Adapted Schmidt Hardness Testing on Large Rock Samples—Kanfanar-South Quarry Case Study

Abstract

This paper deals with the possibility of using the Schmidt hardness test, which does not require much preliminary preparation and is easy to perform, in the production of commercial blocks in a quarry. Previous recommendations for Schmidt hardness testing of rock materials were specifically related to tests performed for geomechanical purposes. They also referred mostly to smaller samples, but testing of commercial blocks has some special features, mainly because they are large samples for which practically larger areas must be tested. This paper presents the testing methodology in terms of the number and position of hammer strikes on a commercial block, as well as the application of corrections to the test results in terms of the way the blocks are cut. The tests were conducted on natural stone blocks from the Kanfanar-South quarry, which is characterized by limestone rocks. The test results show that the test methodology can be applied in quarries with similar geological structure and natural stone mining methodology.

1. Introduction

Hardness is a measure of the resistance of a material to localized elasto-plastic deformation. Early hardness tests were based on the ability of one mineral to scratch another, softer mineral. Over time, the determination of hardness became one of the most used methods for testing the physical and mechanical properties of materials, and many different methods have been developed. Primarily, quantitative hardness testing methods were developed to test metals, and later attempts were made to apply this methodology to other materials [1]. There are several hardness testing methods that have been applied to meet different ore testing requirements. For example, the Vickers test is very suitable for testing the hardness of coal [2,3,4]. However, today, the most common method for determining the hardness of rock materials is the use of the Schmidt hammer. Originally developed to determine the strength of concrete [5,6], the Schmidt hammer has been used to determine the hardness of rock materials since the 1960s [7,8,9,10,11,12,13,14] and to determine the strength of fracture walls [15]. It also has the advantage that it can be used to independently evaluate other rock properties, such as the drillability and cutability of rock. Several models have been proposed to estimate the cutability of rocks and the production rate of chain cutters and diamond wires, which are based on the Schmidt hardness (SH) values [16,17,18,19,20,21].

Ogunsola et al. conducted tests on marble and found a reduction in the values of the Schmidt hardness and other physical and mechanical properties under the influence of weathering [22]. Recently, the effect of weathering [23] on the determination of hardness has been studied, and attempts are being made to use it to determine the age of rock material. Schmidt hardness can be easily determined in any direction, for example, in the direction of movement of mining, cutting, or drilling equipment, as well as for checking the quality of mine roof strata and for the purpose of classifying rock mass excavation opportunities [1,24].

Due to its accessibility and ease of use, the Schmidt hammer is an ideal instrument for testing rock materials, as evidenced by its increasing popularity and wide range of applications, but primarily from a geomechanical perspective and for estimating uniaxial compressive strength [11,12,25,26,27,28,29].

Quarrying has a long tradition in Croatia [30,31], and natural stone is the most valuable mineral resource mined in Croatia today. Therefore, the authors of this paper wanted to investigate the possibility of applying the Schmidt hammer to natural stone from Croatia. Natural stone is an important mineral resource in the economy, and there is a need for sustainable mining methods in response to the challenges of our time [32]. Natural stone quarrying is very specific to other mineral raw materials in terms of the technology of extraction. The products of a natural stone quarry are large blocks extracted from intact stone and are subsequently processed into blocks, slabs, tiles, etc. Therefore, it is expedient to apply a method of excavation that causes the fewest stresses and deformations in the rock mass. For this reason, most natural stone extraction methods avoid the use of explosives, which are used only to remove overburden. The occurrence of deformations and stresses can be reduced using diamond wire saws and chain cutters, making sawing or splitting the most acceptable way of obtaining blocks in natural stone quarries.

The quality of the blocks is mostly related to the physical–mechanical properties and the integrity of the rock mass, whereas the economic value of the blocks is more often determined by the decorativeness of the stone. The properties are determined by the origin of the stone deposit, genetic processes, and post-genetic occurrences [33,34]. However, the appearance and quality of the blocks also depend on the applied technology and cutting method, which is especially evident in sedimentary rocks with prominent anisotropy.

