- Email: [email protected]
Applications of the oedometer, triaxial and resonant column tests to the study of micaceous sands
Engineering Geology 112 (2010) 21–28
Contents lists available at ScienceDirect
Engineering Geology j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / e n g g e o
Applications of the oedometer, triaxial and resonant column tests to the study of micaceous sands Ali Firat Cabalar Department of Civil Engineering, University of Gaziantep, Gaziantep, Turkey
a r t i c l e
i n f o
Article history: Received 11 June 2009 Received in revised form 12 November 2009 Accepted 24 January 2010 Available online 1 February 2010 Keywords: Coarse rotund sand Mica Triaxial test Resonant column test Oedometer test
a b s t r a c t Intergranular void ratio, es, can be used as an alternative indicator to assess the mechanical and dynamic properties of composite matrix of coarse and fine grains. In this paper, an intensive laboratory study of saturated coarse rotund sand (i.e., Leighton Buzzard Sand) and fine (i.e., mica) mixtures with various mix ratios is performed by a series of oedometer, triaxial and resonant column tests. Oedometer tests performed on coarse rotund sand–fine mixtures show that initial conditions, fines percentages, and stress conditions affect the compression behaviours. Tests indicated that, up to a fraction of mica, which is named as transition fines content (FCt), compression behaviour of the mixture is mainly governed by the sand grains. As the percentage of mica exceeds FCt, mica governs the compression. In the triaxial tests, deviatoric stress (q), pore pressure (u), and undrained Young's modulus (Eu) have been obtained in a triaxial testing apparatus. Results revealed that there is a close relationship between the FCt and q, u, and Eu values. In the resonant column tests, a study of dynamic properties of saturated mixtures was investigated by measuring its shear modulus (G) and damping ratio (D), which is harmonic with the oedometer test results. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Most of the investigations concerning the stress–strain and shear strength behaviour of granular soils basically inspected the response of clean sands. However, field observations show that granular soils may contain finer geomaterials having different shape and size properties. The fines should be expected to affect the engineering behaviour of sandy soils. For example, the presence of platy mica, as fine particles, in coarse rotund sands changes the mechanical behaviour of sandy soils. The mechanical response of micaceous sands has been subject to intensive research in soil mechanics (Terzaghi, 1925; Gilboy, 1928; McCarthy and Leonard, 1963; Olson and Mesri, 1970; Mundegar, 1997; Hight et al., 1998; Clayton et al., 2004). As early as 1925, Terzaghi (1925) stated that much more experimental works were required for the foundation settlements prediction, as particle size alone was not enough to estimate a reasonable indication for the foundation settlements prediction. Gilboy (1928) studied the influence of mica content on the compressibility of sand, and concluded that an increase in mica content resulted in an increase in the void ratio of the uncompressed material as well as an increase in compressibility. The observations, first made by Gilboy (1928), that any system of analysis or classification of soil which neglects the presence and effect of the flat-grained constituents will be incomplete and erroneous. Olson and Mesri (1970) concluded that for all apart from the most active of reconstituted clays, mechanical properties
E-mail addresses: [email protected], [email protected]. 0013-7952/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.enggeo.2010.01.004
were the governing factors in determining compressibility. A recent experimental study by Theron (2004) was conducted on mixtures of mica and sand, and demonstrated the enormous impact of particle shape on the mechanical properties. Most current basic soil mechanics texts show that mica particles; (i) cause undrained strength anisotropy from a brittle response in triaxial extension tests to a ductile behaviour in triaxial compression tests (Hight et al., 1998), (ii) decreases strength (Harris et al., 1984), (iii) alters internal shear mechanism (Lupini et al., 1981), (iv) increase compressibility (Clayton et al., 2004). Micaceous sands are deemed unacceptable for earthworks because of these reasons. Actually, a number of slope failures have been attributed to the presence of mica (Harris et al., 1984). The behaviour of micaceous sands was studied in connection with flow slides that occurred during construction of river training for the Jamura Bridge in Bangladesh (Hight et al., 1998), and Merriespruit gold tailings dam in South Africa which failed in such a catastrophic fashion in 1994 (Fourie and Papageorgiou, 2001; Fourie et al., 2001). Interestingly the behaviour of mica is clay-like, but particle size analyses and the origins of the geomaterial provide that they contain little clay-sized material, and do not have colloidallyactive minerals. Researches related to the influence of mica content on liquefaction potential have been also subjected to intensive research in soil mechanics (Georgiannou et al., 1990; Salgado et al., 2000; Naeini and Baziar, 2004). These investigations were done to quantify the effects of mica particles on the liquefaction phenomena. However, it is difficult to say that the influence of fines on liquefaction mechanism, compression, and the other studied behaviours are systematically
22
A.F. Cabalar / Engineering Geology 112 (2010) 21–28
considered in the literature (Lade et al., 1998; Monkul and Ozden, 2005). Therefore, further investigation about the behaviour of mix of soils including two submatrices (i.e. coarser grain and finer grain) is needed in order to gain a detailed insight regarding their influence on compression, dynamic properties, and stress–strain behaviour. The objective of the present study is to systematically investigate the behaviour of the Leighton Buzzard Sand fraction B (coarse rotund sand)–mica (platy fine) mixtures at various ratios in oedometer, triaxial and resonant column testing apparatus in order to have a better understanding on stress–strain behaviour, dynamic properties, and compression behaviour of these soil mixes. 2. Experimental study 2.1. Materials Two different geomaterials were used in the experimental work; these were the Leighton Buzzard Sand and mica. The Leighton Buzzard Sand used in the experiments was a fraction B supplied by the David Ball Group, Cambridge, U.K., confirming to BS 1881–131:1998. Its specific gravity, minimum and maximum dry densities were found to be 2.65, 1.48 g/cm3 and 1.74 g/cm3 respectively. From the sieve analysis results (Fig. 1), it is seen that more than 90% of the coarse sand particles, which are rounded and mainly quartz, are between (around) 0.6 mm and 1.1 mm. Mica used in the experiments was 52–105 μm muscovite minerals supplied by Dean and Tranter Ltd. It's specific gravity, minimum and maximum dry densities were found to be 2.9, 0.725 g/cm3 and 0.916 g/cm3 respectively (Theron, 2004).
Fig. 1. a. Grain size distributions of the Leighton Buzzard Sand and mica. b. Scanning electron micrograph of the Leighton Buzzard Sand with mica.
The Leighton Buzzard Sand and mica were mixed at various percentage of mica. The percentage of mica used in this study refers to the dry weight of mica relative to the total dry weight of the mixture. In the triaxial tests, three mica percentages were considered, 5%, 10%, 15%; in the resonant column tests, two percentages were considered, 5%, 10%; and in the oedometer tests, four percentages were considered, 5%, 10%, 15%, 20%, and then all results were compared with the clean Leighton Buzzard Sand. Fig. 2 shows SEM picture for a Leighton Buzzard Sand–mica mixture. 2.2. Testing apparatus and procedures 2.2.1. Oedometer tests The consolidation test method carried out during the investigation is the standard method of measuring consolidation properties (ASTM D 2435-96), which involves the incremental loading of soil specimens. Incremental loading is to apply daily increments of vertical load to a submerged container in a rigid ring, with draining permitted through porous stones at the bottom and top. Oedometer samples were tested in 7.5 cm diameter rings. Loadings were initiated from 5 kPa, and were doubled each day, that is, the ratio of load increment to existing load is usually 1. 2.2.2. Triaxial tests The tests were performed in a conventional 100-mm-diameter Wykeham Farrance compression triaxial machine. Strain controlled loading was applied using a digitally controlled STALC 4958 type internal load cell at a constant rate of displacement. In order for the cell and the back pressures to be measured, two pressure transducers, PDCR 810 produced by Druck Limited, were used. Pairs of strain gauges were submersible LVDTs produced by R.D.P. Electronics Ltd., which were employed to measure the axial displacement in the middle third of the specimen in diametrically opposite positions. The Leighton Buzzard Sand and mica were mixed manually in a dry state. The dry mixing process continued around 10 min, until the mixtures are observed to be visually homogenous. Then, the mixtures were inundated to the desired water content to produce a uniform paste. Distilled water was used in the experimental program of the study. A cylindrical membrane was attached to the bottom endplate using two o-rings and the split mould was placed around the endplate. Prepared uniform paste was then gently spooned into the split mould on the pedestal. Great care was taken to ensure that no vibration was employed. When the mould was completely filled, the excess sand particles were removed and the weight of the specimen was recorded. The top end plate was attached with two o-rings and 20 kPa suction was applied to the inside of the specimen. The split mould was carefully split to prevent any disturbance to the specimen.
