新莊煤礦1.5 Mta新井設(shè)計含5張CAD圖-采礦工程.zip
新莊煤礦1.5 Mta新井設(shè)計含5張CAD圖-采礦工程.zip,新莊煤礦1.5,Mta新井設(shè)計含5張CAD圖-采礦工程,新莊,煤礦,1.5,Mta,設(shè)計,CAD,采礦工程
英文原文
Effect of strain rate on the mechanical properties of salt rock
1. Introduction
Salt rock is a special material in rock engineering. With itsphysical properties of tight fabric and low permeability, as well as itsmechanical properties of low strength and ductility, it has become afavored medium for waste disposal and oil and gas storage sinceearly in the last century, when it began to be thoroughly studied.More recently, achievements documented in the six conferences on the mechanical behavior of salt since 1984 and other publicationshave greatly increased knowledge of salt rock behavior.
As the strength of salt rock is somewhat low, generally the loading rate for compression tests is also slow. However, loading strain rate (e ) affects the mechanical properties of the salt rock.Studies have been carried out on many different rock types examining the effect of _e on rock strength and deformation characteristics. For example,e effects on dynamic tensile strength of Inada granite and Tage tuff were studied by Cho et al. [1]; loading e values for the granite and the tuff were 4.24–13.18 and 0.46–6.82/s, respectively. It was found that the dynamic tensile strength of the two rock types increased rapidly with e.
After studying the porosity change of Mugla marble by heat treatment, compression tests under e values of 210- 5–5 10-7/s were conducted by Mahmutoglu [2]. It was found that the compressive strength decreased sharply with strain rate decrease. For dry specimens, the percentage of the strength decrease is about 44% whereas the relative strength decrease for saturated specimens is much larger. It was suggested by Qi et al. [3] that the deformation and failure of rock under low strain rates were controlled mainly by thermally activated mechanisms. With strain rate increases, a phonon damping mechanism appears and gradually plays a dominant role in the deformation process. Li and Wang [4] studied the fracture toughness of marble under high loading rates using the Hopkinson bar method and found that the toughness increased significantly with the loading rate. Yang et al. [5] carried out tests on limestone under different loading rates and found that the peak strength increased with the loading rate; at the same time, the strain at peak strength increased linearly with strain rate. He also found that the failure mode of the rock seldom changed with the loading rate.
There are also many other studies [6–13] of the strain rate effect on mechanical properties of rocks; however, reports of strain rate effects on mechanical properties of salt rock are less well documented. To investigate this in the laboratory, we conducted uniaxial compression tests on salt rock (including halite and thenardite) under loading e values of 2 10- 5,2 10- 4, and 2 10- 3/s. These results were generated as part of a more general investigation of the performance of gas storage operations in salt caverns.
2. Sampling and methodology
2.1. Samples
In this paper, ‘‘rock salt’’ refers to rock with a dominant mineral component of halite (NaCl), whereas ‘‘salt rock’’ is a more general designation for all rocks with a dominant mineral component of highly soluble salts generally of evaporitic origin including NaCl, KCl, carnallite, bischofite, tachyhydrite, thenar-dite, mirabilite, glauberite, anhydrite, and so on.
Samples were cored from two evaporite deposits in Jiangsu Province, China. The evaporites are saline lake sediments of Tertiary age. One type of lithology is rock salt (halite), with the major deposit located 900–1100 m deep; the other salt rock lithology is dominated by the mineral thenardite (Na2SO4—sodium sulfate), and is buried about 2000 m deep. Both are of evaporative crystallization origin and have the desired properties of fabric tightness and low permeability for fluid storage or waste disposal. Some silt fillings were occasionally found in the rock salt samples, and the intergranular insoluble silt content is less than 10% in weight; however, the thenardite samples were pure and homogeneous. Samples were transported to the Taiyuan University of Technology (TYUT) laboratory after careful packaging.
In the laboratory, samples were processed according to ISRM suggested methods. Nine cylindrical specimens with aspect ratios of 2:1 were eventually prepared, eight of which were successfully tested, six rock salt specimens and two thenardite specimens. For the six rock salt specimens, uniaxial compression tests with three different loading rates were conducted while the two thenardite specimens were tested with two lower loading rates. Generally, 3–5 specimens are suggested to be prepared for each test to guarantee reliability, but we were hampered by a lack of acceptable quality core. To compare loading rate differences with the limited specimen set, only two specimens for each test could realistically be assigned. The reason we nevertheless choose to present our test results here is that they were found to be reasonably consistent and hence are likely to be reliable.