There are several other scientific works in which the hardness of natural stone has been studied. Pamuk and Büyüksaraç [35] performed non-destructive tests on natural stone from the Ürgüp area. N-type Schmidt hammer tests were performed on 10 × 10 × 10 cm samples collected in the Cappadocia region of Turkey. Vişne, Devetüyü, Gülkurusu, Kirli Beyaz, and Sarı were the natural stone types studied. On each sample, tests were performed on three different sides of the sample, and on each side, 10 strikes were performed on the template in the form of a grid, after which the mean value was calculated. The mean rebound value ranges from 15.8 to 23. Kong et al. [36] presented guidelines for using the Schmidt hammer test to predict uniaxial compressive strength (UCS) in quarrying building stones, and to assess the stability of building structures. Furthermore, they concluded that grain size and anisotropy have a significant effect on the estimation of the correlation between UCS and the Schmidt hammer test. An increase in temperature affects the hardness of Schmidt, and research has identified a critical temperature of 600 °C [37]. After this temperature, the hardness value decreases drastically. Specifically, a decrease in the hardness of the onyx samples by almost 64–69% was found.

From the scientific literature reviewed (

Table 1. A review of the use of Schmidt hardness for various purposes.

Researchers	SH Application	Material
Yaalon and Singer [7]	Variation in strength and porosity	Calcrete
Day and Goudie [8] Day [9]	Geomorphology	Carbonate rocks
Williams and Robinson [10]	Correlation of surface texture	Sedimentary rocks
Katz et al. [12] Hamzaban et al. [28]	Evaluation of mechanical rock properties	Chalk, limestone, sandstone, marble, granite
Cargill and Shakoor [25] Briševac et al. [26] Dinçer et al. [38] Karaman and Kesimal [24] Briševac et al. [27]	Estimating the UCS	Basalt, metabasalt, dacite, limestone, volcanic breccia,
Demirdag et al. [39]	Effect of sample size on SH	Andesite, limestone, marble travertine
Gupta et al. [40]	Surface hardness of natural and modified rocks	Quartz mica gneiss, schist, quartzite, and calc-silicate
Cerna and Engel [41]	SH variation for a granite outcrop	Granite
Copur et al. [16] Tumac et al. [18] Tumac [17] Rasti et al. [19] Mikaeil [20]	Predicting the sawability of carbonate rocks mechanical rock properties	Travertine, marble, limestone
Pamuk and Büyüksaraç [35]	Testing of SH on smaller samples of dimensions 10 × 10 × 10 cm	Natural stone
Ogunsola et al. [22] Karmakar et al. [23]	Effect of weathering on rock properties	Marble
Khoshouei et al. [21]	Correlation with rock acoustic signs	Travertine, limestone, marble, granite, sienite, andesite, biomicritic limestone, sandy limestone
Karakul [42]	Effect of impact direction on SH	Andesite, travertine. limestone marble, tuf, ignimbrite
Kong et al. [36]	Effect of grain size and anisotropy on the correlation between UCS and SH	Fine-grained sandstone,
Özdemir [37]	Effect of high-temperature treatments on rock properties	Limestone, dolomite

SH—Schmidt hardness, UCS—uniaxial compressive strength.

), it can be concluded that the study of Schmidt hardness from the point of view of using the uniformity of the quality of natural stone has not yet been sufficiently studied scientifically. In this sense, the main purpose of this paper is to study the possibility of applying simple non-destructive testing methods, such as the determination of Schmidt hardness, which could be used to determine the uniformity of block quality. For this reason, the main aim of this paper is to investigate the possibility of performing a simple Schmidt hardness test in the production of commercial blocks in a quarry of natural stone and the application of such determined hardness to determine the uniformity of the quality of the blocks that are produced. Larger samples than those tested so far in previously published scientific papers must be tested. Therefore, it is necessary to adapt the methodology of the Schmidt hammer test to larger sample surfaces and study the effects of the test on differently cut surfaces of commercial blocks.

2. Materials and Research Methodology

Field tests to determine Schmidt hardness were carried out at the Kanfanar-South natural stone quarry (Figure 1), which is located on the Croatian peninsula of Istria.