Fig. 2. Variation of void ratio with fines (mica) content and oedometer pressure.
A.F. Cabalar / Engineering Geology 112 (2010) 21–28
23
The test cell was then assembled and filled with water to apply cell pressure. After the test cell was completely assembled, the loading frame was placed. The suction inside the specimen was decreased while gradually increasing the cell pressure by the desired value. A series of isotropically consolidated undrained triaxial compression tests were carried out on the specimens at 100 kPa effective consolidation stress. During the consolidation process, the porepressure, cell pressure, volume, strain measurements were closely examined and recorded. Following the consolidation, the drainage valve to the specimen was closed, and then compressive load was commenced at a constant displacement rate of 0.015 mm/min.
strain amplitude. The applied force was then removed from the top of the specimen after the resonant frequency, the longitudinal displacement and the accelerometer readings were recorded. Subsequently, the applied force was increased to a higher level, and the test procedures described above were repeated. After the test series was completed under a confining pressure, then confining pressure in the chamber was increased to a higher value and the test was repeated.
2.2.3. Resonant column tests A standard Stokoe Resonant Column apparatus employing cylindrical specimens that are fixed at the base with excitation forces applied to set up was used for tests on the different coarser and finer grain matrices. The apparatus is equipped with an electric motor which has eight coils and four magnets. The motor is driven by a power amplifier and a function generator. Both the data acquisition system and function generator are controlled by the experimentalist. A bath of silicon oil is provided around the specimen to avoid any air migration through the membranes. An accelerometer is used to determine the resonance frequency during the testing. The apparatus was enclosed in a gas pressurized outer cell jacket. The apparatus has the capability of applying both longitudinal and torsional excitations; however only the torsional mode was used during the tests. The required amount of the Leighton Buzzard Sand and mica were weighed, mixed with water demand and then placed in a mould on pedestal. The specimens were prepared in a split mould which measures 70 mm in diameter. A cylindrical membrane was attached to the pedestal using o-ring and the mould was placed around the pedestal. The membrane was stretched around the mould. The Leighton Buzzard Sand–mica–water mixes were spooned without vibrating into the mould. When the mould was completely filled, the top endplate was attached with an o-ring and then a 15 kPa vacuum was applied to the inside the specimen, that allows the specimen to hold the same shape. The mould was carefully split to avoid disturbance to the specimen. The resonant column driver was then assembled and filled with air to apply confining pressure to the specimen surrounded by silicon oil from bottom to the top level. The vacuum inside the specimen was decreased while gradually increasing the confining pressure until the desired value was reached. In this investigation, specimens with 5% and 10% mica–70 mm in diameter and approximately 140 mm height were subjected to isotropic consolidated pressures of approximately 50, 100, 150 kPa and then consolidated undrained tests were conducted. The consolidation pressure was maintained on the specimen for approximately 24 h beyond primary consolidation. In the tests, an isotropic stress increment was applied to the specimens and the soil was allowed to undergo approximately 10% primary consolidation. When primary consolidation was completed, steady-state sinusoidal torsional oscillation was applied to the top of the specimen, while the bottom of the specimen remained fixed. Test procedures for each specimen were same. The excitation frequency and driving force were varied to obtain resonant conditions for each amplitude as measured by the accelerometer located in the apparatus near the top of the specimen. The shear modulus and damping were determined by applying a torsional excitation force of very low magnitude to the specimen. In the testing, the confining pressure was brought to the desired magnitude, and then the test was initiated with a very small torsional excitation force, which produced average shear strain amplitude of 10− 3%. The test was started at the lowest confining pressure of 350 kPa and repeated for intermediate confining pressure up to a maximum confining pressure of 450 kPa. Three magnitudes of effective stress, 50, 100, and 150, were chosen to study the shear modulus and damping ratios as a function of shear
The test results on mica and the Leighton Buzzard Sand show that the characteristics of the sands tested is ascribable to the presence of the flat grains in the samples tested in oedometer. From the 1-D compression results, it was found that the presence of platy particles in the specimens tested had a marked effect on the compressibility of the material under load. In the light of the Theron (2004), the author postulates that mica particles occupy the voids between coarse rotund sand particles. Based on the amount of mica particles present, the Leighton Buzzard Sand particles are in contact with each other and the behaviour of the samples tested are controlled by the Leighton Buzzard Sand particles. When the contacts between the Leighton Buzzard Sand particles reduce, the behaviour of the samples becomes to clay like. Global void ratio with fines content and oedometer stress are shown in Fig. 2. As can be seen from the Fig. 2, initial void ratios for the samples are scattered between 0.61 and 0.95 because of their initial conditions. The governing role of either finer or coarser grain matrices on the overall behaviour of the sample should be expected to change during one dimensional compression. The interchange of this governing role can be expressed using intergranular void ratio concept. Monkul and Onal (2006) proposed an equation for calculation of the intergranular void ratio as follows:
3. Results and discussions 3.1. Oedometer tests
es =
G⋅FC Gf ⋅100 : FC 1− 100
e+ G Gs
⋅
ð1Þ
Gs and Gf in the above equation are the specific gravity of coarser and finer grains forming the soil respectively. G is the specific gravity of soil itself, which is assumed to be the weighted average of the specific gravities of the grains forming the mixtures. FC is the fines content in the mixture. Using the equation above, intergranular void ratios for the samples in this study are found to be kept in a larger band (i.e. 0.61–1.39) than the void ratios (Fig. 3).
Fig. 3. Variation of intergranular void ratio with fines (mica) content and oedometer pressure.
24
A.F. Cabalar / Engineering Geology 112 (2010) 21–28
Fig. 6. Variation of intergranular void ratio with fines (mica) content under various oedometer stresses. Fig. 4. Comparison of the results by Monkul and Onal (2006) and SR model.
In this study, as an alternative approach, a stepwise regression (SR) model based on the laboratory study is proposed for the prediction of intergranular void ratio of the composite matrices (Rawlings, 1998; Cevik, 2007). The input variables in the developed SR models are the mica content (FC), global void ratio (e), and specific gravity of soil itself (G), the output is intergranular void ratio (es). As shown in Fig. 4, comparing the results between Monkul and Onal (2006)'s equation and the one proposed here, the performance of accuracy of the proposed SR models are found to be quite satisfactory (R2 = 0.998). Furthermore, the proposed SR model is also presented as a simple explicit mathematical function for further use by researchers. The SR equation is shown as follows: 1:9 : es = 0:15 + e 3⋅FC + G
ð2Þ
Coarser (Leighton Buzzard Sand) and finer (mica) grain particles can rearrange themselves into various modes as shown in Fig. 5 depending on the initial conditions and applied stress. The ‘void ratio’ refers to the space which is not occupied by the mineral grains. However, when such a sandy soil mixture with some amount of fines is examined, the volume of the voids can be thought as voids due to the Leighton Buzzard Sand grains (which is intergranular void ratio, es) and voids due to the mica grains. Thus, intergranular void ratio (es) can be described as the ratio of volume of the intergranular voids to the volume of the Leighton Buzzard Sand grains. From the study by Monkul and Ozden (2007), establishment of direct grain contacts of the coarser grain matrix can be assumed to be initiated when the intergranular void ratio of the mixture becomes equal to the maximum void ratio of the host granular material, which is the Leighton Buzzard Sand in this study (i.e., es = emax). The fines content,
at which this condition occurs, was named as “transition fines content, FCt” by Monkul and Ozden (2005). Variation of intergranular void ratio with fines content under various oedometer stresses is shown in Fig. 6. Transition fines contents (mica contents) varying with effective stresses can be determined by the intersection of the dashed line with curves. Transition fines content (FCt) values vary between 6.1% and 12.5% depending on the applied stress (Table 1). FCt values increase as the effective stress increases. That means, under higher stress, transition arrangement of the Leighton Buzzard Sand grain matrix takes place at higher values of mica content. The same intergranular void ratio (es) can be observed at different combinations of the fines content (FC) and global void ratio (e) (Monkul and Ozden, 2004). In other words, intergranular void ratio value of 0.71 (emax), where transition occurs, can be observed with a higher transition fines content and a lower global void ratio combination with increase in effective stress. Intergranular void ratio, fines content, and applied stress can be seen in a 3-D plot in Fig. 7. Fabric change mechanism during 1-D compression for a sample of certain fines content can be further investigated with Fig. 5. For example, the initial state of the fabric of the sample with 10% mica (shown by triangular points in Fig. 3) can be represented by Fig. 5 a. The Leighton Buzzard Sand grains get closer to each other as the oedometer pressure increases. At the beginning, active contacts among the Leighton Buzzard Sand grains are not generated. Further loading towards 500 kPa should start to generate contact points among the Leighton Buzzard Sand grains. Additional stress increment results in a steady decrease of the intergranular void ratio causing formation of strong contact points as illustrated in Fig. 5 b. The envelope represented by the dashed line in Fig. 3 shows the upper limit, under which the Leighton Buzzard Sand matrix forms Table 1 Transition fines contents for the specimens under different oedometer pressures.
Fig. 5. Different arrangements of matrices under 1-D compression.
Effectiveness stress, (kPa)
FCt (%)
5.88 14.94 28.53 52.53 76.53 124.53 220.53 412.53 604.53 831.83 1059.13 1244.43 1429.73
6.1 6.7 7.0 7.5 7.7 8.2 9.0 9.5 10.4 11.3 11.7 12.2 12.5
A.F. Cabalar / Engineering Geology 112 (2010) 21–28
25
Fig. 7. Variation of intergranular void ratio with effective stress for different fines (mica) contents.
Fig. 9. Variation of compression parameters with fines content under 124 kPa effective stress.
a continuous framework with grain to grain contacts. Hence, the compressional characteristics are also expected to deviate under this limit. In order to better understand the arrangement of Leighton Buzzard Sand grain matrix, the granular compression index (Cc-s) parameter can be utilized (Monkul and Ozden, 2005). The definition of granular compression index is very similar to the definition of compression index (Cc). The Cc-s is expressed as decrease of intergranular void ratio with effective stress, which is shown as below.
fines content, Leighton Buzzard Sand grains become more dispersed so that there is almost no grain contact between them. At this zone, the compressibility of the soil is expected to be mainly governed by the mica grain matrix. Therefore, the difference between granular compression and global compression index increases with an increase in mica content.
Ccs =
Δes Δ logσ ′
ð3Þ
Variation of global compression index (Cc) and granular compression index (Cc−s) with fines (mica) content is shown in Figs. 8 and 9. Cc and Cc-s values are calculated for 28 kPa and 124 kPa effective stresses. Compression indices of both sets of data increase relatively linearly with fines (mica) content. However, granular compression indices of the same set of data show a non-linear behaviour. Compression behaviour of the Leighton Buzzard Sand–mica mixtures can be examined in two stages shown by different zones in Figs. 8 and 9. The dashed lines in the figures show the transition fines contents (FCt) for the pressure ranges (Table 1). Through the zone left of the FCt, Leighton Buzzard Sand grain matrix can be thought to have almost a continuous force chain with grain to grain contacts, and mica grains are mostly located in the intergranular voids. Therefore, a less compressive behaviour is seen. However, initiating from the FCt, values, Leighton Buzzard Sand grain contacts start to decrease because of the filling mica. Therefore, a sharp increase in the slope of granular compression index curve can be observed. With further increase in
Fig. 8. Variation of compression parameters with fines content under 28 kPa effective stress.
3.2. Triaxial tests The interaction between coarser (i.e. Leighton Buzzard Sand) and finer (i.e., mica) grain matrices should also affect the overall strength behaviour. In order to investigate this effect and whether there is a relation with oedometer findings, a series of triaxial tests were conducted. As can be seen from the Fig. 10 denoting deviatoric stress versus strain, there is also a close relationship between transition fines content and the deviatoric stress of mixtures. The vertical dashed lines in Fig. 11 denote the transition fines content (mica) obtained from the oedometer tests for corresponding stress level (100 kPa). Immediately after from transition fines content (8%), the deviatoric stress of the mixtures start to keep a steady state condition for all strain levels. When the fines content of a mixture exceeds its transition content, intergranular void ratio of that mixture becomes greater than the maximum void ratio of the Leighton Buzzard Sand matrix itself. In other words, the infilling mica particles at intergranular voids separate the coarser grains and decrease their contact points. It is the mica particles that mainly control the soil behaviour, as there are no or very few contact between the Leighton Buzzard Sand grains.
Fig. 10. Stress–strain curve for clean loose sand and sand with different proportions of mica.
26
A.F. Cabalar / Engineering Geology 112 (2010) 21–28
Fig. 14. Undrained Young's modulus as a function of local axial strain. Fig. 11. Deviatoric stress change with fines (mica) content.
Therefore, a noticeable decrease in deviatoric stress over all the strain levels can be observed, when the mica content of mixtures exceeds FCt. As the strain level increases, the noticeable decrement trends for the deviatoric stress of the mixtures start to decrease markedly (Fig. 11). Fig. 12 shows the effects of increasing mica content on the pore water pressure–strain behaviour of the material. The clean loose sand specimen behaves as might be expected, but the dilation of 5% and more by weight of mica causes the suppression of any dilation, low undrained shear strengths, and high level of pore pressure generation during shear. Change in pore pressure with the FC over the various strain levels is shown in Fig. 13. The plot area is divided in to two
zones based on the oedometer test results, similar to Fig. 11, which are (i) the zone governed by the Leighton Buzzard Sand, and (ii) the zone governed by mica. From end of the first zone, where the fines content is around 8%, the pore pressure of the mixtures starts to keep a steady state condition for all strain levels. When the fines content of a mixture exceeds the transition zone, intergranular void ratio of that mixture becomes greater than the maximum void ratio of the Leighton Buzzard Sand matrix itself. Therefore, the infilling mica particles at intergranular voids separate the coarser grains and decrease their contact points. As the strain level increases, the noticeable increase trends for the pore pressure of the mixtures start to decrease markedly (Fig. 13). The influence of mica on the small strain stiffness was also investigated by considering undrained secant Young's modulus (Eu) (Fig. 14). Comparing the small strain stiffness behaviour of the micaLeighton Buzzard Sand mixtures at 100 kPa effective consolidation pressure, it is noted that the addition of mica to the Leighton Buzzard Sand particles resulted in reduced specimen stiffness. Fig. 15 shows the variation of Eu with fines content at different strain levels.
3.3. Resonant column tests The interaction between coarser (Leighton Buzzard Sand) and finer (mica) grain matrices should also affect the dynamic parameters of the mixtures. In order to inspect this affect and whether there is a relationship with the oedometer results, a serious of resonant column tests was performed. The Leighton Buzzard Sand and mica were mixed at various percentage of mica. Two mica percentages were considered; namely 5%, 10%, and then compared with the clean Fig. 12. Pore water pressure strain curve for clean loose sand and sand with different proportions of mica.
Fig. 13. Pore pressure generation change with fines (mica) content.
Fig. 15. Undrained Young's modulus change with fines (mica) content.
A.F. Cabalar / Engineering Geology 112 (2010) 21–28
27
Fig. 16. G/Gmax and D/Dmin strain curves for Leighton Buzzard Sand with different proportions of mica at 150 kPa effective consolidation stress.
Fig. 18. Shear modulus and damping ratio curves for Leighton Buzzard Sand with 5% mica at different effective stress values.
