Mechanical properties of rock materials with related to mineralogical characteristics and grain size through experimental investigation: a comprehensive review

Abstract

Mechanical properties of rock materials are related to textural characteristics. The relationships between mechanical properties and textural characteristics have been extensively investigated for differently types of rocks through experimental tests. Based on the experimental test data, single- and multiple- variant regression analyses are conducted among mechanical properties and textural characteristics. Textural characteristics of rock materials are influenced by the following factors: mineral composition, size, shape, and spatial distribution of mineral grains, porosity, and inherent microcracks. This study focuses on the first two: mineral composition and grain size. ƒ

This study comprehensively summarizes the regression equations between mechanical properties and mineral content and the regression equations between mechanical properties and grain size. Further research directions are suggested at the end of this study.

Keywords

Mechanical properties / rock material / texture / mineral characteristics

Introduction

Rock materials are widely used in a large amount of structures in critical infrastructure systems, such as highways, tunnels, critical buildings, dams, etc. Rock materials will significantly affect mechanical properties as the other infrastructure construction materials [1–3]. According to the United States Geological Survey (USGS) [4], various rocks were extracted and crushed into aggregates worth USD11 billion in 2013 in the United States. Mechanical properties of rocks materials have a great influence on service life, reliability, and resilience of the critical infrastructure. Research has shown that mechanical properties of rock materials are closely related to textural characteristics [5–7].

Texture characteristics of rocks are influenced by the following six factors: mineral composition, size, shape, and spatial distribution of mineral grains, porosity, and inherent microcracks [6,8]. Among the aforementioned six factors, this study focuses on the first two: mineral composition and grain size. How mineral composition and grain size influence mechanical properties for rocks through experimental tests is summarized into different regression equations. In experimental tests, mechanical properties of rock materials are measured through laboratory mechanical tests in terms of unconfined compressive strength, tensile strength, Los Angeles (LA) values, and impact value, etc. The mineral composition of rock materials can be determined from scanning electron microscope (SEM) and energy dispersive X-ray (EDX) microanalyzer [9,10]. The size of mineral grain can be measured through microscopic observations and image analysis. Based on the experimental data, the relationships between mechanical properties and mineral properties/grain size are analyzed through single- or multiple- variant regression analysis. Regression equations are summarized. With the regression equations, we can better understand the correlation between mechanical properties and grain size and mineral content quantitatively. These research findings can help to select appropriate rock materials and improve quality control in aggregate production based on textural characteristics according to specific engineering needs.

The aims of this study are as follows: (1) to comprehensively summarize regression equations between mechanical properties and mineral characteristics and regression equations between mechanical properties and grain size; (2) to inspire future research to further investigate on the influence of texture characteristics on mechanical properties for different rocks.

Relationship between mechanical properties and textural characteristics

Increasing attention has been paid to the relationship between mechanical properties and textural characteristics of rock materials. The mechanical properties of aggregates are functions of textural characteristics. The textural characteristics can be represented by mineral characteristics, and the type, size, shape, orientation, interlocking, distribution of mineral grains. The following section discusses how the mechanical properties of rock materials are influenced by mineral characteristics and sizes of mineral grains, respectively.

Mineralogical characteristics

Mineralogical properties have a great influence on mechanical properties of aggregates. Mendes et al. [28] proposed to use mineralogical data to evaluate the quality and mechanical properties of rocks and validated the correlation between mechanical properties and mineralogical properties for granite samples. Since then, there have been many experimental studies investigating the relationship between mechanical properties and mineral content, in terms of mineral content and ratio of dominant minerals.