2.1. Geological Properties of Rock Material

The limestone from the Lower Cretaceous known as Kanfanar, also known as “Istrian yellow” (Ital. Giallo d’Istria), is the most significant and well-known rock quarried nowadays in Croatia. The Kanfanar-South quarry, located in Istria on the territory of the municipality of Kanfanar (Figure 1), i.e., around the surface distribution of the lithostratigraphic unit Kanfanar (lower Apt; K₁⁴), is one of the locations where it is quarried (Figure 2). This unit’s basement is made up of limestones of the Dvigrad Formation, which are morphologically characterized by a stromatolite facies (LLH stromatolites) [43]. The sequence of the Dvigrad Formation deposits ends with the appearance of the first thicker-bedded limestones with mudstone alternations and Bacinella irregularis oncoid floutstones of the roof deposits of the Kanfanar Formation [43].

The area of the former Adriatic carbonate platform emerged regionally in the middle of the Aptian, and the terrestrial phase lasted according to the paleogeographic characteristics of the terrain and the local effects of synsedimentation tectonics. The regional apto-Albanian emersion, which is recognized by an interval more than 1 m thick of greenish-gray, silty, dolomitized, and marly sediments with emersion breccias followed by limestones and dolomites of the Crna Formation, defines the roof of the Kanfanar Formation [43].

For the preparation of this paper, Schmidt hardness tests were carried out in the underground mine workings on the excavation layers, which are locally marked with Roman numerals from “I” to “V”, which correspond to the Kanfanar Formation’s lower lithofacies.

Very large Bacinella irregularis accumulations (up to 5 cm) are prevalent in the zone of exploitation “layers” (I to V), which is vertically bounded by madstone, wackestone, and pekstone alternations (typical “Istrian yellow” Bacinella irregularis oncoid facies). The mentioned interval is characterized by vast Bacinella irregularis accumulations of reticulate structures with or without mantle and dense reticulate stylolites, primarily composed of peloidal–skeletal wackestone to pexton. Small benthic foraminifera and numerous Nubecularia are also present.

In the interval roughly defined by the technical “layers” III to V, the rock is bluish in color (the so-called “blue interval”). Locally, the facies is very similar to the one previously described. The difference is the presence of dark organic material in the pore space (Bacinella irregularis accumulations and fenestrae), a zone protected from oxidation; organic matter and pyrite give this region a bluish color.

2.2. Basic Principles of Schmid Hardness Testing

There are L-type and N-type Schmidt hammers that differ in impact energy, which is 0.735 Nm for the L-type hammer and 2.207 Nm for the N-type hammer. The L-type has a higher sensitivity in a smaller range and gives better results when testing weaker, porous, and weathered rock. In the last decade, a digital version of this test device has been invented. It differs from the classical version in the method of measuring the absorbed kinetic energy. The classic version of the hammer measures how much the weight bounces off its initial position, whereas the digital version of the hammer measures the percentage of energy absorbed as the ratio of the velocity before and after the impact. Both versions of the hammer measure the absorbed kinetic energy, although the results are different for the classic version of the hammer due to the influence of gravity. The fact that the results do not need to be normalized is one of the main advantages of the digital version of the hammer over the classic version of the hammer. The biggest difference and advantage of the digital version of the hammer compared to the analog version is the acquisition, storage, and processing of the data [13,15,44].

Samples to be examined should be intact, petrologically uniform, and representative of the host rock to be examined. Surfaces and contact points should be clean and free of dust and particles. Abrasive material may be used to smooth surfaces during fieldwork if necessary. Fine sandpaper may also be used to smooth the surface of specimens and cores, especially if sawing and drilling cause sharp edges. The minimum specimen size is 84 mm for N-type hammers and 54.7 mm for L-type hammers. The required minimum thickness of the specimen is 100 mm at the contact point. In addition, the contact point must not be too close to the edge of the specimen because of possible cracking or scattering of the shock waves. For these reasons, the test point should be one radius length from the nearest edge of the specimen core and half a thickness length from the edges of the specimen. For example, assuming that the distance between contact points is 2 cm, a core length of 43.5 cm or a block area of 268 cm² (for a thickness of 10 cm) is required to obtain 20 readings. If 10 consecutive readings differ by only 4 (±2), the test can be stopped, and it can be assumed that enough strokes have been performed. The size and arrangement of the grains and the relative hardness of the matrix significantly affect the degree of dispersion of the rebound values. If the tested surface contains grains of a diameter similar to that of the impact piston, the measured values may deviate greatly from the average. In such cases, it is necessary to perform a special test on the contact points where only grains are located and on the part where the matrix is located. Microchanges in the microstructure caused by weathering lead to different rebound values and influence them significantly. For this reason, it is necessary that the samples of the material are uniformly affected and consumed by the atmosphere throughout the volume and that the samples have the same degree of humidity [13].