2.2. Test and equipment
The main purpose of the test was to study the loading rate effect on mechanical properties of salt rock during uniaxial compression. All the experiments were performed on servo-controlled mechan-ical test equipment TYT-600 (Fig. 1), manufactured by Jilin Jinli Test Technology Co. Ltd (Changchun, China). The maximum load capacity of the frame is 6.0 10 4 kg, the minimum displacement rate is 1.0 10 - 3 mm/s and loading rates can be varied with the servo-control system.
2.3. Methodology
For the experiment, three uniaxial compression displacement rates were chosen, leading to e values of 2.0 10 -5 , 2.0 10 - 4 , and 2.0 10 - 3 /s, a factor of 100. As mentioned before, two each of the six rock salt specimens were assigned to be tested under one loading strain rate, and the two thenardite specimens were tested under the two lower strain rates of 2.0 10- 5 and 2.0 10 - 4 /s. Following ISRM suggested methods [14] , the uniaxial compression tests for each strain rate were carried out on the same mechanical test equipment.
3. Experimental results
The uniaxial compression test results with different strain rates are listed in Table 1; Figs. 2–6 are the resulting stress–strain (s–e ) curves from the specimens during the tests. It was found that the s–e curves are ‘‘comb-shaped’’ and that stress fluctuations in the rock salt were observed at the lowest e of 2.0 10 - 5 /s. The two s–e curves for the specimens tested at this lowest e value are shown in Figs. 2 and 3 .
3.1. Relationship between strength and strain rate
From Table 1 and the s–e curves, it is found that the peak strength of rock salt is little affected by the strain rate (within the chosen range for e ). Average UCS values are 13.6, 13.9 and 12.4 MPa, respectively, for _e values of 2.010 - 5 to 2.0 10 -4 and 2.010 - 3 /s. Also, the UCS of the tested rock salt is consistently lower than that of ten natural rock salts in the U.S.A. [15] .
We believe the reason for the lower strength is that the Jiangsu salt deposits were formed in a different style and at a different geologic time. Most salt deposits in North America were formed in the Jurassic, Permian or Devonian periods; some bedded salts have been recrystallized; and, salt from salt domes has undergone extensive recrystallization, deformation, and porosity reduction (domal salts are less porous than bedded salts, and older salts are less porous than younger salts in general). The salt deposits in Jiangsu, China were salt lake deposits formed in the Tertiary (Eocene) and have not undergone repeated recrystallization or large scale creep deformation associated with tectonic mobilization or diapirism. All things being equal, salt from a younger deposit should be somewhat weaker and less rigid than salt from an older deposit. However, the strain rate effect on the salt strength is similar to that reported by Farmer and Gilbert [16] at e values of 5.0 10 - 3 and 2.0 10 -4 /s, where they also report that the s–e curves for their rock salt under confining stresses of 3.5, 7, 14, 35, 42 MPa were quite similar. This is an aspect of rock salt behavior that is substantially different from brittle rocks such as sandstone, marble, and granite over similar strain rates.
For example, Ray et al. [10] found that the UCS of Chunar sandstone increased from 64.0 to 75.1 and 99.5 MPa with e increases from 2.5 10- 1, to 2.5 10 0 and to 2.5 10 1 /s. Fukui et al.[11] also reported that shear strength of Sanjome andesite is dependent on loading rate; if the loading rate was increased by an order of magnitude, the apparent cohesion increased by 6.1%, and this ratio was about the same for the UCS, uniaxial tensile strength, indirect tensile strength, and fracture toughness. Similarly, Jeong et al. [12] note that the UCS of Kumamoto andesite increased linearly with decreasing water vapor pressure for the same _e and with increasing _e in the range 0.95 10 - 6–1.91 10 - 4 /s. Li et al. [13] conducted dynamic triaxial compression tests for the Bukit Timah granite with four strain rates (10 - 4 /s, 10 - 3 /s, 10 - 1 /s, 10 0 /s) and six confining pressures. It was concluded that the compressive strength generally increases with increasing strain rate and confining stress. Also, the rate of increase of compressive strength with strain rate is lower at higher confining pressure.