Table 1 shows the regression equations between mechanical properties and mineral content. In 1979, Hugman and Friedman [11] published their study on compressive test and mineralogical analysis for carbonate rocks (Yule marble, Solenhofen limestone, Hasmark dolomite, and Blair dolomite) with intermediate dolomite and micrite content, and predicted the approximate ultimate strength by the best-fit plane. At a given confining pressure, the ultimate strength of rocks increases with the increase of content of dolomite and microcrystalline carbonate. Brattli conducted laboratory experiments to investigate the influence of minerals on the resistances to fragmentation and abrasion for igneous rocks. Regression analysis showed that the content of feldspar, mica, and amphibole have positive correlations to impact value, abrasion value, and durability value, whereas pyroxene has negative correlations to impact value, abrasion value, and durability value [12]. The feldspar content has a greater influece on impact value than on abrasion value, probably because relatively weak bonding along the cleavage planes in feldspar tend to have a stronger negative effect on brittleness than on abrasion resistance. Lundqvist and Göransson evaluated mechanical properties of Precambrian rocks from Stockholm, Sweden with correlation to mineral content. They found that mica content has a distinct correlation with the abrasion resistance, and iron and magnesium rich aggregates tend to have better resistance to fragmentation and abrasion [13]. Räisänen found the hornblende content has a positive impact on LA values, meaning that aggregates with greater hornblende content are more likely to have less polishing resistance [14]. Miskovsky et al. found that impact value is linearly related to feldspar content and mica content [15]. The abrasion value linearly decreases with the increase of feldspar content and quartz content. These equations validate that rocks with more feldspar and quartz are stronger, having better resistance to abrasion.

Currently, there are conflicting opinions on the relationship between mechanical properties and mineralogical properties of aggregates. Merriam et al. found a definite relationship between quartz content and tensile strength [18]. The strong correlation between quartz content and compressive strength is validated by many researchers [19,20]. However, some other researchers argued that there is no significant relationship between quartz content and strength for sandstones [21–23]. Similar conflicting findings are also reported for mica. For coarse granite aggregates from Swedish, Miskovsky et al. found that a high content of either quartz or feldspar will lead to small abrasion values and a high content of mica will result in great abrasion values, indicating aggregates with more quartz or feldspar and less mica have greater abrasion resistances [15]. However, Åkesson et al. argued that mica itself does not directly influence fragmentation resistance, but mica that forms in plane foliation will interact as large flaws and lead to crack propagation [24].

In addition, relationship between mechanical properties and quartz to feldspar ratio (QFR) are investigated, shown in Table 2. Both uniaxial compressive strength and tensile strength linearly increase with QFR for granitic rocks [20]. Yusof and Zabidi also considered that rocks with greater QFR tend to have greater strength values [25]. However, Sousa considered that there is no obvious proportional relationship between unconfined compressive strength and QFR or quartz content, and it is likely that strength tend to decrease with quartz content [26].

Grain size

Previous studies show that grain size has a great influence on mechanical properties. The relationships between mechanical properties and mean grain size have been extensively investigated through laboratory experiments. Correlations between mechanical properties and mean grain size are found. Simply speaking, mechanical properties decrease with the increase of mean grain size, especially for rocks with the grain size smaller than 1 mm.

Since 1960s, Brace found that rocks with finer mineral grains have better mechanical strengths, indicating that grain size influences mechanical properties [27]. Mendes et al. found that mineralogical properties have a good correlation with mechanical properties of granite samples, and samples with finer grains have greater strengths [28]. Willard and McWilliams reported that the ultimate strength of a rock and the direction of crack propagation are influenced by mineral cleavage, micro-fracture, and grain boundaries [29]. After that, a lot of studies investigate the correlation between mechanical properties and grain size through experiments. Based on experimental results, regression equations are determined to better predict mechanical performance of rock materials according to the grain size consisting minerals. Regression equations mainly fall into four categories: linear, square root, logarithm, and exponential, shown in Table 3.

Linear regression equations are established between grain size and mechanical properties, such as unconfined compressive strength and Los Angles (LA) value. Hugman and Friedman found that ultimate strength is correlated with the weight mean grain size in carbonate [11]. Weighted mean grains size takes into account the grain size of consisting minerals. Onodera and Asoka reported that the strength of granite rocks linearly decreases with the increase of grain size [34]. Tugrul and Zarif reported linear relationships between uniaxial compressive strength and the grain size of different minerals for granite rocks in Turkey [20]. French et al. found a linear relationship between mean grain size and aggregate impact value, and the aggregate impact value increases with mean grain size and decreases with normative hardness [35]. Lundqvist and Göranssion found that the resistances to both wearing and impact forces improve with the decrease of mean grain size for Precambian rocks in Stockholm, Sweden [13]. Räisänen found that both LA value and abrasion value linearly depend on mean grain size, and abrasion values has better correlations with grain size than LA value [14]. The reason is that finer mineral grains have smaller surfaces [36], and that the length-to-width ratio of micro-cracks decreases with the decreases of grain size [12]. Therefore, aggregates with finer mineral grains are more likely to have smooth texture and better mechanical properties.

