ISSN 1062-7391, Journal of Mining Science, 2018, Vol. 54, No. 6, pp. 914–930. © Pleiades Publishing, Ltd., 2018.

Published in Fiziko-Tekhnicheskie Problemy Razrabotki Poleznykh Iskopaemykh, 2018, No. 6, pp. 37–55.

_________________________________ GEOMECHANICS _______________________________

____________________________________________________________________________________________________________________________________ ___________________________________________________________________________________________________________________________

Water Saturation Influences on Engineering Properties

of Selected Sedimentary Rocks of Pakistan
Y. Majeeda* and M. Z. Abu Bakarb
a
Mining Engineering Department, University of Engineering and Technology,
Lahore, 54890 Pakistan
*e-mail: yasirbinmajeed@gmail.com
b
Geological Engineering Department, University of Engineering and Technology, Lahore, Pakistan
Received June 13, 2018
Revised July 12, 2018
Accepted November 28, 2018

Abstract—This study is focused on the evaluation of water saturation influence on the mechanical and
physical properties of 34 sedimentary rock units selected from different geological formations in Pakistan.
The laboratory testing program comprised the determination of physico-mechanical rock properties of both
air-dry as well as fully water saturated rock specimens. Further thin section analyses of all rock samples
were also performed to explain their petrography. According to the statistical analyses overall reductions of
around 40 and 50% were found in UCSsat and BTSsat values, respectively, measured on saturated rock
samples in comparison to their corresponding dry strength values, i.e. UCSdry and BTSdry. Linear
correlations were found between ultrasonic wave velocities as well as density of both dry and saturated rock
samples. In addition, the dry and saturated UCS values were exponentially related with the rock porosity.
Multiple regression technique was also utilized to develop a predictive linear model of UCSsat with
geotechnical rock properties in the dry condition and petrographical characteristics of rock samples. Finally
the validity of multiple regression model developed in this study and an existing correlation for the
conversion of UCSdry into UCSsat was statistically assessed.
Keywords: Uniaxial compressive strength (UCS), Brazilian tensile strength (BTS), rock density, rock
porosity, pore space volume, water saturation, correlation.
DOI: 10.1134/S1062739118065060

INTRODUCTION
The mechanical, physical and petrographical rock properties are frequently utilized by the mining,
civil and geotechnical industries for the design of underground and surface structures in the rock
mass. During construction process, the host rock usually exhibits both dry as well as wet rock
conditions. In specialized applications, including harbor work, dredging and the construction of
structures underneath bodies of water, the ground is fully saturated with water [1]. Therefore, the
presence of water in the rock strata calls for the knowledge about response of rock in the saturated
conditions so as to design a stable structure. Laboratory testing is usually based on air dried rock
samples, which can result in over/under estimation of rock properties, which in turn could affect
overall economics of the project.
In [2] two categories of factors are highlighted which impact the engineering behavior of rocks,
particularly sandstones. The first category includes rock properties namely mineral contents, texture,
porosity, density, in-situ stresses and rock water content. The second category is concerned with the
sample preparation and rock mechanics testing procedures adopted in the laboratory work. Rock
strength is influenced by parameters including mineralogy, grain size, porosity, and micro-fabric [3].
Further in saturated conditions minerals behave differently (i.e. softening, dissolution and corrosion)
depending on their crystal structure and chemical composition. It is also known that the rock behavior
(physico-mechanical properties, cutting forces, specific energy and abrasiveness etc.) changes
at different moisture levels [1, 4].

914
 WATER SATURATION INFLUENCES ON ENGINEERING PROPERTIES 915

Numerous investigations in the past have explored the sensitivity of different rock types under
water saturated conditions. These research studies show significant reduction in uniaxial compressive
strength, as well as Brazilian tensile strength of water saturated rock samples in contrast to the dry
rock samples. According to the hypothesis [5], solids are more susceptible to deformation and
breakdown owing to adhesion from the surrounding medium primarily due to decrease in the surface
energy of newly formed edges in the solid as a result of strain. In [6] they performed UCS tests on
both dry and water saturated rock samples of quartzite shale and quartzite sandstone and found a 50%
reduction in strength of the saturated samples. This decrease in compressive strength was ascribed
to the drop in the surface free energy by the immersing fluid. In [7] the authors carried out strength
tests on rock samples under increased strain rate keeping the pore water pressure at some constant
value and noted an increase in the rock strength values. The research study [8] concluded a reduction
in the tensile strength of rock samples mainly because of increase in surface tension and dielectric
constant of saturating liquid.
In [9] the author noted a decline in the strength of rock under saturation owing to reductions in
rupture energy, capillary tension, internal friction as well as rise in pore pressure accompanied by the
deterioration caused by chemical reaction. In [10] the researcher performed point load strength tests
on all three generic rock types (igneous, sedimentary and metamorphic rocks) on air dried as well
assaturated rock samples and registered a decrease in strength ranging from 20 up to 45%. The
investigations of [2] ended with the conclusion that major decrease in the UCS of sandstone rock
samples was achieved at water contents below 1%. The cited researchers tested three British
sandstones (Penrith, Bunter and Waterstone) at various water contents starting from dry to completely
saturated state. In [11] they conducted a research work on a data set of 35 sandstone rock samples
belonging to various regions of Britain and established that even a slight increase of 1% in the
moisture level of sandstones from the dry state may possibly show a significant influence on their
strength as well as deformability. The cited authors found a decrease in the UCS values of saturated
clay rich and siliceous sandstones to the order of 78 and 8% correspondingly. [12] utilized the
existing data of [11] to investigate the influence of water content on the UCS and deformation
modulus of rocks, they proposed a statistical correlation showing that the strength of saturated rock
samples was about 75.6% of the strength in dry condition.
In [13] there was noticed a substantial drop of 90% in UCS, 93%—in the mean modulus of elasticity
and 90%—in the tensile strength of clay bearing rocks with corresponding increase in the moisture
levels. In [14] the authors conducted laboratory strength tests on an argillaceous quartz sandstone by
changing the water levels starting from 0% (oven dried samples), 0.2% (air dried samples) and 5.2%
(completely saturated samples). They reported noteworthy reductions of 63 and 68% in the UCS at
water levels of 0.2 and 5.2%, respectively, in comparison to the UCS of oven dried state. In [15] the
cited researcher found the strength of rock gypsum to be highly sensitive with variation in the moisture
levels. Meaning a small increase in the moisture level from 1 to 2% causes a considerable drop in the
strength of gypsum. In [16] mechanical property tests (UCS and Young’s modulus) were carried out on
Latrobe valley brown coal specimens by means of three different saturation fluids (water, carbon
dioxide and nitrogen). In case of UCS tests a reduction of 17 and 10% was found for water and carbon
dioxide saturated samples. However, a rise of 2% in strength was registered in the case of coal samples
saturated with nitrogen. In [17] they performed laboratory investigations on coal and associated strata
including stone coal, mudstone and sandstone to quantify the effect of water saturation on their
unconfined compressive strength. They observed that water saturation actually decreased UCS values

JOURNAL OF MINING SCIENCE Vol. 54 No. 6 2018

916 MAJEED, ABU BAKAR

to the order of 9.6, 18.7, 30.3 and 28.8%, respectively. In [18] BTS laboratory tests were conducted on
wet and dry rock specimens of sandstone and limestone. Tensile strength reductions of around 50 and
20% were found for sandstone and limestone rock samples respectively under wet conditions.
As the degree of strength reduction upon water saturation varies considerably among different
types of rocks, therefore this subject has become an active research area in recent years [19–22].
Despite extensive investigations conducted on the impact of moisture level on the mechanical and
physical properties of rocks belonging to various regions around the globe, no research work has yet
been documented on the rocks of Pakistan.
The current investigative study is aimed at correlations development to express the influence
of water saturation on the physical (density, porosity, sonic wave velocity) and mechanical (UCS and
BTS) properties of sedimentary rocks selected from different regions of Pakistan. A multiple
regression model is also proposed to forecast saturated UCS values from the mechanical and physical
properties of dry rocks along with the petrographical parameters.
1. MATERIAL AND RESEARCH METHODS
The current research work utilized a total of 34 sedimentary rock types collected from numerous
stratigraphic formations of Pakistan. The included rock types were selected to incorporate a wide
variety of mechanical and physical properties. Table 1 lists the codes, types, formations and sampling
localities of the collected rock specimens.
In this investigative study a complete set of laboratory experiments incorporating the mechanical
and physical tests (UCS, BTS, sonic wave velocity, density, porosity and pore space volume) on dry
as well as on fully saturated rocks were performed. In addition the petrographic properties of the
included rock samples were also determined. For laboratory testing the selected rock boulders were
cored perpendicularly to the bedding planes by utilizing 54 and 42 mm diamond core cutting bits. The
cylindrical rock samples were prepared in line with the ASTM-D4543 standards [23]. For the
performance of saturated tests the prepared rock specimens were saturated by adopting the
progressive saturation technique (Fig. 1) developed by the US Army Corps of Engineers [24]. Some
past studies [11, 14, 15, 25, 26] report full immersion of rock samples in water to achieve saturation,
but this method of rock saturation generally leaves dry regions inside the specimens primarily owing
to the trapped pore air within these regions. This can be avoided by progressive saturation method,
which starts with the sample base placed in a few centimeters of water. As the saturation front visible
in the rock matrix rises due to capillary action, the water level in the vessel is accordingly raised up to
bring it approximately midway up to the saturation front. This process is reiterated until the saturation
front reaches top of the specimen [1, 4, 27].

Fig. 1. Rock cores placed for full water saturation adopting the progressive saturation technique. The water
saturation front advancing progressively towards the top of the rock core.

JOURNAL OF MINING SCIENCE Vol. 54 No. 6 2018

 WATER SATURATION INFLUENCES ON ENGINEERING PROPERTIES 917

Table 1. List of selected rocks

Sample Rock Rock Rock

Sampling Locality
No. code type formation
1 TBSIS Siltstone Tobra Khewra Gorge, Salt Range Punjab
2 MUSIS Siltstone Murree NelumJehlum Hydro Power Project, Azad Jammu
and Kashmir (Latitude 34°23′ 34″ N; Longitude
73°43′ 08″ E)
3 MUSS-1 Sandstone Murree NelumJehlum Hydro Power Project, Azad Jammu
and Kashmir (Latitude 34°23′ 34″ N; Longitude
73°43′ 08″ E)
4 MUSS-2 Sandstone Murree Herighal Majeed Gala, Hydro Power Project, Azad
Jammu and Kashmir (Latitude 33°56′ 37.03″ N;
Longitude 73°41′ 56.21″ E)
5 TBSS-1 Sandstone Tobra Khewra Gorge, Salt Range Punjab
6 TBSS-2 Sandstone Tobra Tobar Town, Salt Range Punjab
7 KHSS Sandstone Khewra Khewra Gorge, Salt Range Punjab
8 KUSS Sandstone Kussak Khewra Gorge, Salt Range Punjab
9 DDSS Sandstone Dandot Dandot Town, Salt Range Punjab
10 TBSS-3 Sandstone Tobra Zaluch Gorge, Salt Range Punjab
11 CHSS Sandstone Chidru Zaluch Gorge, Salt Range Punjab
12 WASS Sandstone Warchha
Jan-SukhGorge, Warchha, Salt Range Punjab
Sandstone
13 LMSS Sandstone Lumshiwal Makerwal, Surghar Range Punjab
14 HNSS Sandstone Hangu Makerwal, Surghar Range Punjab
15 DTSS Sandstone Datta Zaluch Gorge, Salt Range Punjab
16 WASS-S Sandstone Warchha
Jan-SukhGorge, Warchha, Salt Range Punjab
(Speckled) Sandstone
17 AMSS Sandstone Amb Zaluch Gorge, Salt Range Punjab
18 NGSS Sandstone Nagri Balkasar, Punjab Province
19 WASS-R Sandstone Warchha
Jan-SukhGorge, Warchha, Salt Range Punjab
(Red) Sandstone
20 HZSS Sandstone Hazira Abottabad, Khyber Pakhtunkhwa (KPK)
21 CHCS Chamositic Chichali
Zaluch Gorge, Salt Range Punjab
Siderite
22 KGDL Dolomite Kingriali Zaluch Gorge, Salt Range Punjab
23 JUDL-1 Dolomite Jutana Khewra Gorge, Salt Range Punjab
24 ABTDL Dolomite (Pink) Abottabad Sirban Hill, Abottabad, Khyber Pakhtunkhwa (KPK)
25 JUDL-2 Dolomite Jutana Khewra Gorge, Salt Range Punjab
26 SKLS-1 Limestone Sakesar Limestone Bestway Cement Quarry, Kattas, Punjab (Latitude
32°43′ 13.44″ N; Longitude 72°56′ 6″ E)
27 SKLS-2 Limestone Sakesar Limestone D.G.Khan Cement Quarry, Kattas, Punjab
(Latitude 32°42′ 54″ N; Longitude 72°49′ 30″ E)
28 SSLS-1 Limestone Samana Suk Bestway Cement Quarry, Hattar, Khyber
Pakhtunkhwa Province (Latitude 33°50′ 30.88″ N;
Longitude 72°52′ 24.12″ E)
29 SSLS-2 Limestone Samana Suk Mustahkam Cement Quarry, Hattar, Khyber
Pakhtunkhwa Province (Latitude 33°49′ 19.20″ N;
Longitude 72°50′ 19.68″ E)
30 WRLS Limestone Wargal Limestone Nammal Gorge, Salt Range Punjab
31 NMLS Limestone Nammal Limestone Nammal Gorge, Salt Range Punjab
32 SKLS-3 Limestone Sakesar Limestone Makerwal, Surghar Range Punjab
33 GYPSR Rock Gypsum Salt Range Khewra Gorge, Salt Range Punjab
34 MRLSR Marl Salt Range Khewra Gorge, Salt Range Punjab

JOURNAL OF MINING SCIENCE Vol. 54 No. 6 2018

918 MAJEED, ABU BAKAR

1.1. The Uniaxial Compressive Strength (UCS) Tests

The unconfined compression tests were conducted on trimmed and lapped rock samples on both
dry and saturated cores with length-to-diameter ratios of about 2.0–2.5. The rock specimens were
subjected to compression at a continuous stress rate 0.5–1.0 MPa/s till failure by utilizing Shimadzu
universal testing machine of 200 t peak load capacity. The UCS tests were undertaken in compliance
with the procedures laid down in the suggested methods of ISRM [28].
UCS of the test samples was computed by employing the formula:
σ = P / A0 , (1)
where P is the peak load bearing ability of the rock sample during compression test; A0 is the initial
cross-sectional area of the rock sample.
1.2. The Brazilian Tensile Strength (BTS) Tests
The dry and saturated BTS tests were performed according to the experimental procedures
provided in ISRM suggested methods [29]. The prepared rock discs of NX size (5.4 cm) diameter
with height-to-diameter ratios of approximately 0.5 were pressed across the diametric axis until
failure. The tests were carried out on a Shimadzu 200 t universal testing machine.
BTS of the test specimens was calculated by the formula:
σ t = 0.636 P / Dt , (2)
here D , t are the test sample diameter and thickness.
1.3. The Sonic Velocity Tests
In [30] it was pointed out that the measurement of sonic waves velocity passing through solid
materials (particularly rocks) could be utilized as an indication of material properties including density
and elastic properties as well as the quality of material. In this study both primary sonic wave V p and
secondary sonic wave Vs velocities were determined by using Portable Ultrasonic Non-Destructive
Indication Tester (PUNDIT) according to the ISRM suggested methods [31]. Ultrasonic pulses were
passed through rock cores. PUNDIT transducers were placed at the greased ends of specimens to
measure sonic waves travel time t p and t s for p- and s-waves, respectively. V p and Vs were then
computed by dividing the distance d between the PUNDIT transducers by t p and t s , respectively.
1.4. The Density and Porosity Tests
Dry rock density is the ratio of solid mass portion (excluding pore spaces) of the specimen to its
total volume, while density of saturated rock is the ratio of specimen mass including the mineral
grains, matrix and pore water to its total volume. Rock porosity is defined as the ratio of pore volume
to the total volume of rock sample. Numerous techniques are present to find the density, porosity and
related properties of rock samples. Saturation and caliper method, and saturation and buoyancy
methods are mostly used [32]. In this work the density, porosity and pore space volume tests were
conducted by utilizing the saturation and caliper technique of ISRM suggested methods [33] applied
only to the prepared rock samples. Therefore, rock discs prepared for this purpose were used to
calculate the bulk volume V by averaging out at least three vernier caliper measurements for each
dimension. Then the weight of saturated rock discs were determined to find their saturated surface
dried mass M sat and subsequently the specimens were placed in an oven at a temperature of 105°C for
24 h to determine their dried grain mass M s . Pore space volume Vv , dry density ρ dry , saturated
density ρ sat and porosity n of rock samples were computed by utilizing the standard equations.

JOURNAL OF MINING SCIENCE Vol. 54 No. 6 2018

 WATER SATURATION INFLUENCES ON ENGINEERING PROPERTIES 919

1.5. Petrographic Testing

Petrographic analyses were carried out by preparing thin sections of all 34 rock samples included
in this study to determine their quantitative mineral contents and grain sizes by utilizing a polarizing
microscope (MIN-8, Russia). The grain sizes of minerals were measured by adopting the mean
intersection length technique as recommended by [34]. Similarly their corresponding mineral contents
(in %) were worked out by point counting from thin section under the microscope.
2. RESULTS AND DISCUSSION
This investigative study is chiefly focused on assessing the influence of water saturation on the
physico-mechanical rock properties. A multiple regression generic model is also proposed to predict
saturated uniaxial compressive strength UCSsat of sedimentary rocks from mechanical and physical
rock properties in the dry state as well as from the petrographic parameters. Moreover efforts were
also made to evaluate the utility of regression equations, including the multiple model developed
in this research work as well as already proposed correlation in the literature for the conversion
of UCSdry values into UCSsat values by utilizing the data base produced in this study.
Table 2. Mechanical rock properties tests of dry and saturated rock samples

UCSdry UCSsat Strength BTSdry BTSsat Strength

Sample Code
MPa loss in UCS, % MPa loss in BTS, %
1 TBSIS 57.9 56.1 3.1 9.0 6.8 24.9
2 MUSIS 49.3 17.3 64.9 7.4 2.0 72.2
3 MUSS-1 127.6 85.3 33.1 6.4 2.5 61.5
4 MUSS-2 82.8 62.0 25.1 6.0 3.9 34.4
5 TBSS-1 39.8 19.1 52.0 1.8 2.1 –14.4
6 TBSS-2 26.7 13.6 49.2 1.4 0.9 37.2
7 KHSS 41.6 26.2 36.9 0.5 0.1 76.4
8 KUSS 69.0 43.7 36.7 6.1 3.9 35.9
9 DDSS 44.0 40.3 8.5 2.8 2.7 3.4
10 TBSS-3 109.7 66.3 39.6 6.0 1.2 80.1
11 CHSS 61.5 58.0 5.7 7.3 2.6 65.1
12 WASS 27.1 23.6 12.9 1.6 1.4 12.6
13 LMSS 29.0 14.2 51.3 1.9 1.8 4.6
14 HNSS 16.7 13.0 21.9 0.7 1.0 –42.8
15 DTSS 21.2 19.7 6.8 2.1 2.0 4.3
16 WASS-S 24.2 17.5 27.5 1.3 1.0 25.4
17 AMSS 46.4 30.6 33.9 1.6 1.8 –9.5
18 NGSS 17.1 4.8 71.6 0.9 0.9 –2.3
19 WASS-R 56.8 60.9 –7.3 4.2 1.4 67.4
20 HZSS 129.0 109.7 15.5 22.7 12.8 43.7
21 CHCS 51.7 38.8 25.0 8.1 4.1 48.7
22 KGDL 61.8 33.5 45.8 6.5 3.9 41.0
23 JUDL-1 144.4 67.7 53.1 12.0 8.2 31.2
24 ABTDL 99.9 55.5 44.4 12.5 4.7 62.8
25 JUDL-2 132.7 57.9 56.3 6.7 3.8 42.4
26 SKLS-1 65.3 60.9 6.8 5.0 1.2 75.7
27 SKLS-2 95.8 48.7 49.1 4.6 2.6 42.6
28 SSLS-1 80.7 29.6 63.3 5.6 5.4 3.5
29 SSLS-2 69.9 32.2 53.9 3.3 2.8 16.2
30 WRLS 66.5 37.2 44.0 5.4 3.1 42.2
31 NMLS 20.1 32.1 –60.1 7.8 0.8 90.2
32 SKLS-3 92.7 80.8 12.9 7.9 5.1 34.9
33 GYPSR 13.5 8.5 37.3 1.3 1.4 –2.0
34 MRLSR 5.4 2.1 60.7 0.8 0.3 57.4

JOURNAL OF MINING SCIENCE Vol. 54 No. 6 2018

920 MAJEED, ABU BAKAR

The detailed experimental test results of 34 selected rock samples, incorporating mechanical and
physical properties performed on both dry and saturated rocks are detailed in Table 2 and Table 3,
respectively. The petrographic thin section analyses of the same rock samples including mineral
contents, percentages of rock matrix M and overall mean grain size of minerals Ø are summarized
in Table 4.

Table 3. Physical rock properties tests of dry and saturated rock samples

Vp(dry) Vp(sat) Vs(dry) Vs(sat) ρdry ρsat n Vv

Sample Code
km/s g/cm3 % g/cm3
1 TBSIS 5.60 5.88 2.36 2.58 2.60 2.64 3.63 2.22
2 MUSIS 5.31 5.17 1.90 2.26 2.69 2.71 1.03 0.73
3 MUSS-1 5.85 5.25 2.11 2.33 2.62 2.64 1.17 0.90
4 MUSS-2 5.58 5.84 2.43 2.56 2.63 2.65 2.82 0.94
5 TBSS-1 3.79 4.95 1.69 2.03 2.63 2.70 6.87 3.90
6 TBSS-2 2.64 3.01 1.09 1.25 2.25 2.37 12.22 7.93
7 KHSS 2.30 1.90 1.00 0.85 2.19 2.31 11.74 7.26
8 KUSS 4.39 4.91 1.73 2.19 2.50 2.57 6.82 4.44
9 DDSS 3.38 4.94 1.49 2.02 2.53 2.57 3.12 1.94
10 TBSS-3 4.86 4.44 1.56 1.96 2.568 2.62 4.73 2.95
11 CHSS 6.64 5.87 2.36 2.83 2.57 2.62 4.79 2.99
12 WASS 2.06 1.86 1.16 0.82 2.29 2.40 10.83 6.19
13 LMSS 2.79 2.71 1.13 1.21 2.11 2.26 15.05 9.12
14 HNSS 2.05 2.93 0.89 1.27 2.16 2.26 10.26 7.51
15 DTSS 2.73 3.28 1.27 1.45 2.18 2.31 12.10 7.37
16 WASS-S 2.60 3.17 1.12 1.25 2.28 2.38 10.70 6.76
17 AMSS 4.41 3.86 1.97 1.66 2.22 2.33 11.40 6.72
18 NGSS 1.98 2.58 0.86 1.17 2.48 2.61 12.77 7.43
19 WASS-R 5.12 5.27 2.50 2.35 2.60 2.64 3.61 2.26
20 HZSS 3.66 6.00 1.46 2.55 2.64 2.66 1.63 1.11
21 CHCS 3.61 3.89 1.57 1.70 2.62 2.82 20.02 12.97
22 KGDL 6.29 5.97 2.41 2.43 2.58 2.64 6.15 3.94
23 JUDL-1 5.90 6.48 2.04 2.80 2.77 2.78 1.23 0.83
24 ABTDL 7.26 7.29 2.72 3.05 2.81 2.82 0.38 0.25
25 JUDL-2 5.77 5.77 1.99 2.51 2.49 2.56 6.43 4.11
26 SKLS-1 6.36 6.42 2.71 2.75 2.61 2.64 3.47 0.99
27 SKLS-2 6.12 6.08 2.68 2.58 2.67 2.68 0.79 0.48
28 SSLS-1 6.43 7.00 2.75 3.06 3.17 3.20 2.81 1.57
29 SSLS-2 6.47 6.40 2.75 2.81 2.65 2.66 1.37 0.86
30 WRLS 7.49 6.30 3.26 2.77 2.69 2.70 1.17 0.78
31 NMLS 4.78 5.09 2.14 2.14 2.50 2.57 6.53 4.13
32 SKLS-3 5.41 5.06 2.35 2.24 2.47 2.54 7.15 4.11
33 GYPSR 5.38 5.18 2.23 2.26 2.07 2.25 18.37 10.70
34 MRLSR 2.29 2.06 0.90 0.91 2.00 2.25 24.70 14.95

JOURNAL OF MINING SCIENCE Vol. 54 No. 6 2018

 WATER SATURATION INFLUENCES ON ENGINEERING PROPERTIES 921

Table 4. Petrography of the included rock samples, %

and chlorite
Glauconite

Accessory
Iron oxide
fragments
Feldspars

Dolomite

minerals

Gypsum
Sample

Micrite
Calcite

Ø, mm
Quartz

Matrix

Micas
Code

Rock
1 TBSIS 22.0 35.0 6.0 0.0 2.0 1.5 33.0 0.0 0.0 0.0 0.5 0.0 0.21
2 MUSIS 15.2 13.0 37.0 0.0 5.0 5.0 0.0 24.8 0.0 0.0 0.0 0.0 0.29
3 MUSS-1 64.0 15.0 2.0 9.0 7.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.58
4 MUSS-2 61.0 20.0 0.0 7.0 3.0 2.0 0.0 7.0 0.0 0.0 0.0 0.0 0.25
5 TBSS-1 68.0 9.5 10.5 0.0 4.0 0.0 0.0 2.5 0.0 5.0 0.5 0.0 0.73
6 TBSS-2 78.0 17.0 0.5 0.0 4.0 0.5 0.0 0.0 0.0 0.0 0.0 0.0 0.39
7 KHSS 67.0 20.0 10.0 0.0 1.5 1.0 0.0 0.0 0.0 0.5 0.0 0.0 0.20
8 KUSS 78.0 17.0 0.5 0.0 4.0 0.5 0.0 0.0 0.0 0.0 0.0 0.0 0.25
9 DDSS 62.3 28.0 1.5 4.7 1.5 0.5 0.0 0.0 0.0 0.0 1.5 0.0 0.48
10 TBSS-3 70.1 1.7 1.0 23.9 1.0 0.4 0.0 1.1 0.0 0.5 0.3 0.0 0.81
11 CHSS 67.5 2.1 26.5 2.6 0.0 0.0 0.0 0.0 0.0 0.0 1.3 0.0 0.26
12 WASS 55.0 27.0 16.0 0.0 1.0 0.5 0.0 0.0 0.0 0.5 0.0 0.0 0.38
13 LMSS 78.0 12.0 0.0 9.0 0.5 0.0 0.0 0.0 0.0 0.5 0.0 0.0 0.67
14 HNSS 75.0 4.5 17.0 0.0 2.0 1.0 0.0 0.0 0.0 0.5 0.0 0.0 0.37
15 DTSS 73.0 16.0 5.5 1.5 2.5 1.0 0.0 0.0 0.0 0.0 0.5 0.0 0.59
16 WASS-S 55.5 19.9 13.4 6.7 4.0 0.0 0.0 0.0 0.0 0.0 0.5 0.0 0.82
17 AMSS 77.0 17.5 0.5 0.0 3.7 0.8 0.0 0.0 0.0 0.0 0.5 0.0 0.60
18 NGSS 72.5 18.0 2.5 4.0 2.0 0.5 0.0 0.0 0.0 0.0 0.5 0.0 0.54
19 WASS-R 56.0 28.0 9.0 3.0 1.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.53
20 HZSS 95.0 3.0 0.0 0.0 1.0 0.5 0.0 0.0 0.0 0.5 0.0 0.0 0.26
21 CICS 13.0 0.0 2.0 0.0 1.0 84.0 0.0 0.0 0.0 0.0 0.0 0.0 0.17
22 KGDL 1.0 0.0 0.0 0.0 2.5 0.5 96.0 0.0 0.0 0.0 0.0 0.0 0.17
23 JUDL-1 2.0 0.0 0.0 0.0 3.0 1.0 94.0 0.0 0.0 0.0 0.0 0.0 0.24
24 ABTDL 2.5 0.0 0.0 0.0 1.0 0.0 92.0 4.5 0.0 0.0 0.0 0.0 0.43
25 JUDL-2 10.0 9.0 2.0 0.0 1.0 3.0 75.0 0.0 0.0 0.0 0.0 0.0 0.21
26 SKLS-1 0.0 0.0 0.0 0.0 0.5 0.5 16.0 27.5 55.5 0.0 0.0 0.0 0.49
27 SKLS-2 0.0 0.0 0.0 0.0 0.5 0.5 16.0 28.0 55.0 0.0 0.0 0.0 0.49
28 SSLS-1 0.0 0.0 0.0 0.0 0.0 1.0 36.0 25.0 38.0 0.0 0.0 0.0 1.51
29 SSLS-2 0.0 0.0 0.0 0.0 0.0 1.0 35.0 25.0 39.0 0.0 0.0 0.0 0.75
30 WRLS 0.0 0.0 0.0 0.0 0.5 0.0 0.5 90.0 9.0 0.0 0.0 0.0 0.70
31 NMLS 0.0 0.0 0.0 0.0 0.0 1.0 25.0 20.0 54.0 0.0 0.0 0.0 0.13
32 SKLS-3 0.0 0.0 0.0 0.0 1.0 0.5 0.0 7.5 91.0 0.0 0.0 0.0 0.44
33 GYPSR 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 99.0 0.36
34 MRLSR 21.0 0.0 74.0 0.0 1.0 0.0 4.0 0.0 0.0 0.0 0.0 0.0 0.17

3. INFLUENCE OF MOISTURE CONTENT ON ROCK STRENGTH

Figures 2a and 2b show the regression plots of UCSsat and BTSsat against UCSdry and BTSdry,
respectively. The established correlations suggest overall reductions of about 57% in UCS and 50%
in BTS of the rock specimens when tested under both water saturated and air dried conditions.
Table 2 demonstrates that for majority of the rocks tested, a significant reduction in uniaxial
compressive and Brazilian tensile strengths was observed upon fully saturating rock samples from
their initial air dried condition, and was in total agreement with the results reported in the past
studies [2, 6, 11, 13, 15, 18, 35–49]. This fact is also evident in Figs. 2a and 2b, where most of the
data points appear below the 1:1 line. However, very few cases can be observed where the measured
UCS and BTS values for saturated rocks coincide with or even increase from their corresponding dry
values.

JOURNAL OF MINING SCIENCE Vol. 54 No. 6 2018

922 MAJEED, ABU BAKAR

Fig. 2. (a) Regression plot between UCSsat and UCSdry; (b) regression plot between BTSsat and BTSdry.

One of the basic mechanisms of strength reduction in sedimentary rocks with water saturation is the
degradation or weakening of rock matrix or cements (clay or carbonate) which bind the mineral grains,
including quartz, feldspars, micas, iron oxides and other accessory minerals together [2, 4, 10, 11].
For example, the siltstone of Murree formation (MUSIS) showed significant reduction in both uniaxial
compressive strength (65%) and Brazilian tensile strength (72%) probably due to its reasonably high
clay component of 37%. Whereas, the increase in rock strength of sandstones upon saturation and
corresponding decrease in their sensitivity towards strength reduction, could be attributed primarily
to the mineralogical characteristics, texture and fabric of rock (i.e. cementing material)[2, 11].
4. INFLUENCE OF WATER SATURATION ON PHYSICAL ROCK PROPERTIES
The dry primary Vp and secondary Vs ultrasonic wave velocities range from 1.98 to 7.49 and from
0.86 to 3.26 km/s, respectively, whereas the corresponding values of Vp and Vs for saturated rocks—
from 1.86 to 7.29 and from 0.82 to 3.06 km/s. To clarify the correlation between the measured dry
and saturated primary and secondary ultrasonic wave velocities the regression analyses were
performed (Fig. 3). As expected in majority of the cases the water saturation actually increased both
the Vp and Vs in comparison to their values measured in dry rock condition. This fact can also be
observed from the developed equations in Fig. 3 where Vp(dry)and Vs(dry)are about 85 and 87% of Vp(sat)
and Vs(sat)values, respectively. It is pertinent to mention here that the linear relationship proposed by
[26] for the inter-conversion of dry and saturated sonic wave velocities Vpof travertine, for both
massive and laminated varieties, also indicates an overall increase in the values of sonic wave
velocities through saturated rocks.

Fig. 3. Relationship between (a) primary ultrasonic wave velocities Vp and (b) secondary ultrasonic wave velocities Vs
of saturated and dry rocks.

JOURNAL OF MINING SCIENCE Vol. 54 No. 6 2018

 WATER SATURATION INFLUENCES ON ENGINEERING PROPERTIES 923

Fig. 4. Correlation between the measured dry Fig. 5. Comparison of dry and saturated UCS values
and saturated rock densities. with rock porosity.

Also when air dry and saturated densities of the included rock samples were statistically compared
(Fig. 4) a correlation of exceptional quality (R2 = 0.96) was observed. The established correlation
demonstrates a total rise of about 36% in wet density ρ sat of sedimentary rock specimens as
compared to their density ρ dry in air dried state. Comparable correlation has also been reported [26] to
explain the influence of water content on Hungarian travertine building and dimension stones.
5. EFFECT OF POROSITY ON DRY AND SATURATED UCS OF ROCKS
Another aim of this study was to develop an association between the porosity and UCS of the
included rock samples in water saturated as well as in air dried conditions. Using regression analysis
the best fit negative exponential correlations were developed between the porosity and the uniaxial
compressive strength for both dry and saturated rock conditions (Fig. 5). According to the regression
analysis, correlations of low strength exist between the rock porosity and corresponding dry and
saturated UCS values. The general form of these equations is as follows:
UСS = xe − y ( n ) , (3)
where x, y are the material constants (Table 5).
The findings of this study are confirmed by the past investigations of [50, 51] where the authors
reported exponential relations between UCS and porosity. In [26] a negative power relation between
porosity and UCS of dry and saturated travertines was proposed.
6. MULTIPLE REGRESSION ANALYSIS FOR THE ESTIMATION OF SATURATED UCS
A multiple linear regression model describes the correlations between a dependent variable and the
independent variables. In [52] it was reported that a linear multiple regression equation in its generic
form was as follows:
Yi = β 0 + β1 X i1 + β 2 X i 2 + ... + β p X ip , (4)
Yi is the dependent or response variable; X 1 , X 2 , ..., X p are the independent variables or the regressors;
β 0 is the constant of the equation and represents the interception point of straight line with Y-axis.
Table 5. Material constants

UСS = xe − y ( n )
Rock sample state
X y R2
Air dried 98.57 0.097 0.61
Water saturated 65.57 0.103 0.55

JOURNAL OF MINING SCIENCE Vol. 54 No. 6 2018

924 MAJEED, ABU BAKAR

Table 6. ANOVA for response variable UCSsat

Source Sum of squares df Mean square F Sig
Regression 17221.251 4 4305.313 31.665 0
Residual 3942.928 29 135.963
Total 21164.179 33

In this study SPSS-21.0 statistical software package was used to develop a stepwise multiple linear
regression model for the prediction of UCSsat from the mechanical, physical and petrographic rock
properties generally obtained in routine laboratory testing. Therefore UCSsat was fixed as the
independent variable, whereas UCS, BTS, dry density ρ dry , porosity n, pore space volume Vv, Vp and
Vs in the dry state as well as petrographical parameters, including rock matrix M and overall average
grain size of minerals Ø were taken as independent variables. The stepwise multivariate regression
technique employs forward selection and backward elimination methods to allow regressor variables
to enter in the model turn by turn in descending order of their partial F statistics values as long as they
stay significant at a prefixed significance level of entry SLE. At each step the equation is evaluated
when a variable (from the list of independent variables) is entered in the model with the smallest
p-value for F and this procedure continues. Variables already in the model are deleted if their p-values
become larger than the pre-defined significance level to stay SLS. This iterative process terminates
when the selection of an additional variable does not increase the R2 value by a significant amount at
the pre-selected significance level[53, 54].For UCSsat the best fit empirical model at the selected
significance level α = 0.10 was as follows (R2 = 0.82):
UCSsat = 84.110 + 0.385UCSdry + 2.282BTSdry − 28.727 ρ dry − 1.727Vv . (5)
The validity and significance of the regression model established in (5) was checked by applying
standard test statistics. The coefficient of determination value, calculated for the whole model
(R2 = 0.82) shows that 82% of the variance in UCSsat can be estimated by the regressors UCSdry,
BTSdry, ρdry and Vv. Similarly the computed correlation coefficient (R = 91%) indicates a strong
correlation between the response and independent variables. The standard error of the estimates or
root mean square error explains the standard deviation of the error term [52] which in the case of the
derived model (Equation (5)) is 11.66. Finally the adjusted R2 value specifies that the proposed model
accounts for 79% of total variation in the UCSsat values.
The detailed results of Analysis of Variance ANOVA are presented in Table 6. The hypotheses
testing for the proposed model (Equation 5) was applied with an F-test in ANOVA. The null
hypothesis H0 signifies that there exists no dependence between UCSsat and the four independent
variables. On the other hand, the alternate Ha is otherwise that of null hypothesis. The two hypotheses
H0 and Ha are explained as follows:
if Fmоdel < Fcritical , then H 0 is accepted and H a is rejected,
if Fmоdel > Fcritical , then H 0 is rejected in favor of H a .

Table 7. Results of some parameters affecting the quality of model for response variable UCSsat
Unstandardized coefficient Variance inflation
Model tcoeff ttable
B Std factor VIF
β0 84.110 35.645 2.360 1.699 —
UCSdry 0.385 0.088 4.397 1.699 2.660
BTSdry 2.282 0.658 3.469 1.699 2.075
Density ρdry –28.727 13.576 –2.116 1.699 2.727
Pore volume Vv –1.727 0.893 –1.933 1.699 2.750

JOURNAL OF MINING SCIENCE Vol. 54 No. 6 2018

 WATER SATURATION INFLUENCES ON ENGINEERING PROPERTIES 925

Using F-distribution table at α = 0.1 , d f1 = 4 , d f1 = 29 the Fcritical value of 2.15, was calculated
which is far less than the Fmodel value of 31.67, thus rejecting the null hypothesis. This clearly shows
the existence of a linear relationship between the UCSsat and the four independent variables. Table 7
presents the results (coefficients and summary) of other parameters for the dependent variable UCSsat.
Statistical inference method for an individual variable β j is also applied. For example, the
statistical inference for parameter β 0 was determined as follows: As , ttable = tαn −/ 2k = t0.1/
34 − 5
2 = 1.699 .

Also the value of tcoeff = 2.360 for β 0 is greater than the computed ttable = 1.699, which is an indicator
of its significance at the predefined significance level, i.e. α = 0.10 . When the similar process
is adopted for β1 , β 2 , β3 and β 4 it can be observed that all absolute values of tcoeff are greater than the
ttable values showing their statistical significance and, hence, are the most significant parameters
in explaining the dependency of UCSsat under the given experimental conditions. Meanwhile, the
independent variables n, Vp, Vs, M and Ø are removed from the regression analysis (Equation (5)) owing
to their lack of statistical significance at the pre-fixed level of significance (α = 0.1) .
Table 8. Comparison of actual and predicted UCSsat values
Predicted UCSsat
Actual
Sample Code The data from [12], This study,
UCSsat
UCSsat = 0.759UCSdry, R2 = 0.906 Equation (5)
1 TBSIS 56.071 43.931 48.442
2 MUSIS 17.305 37.419 41.091
3 MUSS-1 85.328 96.848 70.887
4 MUSS-2 61.970 62.824 52.611
5 TBSS-1 19.101 30.208 21.435
6 TBSS-2 13.569 20.288 19.337
7 KHSS 26.199 31.536 25.837
8 KUSS 43.733 52.401 45.199
9 DDSS 40.273 33.398 31.421
10 TBSS-3 66.296 83.285 61.225
11 CHSS 58.031 46.686 45.595
12 WASS 23.602 20.558 21.716
13 LMSS 14.154 22.041 23.171
14 HNSS 13.040 12.668 17.135
15 DTSS 19.737 16.076 21.448
16 WASS-S 17.549 18.368 19.335
17 AMSS 30.648 35.218 30.231
18 NGSS 4.845 12.956 8.590
19 WASS-R 60.900 43.081 36.889
20 HZSS 109.675 98.513 107.951
21 CHCS 38.792 39.252 24.806
22 KGDL 33.504 46.937 42.009
23 JUDL-1 67.750 109.619 85.953
24 ABTDL 55.522 75.845 69.922
25 JUDL-2 57.936 100.722 71.510
26 SKLS-1 60.851 49.530 44.006
27 SKLS-2 48.742 72.694 53.892
28 SSLS-1 29.635 61.251 34.252
29 SSLS-2 32.245 53.047 40.978
30 WRLS 37.209 50.436 43.418
31 NMLS 32.132 15.237 30.750
32 SKLS-3 80.792 70.393 59.828
33 GYPSR 8.475 10.265 14.513
34 MRLSR 2.100 4.061 4.626

JOURNAL OF MINING SCIENCE Vol. 54 No. 6 2018

926 MAJEED, ABU BAKAR

Fig. 6. Comparison of measured and predicted UCSsat values.

Furthermore, the multicollinearity option was also checked in the stepwise regression analysis in the
SPSS software. The multiple regression models usually contain multicollinearity to some extent, which
exists when two or more independent variables are very much correlated and the most commonly used
tools for the detection of severity of multicollinearity is the variance inflation factor VIF [52]. A value
of 10 is recommended as the upper limit for VIF in multiple regression analysis [55, 56]. Table 8 shows
that the VIF values for UCSdry, BTSdry, ρdry and Vv are reasonably lower than the allowed VIF factor of
10 which confirms that Equation (50, does not suffer from a high level of multicollinearity. Finally,
it can be inferred from the proposed model that UCSsat is primarily dependent on UCSdry, BTSdry,
density and pore volume of the sedimentary rocks included in this study.

7. PERFORMANCE ASSESSMENT OF MODELS FOR THE UCS ESTIMATION

OF SATURATED ROCKS
An effort was also made to utilize the mechanical and physical rock properties UCSdry, BTSdry, ρdry, Vv
measured in this study to estimate the saturated UCS values by employing the multivariable
prediction model. Table 8 presents the laboratory measured and estimated UCSsat values using the
selected models (i.e. [12] and Eq. (5) of this study). Figure 6 shows that the plots of actual and
predicted UCSsat based on the multiple regression model are in better agreement with the laboratory
measured UCSsat values.
To further estimation of performance of the selected models (Table 8), the variance account for
VAF and the root mean square RMSE statistical performance indices were utilized:
⎛ variance( yi − yˆi ) ⎞ 1 N
VAF = ⎜1 −
variance( yi ) ⎠
⎟ ⋅100% , RMSE =
N
∑ ( y − yˆ )
i i
2
,
⎝ i =1

where yi , yˆi are the measured and predicted values; N is the total number of samples.

Table 9. Computed values of VAF and RMSE for the models selected

Model VAF, % RMSE

The data from [12] 63.81 16.20
This study, Equation (5) 81.76 10.63

JOURNAL OF MINING SCIENCE Vol. 54 No. 6 2018

 WATER SATURATION INFLUENCES ON ENGINEERING PROPERTIES 927

The higher VAF values indicate that the regression model is providing better prediction outcomes,
whereas, the lower values of RMSE demonstrate the better predictive efficiency of the model.
In contrast to VAF, RMSE also explains a bias in the model [15, 52, 57–59]. Table 9 displays the
computed values of performance indices VAF and RMSE which show better prediction performance
of multiple regression model (Eq. (5)) developed in this study.
CONCLUSIONS
This study evaluates the influence of water saturation on the mechanical and physical rock
properties. A complete suite of laboratory tests were performed comprising petrography and physico-
mechanical rock properties in both dry as well as fully saturated states.
Correlations were developed for the estimation of UCSsat and BTSsat from their corresponding dry
strength values. The equations showed that overall UCS and BTS values of sedimentary rocks
measured in the saturated condition are approximately 0.57 and 0.50 times that of UCS and BTS
values in the dry state respectively. Therefore, practically it is recommended to normalize the UCS
and BTS values obtained on dry rocks to account for the differences in saturated rock when assessing
the rock competence and excavatability in mining and geo-technical projects.
The statistical examination of physical rock properties test results showed linear correlations of
good quality. The developed correlations show that Vp(dry) and Vs(dry) are roughly 0.85 and 0.87 times
the Vp(sat) and Vs(sat), respectively. Similarly saturated density ρ sat of rocks was found to be
approximately 36% higher than the corresponding dry density ρ dry values.
Exponential correlations of the form UСS = xe − y ( n ) have been proposed between UCSdry, UCSsat
and porosity of the rocks.
A multiple linear regression model was also proposed for the prediction of UCSsat from the
geo-mechanical properties UCSdry, BTSdry, ρ dry and Vv of dry sedimentary rock. This model accounts
for 82% of total variation in the dependent variable UCSsat and can be utilized for the prediction of
UCS of sedimentary rocks in the saturated conditions from routine laboratory measurements. In the
past investigations only bivariate correlations, i.e. relationships of UCSsat with UCSdry have been
proposed.
Finally the utility of multiple regression model (5) developed in this study and an existing
correlation proposed by [12] was tested, using the dataset developed in this research. The
performance comparison of these estimation models depicts that the multivariable model developed in
this study is statistically more reliable.

REFERENCES
1. Abu Bakar, M.Z., Majeed, Y., and Rostami, J., Effects of Rock Water Content on CERCHAR
Abrasivity Index, Wear, 2016, vol. 368–369, pp. 132–145. DOI:10.1016/j.wear.2016.09.007.
2. Dyke, C.G. and Dobereiner, L., Evaluating the Strength and Deformability of Sandstones, Quarterly J. of
Eng. Geology and Hydrogeology, 1991, vol. 24, pp. 123–134.
3. Yilmaz, I., Gypsum/Anhydrite: Some Engineering Problems, Bull. Eng. Geol. Env., 2001, vol. 60,
no. 3, pp. 227–230.
4. Abu Bakar, M.Z. and Gertsch, L.S., Evaluation of Saturation Effects on Drag Pick Cutting of a Brittle
Sandstone from Full Scale Linear Cutting Tests, Tunneling and Underground Space Technology, 2013,
vol. 34, pp. 124–134.

JOURNAL OF MINING SCIENCE Vol. 54 No. 6 2018

928 MAJEED, ABU BAKAR

5. Rehbinder, P. and Lichtman, V., Effect of Surface Active Media on Strains and Rupture in Solids, Proc. 2nd
Int. Congress on Surface Activity, 1957, pp. 563–582.
6. Colback, P.S.B. and Wiid, B.L., The Influence of Moisture Content on the Compressive Strength of
Rocks, Proc. of the 3rd Rock Mech. Symp., Toronto, Canada, 1965, pp. 65–83.
7. Brace, W.F. and Martin, R.J., A Test of the Law of Effective Stress for Crystalline Rocks of Low Porosity,
Int. J. Rock Mech. Min. Sci. Geomech.Abstr., 1968, vol. 5, pp. 415–426.
8. Vutukuri, V.S., The Effect of Liquids on the Tensile Strength of Limestone, Int. J. Rock Mech. Min. Sci.
Geomech. Abstr., 1974, vol. 11, pp. 27–29.
9. Van Eeckhout, E.M., The Mechanisms of Strength Reduction Due to Moisture in Coal Mine Shales, Int.
J. Rock Mech. Min. Sci. Geomech. Abstr., 1976, vol. 13, pp. 61–67.
10. Broch, E., Changes in Rock Strength Caused by Water, Proc. of 4th Congress of Int. Society for Rock
Mech., Montreux, Switzerland, 1979, vol. 1, pp. 71–75.
11. Hawkins, A.B. and McConnell, B.J., Sensitivity of Sandstone Strength and Deformability to Changes in
Moisture Content, Quarterly J. Eng. Geology and Hydrogeology, 1992, vol. 25, pp. 115–130.
12. Vasarhelyi, B., Some Observations Regarding the Strength and Deformability of Sandstones in Dry
and Saturated Conditions, Bull. Eng. Geol. Env., 2003, vol. 62, pp. 245–249.
13. Erguler, Z.A. and Ulusay, R., Water Induced Variations in Mechanical Properties of Clay Bearing Rocks,
Int. J. Rock Mech. Min. Sci., 2009, vol. 46, pp. 355–370.
14. Mammen, J., Saydam, S., and Hagan, P.A., Study on the Effect of Moisture Content on Rock Cutting
Performance, Proc. of the Coal Operators Conference, University of Wollongong and the Australian
Institute of Mining and Metallurgy, 2009, pp. 340–347.
15. Yilmaz, I., Influence of Water Content on the Strength and Deformability of Gypsum, Int. J. Rock Mech.
Min. Sci., 2010, vol. 47, pp. 342–347.
16. Perera, M.S.A., Ranjith, P.G., and Peter, M., Effects of Saturation Medium and Pressure on Strength
Parameters of Latrobe Valley Brown Coal: Carbon Dioxide, Water and Nitrogen Saturations, Energy,
2011, vol. 36, pp. 6941–6947.
17. Poulsen, B.A., Shen, B., Williams, D.J., Huddlestone-Holmes, C., Erarslan, N., and Qin, J., Strength
Reduction on Saturation of Coal and Coal Measures Rocks with Implications for Coal Pillar Strength, Int.
J. Rock Mech. Min. Sci., 2014, vol. 71, pp. 41–52.
18. Soni, D.K., Effect of Saturation and Deformation Rate on Split Tensile Strength for Various Sedimentary
Rocks, Int. Conference Data Mining, Civil and Mechanical Engineering, Bali, Indonesia, 2015, pp. 53–55.
19. Li, Z. and Reddish, D.J., The Effect of Ground Water Recharge on Broken Rocks, Int. J. Rock Mech. Min.
Sci., 2004, vol. 41, no. 3, pp. 1B 14.
20. Duperretl, A., Taibil, S., Mortomore, R.N., and Daigneault, M., Effect of Groundwater and Sea Water
Weathering Cycles on the Strength of Chalk Rock from Unstable Coastal Cliffs of NW France, Eng. Geol.,
2005, vol. 78, pp. 321–343.
21. Li, Z., Reddish, D.J., and Sheng, Y., Experimental Investigation of the Effect of Water on the Strength
Evolvement of Fractured Siltstone, Geotech. Spec. Pub. 150 ASCE, 2006, pp. 177–183.
22. Vasarhelyi, B. and Van, P., Influence of Water Content on the Strength of Rock, Eng. Geol., 2006,
vol. 84, pp. 70–74.
23. D4543. Standard Practices for Preparing Rock Core as Cylindrical Test Specimens and Verifying
Conformance to Dimensional and Shape Tolerances, ASTM, 2008.
24. US Army Corps of Engineers. http://www/gsl.erdc.usace.army.mil/SL/MTC/handbook/RT/RTH/116-
95.pdf,1995.Cited July, 9, 2012.

JOURNAL OF MINING SCIENCE Vol. 54 No. 6 2018

 WATER SATURATION INFLUENCES ON ENGINEERING PROPERTIES 929

25. Roxborough, F.F. and Rispin, A., The Mechanical Cutting Characteristics of the Lower Chalk, Tunnels
and Tunnelling, 1973, pp. 45–67.
26. Torok, A. and Vasarhelyi, B., The Influence of Fabric and Water Content on Selected Rock
Mechanical Parameters of Travertine, Examples from Hungary, Eng. Geol., 2010, vol. 115, pp. 237–245.
27. Abu Bakar, M.Z. and Gertsch, L.S., Saturation Effects on Disc Cutting of Sandstone, American Rock
Mech. Association, 45th US Rock Mech., Geomech. Symp., San Francisco, CA, 2011, vol. 254, pp. 1–9.
28. ISRM. Suggested Methods for Determining the Uniaxial Compressive Strength and Deformability of Rock
Materials, Int. J. Rocks Mech. Min. Sci. Geomech. Abstr., 1979a, b, vol. 16, pp. 135–140.
29. ISRM. Suggested Methods for Determining Tensile Strength of Rock Materials, Int. J. Rocks Mech. Min.
Sci. Geomech. Abstr., 1978a, b, vol. 15, pp. 99–103.
30. Khandelwal, M. and Ranjith, P.G., Correlating Index Properties of Rocks with P-Wave Measurements,
J. Applied Geophysics, 2010, vol. 71, pp. 1–5.
31. ISRM. Suggested Methods for Determining Sound Velocity, Int. J. Rocks Mech. Min. Sci. Geomech.
Abstr., 1978, vol. 15, pp. 53–58.
32. Karakus, M., Kumral, M., and Kilic, O., Predicting Elastic Properties of Intact Rocks from Index Tests
Using Multiple Regression Modeling, Int. J. Rocks Mech. Min. Sci., 2005, vol. 42, pp. 323–330.
33. ISRM. Suggested Methods for Determining Water Content, Porosity, Density, Absorption and Related
Properties and Swelling and Slake-Durability Index Properties, Int. J. Rocks Mech. Min. Sci. Geomech.
Abstr., 1979a, b, vol. 16, pp. 141–156.
34. Paschen, D., Petrographic and Geomechanical Characterization of Ruhr Area Carboniferous Rocks for
the Determination of Their Wear Behavior, PhD Dissertation, Technische Unversitat Claustahl, 1980.
35. Wiid, B.L., The Influence of Moisture Content on the Pre-Rupture Fracturing of Two Rock Types, Proc.
of the 2nd Congress of the Int. Society of Rock Mech., Belgrade, 1970, vol. 3, pp. 239–245.
36. Kitaowa, M., Endo, G., and Hoshino, K., Influence of Moisture on the Mechanical Properties of Soft
Rock, Proc. of the 5th National Symp. on Rock Mech., Japan, 1977.
37. Bell, F.G., The Physical and Mechanical Properties of the Fell Sandstones, Northumberland, England,
Eng. Geol., 1987, vol. 12, pp. 1–29.
38. Hassani, F.P., Whittaker, B.N., and Scoble, M.J., Strength Characteristics of Rocks Associated with Open
Cast Coal Mining in the UK, Proc. of the 20th U. S. Symp. on Rock Mech., Austin, 1979, pp. 347–356.
39. Ferreira, R., Monteiro, L.C.C., Peres, J.E., and Prado, Jr.F.A.deA., Analise de Alguns Fatores que
Infleum na Resistencia a Compressao do Arenito Bauru, 3rd Brazilian Congress of Engineering Geology,
ABGE, Itapema, 1981, vol. 3, pp. 89–102.
40. Priest, S.D. and Selvakumar, S., The Failure Characteristics of Selected British Rocks, A Report to the
Transport and Research Laboratory, Department of Environment and Transport, Imperial College,
London, 1982.
41. Koshima, A., Frota, R.G.Q., Lorano, M.H., and Hoshisk, J.C.B. de F., Comportamento e Propriedades
Geomechanicas do Arenito Bauru, Simposio Geotecnico Sobre Bacio Alto Parana, ABGE-ABMS-CBMR,
Sanpaulo, 1983, 2b, pp. 173–189.
42. Pells, P.J.N. and Ferry, M.J., Needless Stringency in Sample Preparation Standards for Laboratory Testing
of Weak Rocks, Proc. of the 5th Congress of the Int. Society of Rock Mech., Melbourne, 1983, pp. 203–207.
43. Dobereiner, L., Engineering Geology of Weak Sandstones, PhD Thesis, Imperial College London, 1984.
44. Dyke, C.G., The Re-Peak Deformation Characteristics of Sandstone at Varying Moisture Contents, M. Sc.
Thesis, Imperial College London, 1984.
45. Gunsallus, K.L. and Kulhawy, F.H., A Comparative Evaluation of Rock Strength Measures, Int. J. Rock
Mech. Min. Sci. Geomech. Abstr., 1984, vol. 21, pp. 233–248.

JOURNAL OF MINING SCIENCE Vol. 54 No. 6 2018

930 MAJEED, ABU BAKAR

46. Denis, A., Durville, J.L., Massieu, E., and Thorin, R., Problemes Poses par un Calcaire Tres Poreux Dans
I’etude de la Stabilite d’une Carrier Souterraine, Proc. of the 5th Congress of the Int. Association of Eng.
Geology, Buenos Aires, 1986, pp. 549–557.
47. Howarth, D.F., The Effect of Pre-Existing Microcavities on Mechanical Rock Performance in Sedimentary
and Crystalline Rocks, Int. J. Rock Mech. Min. Sci. Geomech. Abstr., 1987, vol. 24, pp. 223–233.
48. Pells, P.J.N., Substance and Mass Properties for the Design of Engineering Structures in the Hawkesbury
Sandstone, Aust. Geomech., 2004, vol. 39, pp. 1–21.
49. Hui, Y., Xueliang, J., and Na, L., Experimental Study on Mechanical Property of Peridotite under
Water-Rock Interaction, EJGE, 2014, vol. 19, pp. 1179–1188.
50. Tugrul, A., The Effect of Weathering on Pore Geometry and Compressive Strength of Selected Rock
Types from Turkey, Eng. Geol., 2004, vol. 75, pp. 215–227.
51. Sabatakakis, N., Koukis, G., Tsiambaos, G., and Papanakli, S., Index Properties and Strength Variations
Controlled by Microstructure for Sedimentary Rocks, Eng. Geol., 2008, vol. 97, pp. 80–90.
52. Yilmaz, N.G., Yurdakul, M., and Goktan, R.M., Prediction of Radial Bit Cutting Force in High-Strength
Rocks Using Multiple Linear Regression Analysis, Int. J. Rock Mech. Min. Sci., 2007, vol. 44, pp. 962–970.
53. Samaranayake, V.A., Statistical Data Analysis, STAT-353 Course, Missouri University of Science and
Technology, Rolla, MO, USA, 2009.
54. Majeed, Y. and Abu, Bakar, M.Z., Statistical Evaluation of Cerchar Abrasivity Index (CAI) Measurement
Methods and Dependence on Petrographic and Mechanical Properties of Selected Rocks of Pakistan, Bull.
Eng. Geol. Env., 2016, vol. 75, no. 3, pp. 1341–1360.
55. Kennedy, P., A Guide to Econometrics, 6th Edition Oxford, Willey Blackwell, 2008.
56. Hair, J.F., Black, W.C., Babin, B.J., and Anderson , R.E., Multivariate Data Analysis, 7th Edition, Prentice
Hall, New York, 2009.
57. Grima, M.A. and Babuska, R., Fuzzy Model for the Prediction of Unconfined Compressive Strength of
Rock Samples, Int. J. Rock Mech. Min. Sci., 1999, vol. 36, pp. 339–349.
58. Gokceoglu, C., A Fuzzy Triangular Chart to Predict the Uniaxial Compressive Strength of Ankara
Agglomerates from Their Petrographic Composition, Eng. Geol., 2002, vol. 66, pp. 39–51.
59. Gokceoglu, C. and Zorlu, K., A Fuzzy Model to Predict the Uniaxial Compressive Strength and the
Modulus of Elasticity of a Problematic Rock, Engineering Applications of Artificial Intelligence, 2004,
vol. 17, pp. 61–72.

JOURNAL OF MINING SCIENCE Vol. 54 No. 6 2018