, material infill and roughness.The general characterization of the rock mass jointing for the investigated slopes is presented in Table 3 and these data shall be applied in rock mass and slope mass classifications.

Kinematic stability assessment
Kinematic stability assessment method was applied in order to identify critical planes of weakness in the rock slopes and outline potential danger and likely modes of failure.Measurements of discontinuities were carried out with the use of both scanline and random survey techniques.The slope face orientation was also measured.These analyses were based on Markland's test as described in Hoek and Bray (1981).Pole intensity greater than 4% was regarded as a major discontinuity and assumed friction angle along the discontinuities in the rocks is averaged at 25°.Table 4 shows the summary of the orientations of the major discontinuity planes and slope face orientations.
Table 4. Summary of the orientations of slope locations and major discontinuities.

Uniaxial compressive strength test
Very large rock blocks were collected in the field with use of sledge hammer and chisel for this very test, and the laboratory determination of the strength of the rock samples was carried out using ISRM (1981a) standard.As shown in Fig. 4(a), the rock samples were prepared into right circular cylinders ranging from 12.5 to 15cm in height and 5cm in diameter using diamond embedded rock core drilling machine while Fig. 4(b) shows the set up of this test.The uniaxial compressive strength of the samples was then determined under dry condition which is their ultimate strength.

Uniaxial compressive strength (UCS)
It was observed that the tested samples failed axially following joint trends and fissures, almost perpendicular to foliation.This confirmed field observations whereby failure occurred mostly on outcrops having joints almost perpendicular to foliation.Although some metamorphic rocks tested by Horino and Ellickson (1970) and Broch (1974) showed that uniaxial compressive strength of the rocks are almost the same at different orientations of foliation, but in these tested quartz mica schist samples, there are significant differences in the strength in different orientations of the well developed foliations.Where the orientation of the foliation to the horizontal is less than 45˚ in some quartz mica schist samples, failure occurred along the foliation plane.Several tests showed consistency of failure modes in graphitic schist, whereby majority of the failures occurred along orientations of foliations at 45˚ with minimal oblique joint trends.The determined uniaxial compressive strength of quartz mica schist samples shown in Table 5 ranges from 56MPa to 117MPa, while graphitic schist has value of 87MPa.According to classification of the uniaxial compressive strength of rocks by Deere and Miller (1966), Bieniawski (1978), ISRM (1978 and1981b), the rocks of the study area fall within medium to high strength.This gives insight into the observed failure mechanisms whereby most of the failures were controlled by discontinuities.

J5
Classification of cut slopes in weathered meta-sedimentary bedrocks J5

Rock mass classification
Rock mass classification has been developing over the years and various classification schemes have considered a lot of factors such as water content, discontinuities and rock strength.In this study, the considered rock mass classification schemes include Rock Quality Designation (RQD) and Rock Mass Rating (RMR).

Rock Quality Designation (RQD)
Rock Quality Designation (RQD) index was developed by Deere (1963).It is defined as the percentage of intact rock mass length that are 10cm or longer from borehole drill cores.In absence of drill core logs but where discontinuity traces are visible in rock surface exposures, Palmström (1982) suggested that Rock Quality Designation (RQD) might be estimated from the number of discontinuities per unit volume using below equation: RQD = 115-3.3Jv (1) Where Jv known as the volumetric joint count is the sum of the number of joins per unit length for all joint sets.Using already determined values of Jv and applying equation 1, the RQD values for the rock masses are presented in Table 5.According to Deere (1968) relationship between RQD and the engineering quality of rock mass, the RQD values determined in this work range from 34.5% to 90% covering a broad range of poor, fair and good quality rocks.In view of non-consideration of joint orientation, joint condition, type of joint, infilling and stress condition by RQD index, Singh and Goel (1999) advised that its sole consideration for classification is insufficient to provide adequate description of a rock mass.

Rock Mass Rating (RMR)
Rock Mass Rating (RMR) or Geomechanics classification was initially developed by Bieniawski (1976).The system has evolved due to a better understanding of the importance of the different parameters and increased experience leading to changes to the ratings of parameters.As a result of these advancements, the Bieniawski (1989) version was employed in this study.This scheme uses six parameters: Uniaxial compressive strength of rock material, rock quality designation (RQD), spacing of discontinuities, condition of discontinuities, groundwater conditions and orientation of discontinuities.As explained in Bieniawski (1989), estimation of RMR is the sum of the total ratings of each of the above listed six parameters.Considering the dip angles of the joints in these high cut slopes, -5 rating adjustment for discontinuity orientations was used.As presented in Table 6, the discontinuity condition is the sum of joint persistence, aperture, roughness, infilling and weathering ratings.This RMR scheme has five rock mass classes determined from total ratings ranging from very poor rock to very good rock.
In order to determine the RMR for the investigated extended cut slope, the already determined values of the six parameters were substituted with their individual ratings and the rock masses range from fair to good rocks.The estimation of the RMR of the investigated rock masses are presented in Table 7. Unlike in RQD index where quartz mica schist at location 004 yielded highest value of 90%, in this RMR only quartz mica schist at locations 010 and 014 and graphitic schist at location 015 yielded highest value of 62 and designated good rocks.Although the remaining cut slopes are in quartz mica schist and designated fair rocks, the quartz mica schist at locations 008 and 009 maintained the lowest RMR values of 43 and 44 respectively.The advantage of rock mass classification using RMR is that it incorporates geological, geometric and engineering parameters in arriving at a quantitative value of the rock mass quality.

Slope stability analysis of the rock cuts
The kinematic analyses of the discontinuity sets revealed the possibility of both wedge and planar failures.At location 004, analysis shows intersection of J2 and J3 along 267˚/27˚ line orientation.Four intersections within the critical zone were recorded at location 005; J2 and J3 intersected along 280˚/34˚, J3 and J4 intersected along 265˚/30˚, J2 and J4 intersected along 266˚/46˚, and J3 and J5 intersected along 136˚/54˚ line orientations.Location 006 recorded three intersections; J1 and J4 intersected along 49˚/29˚, J1 and J3 intersected along 128˚/37˚, while J1 and J2 intersected along 155˚/27˚.At location 007, only one intersection between J1 and J3 along 53˚/31˚ was recorded.The same single intersection was also recorded at location 008 between J1 and J3 along 58˚/32˚.J1 and J3 intersected along 242˚/54˚ at location 009.There was also only one intersection between J2 and J4 at location 010 along 48˚/65˚.Only J3 and J4 intersected along 262˚/56˚ at location 011, while J1 and J3 intersected along 222˚/61˚ at location 014.Analysis on the graphitic schist at location 015 yielded only one intersection between J3 and J4 having orientation of 326˚/28˚ line of intersection.As shown in Fig. 5(a), there are possibilities of only wedge failures at locations 004, 010, 011, 014 and 015, while Fig. 5(b) shows possibilities of wedge and planar failures at locations 005, 006, 007, 008 and 009.Based on the kinematic analysis of these discontinuity sets, it is anticipated that all cut slopes in this investigated area are unstable.

Classification of the rock cut slopes
It may be useful to visualize slopes as existing in one of the following three stages: i. Stable -The margin of stability is sufficiently high to withstand all destabilizing forces.
ii. Marginally stable -Likely to fail at some time in response to destabilizing forces reaching a certain level of activity.
ii. Actively unstable -Slopes where destabilizing forces produce continuous or intermittent movements.
Slope Mass Rating (SMR) proposed by Romana (1985) is a method to access the stability of both natural and cut slopes.According to Romana (1993) and Romana et al. (2003), SMR is obtained from RMR of Bieniawski (1989) as shown in equation 2 by adding a factorial adjustment factor which depends on the relative orientation of joint and slope and by adding another factor depending on the method of slope excavation.Rating adjustment for discontinuity orientations in RMR is not considered and this basic RMR is designated RMRb.SMR = RMRb + (F1.F2.F3) + F4 (2) Whereby F1 ranges from 1.0 to 1.5 and depends on the parallelism between discontinuity and it fits into the relationship: and A = the angle between the dip directions of the slope and joint.
For toppling failure, this parameter maintains 1.0 value and F2 thereby fits into the relationship: F2 = (tan βj)² (4) F3 depends on the relationship between dips of slope (βs) and joint (βj).F4 is a correction factor that depends on the excavation method used and it is fixed empirically.
In Table 8, P represents plane failure, αs represents slope dip direction, while αj represents joint dip direction and T represents toppling failure.As proposed by Romana (1993), the adjustment ratings of these factors are presented in Table 8 while the classes of SMR are presented in Table 9.The SMR was carried out using the independent joint sets from each investigated cut slope and finally applying equation 2. By analyzing the joint sets as shown in Table 10, SMR classified the stability of slopes as completely stable, stable, partially stable, unstable and completely stable.Using these five categories of SMR stability by Romana (1993), this study categorized the slopes as existing in three stages: Stable, marginally stable and actively unstable as presented in Table 10.
Table 10.Slope Mass Rating of the investigated cut slopes.

Conclusion
The field and laboratory tests, computation of results and schematic representations were aimed at classifying the stability of the investigated cut slopes.From the uniaxial compressive strength tests, it was observed that the combination of joints and foliations induced structurally controlled failures in the tested samples.RQD and RMR schemes utilized for rock mass classification yielded almost similar poor to good quality ranges for each investigated rock mass.While the RQD values range from 34.5% to 90% covering a broad range of poor, fair and good quality rocks, RMR yielded values from 43 to 62 which cover only fair and good quality rocks.RMR applied in this study is very useful as a tool for the preliminary assessment of slope stability whereby it classified seven of the slopes cut into fair rocks and only three slopes cut into good rocks.Although based on the kinematic analyses carried out in this study, it is anticipated that the entire cut slopes are unstable with possibilities of wedge and planar failures, SMR gave comparative stability classes among these cut slopes, categorizing them actively unstable, marginally stable and stable slopes.Considering the determined unstable slopes, there is need for the use of expensive and wider in-situ and laboratory testing equipment in order to reasonably predict the stability of these cut slopes before any further infrastructural development in the area is carried out.
might probably be of Upper Palaeozoic.In order of abundance the schist consists of quartz mica schist, graphitic schist and quartzite layers.Petrographic analyses by Mohd Asbi & Associates (2005), Andrew Malone Ltd (2007), Nkpadobi (2014) and Nkpadobi et al. (2015) classified the schist as quartz-mica schist, graphitic schist, phyllite and quartzite, and photomicrographs confirm that most of the quartz mica schist are mylonitized.The major identified failure types in the area include planar sliding, wedge failure, and complex failures characterized by slump and earth flow which occurred along the Gunung Pass axis.The geometrical properties of failed slopes are shown in Table1.However, the plausible triggering factors for these different types of failures varied.Figs.2(a) -(c) represent these planar, wedge and complex failures in the area respectively.

Figure 1 .
Figure 1.Geological map of the study area.

Figure 3 .
Figure 3. Field photograph and schematic diagram of weathering profile over quartz mica schist at location 010.

Figure 4 .
Figure 4. (a) Some of the rock core samples.(b) Set up of uniaxial compressive strength test.

Figure 5b .
Figure 5b.Stereographic projections showing possibilities of wedge and planar failures.

Table 1 .
Geometrical properties of failed slopes.

Table 2 .
Summary of spacing and frequency of individual joint sets.

Table 3 .
General characteristics of the rock mass jointing.

Table 6 .
Determination of condition of discontinuities.

Table 7 .
Rock Mass Rating (RMR) of investigated rock masses.