Leighton Buzzard Sand. Specimens were tested under the pressures of approximately 50, 100, 150 kPa. The effects of finer grains at different mix ratios on the smallstrain shear modulus, G/Gmax, and material damping ratio, D/Dmin are shown in Figs. 16–18. The general effect is that G/Gmax decreases and D/Dmin increases with mica content regardless of confining pressure. All reported values of G/Gmax and D/Dmin were measured at 50 kPa, 100 kPa, and 150 kPa effective stresses. The applied back pressure was kept constant at 300 kPa for all the tests. The influence of effective stress on G/Gmax and D/Dmin for the Leighton Buzzard Sand with 5% mica is shown in Fig. 18. It can be seen from the Fig. 18 that all the mixtures have shown an overall increase in G/Gmax and decrease in D/Dmin values as the effective stress increases. Figs. 16 and 17 show the effects of mica content on the smallstrain shear modulus and material damping ratio of the material, under isotropically-consolidated undrained triaxial compression at 150 and 100 kPa effective stresses, respectively. The dashed and solid lines in the figures show the test results for the data from the clean Leighton Buzzard Sand and the Leighton Buzzard Sand with mica at various mix ratios respectively. This kind of comparison clearly shows that the stress–strain response to torsional oscillation is getting softer as the mica content increases. It is seen that the greater the proportion of finer grain matrices (mica) in coarser sand (Leighton Buzzard Sand), the greater is the damping ratio, and the
lower is the shear modulus. Moreover, as an overall view within a given strain level, G decreases with an increase of the strain as D increases (Figs. 16–18). The response of soil with less mica exhibit more linear behaviour at smaller strain levels. As shearing increases, non-linearity in G/Gmax and D/Dmin values increase. The curves, which are quite close at very small strains, tend to diverge for increasing shear strain level. According to the study by Hardin and Blandford (1989), the maximum shear modulus is primarily dependent on the confining pressure. Fig. 18 shows the measured shear moduli at small strains as a function of effective stress while the pore pressure was kept at the same value (i.e., 300 kPa). The influence of effective stress on shear modulus is that the higher the confining pressure, the higher the shear modulus. Damping ratios calculated for the specimens tested at given shear strain amplitudes are also plotted in Fig. 18 as a function of effective stress for the Leighton Buzzard Sand with 5% mica. A trend for the damping ratio to decrease with increasing confining pressure is evident from Fig. 18. The results show the shear modulus remaining near the maximum value through 10− 2% strain before deviating significantly. Moreover, the damping ratio results show the damping ratio remaining about 1% throughout 7 × 10− 3% strain before increasing significantly. Fig. 19 shows the D/Dmin and G/Gmax with fines (mica) content for the specimen tested under 150 kPa effective stress. The FCt value for this effective stress was found to be 8.5% (Table 1).
Fig. 17. G/Gmax and D/Dmin strain curves for Leighton Buzzard Sand with different proportions of mica at 100 kPa effective consolidation stress.
Fig. 19. Changes in D/Dmin and G/Gmax with fine (mica) content for 150 kPa effective stress.
28
A.F. Cabalar / Engineering Geology 112 (2010) 21–28
4. Conclusions Engineering behaviour of micaceous sand was examined with the help of submatrices forming them (Leighton Buzzard Sand and mica). The effectiveness of the Leighton Buzzard Sand grain matrix on compressional behaviour was determined by oedometer tests by using intergranular void ratio instead of global void ratio. The mica content where the intergranular void ratio of the mixture and the maximum void ratio of the Leighton Buzzard Sand are equal to each other, has been termed the transition fines content. The transition fines content increases with increasing effective stresses. The 1-D compression behaviour of micaceous sands is mainly governed by the Leighton Buzzard Sand up to the transition fines content, whereas, it is governed by the mica grain matrix after the transition fines content. It is indicated that the granular compression index can be used as an important parameter to assess the role of the mica grain matrix on overall compression behaviour. Results of triaxial tests showed a close relationship between the transition fines content and the mechanical properties of mixtures. Exceeding a transition fines content value under the specified stress condition, the deviatoric stress, pore water pressure generation and undrained Young's modulus of the mixtures tend to be a steady state. Results of resonant column tests also showed a close relationship between the transition fines content and the dynamic properties of mixtures. Exceeding a transition fines content value for a given stress condition, the damping ratio and shear modulus of the mixtures tend to a steady state. This study investigated the transition fines content and its effect on 1-D compression behaviour of micaceous soils, and its effect on mechanical and dynamic properties. The concept of the transition fines content can be useful in various geotechnical and geoenvironmental engineering applications, where naturally sandy soils are treated, to estimate whether the fine or coarser matrix governs the overall engineering properties. It is believed that further macro and microfabric studies, and studies of the mineralogy of the mixture, chemical state of minerals, pore fluid properties, and gradation of the materials (both finer and coarser) are needed in order to better understand the arrangement of the Leighton Buzzard Sand and mica grain matrices with respect to each other, and to model the mechanical and dynamic properties of micaceous sands. Acknowledgements The author would like to thank Prof. C.R.I. Clayton, and Dr. J. Priest of the University of Southampton for their invaluable help.
References Cevik, A., 2007. Unified formulation for web crippling strength of cold-formed steel sheeting using stepwise regression. Journal of Steel Research 63, 1305–1316. Clayton, C.R.I., Theron, M., Vermeulen, N.J., 2004. The effect of particle shape on the behaviour of gold tailings. Advances in Geotechnical Engineering: The Skempton Conference. Thomas Telford, London, pp. 393–404. Fourie, A.B., Papageorgiou, G., 2001. Defining an appropriate steady state line for Merriespruit gold tailings. Canadian Geotechnical Journal 38, 695–706. Fourie, A.B., Blight, G.E., Papageorgiou, G., 2001. Static liquefaction as a possible explanation for the Merriespruit tailings dam failure. Canadian Geotechnical Journal 38, 707–719. Georgiannou, V.N., Burland, J.B., Hight, D.W., 1990. The undrained behaviour of clayey sands in triaxial compression and extension. Géotechnique 40 (3), 431–449. Gilboy, G., 1928. The compressibility of sand–mica mixtures. Proceedings of the A.S.C.E. 2, 555–568. Hardin, B.O., Blandford, G.E., 1989. Elasticity of particulate materials. Journal of Geotechnical Engineering, ASCE 115 (6), 788–805. Harris, W.G., Parker, J.C., Zelazny, L.W., 1984. Effects of mica content on the engineering properties of sand. Soil Science Society of America Journal 48 (3), 501–505. Hight, D.W., Georgiannou, V.N., Martin, P.L., Mundegar, A.K. 1998. Flow slides in micaceous sand. Problematic soils, Yanagisawa, E., Moroto, N. and Mitachi, T. eds., Balkema, Rotterdam, Sendai, Japan, 945–958. Lade, P.V., Liggio Jr., C.D., Yamamuro, J.A., 1998. Effects of nonplastic fines on minimum and maximum void ratios of sand. Geotechnical Testing Journal, ASTM 21 (4), 336–347. Lupini, J.F., Skinner, A.E., Vaughan, P.R., 1981. The drained residual strength of cohesive soils. Géotechnique 31 (2), 181–213. McCarthy, D.F., Leonard, R.J., 1963. Compaction and compression characteristics of micaceous fine sands and silts. Highway Research Record 22. Transportation Research Board, Washington, D.C, pp. 23–37. Monkul, M.M., Ozden, G., 2004. Intergranular and interfine void ratio concepts. 10th Turkish Congress on Soil Mechanics and Foundation Engineering /ZM10), TNCSMFE, Istanbul, pp. 3–12. in Turkish. Monkul, M.M., Ozden, G., 2005. Effect of intergranular void ratio on one-dimensional compression behaviour. Proceedings of International Conference on Problematic Soils, International Society of Soil Mechanics and Geotechnical Engineering, Famagusta, Turkish Republic of Northern Cyprus 3, 1203–1209. Monkul, M.M., Onal, O., 2006. A visual basic program for analyzing oedometer test results and evaluating intergranular void ratio. Computers and Geosciences, Elsevier Science 32, 696–703. Monkul, M.M., Ozden, G., 2007. Compression behavior of clayey sand and transition fines content. Engineering Geology 89, 195–205. Mundegar, A.K., 1997. An investigation into the effects of platy mica particles on the behaviour of sand. M.Sc, Thesis, Imperial College, London. Naeini, S.A., Baziar, M.H., 2004. Effect of fines content on steady state strength of mixed and layered samples of sand. Journal of Soil Dynamics and Earthquake Engineering 24, 181–187. Olson, R.E., Mesri, G., 1970. Mechanisms controlling the compressibility of clay. Journal of the Soil Mechanics and Foundations Division, A.S.C.E., 96 (SM6), Proc. Paper 7649, November, pp. 1863–1878. Salgado, R., Bandini, P., Karim, A., 2000. Shear strength and stiffness of silty sand. Journal of Geotechnical and Geoenvironmental Engineering, ASCE 126 (5), 451–462. Rawlings, J.O., 1998. Applied regression analysis: a research tool. Springer-Verlag, New York. Terzaghi, K., 1925. Erdbaumechanik auf bodenphysikalischer grundlage. Deuticke, Leipzig/Vienna. Theron, M. 2004. The Effect of Particle Shape on the Behaviour of Gold Tailings. PhD thesis, University of Southampton, U.K.