The e -effect can be explained by energy theory. Damage of material is the result of development of through-going micro-fractures, and during a short but rapid loading process with a large e , the development of micro-fractures in the material lags the increment of loading. This means that the response and deformation speed of grains in the material is slower than the faster loading rate, just as we note with deformation hysteresis when we compare rock response with the rate of load decrease during the unloading process of a compressed material. Because of this ‘‘hysteresis’’, the energy absorbed by the material cannot be consumed or released fully in the short time span by development of micro-fractures, and it is temporarily stored as material compression, thus the strength of the material is somewhat enhanced.
In comparison to igneous rocks and many other sedimentary rocks, salt rock is classified as a ‘‘soft rock’’, and it possesses characteristic mechanical properties of low strength and strong rheology. Many studies [18–22] demonstrate that the UCS of salt rock is in range 12.0–20.0 MPa, and the rheology of salt rock typically arises as a combination of elastic and viscoplastic deformation [23–27] , the ratio of which is a function of the loading rate and confining stress. Large viscous strains (deforma-tions) can be achieved with long load times, even under the action of relatively modest deviatoric stresses. This large viscous strain includes axial and lateral deformations such that under slow deformation conditions the material behaves incompressibly (non-dilatant). In the non-dilatant viscoplastic strain rate regime, the Poisson’s ratio of rock salt is essentially 0.5, almost an ideal plastic deformation condition with a zero plasticity angle. Of course, at higher strain rates, salt displays dilatancy, the magnitude of which is also a function of e and the confining stress.
In any case, the UCS of salt rock is not enhanced by an increasing strain rate as much as for brittle rocks. The energy accumulated in salt rock during rapid compression during our experiments is absorbed and consumed by its strong rheological deformation (i.e. energy dissipation or relaxation mechanisms), and the pseudo-confining stress effect noted in brittle rock is absent or weak. This seems to be a rational interpretation for the near-invariability of rock salt strength we observed with a strain rate increase (limited to our range of e ). There are other factors which also lead to the different mechanical response of salt rock compared to brittle rocks in the rapid loading domain. For example, effects of mineralogy, mineral grain size, and mineral fabric differences arise in siliceous and igneous rocks, and though important, these additional factors are not discussed extensively herein.
The elastic modulus of the rock salt specimens is calculated to be 2.6, 3.0, and 3.4 GPa, respectively, with _e increasing from 2.0 10 - 5 to 2.0 10 - 4 and 2.0 10 - 3 /s. The elastic modulus seems to show a trend of slow increase with strain rate. This trend is similar to that of Three Gorges granite [17] , for example.
The UCS of thenardite is somewhat larger than that of rock salt under the same strain rate conditions. Apart from the difference of crystal lattice arrangement, in our opinion it is mainly the consequence of the grain size difference of the two crystallites. Fig. 7 shows damaged specimens of halite and thenardite where we can see that the halite grain size is - 1 - 10 mm, whereas the thenardite grain size is 1 - 3 mm. The smaller the grain size, the more difficult it is to propagate planar intercrystalline micro-fractures to achieve the critical fracture size associated with peak strength. In addition, the two thenardite specimens in the experiment are pure and contain no impurities, but the rock salt specimens contain silt at the grain boundaries. The different stiffness of the impurity represented by the silt will act as a stress concentrator for shear, and the weaker bonding between the silt and the rock salt makes tensile parting easier; these are important factors affecting the UCS. Finally, the thenardite was buried 1000 m deeper than the rock salt, and was thus more compressed (i.e. lower porosity) and hence stronger.
3.2. Relationship between deformation/failure and strain rate
The strain rate also has an evident relationship to the strain-to-failure of the rock salt: with an increasing e value, the strain at peak strength is less. For example, the strain-to-failure is 1.3–1.7% (specimens # 1, # 2) when _e ? 2:0 1 - 5 =s; however, it decreases to 0.3–0.7% (specimens # 5, # 6) when _e is increased to 2.0 10 - 3 /s. If we define a normalized secant deformation modulus 0 as the value of axial strain at the point of peak strength divided by the peak strength, then on average E 0 ? 0.93, 1.50, and 1.92 GPa for the three progressively more rapid _e values. Obviously, the deformation modulus is increasing with _e in a manner similar to the elastic modulus. However, the deformation modulus increment with strain rate increase is larger than that of the elastic modulus. The deformation modulus increases by 52% and 28% with _e increasing from 2.0 10 - 5 to 2.0 10 - 3 /s, whereas elastic modulus increases by 15% and 14%, respectively.