The other types of regression equations are inverse square root, logarithm, and exponential. Olsson found that the yield stress of marble linearly increases with the inverse square root of mean grain size for marble [31]. Later Wong et al. validated the inverse square root relationship for Yuen Long marble [30]. On the other hand, based on laboratory experiments of granite rocks, Prikryl found a logarithm relationship between uniaxial compressive strength and the mean grain size of consisting minerals for granite [32]. Hareland et al. reported an exponential relationship between confined compressive strength and grain size for shales [33].

Although there are obvious correlations between mechanical properties and the size of mineral grains, there are some researchers arguing that stresses may not strongly depend on grain size. Hatzor et al. found that crack initiations and peak stresses have a very weak dependence on the mean grain size and no dependence on the maximum grain size [37]. Through laboratory experiments on crystalline rocks with similar minerals but different grain sizes, Eberhardt et al. found that at the initial stages of cracking, the feldspar and quartz mineralogy play the most critical role, whereas the grain size effect plays a secondary role [38]. This ongoing debate can help to motivate more experimental investigations and numerical simulations to figure out how the grain size influence on mechanical properties for different rocks.

Texture characteristics

Texture characteristics of rock materials depend on the geometric relationship of mineral grains and matrix [7]. Rocks consisting of hard minerals are more likely to have rough surface texture higher strength properties [1]. Rock texture characteristics can be described by one index, such as texture coefficient, micropetrographic index and geomechanical order. Irfan and Dearman found that there are good correlations between micropetrographic index and geomechanical properties of granitic rocks [39]. Geomechanical order was defined by Hecht et al. [40] as a function of microstructural properties, such as grain size, grain shape, packing density, compositional order and cementation grade. Hecht et al. established the correlation between mechanical properties and geomechanical order for coarse grained sedimentary rocks (sandstones and conglomerates) of Permocarboniferous age [40].

Texture coefficient has widely used to quantify rock texture characteristics. Texture coefficient was defined by Howarth and Rowlands to comprehensively consider influencing factors on morphological characteristics, including grain size, grain shape, grain orientation, porosity and matrix materials [8,41], by the following equation.

TC=AW\times [\left({\displaystyle \frac{{{\displaystyle N}}_{\mathrm{0}}}{{{\displaystyle N}}_{\mathrm{0}}+{{\displaystyle N}}_{\mathrm{1}}}}\times {\displaystyle \frac{\mathrm{1}}{F{{\displaystyle F}}_{\mathrm{0}}}}\right)

+\left({\displaystyle \frac{{{\displaystyle N}}_{\mathrm{0}}}{{{\displaystyle N}}_{\mathrm{0}}+{{\displaystyle N}}_{\mathrm{1}}}}\times {{\displaystyle AR}}_{\mathrm{1}}\times {{\displaystyle AF}}_{\mathrm{1}}\right)

where TC is texture coefficient; AW is grain packing weighting; {{\displaystyle N}}_{\mathrm{0}} is the number of grains whose aspect ratio is below a pre-set discrimination level; {{\displaystyle N}}_{\mathrm{1}} is the number of grains whose aspect ratio is above a pre-set discrimination level; {{\displaystyle FF}}_{\mathrm{0}} the is arithmetic mean of discriminated form factors; {{\displaystyle AR}}_{\mathrm{1}} is the arithmetic mean of discriminated aspect ratios; and {{\displaystyle AF}}_{\mathrm{1}} is the angle factor to quantify grain orientation.

Table 4 shows the linear regression equations between unconfined compressive strength and texture coefficient of rocks. As shown in this table, unconfined compressive strength linearly increases with the increase of texture coefficient for various types of rocks except fault breccia. Ersoy and Waller found a positive correlation coefficient of 0.621 between unconfined compressive strength and texture coefficient for various sedimentary and igneous rock (limestone, sandstone, siltstone, granite, and diorite) [42]. Ozturk et al. validated such a strong correlation for sandstone, siltstone, marl, shale, and limestone [7].

