Page 1

Prediction of Side Resistance in Poor Quality Rock: RQD vs. GSI

Scott M. Mackiewicz, PhD, PE1 and Arlan Rippe, PE2

1 Sr. Principal Geotechnical Engineer, Kleinfelder, 7802 Barton, Lenexa, Kansas, 66214, USA, (913)962-0909, smackiewicz@kleinfelder.com 2 Senior Principal Consultant, Kleinfelder, 15050 SW Koll Pkwy # L, Beaverton, OR 97006, USA, (503)644-9447, arippe@kleinfelder.com ABSTRACT

There are many factors that influence the side resistance prediction of drilled shafts in poor quality rock. As part of a bridge replacement project in Oregon, we evaluated the performance of various design methodologies for predicting the side resistance in poor quality rock. In particular, we looked at the side resistance predicted in poor quality rock from the rock mass strength developed using Rock Quality Designation (RQD) as well as the Geological Strength Index (GSI) developed by Hoek and Brown.

We predicted the capacities of drilled shafts at several bents along the bridge

alignment and compared the estimated bedrock socket required to resist the structural loading. This paper summarizes the determination of rock mass quality using the RQD and GSI methods, describes the prediction methods used for determining the side resistance in poor quality rock, presents the comparison of RQD versus GSI predicted design values, and evaluates the performance of the predictive methods and discusses the difference in cost associated with the drilled shaft lengths predicted by each method. Based on the results of this study, using the RQD methods without any specific laboratory strength or deformation test results may under predict the ultimate shaft resistance within poor quality rock, i.e., RQD < 60 percent. INTRODUCTION The assessment of overall rock characteristics is key to designing deep foundations in poor quality rock. Several methods of rock characterization have been established including RQD (Deere, 1964), Rock Mass Rating (RMR) using the Geomechnics System (Bieniawski, 1989) and Geologic Strength Index (GSI), (Hoek and Brown 1997, 2002). For this paper, we have evaluated the performance of predicting shaft resistance in rock using current American Association of State Highway and Transportation Officials (AASHTO) methods with only RQD data and compared the result with the GSI method from laboratory data. During our evaluation, we found that the rock shaft resistance predicted using the RQD method was lower than that

Page 2

predicted by the GSI method for poor to fair quality rock with RQD’s less than 60 percent. This paper describes the methods used to predict the shaft resistance, evaluates the performance of the predictive methods and presents a cost analysis of drilled shafts designed with varying shaft resistance values. PROJECT SITE

The project site is located in Lane County, Oregon and generally extends along I-5 from the Canoe Canal bridges on the north to just south of Franklin Boulevard (see Figure 1). I-5 generally trends north-south through most of the project area; near the southern end of the project alignment, I-5 curves southeasterly. The Willamette River flows westerly beneath I-5 and generally separates the project area into a northern and southern portion.

Figure 1. Project Location

GEOLOGIC FORMATION

Lithologic units underlying the project area include (in order of increasing depth and age) fill, alluvial deposits, and Eocene-Oligocene bedrock. Bedrock beneath the project area consists primarily of marine siltstone and sandstone of the Eugene Formation, with local occurrences of basalt and lapilli tuff of the Fisher Formation as mapped by Madin and Murray (2006) on the southern end of the project site. Generalized subsurface conditions are illustrated on Geologic Cross Section in Figure 2.

Figure 2. Geologic Cross Section

Project Site

Page 3

ROCK MASS CHARACTERIZATION Rock Quality Designation (RQD): One indicator of rock quality and the degree of fracturing is a simple characterization known as the Rock Quality Designation (RQD). The RQD is measured when coring rock, and is defined as the sum of the lengths of rock core longer than 4 inches, divided by the total length of the core drilled (in percent). Breaks obviously caused by the drilling operation are ignored. Table 1 below shows the correlation between RQD and apparent rock quality that is typically used (Deere 1964).

Geological Strength Index (GSI): The Geological Strength Index (GSI) developed by Hoek in 1997 and updated in 2000 (Marinos and Hoek, 2000) provides a better estimation of the rock mass characteristics than the RQD system, see Table 2. The GSI is based on a comprehensive engineering geology description of the rock mass using the lithology, structure, condition of disconti-nuity surfaces observed in the bedrock cores.

ROCK ENGINEERING PARAMETERS Rock engineering parameters were developed for the materials encountered in our borings from geologic logging, percent core recovery, RQD, GSI and laboratory rock strength testing. Rock cores in the bedrock formations were obtained using HQ (60mm) wireline rock coring techniques. The strength of intact rock specimens were evaluated in the laboratory unconfined compression (ASTM D7012) and point load testing (ASTM D5731). On 27 specimens, the elastic modulus of the intact rock core, Ei, and the unconfined compressive strength of the intact rock, qui, was established by performing unconfined compression strength testing in accordance with ASTM D7012. On an additional, 14 samples, point load testing was performed and the point load index was

Table 2. Geological Strength Index for Jointed Rock (Marinos and Hoek, 2000)

RQD (%) Rock Quality90 to 100 Excellent75 to 90 Good50 to 75 Fair25 to 50 Poor0 to 25 Very Poor

Table 1. Correlation between RQD and Rock Quality (after Deere, 1964)

Page 4

corrected to the standard 50mm size and correlated to the unconfined compressive strength using a “C” value of 24.5 based on the average values from the International Society of Rock Mechanics, ISRM, 1985, Suggested Method for Determining Point Load Strength. A summary of the established intact rock parameters obtained from laboratory tests is shown in Table 3.

Table 3. Summary of Laboratory Test Results

RQD Method: Using the guidelines set out in AASHTO 2008 the elastic modulus of the rock mass was estimated using Table 4. This table was developed by O’Neill and Reese (1999) for establishing the shaft resistance in rock and was based on data presented by Carter and Kulhawy, 1988. For the purpose of this study, the rock samples were estimated to have closed joints.

Ei-ksi qui-ksi

Rock Unit No. Mean Std. Dev No. Mean Std. DevSandstone 4 1230 258 4 5.49 1.35Siltsone 16 1430 1450 34 4.60 3.76Tuff 7 1370 445 13 4.39 1.30

RQD Em/Ei(percent) Closed Joints Open Joints

100 1.00 0.6070 0.70 0.1050 0.15 0.1020 0.05 0.05

Table 4. Estimation of Em based on RQD (O’Neill and Reese, 1999) AASHTO 2008, Table 10.4.6.5-1

22

107

9

30

2

41

8

00

5

10

15

20

25

30

35

40

45

80-100 60-80 40-60 20-40 0-20RQD or GSI Value

Num

ber

of B

edro

ck S

ampl

es

RQDGSI

Figure 3. RQD/GSI Value Distribution

Page 5

GSI Method: Using the Hoek-Brown criterion originally presented in Hoek and Brown, 1997 and updated in 2002 (Hoek, et. al., 2002) with the GSI, a practical estimate of the rock mass deformation modulus can be established using Equation 1.

��

���

� −−

− = 4010

105.14

145GSI

ksiuiksim

qE EQ 1.

Where: Em-ksi=Modulus of the Rock Mass (ksi) qui-ksi=Unconfined Compression Strength of Intact Rock Core (ksi) GSI=Geological Strength Index (see Table 2) SHAFT RESISTANCE PREDICTION METHODS

The average ultimate shaft resistance, fsu, was estimated using correlations to the strength of intact rock material, which is most often measured taken as the unconfined compressive strength, qui. The correlations, which are numerous, were developed using full scale load test results and relating the fsu to the qui. For the purpose of this study, we used the method developed by Horvath and Kenney, 1979, as presented in AASHTO 2008, 10.8.3.5.4b for determining the side resistance of drilled shafts socketed into rock (see below). This method includes a reduction factor, αe, to account for discontinuities in the rock mass as estimated by O’Neill and Reese, 1999, which is shown in Table 5. This study was based on the assumption of smooth shafts and no artificial rouging of the rock socket side walls was include in the prediction methods.