2.3. Description of Research Methodology

The research methodology had to be adapted to the size of the samples, which are not small but large. For this reason, a procedure was developed that involved several steps. The first step is to select enough suitable blocks for testing. Nine blocks from each of the five layers were tested, and their surfaces had to be cut using different machines. For the second step, a methodology with bulk and scatter testing procedures was developed so that different surfaces could be tested using these procedures. The third step is to analyze the results obtained. The last step is the selection of the most appropriate test method. Figure 3 illustrates the research methodology in this paper.

Quarrying of natural stone in the Kanfanar-South quarry is carried out underground [37], thus, blocks of stone with a maximum weight of up to 30 tons are produced with a chain saw. The shape of the block is such that the maximum length is up to 3.5 m, the maximum width is up to 2.0 m, and the maximum height is up to 2.0 m. Realistically, the dimensions of the block depend on the thickness of each geological layer, as shown in Figure 4.

The chain cutter can cut four sides, and the back side breaks off so that the block can be pulled out. This back side is then sawed using diamond wire, which leaves a smooth surface finish. This quarrying technique makes it possible to study the Schmidt hardness of blocks with different cutting surfaces.

After the blocks were extracted from the rock mass, they were marked as to which layer they belonged to and transported from the underground to the surface test site. A RockSchmidt digital hammer was used for the hardness test, and a test form was prepared for each test (Figure 5).

To investigate the influence of surface treatment, a test was performed on blocks whose sides were cut with a chain saw and a diamond wire saw (Figure 6). In this way, a realistic comparison of the test results on two differently treated surfaces was possible. The number of strikes for each test is 20 and is identical to the ISRM recommendations. The principle of operation is very simple. The probe of the hammer is placed on a certain spot and pressed (Figure 7a). Then, the device performs the impact against a hard surface, whereupon its inner impact element bounces to a certain height, which is read either analogously or on the display of the device.

However, it was decided that data collection should be conducted in two ways. The first was called the bulk method, in which the Schmidt hammer strikes were placed on a 10 cm × 12 cm area (Figure 7a). This method is identical to all previous Schmidt hardness tests. The second method of data collection represents a step toward testing large specimens; in this method, 20 hammer strikes are collected scattered over the surface of the entire commercial block (Figure 7b).

For each test, the average value of 20 strikes was calculated according to Equation (1):

$$\overline{x}=\frac{\left({{\displaystyle \sum }}_i^{20}{x}_i\right)}{20},$$

(1)

where “$\overline{x}$” stands for the mean values, and “x_i” for the individual x-values. Then, the results were grouped according to the test method (bulk and scatter) and according to the production of the test surface (chain saw machine and diamond wire saw machine). For these groups, the mean, standard deviation (2), and coefficient of variation (3) were calculated according to the following equations:

$$s=\sqrt{\frac{1}{n}{\displaystyle \sum }_{i=1}^n{\left(x_i-\overline{x}\right)}^2}$$

(2)

$$cv=\frac{s}{\overline{x}}100$$

(3)

where “s” stands for the standard deviation, “n” for the number of tests (in this case 40), “x_i” for the individual x-values, “$\overline{x}$” for the mean of all adjusted values, and “cv” for the coefficient of variation.

3. Results

A total of 40 commercial blocks of natural stone were examined. Blocks were selected, and the individual faces of the blocks were cut with diamond wire saws and chain saws. After the commercial blocks were examined, the test results were statistically analyzed. Descriptive statistics values were calculated: mean, minimum value, maximum value, standard deviation, and coefficient of variation.

Table 2. Results according to the test method.

	Bulk Method	Scattered Method
Number of tests	40	40
Mean value	74.33	73.86
Minimum value	67.33	67.53
Maximum value	79.00	78.53
Standard deviation	3.11	3.06
Coefficient of variation	4.18	4.15

and Figure 8a show a comparison of the results according to the test method.

Table 3. Results according to the type of machine used to cut the tested surfaces.

	CSM	DWSM
Number of tests	40	40
Mean value	71.48	76.62
Minimum value	67.33	70.35
Maximum value	75.53	79.00
Standard deviation	2.06	1.27
Coefficient of variation	2.89	1.66

CSM—Chain Saw Machine; DWSM—Diamond Wire Saw Machine.

and Figure 8b show a comparison according to the type of machine used to cut the tested surfaces.

The mean values of the exploitation layers from I to V were also calculated, and the results are shown in Figure 9.

Table 4. Descriptive statistics of test results of layers on cuts made with a chain saw machine.

Layers	I	II	III	IV	V
Number of tests	16	16	16	16	16
Mean value	71.41	72.74	70.84	72.17	70.26
Minimum value	68.73	69.03	67.53	69.75	67.33
Maximum value	74.58	75.53	74.73	74.50	75.53
Standard deviation	1.57	1.87	2.04	1.72	2.22
Coefficient of variation	2.19	2.57	2.88	2.38	3.16

shows a detailed presentation of the descriptive statistics of the Schmidt hardness test on the surface cut with the chain saw. Similarly,

Table 5. Descriptive statistics of layer results on cuts made with a diamond wire saw machine.

Layers	I	II	III	IV	V
Number of tests	16	16	16	16	16
Mean value	76.34	76.62	76.54	77.08	76.92
Minimum value	74.13	74.45	73.98	75.43	68.45
Maximum value	78.03	77.53	77.93	79.00	78.80
Standard deviation	0.92	1.07	1.11	0.99	3.52
Coefficient of variation	1.21	1.39	1.45	1.28	4.57

CSM—Chain Saw Machine; DWSM—Diamond Wire Saw Machine.

shows the descriptive statistics of the test on the surface cut with a diamond wire saw.

4. Discussion

To determine a representative result, 40 larger blocks in total were examined, the joint surfaces of which were cut with both a chain saw and a diamond wire saw. The surface thus prepared was tested using the bulk and scattered data collection method. In this way, a total of 160 tests were performed, and the result of each test was calculated based on the average of 20 strikes with a Schmidt hammer. In this way, the issue of an insufficient number of tests, which occurred in the first investigation [45], is overcome, and the results can be considered relevant.

Most scientific papers treat hardness according to the recommendations of ISRM [15] and ASTM standards [14]. Demirdag et al. [39] compared these two methods in terms of the smallest sample size and concluded that the edge dimension of the cubic block should be at least 11 cm to obtain a uniform hardness value. However, the authors of this paper have not found any papers dealing with the study of large specimens of which the edges are several meters in size. Therefore, it was necessary to develop a methodology for studying such large samples in the form of a scattering method of data collection. Looking at the results of data collection using the scatter or bulk method (Table 2 and Figure 8a) revealed that all statistical parameters of result processing are almost the same, as the mean value differs only by 0.47. Because the bulk method is a common and already established method of data collection, we can consider it a control method in this study, i.e., we can check how successful the scatter method is. Because the results are practically the same, it can be assumed that the scatter method is a valid data collection method for large rock samples. At the very least, both data collection methods can be used successfully and completely equivalently.

Standard determination of Schmidt hardness requires sanding the surface of the material to be tested [14,15], but tests have also been performed on untreated and treated sample surfaces of certain materials, for example, granite [41], quartz mica gneiss, schist, quartzite, and calc-silicate [40]. All studies show an increase in Schmidt hardness values for sanded surfaces. Following these trends, this study examined variously processed block surfaces. A comparison of the test results of surfaces produced with different cutting machines shows a larger deviation (Table 3 and Figure 8b). It is already known that the surface should be treated with an abrasive before testing, so a difference between the results on the rough surface caused by the chain saw machine and the smooth surface of the wire saw machine is to be expected. Figure 8b shows that the Schmidt hardness is 7% lower on a rough surface. It was necessary to investigate the difference in hardness depending on the layers (Figure 9) from which the blocks were taken, so the difference was processed separately and is presented in

Table 6. Differences in hardness according to the layers.

Layers	I	II	III	IV	V
DWSM	76.34	76.62	76.54	77.08	76.92
CSM	71.41	72.74	70.84	72.17	70.26
Difference	4.93	3.88	5.7	4.91	6.66
Difference in %	6.5	5.1	7.4	6.4	8.7

CSM—Chain Saw Machine; DWSM—Diamond Wire Saw Machine.

. The smallest decrease is in layer II and is 5.1%; the largest, which is in layer V, is 8.7%. It can be concluded that the test surfaces made with the chain saw machine need to be treated with an abrasive material before the test, whereas the surfaces made with a diamond wire saw do not need to be additionally treated with an abrasive material before the test.

The results of this study were compared with the hardness of the stone material, which was determined using small samples. This was possible because in the previous paper [46], four smaller blocks measuring 23 × 30 × 23 cm from layers I, II, IV, and V were studied.

Table 7. Comparison of hardness determined on small samples and commercial blocks from the Kanfanar-South quarry.

Layers	I	II	IV	V
Min value according to [46]	55.5	60.0	61.0	59.5
Max value according to [46]	66.5	67.0	68.5	69.5
DWSM surface in this study	76.34	76.62	77.08	76.92
CSM surface in this study	71.41	72.74	72.17	70.26

CSM—Chain Saw Machine; DWSM—Diamond Wire Saw Machine.

shows that all the results obtained in this study are above the maximum values obtained in previous studies. This is true for both types of surfaces studied, which were cut with different tools. Because this study examined many more samples in the form of commercial blocks, the results are more relevant and should serve as a basis for future field tests.

Every company that produces natural stone is obliged to test the properties of the stone for the purpose of laying it according to European standards within certain periods of time. The aim of this research is not to replace these tests, but the authors of this paper recommend using Schmidt hardness as a simple indicator of the uniform quality of each commercial block cut in a quarry, because the quality of blocks in production is currently determined only visually. Future research should focus on determining the correlation between Schmidt hardness and more important physical–mechanical properties, such as compressive strength, flexural strength, wear resistance, and water absorption.

In previous studies [12,25,26,27,28,29,38,42] on the estimation of uniaxial compressive strength using Schmidt hardness, the determination of uniaxial compressive strength was usually performed according to the recommendations of ISRM, i.e., from a geomechanical point of view. However, the authors of this paper believe that the models for estimation should be based on a specific uniaxial compressive strength according to the standards of EN [47], i.e., from the perspective of the use of natural stone in construction. Due to the different dimensions of specimens in tests according to EN and ISRM, the results for a specific strength differ, so previous research and estimation models cannot be applied to the production of natural stone because uniaxial strength according to the European standards is relevant for the production and sale of natural stone.

From a global perspective, the energy crisis requires solutions in all sectors of society, including natural stone production, and this industry must use processes that require less energy in all phases of production. In this sense, this paper contributes to the application of a simple device (Schmidt hammer) that can be used easily and reliably on the cut surfaces of a commercial block without additional sanding, which makes it easier to determine the uniform quality of the blocks produced. In this way, a small step is taken towards more sustainable production.

5. Conclusions

After the case study of the Kanfanar-South quarry and the detailed analysis of the results and discussion, conclusions were reached on the possibility of using the Schmidt hammer for testing large natural stone samples for the purpose of determining the uniformity of the quality of commercial blocks.

• The commercial blocks in the quarry have a uniform hardness, and no major variations were found in the blocks obtained from different quarrying layers. The average tested Schmidt hardness by mining layer was 76.34 for layer I, 76.62 for layer II, 76.54 for layer III, 77.08 for layer IV, and 76.92 for layer IV.
• The scatter method is a valid data collection method for determining Schmidt hardness on large rock samples. It is more suitable for checking the uniformity of the quality of natural stone blocks than the bulk method, which is commonly used to determine Schmidt hardness.
• The proposed method considers the sustainability of the test and does not promote an increase in energy consumption, so the test surface does not need to be sanded before the hardness test. However, it should be noted that a lower value of up to 8.7% can be expected if the test is performed on a surface that was cut with a chain cutter in the Kanfanar-South quarry.
• By applying the Schmidt hammer to check the uniformity of the quality of natural stone blocks, sustainable production can be established, which is what global trends strive for. However, because this method has never been used before, it needs to be tested for some time in several other quarries to fully verify it under different mining conditions and finally write a technical note.
• Further research should be directed towards the creation of a model for the estimation of uniaxial compressive strength determined in accordance with the European standard that specifies the method for determining the uniaxial compressive strength of natural stone.