Although most studies found that unconfined compressive strength linearly increases with the increase of texture coefficient for different types of rocks [7,42–45]. However, there are contradictory correlations between TC and strength [46]. Possible reasons for the conflicting correlations might be as follows: complicated texture characteristics of rocks may not be able well represented by one index as texture coefficient; fault breccia consisting of broken fragment of mineral grains filling with a fine-grained matrix may show different mineral characteristics from the other type or rocks, such as sandstone, limestone, granite, etc., resulting in different relationship between unconfined compressive strength and grain size.

Validation and application

Effective analyses on influence of mineral characteristics and grain size on mechanical properties of rocks materials require sufficient up-to-date data and comprehensive correlation analysis of the data sets. There are similarities and overlaps on the types and mechanical properties of rocks in current studies. However, these studies are lack of validating by sufficient data for validation among rocks of different types and origins, and there is not enough fundamental mechanical explanation from the view of consisting materials at microscale. Without appropriate validations and fundamental explanations, we may not be confident to directly apply the regression equations collected from different studies into practice, i.e., prediction of mechanical properties from textural characteristics.

To validate the influence of mineral characteristics and grain size, as well as other texture characteristics on mechanical properties of various rocks, a large amount of experimental tests should be further conducted on various types of rock materials from different origins. The texture characteristics should be well identified. And then the correlation between mechanical properties and textural characteristics should be investigated not only limited to statistical one- or multi- variable regression analysis. More laboratory experiments should be conducted on rocks of different materials from different origins. In addition, the relationships between mechanical properties and textural characteristics should be investigated at microscale and atomic-scale through both experimental test and numerical simulations.

To support decision-making on the selection of rock materials and quality control of aggregate production, further research should be conducted in the following aspects: databases, guidelines, and decision-making tools. The databases should be comprehensively collect mechanical properties of different rocks from various origins and the corresponding textural characteristics. The guidelines should suggest measurement methods and correlation analysis methods. The decision-making tools should developed to apply the research findings to making more accurate predictions of mechanical properties from textural characteristics and grain size.

Conclusions

Analyses on the relationships between mechanical properties and mineral characteristics and grain size can identify critical mineral components, and optimize the selection of rock materials. This study presents the state-of-the-art of the influence of mineral characteristics and grain size on mechanical properties of rock materials. It gives a good introduction to researchers who are new to this field. Even though there are a lot of research on regression analysis between experimental data sets of mechanical properties and textural characteristics, there are still many challenges left. Additional efforts should be made the following aspects: (1) develop models and equations to reflect the comprehensive influencing effects of textural characteristics on mechanical properties through both experimental test and numerical simulations, (2) validate these models and equations for various rocks from different origins, and (3) further apply these findings to more accurately predict mechanical properties.

References

[1] Räisänen M ,  Kupiainen K ,  Tervahattu H . The effect of mineralogy, texture and mechanical properties of anti-skid and asphalt aggregates on urban dust. Bulletin of Engineering Geology and the Environment, 2003, 62(4): 359–368

[2] Hou Y, Wang  L, Yue P ,  Pauli T ,  Sun W. Modeling Mode I Cracking Failure in Asphalt Binder by Using Nonconserved Phase-Field Model. Journal of Materials in Civil Engineering, 2014, 26(4): 684–691

[3] HouY, SunW, HuangY, Ayatollahi MR, WangL , ZhangJ. Diffuse Interface Model to Investigate the Asphalt Concrete Cracking Subjected to Shear Loading at Low Temperature.Journal of Cold Regions Engineering, 2016, 31(3): 04016009.

[4] United States Geological Survey (USGS). USGS Minerals Information: Crushed Stone. Retrieved 2014-8-11.