The difference is because of plastic deformation of the specimen; the more rapid the strain rate, the less the plastic deformation compared to elastic deformation. Empirically, using a logarithmic expression to link the strain rate and deformation modulus,E 0 ? 0 :2ln _e t 3:2. This equation is sketched in Fig. 8 , giving a reasonable fit over the _e range used.
Poisson’s ratio relationships are less clear; n is found to decrease with strain rate with the exceptions of specimens # 1 and # 2 which were tested at _e ? 2:0 10 - 5 =s. On average, n in the rock salt specimens decreases from 0.26 to 0.16 as e increases from 2.0 10 - 4 to 2.0 10 - 3 /s, and a similar trend is found for thenardite ( n decreases from 0.30 to 0.15 as e increases from 2.0 10 - 5 to 2.0 10 - 3 /s). As these are low values compared to the literature (0.30–0.40), they likely reflect crack closure processes, always an important issue in polycrystalline rocks tested at low confining stress. Also, with lower e values, there is more time for viscoplastic processes to act, giving more lateral deformation, thus a higher Poisson’s ratio (as stated previously, at very low _e values and under some confining stress, salt is non-dilatant as it undergoes viscous deformation).
The failure mode of the specimens was found to be consistent over the strain rate range (similar to results in Ref. [28] ). The failure of the rock salt is a clear brittle fracture process along grain boundaries with little shear; for thenardite, shear develops along a relatively planar surface (Fig. 7 ). The failure style difference is because of different rock fabrics; the grain size of the rock sal crystallite is larger than that of the thenardite, and the presence of intercrystalline silt in the rock salt weakens the bonding between adjacent halite crystals and tensile parting can happen more easily in orientations close to 901 to s 3. Compared with the homogeneous thenardite with small pure grains, impure rock salt breaks much more easily with a tensile brittle failure demon-strated as the development of approximately columnar surfaces in the unconfined condition.
3.3. Volume dilation stress during compression
Generally, the compressive s –e curve of these rocks can be divided into five stages ( Fig. 9 ). Stage I is closing of pre-existing microcracks or pore space oriented at suitable angles to the applied stress. This stage is less obvious for salt rock because of its low porosity, but there is usually intercrystalline extensional damage from the sampling and de-stressing process. Stage II is elastic and largely recoverable deformation.
Stage III is characterized by the onset of dilation and by a near-linear increase in volume, which is offset by the continuing compression. It is supposed that microcrack propagation occurs in a stable manner during this stage. Its upper boundary is the point of maximum compaction and zero volume change, and the stress at the point of zero volume change has been called the critical stress for volume dilation; it is about 80% of peak stress for most rock specimens. If the stress increases continuously after the critical stress is passed, stage IV, characterized by positive dilation, takes place with an acceleration of microcracking, and the internal structure is rendered more porous and debonded until stage V, the cessation of dilation occurs. What we will examine here is the critical stress for volume dilation of the specimen under different strain rates.
It can be shown from diagrams of volumetric strain versus stress (Figs. 10 and 11 ) that the critical stress of volume dilation is different for the two groups of specimens which were compressed under _e ? 10 - 4 and 10 - 3 =s, respectively. For the specimens compressed under _e ? 10 -4 =s, the critical stress for volume dilation is 9.6 and 8.6 MPa for specimens # 3 and # 4, about 63% and 69% of their peak stress. However, this increases to 94% and 96% for specimens # 5 and # 6 when e increases to 10 - 3 /s; obviously, the stress triggering fast microcrack development is less under a low strain rate than at a high strain rate. This may be explained by the deformation or response hysteresis effect at a rapid strain rate discussed previously. It can be concluded that the higher the strain rate, the larger the critical stress needed to dilate the rock salt.
4. ‘‘Stress fluctuation’’ phenomenon under low strain rate
During the uniaxial compression test for halite at e ? 2: 0 10 - 5 =s, a ‘‘stress fluctuation’’ phenomenon was observed. When the loading stress reaches about 8.0 MPa ( Figs. 2 and 3), about 60% of the peak strength, plastic deformation began in the specimen, defined as the threshold value of plastic deformation. When the loading stress rose to about 10.0 MPa, about 75% of the peak strength, the stress suddenly dropped to a value around the threshold of the plastic deformation and then recovered to the original stress or slightly higher; this ‘‘stress fluctuation’’ phenomenon persisted till the specimen underwent strain-weakening.
This phenomenon was also noted in Ref. [29] during mechan-ical tests on specimens of salt rock comprised of laminated halite and non-salt interlayers. At e ? 1 10 - 4 =s
收藏