Contents lists available at ScienceDirect
Engineering Geology j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / e n g g e o
Applications of the oedometer, triaxial and resonant column tests to the study of micaceous sands Ali Firat Cabalar Department of Civil Engineering, University of Gaziantep, Gaziantep, Turkey
a r t i c l e
i n f o
Article history: Received 11 June 2009 Received in revised form 12 November 2009 Accepted 24 January 2010 Available online 1 February 2010 Keywords: Coarse rotund sand Mica Triaxial test Resonant column test Oedometer test
a b s t r a c t Intergranular void ratio, es, can be used as an alternative indicator to assess the mechanical and dynamic properties of composite matrix of coarse and fine grains. In this paper, an intensive laboratory study of saturated coarse rotund sand (i.e., Leighton Buzzard Sand) and fine (i.e., mica) mixtures with various mix ratios is performed by a series of oedometer, triaxial and resonant column tests. Oedometer tests performed on coarse rotund sand–fine mixtures show that initial conditions, fines percentages, and stress conditions affect the compression behaviours. Tests indicated that, up to a fraction of mica, which is named as transition fines content (FCt), compression behaviour of the mixture is mainly governed by the sand grains. As the percentage of mica exceeds FCt, mica governs the compression. In the triaxial tests, deviatoric stress (q), pore pressure (u), and undrained Young's modulus (Eu) have been obtained in a triaxial testing apparatus. Results revealed that there is a close relationship between the FCt and q, u, and Eu values. In the resonant column tests, a study of dynamic properties of saturated mixtures was investigated by measuring its shear modulus (G) and damping ratio (D), which is harmonic with the oedometer test results. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Most of the investigations concerning the stress–strain and shear strength behaviour of granular soils basically inspected the response of clean sands. However, field observations show that granular soils may contain finer geomaterials having different shape and size properties. The fines should be expected to affect the engineering behaviour of sandy soils. For example, the presence of platy mica, as fine particles, in coarse rotund sands changes the mechanical behaviour of sandy soils. The mechanical response of micaceous sands has been subject to intensive research in soil mechanics (Terzaghi, 1925; Gilboy, 1928; McCarthy and Leonard, 1963; Olson and Mesri, 1970; Mundegar, 1997; Hight et al., 1998; Clayton et al., 2004). As early as 1925, Terzaghi (1925) stated that much more experimental works were required for the foundation settlements prediction, as particle size alone was not enough to estimate a reasonable indication for the foundation settlements prediction. Gilboy (1928) studied the influence of mica content on the compressibility of sand, and concluded that an increase in mica content resulted in an increase in the void ratio of the uncompressed material as well as an increase in compressibility. The observations, first made by Gilboy (1928), that any system of analysis or classification of soil which neglects the presence and effect of the flat-grained constituents will be incomplete and erroneous. Olson and Mesri (1970) concluded that for all apart from the most active of reconstituted clays, mechanical properties
E-mail addresses: [email protected], [email protected]. 0013-7952/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.enggeo.2010.01.004
were the governing factors in determining compressibility. A recent experimental study by Theron (2004) was conducted on mixtures of mica and sand, and demonstrated the enormous impact of particle shape on the mechanical properties. Most current basic soil mechanics texts show that mica particles; (i) cause undrained strength anisotropy from a brittle response in triaxial extension tests to a ductile behaviour in triaxial compression tests (Hight et al., 1998), (ii) decreases strength (Harris et al., 1984), (iii) alters internal shear mechanism (Lupini et al., 1981), (iv) increase compressibility (Clayton et al., 2004). Micaceous sands are deemed unacceptable for earthworks because of these reasons. Actually, a number of slope failures have been attributed to the presence of mica (Harris et al., 1984). The behaviour of micaceous sands was studied in connection with flow slides that occurred during construction of river training for the Jamura Bridge in Bangladesh (Hight et al., 1998), and Merriespruit gold tailings dam in South Africa which failed in such a catastrophic fashion in 1994 (Fourie and Papageorgiou, 2001; Fourie et al., 2001). Interestingly the behaviour of mica is clay-like, but particle size analyses and the origins of the geomaterial provide that they contain little clay-sized material, and do not have colloidallyactive minerals. Researches related to the influence of mica content on liquefaction potential have been also subjected to intensive research in soil mechanics (Georgiannou et al., 1990; Salgado et al., 2000; Naeini and Baziar, 2004). These investigations were done to quantify the effects of mica particles on the liquefaction phenomena. However, it is difficult to say that the influence of fines on liquefaction mechanism, compression, and the other studied behaviours are systematically
22
A.F. Cabalar / Engineering Geology 112 (2010) 21–28
considered in the literature (Lade et al., 1998; Monkul and Ozden, 2005). Therefore, further investigation about the behaviour of mix of soils including two submatrices (i.e. coarser grain and finer grain) is needed in order to gain a detailed insight regarding their influence on compression, dynamic properties, and stress–strain behaviour. The objective of the present study is to systematically investigate the behaviour of the Leighton Buzzard Sand fraction B (coarse rotund sand)–mica (platy fine) mixtures at various ratios in oedometer, triaxial and resonant column testing apparatus in order to have a better understanding on stress–strain behaviour, dynamic properties, and compression behaviour of these soil mixes. 2. Experimental study 2.1. Materials Two different geomaterials were used in the experimental work; these were the Leighton Buzzard Sand and mica. The Leighton Buzzard Sand used in the experiments was a fraction B supplied by the David Ball Group, Cambridge, U.K., confirming to BS 1881–131:1998. Its specific gravity, minimum and maximum dry densities were found to be 2.65, 1.48 g/cm3 and 1.74 g/cm3 respectively. From the sieve analysis results (Fig. 1), it is seen that more than 90% of the coarse sand particles, which are rounded and mainly quartz, are between (around) 0.6 mm and 1.1 mm. Mica used in the experiments was 52–105 μm muscovite minerals supplied by Dean and Tranter Ltd. It's specific gravity, minimum and maximum dry densities were found to be 2.9, 0.725 g/cm3 and 0.916 g/cm3 respectively (Theron, 2004).
Fig. 1. a. Grain size distributions of the Leighton Buzzard Sand and mica. b. Scanning electron micrograph of the Leighton Buzzard Sand with mica.
The Leighton Buzzard Sand and mica were mixed at various percentage of mica. The percentage of mica used in this study refers to the dry weight of mica relative to the total dry weight of the mixture. In the triaxial tests, three mica percentages were considered, 5%, 10%, 15%; in the resonant column tests, two percentages were considered, 5%, 10%; and in the oedometer tests, four percentages were considered, 5%, 10%, 15%, 20%, and then all results were compared with the clean Leighton Buzzard Sand. Fig. 2 shows SEM picture for a Leighton Buzzard Sand–mica mixture. 2.2. Testing apparatus and procedures 2.2.1. Oedometer tests The consolidation test method carried out during the investigation is the standard method of measuring consolidation properties (ASTM D 2435-96), which involves the incremental loading of soil specimens. Incremental loading is to apply daily increments of vertical load to a submerged container in a rigid ring, with draining permitted through porous stones at the bottom and top. Oedometer samples were tested in 7.5 cm diameter rings. Loadings were initiated from 5 kPa, and were doubled each day, that is, the ratio of load increment to existing load is usually 1. 2.2.2. Triaxial tests The tests were performed in a conventional 100-mm-diameter Wykeham Farrance compression triaxial machine. Strain controlled loading was applied using a digitally controlled STALC 4958 type internal load cell at a constant rate of displacement. In order for the cell and the back pressures to be measured, two pressure transducers, PDCR 810 produced by Druck Limited, were used. Pairs of strain gauges were submersible LVDTs produced by R.D.P. Electronics Ltd., which were employed to measure the axial displacement in the middle third of the specimen in diametrically opposite positions. The Leighton Buzzard Sand and mica were mixed manually in a dry state. The dry mixing process continued around 10 min, until the mixtures are observed to be visually homogenous. Then, the mixtures were inundated to the desired water content to produce a uniform paste. Distilled water was used in the experimental program of the study. A cylindrical membrane was attached to the bottom endplate using two o-rings and the split mould was placed around the endplate. Prepared uniform paste was then gently spooned into the split mould on the pedestal. Great care was taken to ensure that no vibration was employed. When the mould was completely filled, the excess sand particles were removed and the weight of the specimen was recorded. The top end plate was attached with two o-rings and 20 kPa suction was applied to the inside of the specimen. The split mould was carefully split to prevent any disturbance to the specimen.
Fig. 2. Variation of void ratio with fines (mica) content and oedometer pressure.
A.F. Cabalar / Engineering Geology 112 (2010) 21–28
23
The test cell was then assembled and filled with water to apply cell pressure. After the test cell was completely assembled, the loading frame was placed. The suction inside the specimen was decreased while gradually increasing the cell pressure by the desired value. A series of isotropically consolidated undrained triaxial compression tests were carried out on the specimens at 100 kPa effective consolidation stress. During the consolidation process, the porepressure, cell pressure, volume, strain measurements were closely examined and recorded. Following the consolidation, the drainage valve to the specimen was closed, and then compressive load was commenced at a constant displacement rate of 0.015 mm/min.
strain amplitude. The applied force was then removed from the top of the specimen after the resonant frequency, the longitudinal displacement and the accelerometer readings were recorded. Subsequently, the applied force was increased to a higher level, and the test procedures described above were repeated. After the test series was completed under a confining pressure, then confining pressure in the chamber was increased to a higher value and the test was repeated.
2.2.3. Resonant column tests A standard Stokoe Resonant Column apparatus employing cylindrical specimens that are fixed at the base with excitation forces applied to set up was used for tests on the different coarser and finer grain matrices. The apparatus is equipped with an electric motor which has eight coils and four magnets. The motor is driven by a power amplifier and a function generator. Both the data acquisition system and function generator are controlled by the experimentalist. A bath of silicon oil is provided around the specimen to avoid any air migration through the membranes. An accelerometer is used to determine the resonance frequency during the testing. The apparatus was enclosed in a gas pressurized outer cell jacket. The apparatus has the capability of applying both longitudinal and torsional excitations; however only the torsional mode was used during the tests. The required amount of the Leighton Buzzard Sand and mica were weighed, mixed with water demand and then placed in a mould on pedestal. The specimens were prepared in a split mould which measures 70 mm in diameter. A cylindrical membrane was attached to the pedestal using o-ring and the mould was placed around the pedestal. The membrane was stretched around the mould. The Leighton Buzzard Sand–mica–water mixes were spooned without vibrating into the mould. When the mould was completely filled, the top endplate was attached with an o-ring and then a 15 kPa vacuum was applied to the inside the specimen, that allows the specimen to hold the same shape. The mould was carefully split to avoid disturbance to the specimen. The resonant column driver was then assembled and filled with air to apply confining pressure to the specimen surrounded by silicon oil from bottom to the top level. The vacuum inside the specimen was decreased while gradually increasing the confining pressure until the desired value was reached. In this investigation, specimens with 5% and 10% mica–70 mm in diameter and approximately 140 mm height were subjected to isotropic consolidated pressures of approximately 50, 100, 150 kPa and then consolidated undrained tests were conducted. The consolidation pressure was maintained on the specimen for approximately 24 h beyond primary consolidation. In the tests, an isotropic stress increment was applied to the specimens and the soil was allowed to undergo approximately 10% primary consolidation. When primary consolidation was completed, steady-state sinusoidal torsional oscillation was applied to the top of the specimen, while the bottom of the specimen remained fixed. Test procedures for each specimen were same. The excitation frequency and driving force were varied to obtain resonant conditions for each amplitude as measured by the accelerometer located in the apparatus near the top of the specimen. The shear modulus and damping were determined by applying a torsional excitation force of very low magnitude to the specimen. In the testing, the confining pressure was brought to the desired magnitude, and then the test was initiated with a very small torsional excitation force, which produced average shear strain amplitude of 10− 3%. The test was started at the lowest confining pressure of 350 kPa and repeated for intermediate confining pressure up to a maximum confining pressure of 450 kPa. Three magnitudes of effective stress, 50, 100, and 150, were chosen to study the shear modulus and damping ratios as a function of shear
The test results on mica and the Leighton Buzzard Sand show that the characteristics of the sands tested is ascribable to the presence of the flat grains in the samples tested in oedometer. From the 1-D compression results, it was found that the presence of platy particles in the specimens tested had a marked effect on the compressibility of the material under load. In the light of the Theron (2004), the author postulates that mica particles occupy the voids between coarse rotund sand particles. Based on the amount of mica particles present, the Leighton Buzzard Sand particles are in contact with each other and the behaviour of the samples tested are controlled by the Leighton Buzzard Sand particles. When the contacts between the Leighton Buzzard Sand particles reduce, the behaviour of the samples becomes to clay like. Global void ratio with fines content and oedometer stress are shown in Fig. 2. As can be seen from the Fig. 2, initial void ratios for the samples are scattered between 0.61 and 0.95 because of their initial conditions. The governing role of either finer or coarser grain matrices on the overall behaviour of the sample should be expected to change during one dimensional compression. The interchange of this governing role can be expressed using intergranular void ratio concept. Monkul and Onal (2006) proposed an equation for calculation of the intergranular void ratio as follows:
3. Results and discussions 3.1. Oedometer tests
es =
G⋅FC Gf ⋅100 : FC 1− 100
e+ G Gs
⋅
ð1Þ
Gs and Gf in the above equation are the specific gravity of coarser and finer grains forming the soil respectively. G is the specific gravity of soil itself, which is assumed to be the weighted average of the specific gravities of the grains forming the mixtures. FC is the fines content in the mixture. Using the equation above, intergranular void ratios for the samples in this study are found to be kept in a larger band (i.e. 0.61–1.39) than the void ratios (Fig. 3).
Fig. 3. Variation of intergranular void ratio with fines (mica) content and oedometer pressure.
24
A.F. Cabalar / Engineering Geology 112 (2010) 21–28
Fig. 6. Variation of intergranular void ratio with fines (mica) content under various oedometer stresses. Fig. 4. Comparison of the results by Monkul and Onal (2006) and SR model.
In this study, as an alternative approach, a stepwise regression (SR) model based on the laboratory study is proposed for the prediction of intergranular void ratio of the composite matrices (Rawlings, 1998; Cevik, 2007). The input variables in the developed SR models are the mica content (FC), global void ratio (e), and specific gravity of soil itself (G), the output is intergranular void ratio (es). As shown in Fig. 4, comparing the results between Monkul and Onal (2006)'s equation and the one proposed here, the performance of accuracy of the proposed SR models are found to be quite satisfactory (R2 = 0.998). Furthermore, the proposed SR model is also presented as a simple explicit mathematical function for further use by researchers. The SR equation is shown as follows: 1:9 : es = 0:15 + e 3⋅FC + G
ð2Þ
Coarser (Leighton Buzzard Sand) and finer (mica) grain particles can rearrange themselves into various modes as shown in Fig. 5 depending on the initial conditions and applied stress. The ‘void ratio’ refers to the space which is not occupied by the mineral grains. However, when such a sandy soil mixture with some amount of fines is examined, the volume of the voids can be thought as voids due to the Leighton Buzzard Sand grains (which is intergranular void ratio, es) and voids due to the mica grains. Thus, intergranular void ratio (es) can be described as the ratio of volume of the intergranular voids to the volume of the Leighton Buzzard Sand grains. From the study by Monkul and Ozden (2007), establishment of direct grain contacts of the coarser grain matrix can be assumed to be initiated when the intergranular void ratio of the mixture becomes equal to the maximum void ratio of the host granular material, which is the Leighton Buzzard Sand in this study (i.e., es = emax). The fines content,
at which this condition occurs, was named as “transition fines content, FCt” by Monkul and Ozden (2005). Variation of intergranular void ratio with fines content under various oedometer stresses is shown in Fig. 6. Transition fines contents (mica contents) varying with effective stresses can be determined by the intersection of the dashed line with curves. Transition fines content (FCt) values vary between 6.1% and 12.5% depending on the applied stress (Table 1). FCt values increase as the effective stress increases. That means, under higher stress, transition arrangement of the Leighton Buzzard Sand grain matrix takes place at higher values of mica content. The same intergranular void ratio (es) can be observed at different combinations of the fines content (FC) and global void ratio (e) (Monkul and Ozden, 2004). In other words, intergranular void ratio value of 0.71 (emax), where transition occurs, can be observed with a higher transition fines content and a lower global void ratio combination with increase in effective stress. Intergranular void ratio, fines content, and applied stress can be seen in a 3-D plot in Fig. 7. Fabric change mechanism during 1-D compression for a sample of certain fines content can be further investigated with Fig. 5. For example, the initial state of the fabric of the sample with 10% mica (shown by triangular points in Fig. 3) can be represented by Fig. 5 a. The Leighton Buzzard Sand grains get closer to each other as the oedometer pressure increases. At the beginning, active contacts among the Leighton Buzzard Sand grains are not generated. Further loading towards 500 kPa should start to generate contact points among the Leighton Buzzard Sand grains. Additional stress increment results in a steady decrease of the intergranular void ratio causing formation of strong contact points as illustrated in Fig. 5 b. The envelope represented by the dashed line in Fig. 3 shows the upper limit, under which the Leighton Buzzard Sand matrix forms Table 1 Transition fines contents for the specimens under different oedometer pressures.
Fig. 5. Different arrangements of matrices under 1-D compression.
Effectiveness stress, (kPa)
FCt (%)
5.88 14.94 28.53 52.53 76.53 124.53 220.53 412.53 604.53 831.83 1059.13 1244.43 1429.73
6.1 6.7 7.0 7.5 7.7 8.2 9.0 9.5 10.4 11.3 11.7 12.2 12.5
A.F. Cabalar / Engineering Geology 112 (2010) 21–28
25
Fig. 7. Variation of intergranular void ratio with effective stress for different fines (mica) contents.
Fig. 9. Variation of compression parameters with fines content under 124 kPa effective stress.
a continuous framework with grain to grain contacts. Hence, the compressional characteristics are also expected to deviate under this limit. In order to better understand the arrangement of Leighton Buzzard Sand grain matrix, the granular compression index (Cc-s) parameter can be utilized (Monkul and Ozden, 2005). The definition of granular compression index is very similar to the definition of compression index (Cc). The Cc-s is expressed as decrease of intergranular void ratio with effective stress, which is shown as below.
fines content, Leighton Buzzard Sand grains become more dispersed so that there is almost no grain contact between them. At this zone, the compressibility of the soil is expected to be mainly governed by the mica grain matrix. Therefore, the difference between granular compression and global compression index increases with an increase in mica content.
Ccs =
Δes Δ logσ ′
ð3Þ
Variation of global compression index (Cc) and granular compression index (Cc−s) with fines (mica) content is shown in Figs. 8 and 9. Cc and Cc-s values are calculated for 28 kPa and 124 kPa effective stresses. Compression indices of both sets of data increase relatively linearly with fines (mica) content. However, granular compression indices of the same set of data show a non-linear behaviour. Compression behaviour of the Leighton Buzzard Sand–mica mixtures can be examined in two stages shown by different zones in Figs. 8 and 9. The dashed lines in the figures show the transition fines contents (FCt) for the pressure ranges (Table 1). Through the zone left of the FCt, Leighton Buzzard Sand grain matrix can be thought to have almost a continuous force chain with grain to grain contacts, and mica grains are mostly located in the intergranular voids. Therefore, a less compressive behaviour is seen. However, initiating from the FCt, values, Leighton Buzzard Sand grain contacts start to decrease because of the filling mica. Therefore, a sharp increase in the slope of granular compression index curve can be observed. With further increase in
Fig. 8. Variation of compression parameters with fines content under 28 kPa effective stress.
3.2. Triaxial tests The interaction between coarser (i.e. Leighton Buzzard Sand) and finer (i.e., mica) grain matrices should also affect the overall strength behaviour. In order to investigate this effect and whether there is a relation with oedometer findings, a series of triaxial tests were conducted. As can be seen from the Fig. 10 denoting deviatoric stress versus strain, there is also a close relationship between transition fines content and the deviatoric stress of mixtures. The vertical dashed lines in Fig. 11 denote the transition fines content (mica) obtained from the oedometer tests for corresponding stress level (100 kPa). Immediately after from transition fines content (8%), the deviatoric stress of the mixtures start to keep a steady state condition for all strain levels. When the fines content of a mixture exceeds its transition content, intergranular void ratio of that mixture becomes greater than the maximum void ratio of the Leighton Buzzard Sand matrix itself. In other words, the infilling mica particles at intergranular voids separate the coarser grains and decrease their contact points. It is the mica particles that mainly control the soil behaviour, as there are no or very few contact between the Leighton Buzzard Sand grains.
Fig. 10. Stress–strain curve for clean loose sand and sand with different proportions of mica.
26
A.F. Cabalar / Engineering Geology 112 (2010) 21–28
Fig. 14. Undrained Young's modulus as a function of local axial strain. Fig. 11. Deviatoric stress change with fines (mica) content.
Therefore, a noticeable decrease in deviatoric stress over all the strain levels can be observed, when the mica content of mixtures exceeds FCt. As the strain level increases, the noticeable decrement trends for the deviatoric stress of the mixtures start to decrease markedly (Fig. 11). Fig. 12 shows the effects of increasing mica content on the pore water pressure–strain behaviour of the material. The clean loose sand specimen behaves as might be expected, but the dilation of 5% and more by weight of mica causes the suppression of any dilation, low undrained shear strengths, and high level of pore pressure generation during shear. Change in pore pressure with the FC over the various strain levels is shown in Fig. 13. The plot area is divided in to two
zones based on the oedometer test results, similar to Fig. 11, which are (i) the zone governed by the Leighton Buzzard Sand, and (ii) the zone governed by mica. From end of the first zone, where the fines content is around 8%, the pore pressure of the mixtures starts to keep a steady state condition for all strain levels. When the fines content of a mixture exceeds the transition zone, intergranular void ratio of that mixture becomes greater than the maximum void ratio of the Leighton Buzzard Sand matrix itself. Therefore, the infilling mica particles at intergranular voids separate the coarser grains and decrease their contact points. As the strain level increases, the noticeable increase trends for the pore pressure of the mixtures start to decrease markedly (Fig. 13). The influence of mica on the small strain stiffness was also investigated by considering undrained secant Young's modulus (Eu) (Fig. 14). Comparing the small strain stiffness behaviour of the micaLeighton Buzzard Sand mixtures at 100 kPa effective consolidation pressure, it is noted that the addition of mica to the Leighton Buzzard Sand particles resulted in reduced specimen stiffness. Fig. 15 shows the variation of Eu with fines content at different strain levels.
3.3. Resonant column tests The interaction between coarser (Leighton Buzzard Sand) and finer (mica) grain matrices should also affect the dynamic parameters of the mixtures. In order to inspect this affect and whether there is a relationship with the oedometer results, a serious of resonant column tests was performed. The Leighton Buzzard Sand and mica were mixed at various percentage of mica. Two mica percentages were considered; namely 5%, 10%, and then compared with the clean Fig. 12. Pore water pressure strain curve for clean loose sand and sand with different proportions of mica.
Fig. 13. Pore pressure generation change with fines (mica) content.
Fig. 15. Undrained Young's modulus change with fines (mica) content.
A.F. Cabalar / Engineering Geology 112 (2010) 21–28
27
Fig. 16. G/Gmax and D/Dmin strain curves for Leighton Buzzard Sand with different proportions of mica at 150 kPa effective consolidation stress.
Fig. 18. Shear modulus and damping ratio curves for Leighton Buzzard Sand with 5% mica at different effective stress values.
Leighton Buzzard Sand. Specimens were tested under the pressures of approximately 50, 100, 150 kPa. The effects of finer grains at different mix ratios on the smallstrain shear modulus, G/Gmax, and material damping ratio, D/Dmin are shown in Figs. 16–18. The general effect is that G/Gmax decreases and D/Dmin increases with mica content regardless of confining pressure. All reported values of G/Gmax and D/Dmin were measured at 50 kPa, 100 kPa, and 150 kPa effective stresses. The applied back pressure was kept constant at 300 kPa for all the tests. The influence of effective stress on G/Gmax and D/Dmin for the Leighton Buzzard Sand with 5% mica is shown in Fig. 18. It can be seen from the Fig. 18 that all the mixtures have shown an overall increase in G/Gmax and decrease in D/Dmin values as the effective stress increases. Figs. 16 and 17 show the effects of mica content on the smallstrain shear modulus and material damping ratio of the material, under isotropically-consolidated undrained triaxial compression at 150 and 100 kPa effective stresses, respectively. The dashed and solid lines in the figures show the test results for the data from the clean Leighton Buzzard Sand and the Leighton Buzzard Sand with mica at various mix ratios respectively. This kind of comparison clearly shows that the stress–strain response to torsional oscillation is getting softer as the mica content increases. It is seen that the greater the proportion of finer grain matrices (mica) in coarser sand (Leighton Buzzard Sand), the greater is the damping ratio, and the
lower is the shear modulus. Moreover, as an overall view within a given strain level, G decreases with an increase of the strain as D increases (Figs. 16–18). The response of soil with less mica exhibit more linear behaviour at smaller strain levels. As shearing increases, non-linearity in G/Gmax and D/Dmin values increase. The curves, which are quite close at very small strains, tend to diverge for increasing shear strain level. According to the study by Hardin and Blandford (1989), the maximum shear modulus is primarily dependent on the confining pressure. Fig. 18 shows the measured shear moduli at small strains as a function of effective stress while the pore pressure was kept at the same value (i.e., 300 kPa). The influence of effective stress on shear modulus is that the higher the confining pressure, the higher the shear modulus. Damping ratios calculated for the specimens tested at given shear strain amplitudes are also plotted in Fig. 18 as a function of effective stress for the Leighton Buzzard Sand with 5% mica. A trend for the damping ratio to decrease with increasing confining pressure is evident from Fig. 18. The results show the shear modulus remaining near the maximum value through 10− 2% strain before deviating significantly. Moreover, the damping ratio results show the damping ratio remaining about 1% throughout 7 × 10− 3% strain before increasing significantly. Fig. 19 shows the D/Dmin and G/Gmax with fines (mica) content for the specimen tested under 150 kPa effective stress. The FCt value for this effective stress was found to be 8.5% (Table 1).
Fig. 17. G/Gmax and D/Dmin strain curves for Leighton Buzzard Sand with different proportions of mica at 100 kPa effective consolidation stress.
Fig. 19. Changes in D/Dmin and G/Gmax with fine (mica) content for 150 kPa effective stress.
28
A.F. Cabalar / Engineering Geology 112 (2010) 21–28
4. Conclusions Engineering behaviour of micaceous sand was examined with the help of submatrices forming them (Leighton Buzzard Sand and mica). The effectiveness of the Leighton Buzzard Sand grain matrix on compressional behaviour was determined by oedometer tests by using intergranular void ratio instead of global void ratio. The mica content where the intergranular void ratio of the mixture and the maximum void ratio of the Leighton Buzzard Sand are equal to each other, has been termed the transition fines content. The transition fines content increases with increasing effective stresses. The 1-D compression behaviour of micaceous sands is mainly governed by the Leighton Buzzard Sand up to the transition fines content, whereas, it is governed by the mica grain matrix after the transition fines content. It is indicated that the granular compression index can be used as an important parameter to assess the role of the mica grain matrix on overall compression behaviour. Results of triaxial tests showed a close relationship between the transition fines content and the mechanical properties of mixtures. Exceeding a transition fines content value under the specified stress condition, the deviatoric stress, pore water pressure generation and undrained Young's modulus of the mixtures tend to be a steady state. Results of resonant column tests also showed a close relationship between the transition fines content and the dynamic properties of mixtures. Exceeding a transition fines content value for a given stress condition, the damping ratio and shear modulus of the mixtures tend to a steady state. This study investigated the transition fines content and its effect on 1-D compression behaviour of micaceous soils, and its effect on mechanical and dynamic properties. The concept of the transition fines content can be useful in various geotechnical and geoenvironmental engineering applications, where naturally sandy soils are treated, to estimate whether the fine or coarser matrix governs the overall engineering properties. It is believed that further macro and microfabric studies, and studies of the mineralogy of the mixture, chemical state of minerals, pore fluid properties, and gradation of the materials (both finer and coarser) are needed in order to better understand the arrangement of the Leighton Buzzard Sand and mica grain matrices with respect to each other, and to model the mechanical and dynamic properties of micaceous sands. Acknowledgements The author would like to thank Prof. C.R.I. Clayton, and Dr. J. Priest of the University of Southampton for their invaluable help.
References Cevik, A., 2007. Unified formulation for web crippling strength of cold-formed steel sheeting using stepwise regression. Journal of Steel Research 63, 1305–1316. Clayton, C.R.I., Theron, M., Vermeulen, N.J., 2004. The effect of particle shape on the behaviour of gold tailings. Advances in Geotechnical Engineering: The Skempton Conference. Thomas Telford, London, pp. 393–404. Fourie, A.B., Papageorgiou, G., 2001. Defining an appropriate steady state line for Merriespruit gold tailings. Canadian Geotechnical Journal 38, 695–706. Fourie, A.B., Blight, G.E., Papageorgiou, G., 2001. Static liquefaction as a possible explanation for the Merriespruit tailings dam failure. Canadian Geotechnical Journal 38, 707–719. Georgiannou, V.N., Burland, J.B., Hight, D.W., 1990. The undrained behaviour of clayey sands in triaxial compression and extension. Géotechnique 40 (3), 431–449. Gilboy, G., 1928. The compressibility of sand–mica mixtures. Proceedings of the A.S.C.E. 2, 555–568. Hardin, B.O., Blandford, G.E., 1989. Elasticity of particulate materials. Journal of Geotechnical Engineering, ASCE 115 (6), 788–805. Harris, W.G., Parker, J.C., Zelazny, L.W., 1984. Effects of mica content on the engineering properties of sand. Soil Science Society of America Journal 48 (3), 501–505. Hight, D.W., Georgiannou, V.N., Martin, P.L., Mundegar, A.K. 1998. Flow slides in micaceous sand. Problematic soils, Yanagisawa, E., Moroto, N. and Mitachi, T. eds., Balkema, Rotterdam, Sendai, Japan, 945–958. Lade, P.V., Liggio Jr., C.D., Yamamuro, J.A., 1998. Effects of nonplastic fines on minimum and maximum void ratios of sand. Geotechnical Testing Journal, ASTM 21 (4), 336–347. Lupini, J.F., Skinner, A.E., Vaughan, P.R., 1981. The drained residual strength of cohesive soils. Géotechnique 31 (2), 181–213. McCarthy, D.F., Leonard, R.J., 1963. Compaction and compression characteristics of micaceous fine sands and silts. Highway Research Record 22. Transportation Research Board, Washington, D.C, pp. 23–37. Monkul, M.M., Ozden, G., 2004. Intergranular and interfine void ratio concepts. 10th Turkish Congress on Soil Mechanics and Foundation Engineering /ZM10), TNCSMFE, Istanbul, pp. 3–12. in Turkish. Monkul, M.M., Ozden, G., 2005. Effect of intergranular void ratio on one-dimensional compression behaviour. Proceedings of International Conference on Problematic Soils, International Society of Soil Mechanics and Geotechnical Engineering, Famagusta, Turkish Republic of Northern Cyprus 3, 1203–1209. Monkul, M.M., Onal, O., 2006. A visual basic program for analyzing oedometer test results and evaluating intergranular void ratio. Computers and Geosciences, Elsevier Science 32, 696–703. Monkul, M.M., Ozden, G., 2007. Compression behavior of clayey sand and transition fines content. Engineering Geology 89, 195–205. Mundegar, A.K., 1997. An investigation into the effects of platy mica particles on the behaviour of sand. M.Sc, Thesis, Imperial College, London. Naeini, S.A., Baziar, M.H., 2004. Effect of fines content on steady state strength of mixed and layered samples of sand. Journal of Soil Dynamics and Earthquake Engineering 24, 181–187. Olson, R.E., Mesri, G., 1970. Mechanisms controlling the compressibility of clay. Journal of the Soil Mechanics and Foundations Division, A.S.C.E., 96 (SM6), Proc. Paper 7649, November, pp. 1863–1878. Salgado, R., Bandini, P., Karim, A., 2000. Shear strength and stiffness of silty sand. Journal of Geotechnical and Geoenvironmental Engineering, ASCE 126 (5), 451–462. Rawlings, J.O., 1998. Applied regression analysis: a research tool. Springer-Verlag, New York. Terzaghi, K., 1925. Erdbaumechanik auf bodenphysikalischer grundlage. Deuticke, Leipzig/Vienna. Theron, M. 2004. The Effect of Particle Shape on the Behaviour of Gold Tailings. PhD thesis, University of Southampton, U.K.