[5] Little D, Button  J, Jayawickrama P,  Solaimanian M , and  Hudson B.  Uantify shape, angularity and surface texture of aggregates using image nalaysis and study their effect on performance. FHWA/TX-06/0-1707-4, 2003

[6] Liu H, Kou  S, Lindqvist P.-A. ,  Lindqvist J.E. , and  Akesson U . Microscope rock texture characterization and simulation of rock aggregate properties. SGU project 60-1362/2004, 2005

[7] Ozturk C A, Nasuf  E, Kahraman S . Estimation of rock strength from quantitative assessment of rock texture. Journal of the South African Institute of Mining and Metallurgy, 2014, 114: 471–480

[8] Howarth D F, Rowlands  J C. Development of an index to quantify rock texture for qualitative assessment of intact rock properties. Geotechnical Testing Journal, 1986, 9(4): 169–179

[9] SunW, WeiY, WangD, Wang L. Review of Multiscale Characterization Techniques and Multiscale Modeling Methods for Cement Concrete: From Atomistic to Continuum.Multi-Scale Modeling and Characterization of Infrastructure Materials, 2013, 8: 325–341.

[10] YusofNQAM, ZabidiH. Correlation of Mineralogical and Textural Characteristics with Engineering Properties of Granitic Rock from Hulu Langat, Selangor.Procedia Chemistry, 2016, 19: 975–980.

[11] Hugman R H H ,  Friedman M . Effect of texture and composition on mechanical behaviour of experimentally deformed carbonate rocks. American Association of Petroleum Geologists Bulletin, 1979, 63(9): 1478–1489

[12] Brattli B. The influence of geological factors on the mechanical properties of basic igneous rocks used as road surface aggregates. Engineering Geology, 1992, 33(1): 31–44

[13] Lundqvist S, Göransson  M. Evaluation and interpretation of microscopic parameters vs. mechanical properties of Precambrian rocks from the Stockholm region, Sweden. Proceedings of the 8th Euroseminar Applied to Building Materials, Athens, 13–20, 2001

[14] Räisänen M . Relationships between texture and mechanical properties of hybrid rocks from the Jaala–Iitti complex, southeastern Finland. Engineering Geology, 2004, 74(3-4): 197–211

[15] Miskovsky K, Duarte  M K, Kou  S Q, Lindqvist  P A. Influence of the mineralogical composition and textural properties on the quality of coarse aggregates. Journal of Materials Engineering and Performance, 2004, 13(2): 144–150

[16] Erichsen E, Ulvik  A, Wolden K ,  Neeb P R . Aggregates in Norway—Properties defining the quality of sand, gravel and hard rock for use as aggregate for building purposes. In Slagstad, T. (ed.) Geology for Society, Geological Survey of Norway, Special Publication, 2008, 11, 37–46.

[17] Ündül O. Assessment of mineralogical and petrographic factors affecting petro-physical properties, strength and cracking processes of volcanic rocks. Engineering Geology, 2016, 210: 10–22

[18] Merriam R, Rieke  H H III, Kim  Y C. Tensile strength related to mineralogy and texture of some granitic rocks. Engineering Geology, 1970, 4(2): 155–160

[19] Gunsallus K L ,  Kulhawy F H . A comparative evaluation of rock strength measurements. International Journal of Rock Mechanics and Mining Sciences & Geomechanics Abstracts, 1984, 21(5): 233–248

[20] Tuğrul A, Zarif  I H. Correlation of mineralogical and textural characteristics with engineering properties of selected granitic rocks from Turkey. Engineering Geology, 1999, 51(4): 303–317

[21] Bell F G. The physical and mechanical properties of the Fell Sandstones, Northumberland, England. Engineering Geology, 1978, 12: 1–29

[22] Fahy M P, Guccione  M J. Estimating the strength of sandstone using petrographic thin section data. Bull. Int. Assoc. Eng. Geol., 1979, 16(4): 467–485

[23] Shakoor A, Bonelli  R E. Relationship between petrographic characteristics, engineering index properties, and mechanical properties of selected sandstone. Bull. Int. Assoc. Eng. Geol., 1991, XXVIII(1): 55–71

[24] Åkesson U, Stigh  J, Lindqvist J E ,  Göransson M . The influence of foliation on the fragility of granitic rocks, image analysis and quantitative microscopy. Engineering Geology, 2003, 68(3–4): 275–288

[25] Yusof N Q A M ,  Zabidi H . Correlation of mineralogical and textural characteristics with engineering properties of granitic rock from Hulu Langat, Selangor. Procedia Chemistry, 2016, 19: 975–980

[26] Sousa L M O . The influence of the characteristics of quartz and mineral deterioration on the strength of granitic dimensional stones. Environmental Earth Sciences, 2013, 69(4): 1333–1346

[27] Brace W F. “Dependence of fracture strength of rocks on grain size.” Bulletin of Mineral Industries Experiment Station, Mining Engineering Series. Rock Mechanics, 1961, 76: 99–103

[28] Mendes F M, Aires-Barros  L, Rodrigues F P . The use of modal analysis in the mechanical characterization of rock masses. In: Proc 1st Int. Cong. Rock Mech. Lisbon, 1966, 1, 217–223.