EQ 2

Where: fsu = average ultimate shaft resistance qui = unconfined compressive strength of rock pa = atmospheric pressure αe = reduction factor f’c = concrete compressive strength RQD Method: The RQD method used the laboratory unconfined compressive strength, qui, results and the reduction factor, αe, based on the RQD values established from our logging of the rock core and the ratio of the modulus of the rock mass to modulus of the intact rock ratio (Em/Ei) outlined in Table 4. GSI Method: The GSI method used the laboratory unconfined compressive strength, qui, results and the reduction factor, αe, based on the GSI values established from our logging of the rock core. The ratio of the modulus of the rock mass to modulus of the intact rock ratio (Em/Ei) was established using the rock mass modulus established using the GSI values and Equation 1 divided by the intact rock modulus established

5.05.0'

8.765.0 ���

����

�<��

�

����

�=

a

ca

a

uiaesu p

fp

pq

pf α

Em/Ei ααααεεεε

1.0 1.00.5 0.80.3 0.70.1 0.55

0.05 0.45

Table 5. Estimation of ααααe (O’Neill and Reese, 1999) AASHTO 2008 Table 10.8.3.5.4b-1

Page 6

from our laboratory strength testing. For cases where only a point load test was performed, the rock modulus ratio (Em/Ei) was estimated using the GSI to established the Em and the Ei was assumed to be the average intact modulus from our laboratory strength testing. Figure 4 shows the computed modulus ratio and trend lines with respect to RQD for this study as well as other known references including O’Neill and Reese, 1999 and Zhang & Einstein, 2000. The reduction factor, αe, was established using the computed (Em/Ei) and Table 5.

FINDINGS

The ultimate shaft resistances computed using the RQD and GSI methods, described above were compared to establish the efficiency of the predictive methods. This evaluation revealed that the two methods appear to vary significantly for poor to fair quality rock with RQD values of less than 60 percent versus fair to good quality rock with RQD values > 60 percent. The predictive difference is evident in Figures 5 and Figure 6, which plots the computed shaft resistance results of samples with RQD’s less than 60 and RQD’s greater than 60.

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

0 20 40 60 80 100

RQD

Em

/Ei

O'Neil&Reese1999 Em/Ei (closed)O'Neil&Reese1999 Em/Ei (open)RQD > 60RQD < 60Zhang & EinsteinExpon. (RQD < 60)

Figure 4. Modulus Ratio (Em/Ei) vs. RQD

Page 7

y = 1.4101xR2 = 0.8519

0

5

10

15

20

25

30

35

0 5 10 15 20 25 30 35

fs-(ksf) using RQD

fsu-

(ksf

) usin

g G

SI

y = 0.9826xR2 = 0.9404

0

5

10

15

20

25

30

35

0 5 10 15 20 25 30 35

fs-(ksf) using RQD

fs-(k

sf) u

sing

GSI

Figure 5. Ultimate Shaft Resistance: RQD < 60 %

Figure 6. Ultimate Shaft Resistance: RQD > 60 %

Page 8

As indicated in Figure 5, the shaft resistance predicted using the GSI method is on

the order of 40 percent greater than computed using the RQD method. Figure 6 indicates that for samples with RQD’s greater than 60 percent the predictive methods are about equal. COST ANALYSIS

As discussed earlier, the study was performed in conjunction with a bridge replacement project and the intermediate piers were to be supported on groups of drilled shafts. We have provided an example of the cost difference of installation for a drilled shafts installed in poor quality rock, with an RQD<60 percent. For this example, we have assumed that the average ultimate shaft resistance of the rock would be 10 ksf for the RQD method and 14 ksf for the GSI method.

Based on an anticipated shaft loading of 4,500 kips-Strength I and using a resistance factor of 0.55 (AASHTO 2008), we estimated that an 8-foot diameter shaft extending to depths of 32 and 23.5 feet would be required using the RQD and GSI method, respectively. Assuming an installation cost of about $1,220 per lineal foot of shaft, a cost savings due to a reduction in shaft length of about $10,000/shaft could be obtained using the GSI method for the shafts installed in poor quality rock. We would anticipate that the shaft lengths could be further reduced if full scale load testing would be performed at the project site, thereby allowing the use of a higher resistance factor, and based upon the results of the load test. CONCLUSIONS

Based on the results of this study, using the RQD methods without any specific laboratory strength or deformation test results may under predict the ultimate shaft resistance within poor quality rock, i.e., RQD < 60 percent. Our study also indicates that by using the GSI method and laboratory strength/modulus data, the ultimate shaft resistance predicted for poor quality rock may be up to 40 percent greater than that predicted using only RQD’s.

This study indicates that the shaft resistance in poor quality rock can be very

difficult to estimate with commonly used methods and guidelines. Thus, by performing additional laboratory testing and more detailed classification of the rock mass (i.e., GSI method), the designer will have better information to predict shaft resistance estimates and, in turn reduce project costs. The economic example contained herein also shows that a considerable cost savings could be obtained on projects with a large number of shafts. We anticipate that the cost of the additional testing and classification could be well offset by the foundation installation savings on large projects. As indicated by Although GSI is not currently widely used at this time, Turner, 2006, suggests that it will likely become a standard in rock mass characterization for rock socket design.

Page 9

REFERENCES AASHTO LRFD Bridge Design Specifications (2008). Section 10, 2008 Interim,

American Association of State Highway and Transportation Officials, Washington, D.C.

ASTM D 7012 – 07. (2007) Standard Test Method for Compressive Strength and Elastic Moduli of Intact Rock Core Specimens under Varying States of Stress and Temperatures, ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959, United States.

ASTM D 5731 – 07. (2007). Standard Test Method for Determination of the Point Load Strength Index of Rock and Application to Rock Strength Classifications, ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959, United States.

Bieniawski, Z.T., (1989) Engineering Rock Mass Classifications, Wiley, New York. Carter, J. P. and F. H. Kulhawy, “Analysis and Design of Drilled Shaft Foundations

Socketed into Rock, “Report ELL-5918, Electric Power Research Institute, Pal Alto, California, 1988, 188pp.

Deere, D. U., (1964) “Technical Description of Rock Cores for Engineering Purposes,” Rock Mechanics and Engineering Geology, Vol.1, No. 1, pp. 17-22.

Hoek, E. And E.T. Brown, (1997). “Practical Estimates of Rock Mass Strength,” International Journal of Rock Mechanics and Mineral Science, Vol. 34, No. 8, pp. 1165-1180.

Hoek, E., C. Carranza-Torres, and B. Corkum, (2002). “Hoek-Brown Failure Criterion-2002 Edition,” Proceedings, North American Rock Mechanics Society Meeting, July 8-10, Toronto, ON, Canada.

Horvath, R.G. and T. C. Kenney,(1979). “Shaft Resistance in Rock Socketed Drilled Piers,” Proceedings, Symposium on Deep Foundations, ASCE, New York, N.Y., pp. 182-214.

Madin, I.P. and Murray, R.B. (2006). Preliminary Geologic Map of the Eugene East and Eugene West 7.5 Quadrangle, Lane County, Oregon: State of Oregon, Department of Geology and Mineral Industries, Open File Report O-06-17.

Marinos, P. and E. Hoek (2000). “GSI:A Geologically Friendly Tool for Rock Mass Strength Estimation,” Proceedings Geo-Engineering 2000, International Conference on Geotechnical and Geological Engineering, Melbourne, Australia, 2000, pp.1422-1440.

O’Neill, M.W. and L.C. Reese (1999). Drilled Shafts: Construction Procedures and Design Methods, Report FHWA-IF-99-025, Federal Highway Administration, Washington, D.C., 758 pp.

Turner, J. (2006). NCHRP Synthesis 360: Rock Socketed Shafts for Highway Structure Foundations, Transportation Research Board, Washington, D.C.

Zhang, L. and H. H. Einstein (2000). “Estimating the deformation modulus of rock masses.” Proceedings, 4th North American Rock Mechanics Symposium, Pacific Rocks 2000: Rock around the Rim, Seattle, Washington.