International Journal of Rock Mechanics & Mining Sciences 38 (2001) 1113–1119
Draft ISRM suggested method for determining block punch strength index (BPI) R. Ulusay*, C. Gokceoglu, S. Sulukcu Department of Geological Engineering, Faculty of Engineering, Applied Geology Division, Hacettepe University, 06532 Beytepe, Ankara, Turkey Accepted 26 November 2001
1. Introductio Introduction n 1.1. Rock strength, particularly the uniaxial compressive strength (UCS) is an important parameter in rock mass mass clas classi sific ficat atio ion n meth method odss and and in vari variou ouss rock rock engineering engineering design approaches. approaches. Measurement Measurement of rock strength requires testing which must be undertaken on test specimens of particular sizes in order to fulfill testing stand standar ards ds.. Howe However ver,, ther theree are are some some shor shortc tcom omin ings gs associ ass ociate ated d with with these these conven conventio tional nal tests. tests. When When rock rock core coress are are only only divi divide ded d into into smal smalll disc discs, s, due due to the the presence of thin bedding or schistosity planes, the core length length may be too short to allow allow preparati preparation on of the specimens long enough even for the point load strength index test. 1.2. To overcome the above-mentioned difficulty, the possibility of using relatively short samples for a rock strength or index test has always been attractive. The block punch strength index (BPI) test apparatus, which was similar to that used for the measurement of direct shear strength of a thin plate of rock [1–2], has been developed in Delft University, The Netherlands, as an inde index x test test in dire direct ctly ly asse assessi ssing ng UCS UCS by Schr Schrie ierr [3]. [3]. However, in the previous studies, rock-disc specimens of about about 40 mm in diamet diameter er and 10 mm in thickness thickness were were tested, and the size effect of the test specimens and the use of the BPI test in rock engineering have not been considered. The studies by Ulusay and Gokceoglu [4–6] indicated that size correction was indispensable in the BPI test and the use of a generalized size correction factor established from the experimental data should be used. A considerably important correlation found between UCS and BPI indica indicates tes that that BPI tests tests lead lead to insign insignific ificant ant errors errors in determ determini ining ng UCS UCS when when compar compared ed to those those obtain obtained ed *Corresponding *Corresponding author. Tel.: +90-312-297-7767; +90-312-297-7767; fax: +90-312-299+90-312-2992034. E-mail address:
[email protected] (R. Ulusay). 1365-160 1365 -1609/01 9/01/$ /$ - see front front matter matter r 2002 Published by Elsevier Science Ltd. PII: S 1 3 6 5 - 1 6 0 9 ( 0 1 ) 0 0 0 7 8 - 8
from point load testing, particularly for laminated weak rocks [4–7]. It was also suggested that BPI be used as an alternative input parameter for intact rock strength in rock mass classification and as a measure of anisotropy using oriented disc samples [4–7]. 1.3. The BPI test described in this suggested method is applied applied to the rock-disc specimens, specimens, and involves the use of size correction, and determination of the strength in the strong strongest est direct direction ion where where only only core core sample sampless from from boreholes drilled at any angle to the weakness planes are available. 1.4. In this this sugges suggested ted method method,, the appara apparatus tus and operating operating procedure are described described together with data evaluation. There is an explanation for the presentation of the result results. s. The empiri empirical cal relati relations onship hipss to predic predictt some strength parameters from BPI are also presented in the last chapter.
2. Scope Scope 2.1. The block punch strength index test is intended as an inde index x test test for for the the stre streng ngth th clas classi sific ficat atio ion n of rock rock materials. It is also be used to predict other strength parame parameter terss with with which which it is correl correlate ated, d, for examp example le uniaxial uniaxial compressive compressive and tensile tensile strength. strength. 2.2. The test measures the size-corrected block punch strength strength index ðBPI s Þ of rock rock spec specim imen ens, s, and and thei theirr strength index in the strongest direction ðBPI s90 s90 Þ which is calc calcul ulat ated ed from from the the mult multip ipli lica cati tion on of a stre streng ngth th anisotropy anisotropy transformati transformation on factor with the BPI s value of the specimens obtained from cores inclined at any angle to the weakness planes. 2.3. Rock specimens in the form of thin cylindrical discs prepared from cores or blocks are placed into an apparatus which is designed to fit the point load device, and and are are brok broken en by the the appl pplicati cation on of loa load by a rectangular rigid punching block.
1114
R. Ulusay et al. / International Journal of Rock Mechanics & Mining Sciences 38 (2001) 1113–1119
2.4. The test can be performed with a portable apparatus and point load device, and so may be conducted in the laboratory. However, it can also be performed in the field if the facilities for cutting the specimen into small discs are available.
3. Apparatus 3.1. There are no published standards for construction of the apparatus for a block punch index test, and since this test apparatus is not commercially available, it has to be designed and fabricated in-house. The end result of the design and fabrication process is a unit consisting of two major parts: a lower platen (base support) and an upper platen (punching block) as can be seen in Fig. 1a. Both these platens should be machined from hardened tool steel with a Rockwell hardness of 40
Fig. 2. BPI test device fitted into a point load testing frame (PB: punching block; BS: base support; R: ram).
Fig. 1. (a) A general view from the BPI test apparatus consisting of base support, steel bars and punching block; (b) a plan view from the base support before clamping of the specimen; (c) a perspective view of the base support after the specimen is fixed; and (d) a schematic view from the punching canal of the base support.
in order to withstand the high stresses generated during the test. 3.2. The base support is fitted to the columns of the point load test frame through the holes (Fig. 1b) at its both ends and then it is attached to the ram of the frame by means of a block with a hole at its bottom (Fig. 1c) as shown in Fig. 2. Because the punching block is designed to thread into the base support to allow sandwiching of the rock-disc specimen, the base support should have a rectangular canal along the centre of its axis through which the punching block passes (Fig. 1b). The disc specimen placed on the base support (Fig. 1b) is clamped from its two ends by means of clamping bars which are screwed down as shown in Fig. 1c. The dimensions and tolerances of the base support are not given here specifically, because they depend on the type and size of the point load-testing device, particularly diameter of its reaction rods (columns). However, the width of the base canal can be taken as 19.75 mm (Fig. 1c). Some of the information is in Refs. [4–7] and further information can be obtained from Professor Ulusay and Asistant Professor Gokceoglu. 3.3. The second part of the device forms the rectangular rigid punching block, which transfers the load onto the specimen. It is designed to thread into the canal along the axis of the base support. Therefore, it should easily pass between the walls of the canal with a clearance of approximately 0.25 mm. Several views from the punching block and its dimensions are shown
R. Ulusay et al. / International Journal of Rock Mechanics & Mining Sciences 38 (2001) 1113–1119
1115
Fig. 3. Dimensions of the punching block of the BPI apparatus.
in Fig. 3. This part of the apparatus is attached to the upper rigid block of the point load-testing device by means of a long screw as can be seen in Figs. 1a and 2. 3.4. The load is provided by a conventional portable point load-testing device comprising a hydraulic ram and a manual hydraulic pump equipped with a pressure gauge. Spherically truncated conical platens of the point load-testing device are removed during the BPI test. To apply a load approximately at a given rate, the hydraulic pump is manually operated while simultaneously both the pressure gauge and a stopwatch are monitored. 3.5. An instrument, such as caliper is required to measure diameter D and thickness t of the specimens. When the test is carried out on cores from boreholes at any angle to the weakness planes, a device such as a goniometer should be used to measure the inclination of the weakness planes.
4. Procedure
4.1. Specimen preparation 4.1. Test specimens should be right cylindrical thin discs. For the purpose, the cores are cut into discs of various raw thicknesses ranging between 5 and 15 mm using a diamond saw perpendicularly to the core axis. The diameter of the disc specimens should preferably be not less than BX core size approximately 42 mm. 4.2. Although nearly all of the specimens are prepared without special treatment, care should be taken to ensure that the disc faces are as parallel as possible and the sides of the specimens are smooth and free of abrupt irregularities. However, if it is required, a surfacegrinding machine can be used to smooth the end faces of the discs.
4.3. The use of capping material or end surface treatments between the upper surface of the specimen and the punching block is not permitted. 4.4. The diameter D of the test specimen should be measured to the nearest 0.1 mm by averaging two diameters measured at right angles to each other at about the mid-height of the specimen. The thickness of the specimen should also be determined to the nearest 0.1 mm by averaging two thicknesses measured at right angles to each other. The average values of diameter and thickness are later used in any subsequent calculations. 4.5. For routine testing and classification, specimens should be tested either at their natural water content or air dried. Samples should be stored, for no longer than 30 days, in such a way as to preserve their natural water content, as far as possible, and tested in that condition. This moisture content should be reported in accordance with ‘‘Suggested method for determination of the water content of a rock sample’’, Method 1, ISRM Committee on Laboratory Tests [8]. 4.6. If the BPI test has to be carried out to measure the strength anisotropy (i.e. to estimate the strength in the strongest direction from the specimens obtained from the cores inclined at any angle to the weakness planes), the inclination of the weakness plane a should be measured by a goniometer to the nearest 1 . 4.7. The number of specimens tested under a specified set of conditions shall be governed by practical considerations, but at least five are preferred. 1
4.2. Testing 4.8. The base support of the BPI apparatus is mounted onto the ram of the point load device of which conical platens have been removed. The punching block is fixed to the upper block of the device by means of a
1116
R. Ulusay et al. / International Journal of Rock Mechanics & Mining Sciences 38 (2001) 1113–1119
Fig. 4. A view from the block punch index testing in point load test device.
screw. The specimen is then centered on the base support of the test apparatus (see Fig. 1d) and clamped to be sure that it does not move and is tightly fixed (see Fig. 1b). By using the hand pump, the base support is risen up until the punching block is nearly touching the specimen. 4.9. The load is then gradually applied to the specimen at a constant rate such that failure occurs within 10–60 s as suggested by ISRM [8] for point load strength (Fig 4). Fracturing is thus forced to take place along two parallel planes on which the normal stress is considered to be zero while the tensile stresses caused by bending are reduced. The load F t;D which is the load required for the failure of a specimen of any diameter and any thickness, is recorded. After failure, theoretically, the specimen is broken into three parts, the two ends which are fixed in the apparatus and the middle part of the specimen which is punched out (Fig. 5). The test should be rejected as invalid if the parallel fracture planes are either absent or not fully developed (irregular failure) or cross joints develop as shown in Fig. 6. 4.10. The procedure (4.8) through (4.9) given above is repeated for the remaining tests in the sample.
Fig. 5. Schematic illustrations of the BPI test specimen before and after failure.
Fig. 6. Views from the specimens after BPI test, and the failure patterns for valid and invalid tests.
where F t;D is the failure load recorded from the gauge in kN (and converted to MN by the multiplication of 103), and A is the area (in m2) through which the shearing takes place. The formula quoted below is for the area
5. Calculations
A ¼ 4tðr2 95:1Þ0:5 106 ðm2 Þ;
5.1. Uncorrected block punch strength index
where t and r are the thickness and radius of a disc specimen (in mm), respectively (Fig. 7).
5.1. The uncorrected block punch strength index BPI (in MPa) is calculated from the following equation:
5.2. Size correction
BPI ¼
103 F t;D ; A
ð1Þ
ð2Þ
5.2. BPI varies as a function of D and t [4–7], so that a size correction must be applied to obtain a unique block
R. Ulusay et al. / International Journal of Rock Mechanics & Mining Sciences 38 (2001) 1113–1119
1117
Fig. 7. Calculation steps of the area ðAÞ of the failure surface in the BPI test.
punch strength index value for the rock sample and one that can be used for purposes of rock strength classification. 5.3. The size-corrected block punch strength index (BPI 10;50 or BPI s ) of a rock specimen is defined as the value of BPI that would have been calculated from a failure load converted to a corrected load for a nominal 50 mm diameter and 10 mm thickness by multiplying BPI with the constants K t and K D ; representing correction factors for thickness and diameter, respectively. Because the load at failure is converted to a corrected BPI value for a equivalent size (D ¼ 50 mm, t ¼ 10mm), the area of the surface through which shearing takes place used in calculation of the corrected BPI should be expressed in terms of equivalent specimen dimensions. The equivalent area ðA10;50Þ is 6 2 921 10 m . When testing single-sized disc specimen with a diameter and thickness other than 50 mm and 10 mm, respectively, the size correction is accomplished using the formula BPI 10;50 ðBPI s Þ ¼
F 10;50 F t;D 103 K t K D ¼ ðMPaÞ: A10;50 921 106 ð3Þ
The correction factors K t and K D can be obtained from the charts in Fig. 8 or from the expressions: 1:1265
K t ¼ 13:74
ðthickness correction factorÞ;
K D ¼ 234:53D1:3926 ðdiameter correction factorÞ:
ð4aÞ
Fig. 8. Charts for the size correction factors to be used in the calculation of the corrected BPI [7].
the corrected BPI value without considering the failure area BPI c ¼ 3499D1:3926 t1:1265 F t;D ;
ð5Þ
where D and t are in mm and F t;D is in kN. 5.3. Strength index in the strongest direction 5.4. In the case of a testing, which is carried out on specimens prepared from cores from boreholes inclined at any angle to the weakness planes, if determination of the strength index in the strongest direction (i.e. loading perpendicular to the weakness plane) is considered, an additional conversion on BPI s should be done. For the purpose, a strength anisotropy transformation factor of K was suggested by Ulusay and Gokceoglu [4–6] a
K ¼ a
BPI s90 ; BPI s
ð6Þ
a
ð4bÞ
Alternatively, the following equation derived from the combination of Eqs. (3), (4a) and (4b) is used to obtain
where BPI s90 is the BPI s of the specimens obtained from boreholes perpendicular to the weakness planes (strongest direction), and BPI s is the BPI s of the specimens a
1118
R. Ulusay et al. / International Journal of Rock Mechanics & Mining Sciences 38 (2001) 1113–1119
from boreholes inclined at any angle to the weakness planes. The relationship between the values of K and the angle a (the angle in degrees between the core axis and the weakness plane) (Fig. 9) is given by the following expression:
6. Presentation of results
K ¼ 4:24e0:0156 :
(a) Lithologic description of the rock. (b) Orientation of the axis of loading with respect to specimen anisotropy, e.g. bedding planes, foliation, etc. (angle a). (c) The sample number, source location and sampling depth. (d) Number of specimens tested. (e) Water content at time of test (air dried, oven dried or value of water content in per cent). (f) Date of testing. (g) Failure pattern. (h) A tabulation of the values of diameter and thickness of the specimens, failure load and corrected block punch strength index, and strength index in the strongest direction if the angle between the direction of loading and weakness planes is o90 . All BPI c values should be expressed to three significant figures.
a
a
ð7Þ
a
Corrected BPI value in the strongest direction is obtained from the expression, which is the combination of Eqs. (6) and (7) BPI s90 ¼ 4:24 e0:0156 BPI s : a
ð8Þ
a
6.1. Results for BPI test should be tabulated (see typical results shown in Fig. 10). The report should contain at least the following information for each specimen tested:
1
7. Notes
Fig. 9. Strength anisotropy transformation factor ðK Þ as a function of the angle ð aÞ between the weakness plane and loading direction in the BPI test [4]. a
7.1. When first introduced, the block punch strength index test, without application of any size correction, was used to predict uniaxial compressive strength [3]. Then, it was experimentally shown [4–7] that the BPI test could be more preferable in the estimation of the uniaxial compressive strength (UCS), because the BPI
Fig. 10. Typical results for the BPI test.