[29] Willard R J, McWilliams  J R. Microstructural techniques in the study of physical properties of rocks. International Journal of Rock Mechanics and Mining Sciences, 1969, 6(1): 1–12

[30] Wong R H C ,  Chau K T ,  Wang P. Microcracking and grain size effect in Yuen Long marbles. International Journal of Rock Mechanics and Mining Sciences & Geomechanics Abstracts, 1996, 33(5): 479–485

[31] Olsson W A. Grain size dependence of yield stress in marble. Journal of Geophysical Research, 1974, 79(32): 4859–4862

[32] Přikryl R. Some microstructural aspects of strength variation in rocks. International Journal of Rock Mechanics and Mining Sciences, 2001, 38(5): 671–682

[33] Hareland G, Polston  C E, White  W E. Normalized rock failure envelope as a function of rock grain size. International Journal of Rock Mechanics and Mining Sciences & Geomechanics Abstracts, 1993, 30(7): 715–717

[34] Onodera T F, Asoka  K H M. Relationship between texture and mechanical properties of crystalline rocks. Bull Int Assoc Eng Geol, 1980, 22: 173–177

[35] French W J, Kermani  S, Mole C F . Petrographic evaluation of aggregate parameters. In: Proceeding of the 8th Euroseminar on microscopy applied to building materials, Athens, 2001, 557–564.

[36] Åkesson U, Lindqvist  J E, Göransson  M, Stigh J . Relationship between texture and mechanical properties of granites, central Sweden, by use of image-analysing technique. Bulletin of Engineering Geology and the Environment, 2001, 60(4): 277–284

[37] Hatzor Y H, Zur  A, Mimran Y . Microstructure effects on microcracking and brittle failure of dolomites. Tectonophysics, 1997, 281(3–4): 141–161 

[38] Eberhardt E, Stimpson  B, Stead D . Effects of grain size on the initiation and propagation thresholds of stress-induced brittle fractures. Rock Mechanics and Rock Engineering, 1999, 32(2): 81–99

[39] Irfan T Y, Dearman  W R. Engineering classification and index properties of a weathered granite. Bulletin of the International Association of Engineering and Geology, 1978, 17(1): 79–90

[40] Hecht C A, Bönsch  C, Bauch E . Relations of rock structure and composition to petrophysical and geomechanical rock properties: examples from Permocarboniferous red-beds. Rock Mechanics and Rock Engineering, 2005, 38(3): 197–216

[41] Howarth D F, Rowlands  J C. Quantitative assessment of rock texture and correlation with drillability and strength properties. Rock Mechanics and Rock Engineering, 1987, 20(1): 57–85

[42] Ersoy A, Waller  M D. Textural characterization of rocks. Engineering Geology, 1995, 39(3–4): 123–136

[43] Azzoni A, Bailo  F, Rondena E ,  Zaninetti A . Assessment of texture coefficient for different rock types and correlation with uniaxial compressive strength and rock weathering. Rock Mechanics and Rock Engineering, 1996, 29(1): 39–46

[44] Ozturk C A, Nasurf  E, Bilgin N . The assessment of rock cutability, and physical and mechanical rock properties from a texture coefficient. Journal of the South African Institute of Mining and Metallurgy, 2004, 7: 397–403

[45] Přikryl R. Assessment of rock geomechanical quality by quantitative rock fabric coefficients: Limitations and possible source of misinterpretations. Engineering Geology, 2006, 87(3): 149–162

[46] Alber M, Kahraman  S. Predicting the uniaxial compressive strength and elastic modulus of a fault breccia from texture coefficient. Rock Mechanics and Rock Engineering, 2009, 42(1): 117–127

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg