Durability Assessment of Weak Rock by Using Jar Slaking Test

Durability Assessment of Weak Rock by Using Jar Slaking Test Edy Tonnizam Mohamad Senior Lecturer, Faculty of Civil Engineering, Universiti Teknologi Malaysia edy@utm.my Rosli Saad School of Physics, Universiti Sains Malaysia, Penang rosli@usm.my Seyed Vahid Alavi Nezhad Khaili Abad Researcher, Department of Geotechnics and Transportation, Faculty of Civil Engineering, Universiti Teknologi Malaysia v_alavi_59@yahoo.com ABSTRACT Jar test is a simple test developed to determine the reaction of weak rock material to water during a certain period of time which can be tested on irregular bulk samples. It indicates the porosity, grain interactions and density of the material. The total index observed during a 10 minutes, 15 minutes, 30 minutes and 60 minutes sequence was found to be a good indicator for the classification. This should enable the classifying of the weak rock material to be carried out based on the slaking index. The reaction of the rock matter was inevitable after immersing the sample in the jar with water. The slaking index or changes in the sample also depend on the duration of the immersion time. By the observation of the changes caused by the immersing process the rock can be classified and the description for shale and sandstone. The jar slaking test was found to be more suitable to measure durability of weak rock especially in highly weathered (grade IV) and completely weathered (grade V) states as compared to slake durability index (Id 1 or Id 2 ) which was found to be unsuitable to test the durability of such weak rock. KEYWORDS: Jar Test, Porosity, Grain Interaction, Density, Slaking Index, Durability INTRODUCTION The assessment of reaction of rock material to water is very essential in determining the effects of weathering on the rock material; thus are the estimates the failure mode of the rock material. The resistance of a rock to short-term weathering is described as durability of the material. Thus, durability is an important engineering parameter, particularly for weathered rocks. This non-durable behaviour of these rocks is responsible for loss of strength especially when influenced by water. Two approaches were used in this study: slake durability and the jar test to evaluate the suitability in assessing durability of weak rock. The slake durability test aims to assess the resistance of rock material to weathering and disintegration when subjected to two standard cycles of drying and wetting (Id 1 and Id 2 ). While, the jar test was used by immersing the rock samples in water for a certain period of time and the changes of appearance will be recorded. - 1319 -

Vol. 16 [2011], Bund. O 1320 Wood & Deo (1975) suggested the jar test should be completed at both 30 minutes and 24 hours by comparing the test results with the slaking index. Santi (1995) proposed to use a 30- minute period to classify the index of the material. However, these procedures are found to be less effective in very weak rock. Thus, a modification of this test was carried out by observing the samples at 4 time intervals: 10, 15, 30 and 60 minutes. Each class of rock material has a different reaction to water or moisture (Edy Tonnizam et al., 2011). This test enables us to observe and document the behaviour of these rock materials when it is immersed into water. As stated in previous assessments, the strength of rock material is also affected by presence of moisture in the rock material. The rock material has variable grade of porosity and this is different between sandstone and shale. Both of these types of rock material give certain effects when immerse in water and sandstone may decompose faster than shale as it has a larger amount of pores that can contain water. But still, shale may lose more strength than sandstone when immersed in water becausee of its soft and fine particles that can break down in water. Santi (1998a,b) classified the behavior of shale into six slaking indices. These indices were also used by Zainab Mohamed (2004). However, her assessments are mainly concentrated on the reactions and behaviour of materials to water after 30 minutes. However, in this study, the sandstone and shale are classified into six slaking indices and the samples are observed for 10, 15, 30 and 60 minutes. Index classification for shale and sandstone are shown in Figure 1 and 2 respectively. Figure 1: Index Classification and Slake Jar Test for Shale (Santi, 1998a,b)

Vol. 16 [2011], Bund. O 1321 Figure 2: Slaking Index for Sandstone

Vol. 16 [2011], Bund. O 1322 TESTING PROCEDURES i. In jar test, four samples are needed for each type of rock material. The weathering grade of rock material is classified based on Table 1. ii. iii. iv. Each sample is then trimmed to a suitable size (about 40 mm). Tap water was filled into 4 separate jars. The samples were then immersed into water for a period of 10, 15, 30 and 60 minutes (Figure 3). v. After the end of each period, the sample s behaviour was observed and recorded. The samples that have been tested are irregular bulk samples. A slaking index was given to each sample for the stated period by referring to the table of slaking indices. Through observation of the changes caused by the immersing process, the rock materials were then classified according to the slaking index for shale and sandstone. The slake durability test was originally developed by Franklin and Chandra (1972), recommended by the International Society for Rock Mechanics (ISRM, 1981) and standardized by the American Society for Testing and Materials (ASTM, 1990). Figure 3: Jar Slaking Test

Vol. 16 [2011], Bund. O 1323 Table 1: Weathering classification used to describe rock mass (modified from Ibrahim Komoo, 1995b) MATERIAL MASS DESCRIPTION ZONE Colour Texture Slaking Structure Ironrich In By Conditi Changes water hand on layer Strength (Schmidt hammer) Residual soil Completely weathered VI b V a Completely changed (homogeny) Completely changed (homogeny) Destroyed Half remains unchanged disintegrate disintegrate 100% destroyed <25% remains Completely changed None Normally exist None Highly weathered Moderately weathered b IV a III Completely discoloured Slightly discoloured unchanged Becomes flakes or small pieces Becomes flakes or small pieces Edges can be broken >50-75% remains Iron-rich filling in discontinuity May exist Less than 25 Slightly weathered II No changes Remains as Mass Edges unbroken 100% intact Discolorations along discontinuity none Exceeds 25 Unweathered I No changes COMPARISON OF SLAKE DURABILITY AND JAR TEST RESULTS The results of slake durability and jar slake test are presented in this section. Slake Durability Results Tables 2 and 3 show the results of Id 1 and Id 2 by weathering grades.

Vol. 16 [2011], Bund. O 1324 Table 2: Summary of slake durability Id 1 result for respective weathering grade Table 3: Summary of slake durability Id 2 result for respective weathering grade The mean values for the Id 1 and Id 2 are shown in Table 4 and 5 respectively.

Vol. 16 [2011], Bund. O 1325 Table 4: Mean value of Id 1 for respective weathering grade Table 5: Mean value of Id 2 for respective weathering grade Slake Durability Index Id 1 and Id 2 are determined and presented in the graphs as shown in Figures 4 and 5 respectively. The durability index Id 1 and Id 2 generally shows increase of deterioration percentage with increase of weathering grade. As for Id 1, some materials may be difficult to distinguish whether in grade II, III or IVa by just depending on values from 1 cycle. The values for grade II in the range of 94 percent 99 percent with a mean value of 97 percent while the value for grade III, are in range of 90 percent to 99 percent with a mean value of 94 percent. By using Id 1 as the parameter, the division of grades are clear for grade IVa and IVb. Values for Grade IVa are lower in ranges of 60 percent to 98 percent with a mean value of 80 percent. There seems to be a gap of values of more than 20 percent between grade IVa and IVb, where the values for IVb are in ranges 10 percent to 40 percent with a mean value of 31 percent. Id 1 values cannot be used to distinguish grade Va and Vb materials. Grade Va to Vb has 0 percent value which means that these samples are totally destroyed in the test. Figure 4 shows the result of Id 1 that is graphically illustrated. It shows a rapid decrease in values from grade IVa to IVb. This shows that shale materials in grade Va and Vb could not even retain their structure in the first cycle of slake durability test. As samples are further tested in the second cycles (Id 2 ), results show a clearer division for samples in grade II, III and IVa. However, samples in grade IVb, Va and Vb will further

Vol. 16 [2011], Bund. O 1326 destroyed in the second cycle. For both Id1 and Id2, the alluvium samples in Desa Tebrau for grade IVa show a difference in the result for its different grain size. The coarse material shows lower values as compared to the finer ones. This signifies that coarser material is destroyed faster than the finer grain size. Figure 4: Box-plot of Id 1 Versus Weathering Grade The Id 2 results are shown graphically in Figure 5. From the boxplot, we can see that grade IVb to Vb materials could not survive the second cycle of the test and has a zero value. The percentage value can only be obtained until grade IVa. Values for grade II are in the range of 90 percent to 95 percent with a mean value of 93 percent; grade III is in the range of 80 percent to 93 percent with a mean value of 86 percent and grade IVa is in the range of 30 percent to 90 percent with a mean value of 56 percent. It was found that slake durability test is primarily influenced by rock properties which allow ingress of water into the rock material. The presence of clay minerals enhanced rock susceptibility to slaking as can be seen in the higher weathering grades materials.

Vol. 16 [2011], Bund. O 1327 Figure 5: Box-plot of Id 2 Versus Weathering Grade Jar Slaking Results The objective of Jar test is to observe the reactions of the rock material to water in terms of weathering (Santi, 1998 a,b). The test is done by immersing the samples. The samples are placed in jars filled with tap water for a period of 10 minutes, 15 minutes, 30 minutes and 60 minutes. Slaking index is given to each sample based on the behaviour of samples after each period of time. This test can be carried out at the site or in the laboratory. The rock samples for the test were classified as their respective weathering grade as what have been determined at the site. The rock samples were immersed in water and their slaking behaviours were noted. For each observation, an index was given based on Figures 1 and 2. Test results for jar slaking are presented in Figure 6 until Figure 11. Figure 6 and 7 shows the jar slaking index for Bukit Indah and Mersing sandstone respectively. Grade II samples were found intact for the first 10 minutes. After 30 minutes of immersion, the samples deteriorated to index number 5 and few samples were still intact as index 6. At the end of the 60 minutes immersion, the samples showed index 4 as the lowest index. As for grade III sandstone, the samples were observed to be of index 5 after 10 min of immersion in water. As the immersion was prolonged for another 20 minutes, the samples broke down to index no 4. At 60 minutes, the samples were observed to show number 2 as the lowest index.

Vol. 16 [2011], Bund. O 1328 Figure 6: Test Results of Jar Slaking Index for Bukit Indah Sandstone Figure 7: Test Results of Jar Slaking Index for Mersing Sandstone Grade IVa samples showed index numbers 2, 3 and 4 for the first 10 minutes of immersion. The range of slaking index for sandstone grade IVa after 30 minutes was observed to be 2, 3 or 4. At the end of the test (60 minutes) all samples showed index of number 1. For grade IVb samples, they were observed to show index number of 3 or 4 after 10 minutes of immersion in water. All the samples broke down to index number 1 after 30 minutes. Fine and coarse grained materials from Desa Tebrau (grade IVa) showed index number of 4 after 10 minutes of immersion in water (Figure 8). The coarser grained materials broke down faster than the finer grained ones due to the higher porosity that was detected after 15 minutes. Subsequently, the coarse grained materials showed index numbers of 1 and 2 after 60 minutes while the fine grained showed index number of 2 and 3. As for grade Va samples, it showed index number 2 after 10 minutes and all samples

Vol. 16 [2011], Bund. O 1329 showed index number 1 after 15 minutes. The same results were also observed for Kempas materials as shown in Figure 9. As for the grade Vb sandstone, the materials broke down to the lowest index of number 1 after 10 minutes. Figure 8: Test Results of Jar Slaking Index for Desa Tebrau Figure 9: Test Results of Jar Slaking Index for Kempas Figures 10 and 11 show the jar test result for shale. Grade II shale showed index number 6 after 10 minutes of immersion in water for both Bukit Indah and Mersing materials. The index dropped to number 5 after 15 minutes and maintained at the same index after 30 minutes immersed in water. At the end of 60 minutes, the materials showed index number 4. As for grade III shale, samples from Bukit Indah and Mersing showed index number 5 after 10 minutes. After 15 minutes, they showed index number 4 and after 60 minutes the samples finalized at index number 2, 3 and 4. The grade IVa materials showed a lower index when observed after 10 minutes which was index number 4. In 15 minutes, the materials broke down to index number 2, 3 or 4 depending on the porosity of the samples and their strength. Higher porosity samples with lower strength broke down faster than high strength lower porosity samples. After 30 minutes,

Vol. 16 [2011], Bund. O 1330 samples showed index number 2 and 3. At the end of the test period (60 minutes), the samples showed index number 1. Grade IVb samples showed index number 3 after 10 minutes and subsequently after 15 minutes immersion, the samples broke down to index number 1. The shale of grade IVb broke down faster as compared to sandstone in the same grade as the shale has clay constituents that swelled when immersed in water. The findings in the rate of deterioration in shale were also reported by Santi (1997). Samples Va showed index number 2 after 10 minutes and completely broke down to index number 1 in 30 minutes. Samples Vb showed index number 1 after 10 minutes of immersion. Table 6 and 7 show the summary of the results and the lowest index observed during the study respectively. Figure 10: Test Results of Jar Slaking Index for Bukit Indah Shale Figure 11: Test Results of Jar Slaking Index for Mersing Shale

Vol. 16 [2011], Bund. O 1331 Table 6: Summary of jar test results for respective weathering grade Table 7: Summary of the lowest index observed The indices were used to produce the total jar slake index by adding the index observed in 10, 15, 30 and 60 minutes. The result of the jar slake index with regard to the weathering grade is shown by box-plot in Figure 12. The boxplot showed the decrease of total jar slake index with the increase of weathering grade and the mean value for respective weathering grade is shown in Table 8. Figure 13 (a) to (f) show the typical slaking indices observed during the test. Table 8: Mean value of jar slaking index for respective weathering grade

Vol. 16 [2011], Bund. O 1332 Figure 12: Boxplot of slaking index versus weathering grade

Vol. 16 [2011], Bund. O 1333 Figure 13: Typical jar slake result: (a) index no. 6 (sample R8LN6R2S- 15 min), (b)- index no 5 (sample RL1L5-15 min), (c)- index no. 4 (B1L3-10 min), (d)- index no. 3 (B7L2-10 min), (e)-index no. 2 (sample B1L3-30 min), (f)- index no 1 (sample B6L2-10 min) CONCLUSIONS These modes are limited to the degrading of rock material caused by weathering process before forming residual soil. As stated, shale has different reactions with water apart from sandstone. Shale usually turns into flakes and sandstone would break into smaller pieces. It was found that slake durability test is primarily influenced by rock properties which allow ingress of water into the rock material. The presence of clay minerals enhanced rock susceptibility to slaking as can be seen in the higher weathering grades materials.

Vol. 16 [2011], Bund. O 1334 The test results indicate that the strength, durability and density of rock materials deteriorate with the increase of weathering grade. As strength of rock material is a function of several properties including the hardness of the mineral constituents, degree of compactness, texture and inter-granular bonding material, their inter relation can be expected. Petrographic analysis revealed that feldspar and biotite minerals decomposed to clay as weathering takes place. Sandstone and shale have different orders of slaking index. These differences are caused by the variable amount of pores and particles in each grade of rock material. Various tests were adopted in this study to determine the material properties. However, it was found that certain tests were only suitable to be adopted for certain weathering grades. Id 1 were found suitable to be used to test grade II, III and IVa materials. This was because weaker samples of grade IVb, Va and Vb easily broke down during the first cycle of slake durability test. Thus, Id 1 cannot be measured. The Id 2 was found only suitable to measure stronger rock materials in grades II and III, which can sustain the first cycle (Id 1 ). On the other hand, jar slaking test can be carried out in the field for fast identification of the material properties. The jar slaking test was found to be suitable to measure weak rock materials especially in grade IV and V. The total index observed during the sequence: 10 minutes, 15 minutes, 30 minutes and 60 minutes, was found to be a good indicator for the classification. ACKNOWLEDGEMENTS Authors would like to extent sincere gratitude and appreciation to Research Management Centre, UTM and the Government of Malaysia for the research grant and making the study a success. REFERENCES 1. Franklin, J.A. and A. Chandra (1972) The Slake-Durability Test. Int. J. Rock. Mech. Min. Sci. 9: 325-341. 2. Edy Tonnizam Mohamad, Mohamed Fauzi Md Isa, Mohd For Mohd Amin, Ibrahim Komoo, Nurly Gofar & Rosli Saad (2011) Effect of Moisture Content on the Strength of Various Weathering Grades of Granite, Electronic Journal of Geotechnical Engineering, Vol.16, Bund. H, pp. 863-886 3. Ibrahim Komoo (1995) Syarahan Perdana Geologi Kejuruteraan Perspektif Rantau Tropika Lembap,: Universiti Kebangsaan Malaysia., Universiti Kebangsaan Malaysia. 4. Santi, P.M. (1995) Assessing the Strength and Durability Properties of Shales in Keefer, D.K., Ho, C.L.ed. Landslides Under Static and Dynamic Condition- Analysis, Monitoring and Mitigation, ASCE Geotechnical Special Publication. 52 : 37-55. 5. Santi, P.M. (1997) Comparison of Weak and Weathered Rock Classification Systems, Santi, P.M., and Shakoor, A. (eds). Characterization of Weak and Weathered Rock Masses: Association of Engineering Geologist Special Publication, v. 9 : 139-159. 6. Santi, P.M. (1998a) Refined field methods for identifying, describing and testing shale and weak rock., Proceedings of the 29th Ohio River Valley Soil Seminar: Louisville, Kentucky. U.S.A.

Vol. 16 [2011], Bund. O 1335 7. Santi, P.M. (1998b) Improving the Jar Slake, Slake Index, and Slake Durability Tests for Shales, Environmental & Engineering Geoscience, Vol. IV, No. 3, Fall 1998 : 385-396. 8. Wood, L.E. and Deo, P. (1975) A Suggested System for Classifying Shale Materials for Embankment. Bulletin of the Association of Engineering Geologists. Vol. 12, no. 1 : 39-55. 9. Zainab Mohamed (2004) Pencirian Kejuruteraan Batuan Sedimen Terluluhawa Untuk Kerja Kejuruteraan. Universiti Kebangsaan Malaysia: PhD Thesis, Unpublished. 10. Zainab Mohamed, Ibrahim Komoo, Abd Ghani Rafek and Nurul Aini Osman (2004) A comparative study for strength determination of weak rock. In Fatt, I.D.C.S. ed., Proceedings of the Malaysian Geotechnical Conference 2004 : 255-260. 2011 ejge