R. Ulusay et al. / International Journal of Rock Mechanics & Mining Sciences 38 (2001) 1113–1119
1119
Table 1 Classification of block punch strength index [7] BPIs (MPa)
Strength class
1 1–5 5–10 10–20 20–50 >50
Very weak Weak Moderate Medium High Very high
o
Acknowledgements
Fig. 11. Rating chart of the block punch strength index and uniaxial compressive strength for RMR and M-RMR rock mass classification systems [7].
tests lead to insignificant errors in determining UCS when compared with those obtained from point load testing which yields a multiplying factor of k to predict UCS ranging between 15 and 50 depending on rock type. The following relation between the UCS and the corrected BPI was obtained by regression analysis with a statistically significant correlation of 0.90 [7]: UCS ¼ 5:1BPI s :
ð9Þ
7.2. Assuming a mean of UCS =BPI s ; the ratio of 5.1 leads to errors of maximum 20 per cent in estimations of the UCS from BPI s [7]. This may be sufficiently accurate for using BPI as an index for intact rock strength in rock mass classification. Therefore, the BPI can be introduced into rock mass classification systems as an alternative strength index input parameter, especially for weak rocks where obtaining a standard specimen is rather difficult. If the ranges of UCS used in Bieniawski’s Geomechanical Classification (RMR) System [9], and M-RMR System [10–11], which is a modification of RMR, are divided by the strength conversion factor of 5.1, and decimals are avoided, perhaps a more realistic scale for BPI s can be obtained. For this purpose, a combined chart (Fig. 11) which considers both rock mass classification systems showing the variation of the rating both for block punch strength index and UCS, and the BPI classification (Table 1) can be used. 7.3. BPI is approximately 0.68 times of the indirect tensile or Brazilian tensile strength [7].
The authors wish to acknowledge the encouragement and support given by Professor J.A. Hudson of Imperial College of Science, Technology and Medicine in the UK. The co-ordinators are also most grateful to Professor K. Sugawara of Department of Civil Engineering, Kumamoto University in Japan for his kind interest at the beginning of the studies on the suggested method.
References [1] Mazanti BB, Sowers GF. Laboratory testing of rock strength. In: Proceedings of the International Symposium on Testing Techniques for Rock Mechanics, Seattle, Washington, 1965. p. 207–27. [2] Stacey TR. A simple device for the direct shear strength testing of intact rock. J SA Inst Min Metall 1980;80(3):129–30. [3] Schrier van der JS. The block punch index test. Bull Int Assoc Eng Geol 1988;38:121–6. [4] Ulusay R, Gokceoglu C. The modified block punch index test. Can Geotech J 1997;34:991–1001. [5] Ulusay R, Gokceoglu C. An experimental study on the size effect in block punch index test. Int J Rock Mech Min Sci 1998;35(4– 5):628–9 (In: NARMS’98 ISRM International Symposium, Cancun, Mexico, Paper No. 008). [6] Ulusay R, Gokceoglu C. A new test procedure for the determination of the Block Punch Index and its possible uses in rock engineering. ISRM News J 1999;6(1):50–4. [7] Sulukcu S, Ulusay R. Evaluation of the block punch index test with prime consideration on size effect, failure mechanism and its effectiveness in predicting rock strength. Int J Rock Mech Min Sci 2001;38(8):1091–1111. [8] I.S.R.M. Rock characterization. In: Brown ET, editor. Testing and monitoringFISRM suggested methods. Oxford, UK: Pergamon Press, 1981; 211p. [9] Bieniawski ZT. Engineering rock mass classification. New York: McGraw Hill, 1989. 237p. [10] Unal E, Ozkan I. Determination of classification parameters for clay bearing and stratified rock masses. In: Peng S, editor. Proceedings of the Ninth International Conference on Ground Control in Mining, West Virginia University, 1990. p. 250–9. [11] Ulusay R, Unal E, Ozkan I. Characterization of weak, stratified and clay-bearing rock masses for engineering applications. In: Myer LR, Cook NGW, Goodman RE, Tsans CF, editors. Proceedings of the Conference on Fractured and Jointed Rock Masses, Lake Tahoe, California, 1995. p. 229–35.
