Introduction

In civil engineering, rock aggregates are essential materials applied in concrete, mortar, road, and railway infrastructures (Afolagboye et al. 2016, 2017; Kamani and Ajalloeian 2019). The evaluation of the durability of aggregates is an essential task in designing. One of the most important properties of crushed stones is the resistance to mechanical degradation which phenomenon can take place when the raw materials are deposited, transported, subjected to loading, and constructed. The different effects result in different types of stress, just as abrasion, crushing, impacting, and disintegration (Kamani and Ajalloeian 2019). Numerous tests have been developed to understand the mechanical properties of aggregates. The most widely used aggregate degradation tests are the following:

  • micro-Deval Abrasion Test,

  • Los Angeles Abrasion Test,

  • Impact Test (EN),

  • Impact Test (BS),

  • Crushing Test.

Although these tests are very common, they are performed according to different standards in different parts of the world. Which fact makes it difficult to compare or use the previous research results. Furthermore, the lack of uniform terminology makes it difficult to interpret the results and understand the physical processes. In the international literature on mechanical weathering of aggregates, there is a wide variety of word usage for wear. A good example of this is the use of the terms of attrition and abrasion. In most cases, these words are used as synonyms for each other. The subdivision of Bach (2013) deviates from this trend. In his typology, the main mechanical weathering group is named attrition and divided into three subgroups: fragmentation of grains, rounding (or “chipping”), and abrasion (or “polishing”) (see Fig. 1). The authors of this article intend to follow this typology.

The authors consider it important to note that there are also large-scale tests in special laboratory conditions, which are better corresponding with the real circumstances’ that effect railway ballast aggregate behavior (Alabbasi and Hussein 2019). Such large-scale experimental tests are the following:

  • different scale of ballasted track models (Al-Shaer et al. 2008; Nålsund et al. 2013; Koike et al. 2014; Kumara and Hayano 2016).

  • large-scale box tests (Stewart et al. 1985; Raymond and Bathurst 1987; Selig and Waters 1994; Indraratna et al. 2000; Hussaini et al. 2016; Fischer 2017; Navaratnarajah et al. 2018; Esmaeili et al. 2022; Esmaeili and Askari 2023),

  • large-scale direct shear tests (Huang et al. 2009; Stark et al. 2014; Estaire and Santana 2018; Jing et al. 2020),

  • large-scale oedometer tests (McDowell et al. 2003),

  • large-scale shaker tests (Esmaeili et al. 2016).

  • and large-scale triaxial tests (Indraratna et al. 1998, 2005, 2012, 2015; Anderson et al. 2001; Fair 2004; Selig and Frenkel 2005; Lackenby et al. 2007; Anderson and Fair 2008; Sevi et al. 2009; Aursudkij et al. 2009; Nålsund 2010; Trinh et al. 2012; Nålsund et al. 2013; Mishra et al. 2013; Qian et al. 2014; Sun et al. 2014, 2016, 2017, 2018; Bian et al. 2016; Sun and Zheng 2017; Alabbasi and Hussein 2019; Xie et al. 2024).

In recent decades, the number of research in which the mechanical behavior of crushed stones has been investigated with the help of discrete element modeling has increased. During these endeavors, the models were often calibrated based on some kind of laboratory test. The international literature provides examples of the use of both conventional testing, such as micro-Deval test (Deiros et al. 2016; Wu et al. 2018) and large-scale testing, such as box test (Lim 2004), direct shear test (Wang et al. 2015; Guo et al. 2020; Liu et al. 2022), triaxial test (Qian et al. 2013, 2018; Ngo et al. 2016), oedometer test (Lim 2004), Hummel test (Orosz et al. 2017, 2019).

Despite the popularity of the complex modeling procedures mentioned above increasing in parallel with the development of computer technology, traditional degradation tests are still used in many cases. The reason for this is that the standards that form the basis of these laboratory tests are internationally accepted, and the qualification of construction aggregates - and thus their install ability - is based on the results of the tests described in the standards.

Due to the decrease in raw material sources, it is becoming more and more challenging to meet the demand for large quantities of aggregates in the construction industry. Secondary areas and naturally occurring aggregate deposits are exploited to satisfy these needs, and the concept of recycling also appeared. The latter is also favorable from a nature conservation point of view, as no new mines have to be opened, and it reduces construction waste (Bai et al. 2020). The topic researched in the international literature is the applicability of slag ballast. In these mixtures, all or a part of the rock aggregate is replaced with industrial slag (Jing et al. 2020; Esmaeili and Askari 2023). As a result, much new research is being carried out, which examines stone aggregate materials from alternative sources (Çavdar and Çavdar 2024), recycled aggregates (Adomako et al. 2023), or aggregate mixtures – such as steel slag aggregates (Hussain and Hussaini 2022; Paixão and Fortunato 2023; Olagunju et al. 2023) – with including traditional laboratory tests. This is because these tests are relatively easy to implement and beneficial from an economic point of view, as well as the mentioned qualification expectations.

The application of machine learning, a subfield of AI, is also being researched to predict the mechanical degradation of aggregates, which is based on the Impact Test (Koohmishi and Guo 2023).

The authors aimed to present studies on the forms of mechanical degradation, their measurement, and quantification. Some papers in the literature deal with one smaller group of the five laboratory tests, thus providing a moderate amount of literature review (Al-Harthi 2001; Erichsen et al. 2011; Palassi and Danesh 2016; Agyeman et al. 2019). This article aims to provide an extensive overview of mechanical degradation tests, related standards, and previous research results, such as regression relationships among various degradation test outcomes – regarding natural rock materials. The authors summarized the key information regarding the standards in tables to aid understanding and collected the formulas of existing regression equations from the literature. To illustrate the relationships between different degradation parameters, the measured results and the established regression curves were depicted on graphs that demonstrate the diversity and the contradictions of the published results.

Fig. 1
figure 1

Mechanical weathering typology by Bach (2013). The three subgroups of attrition, fragmentation of grains, rounding (or “chipping”), and abrasion (or “polishing”) are depicted supplemented with stresses that trigger them and the effect the different attrition types cause

Full size image

Different types of aggregate test methods

Micro-Deval abrasion test

The micro-Deval Test is one of the most widely used methods for evaluating aggregates in Europe but also American Standard exists. In general, this test determines a coefficient called micro-Deval Coefficient (EN 1097-1:2011) or micro-Deval Abrasion Loss (ASTM D6928-17:2017) which is the percentage of the initial sample reduced to a size smaller than 1.6 mm (EN), or 1.18 mm (ASTM) due to rolling. In the European standard, this test was also extended to railway ballast aggregates.

After the preparation of the test sample, it is placed into the unique drum with the required water rotated for the required revolutions. The preparation of the test specimens and the test procedure depend on the fraction size of the aggregates, and it is different from standard to standard (see Table 1).

Table 1 Properties of the micro-Deval Abrasion Test method prescribed by various standards (EN 1097-1:2011; ASTM D6928-17:2017)
Full size table

Los Angeles abrasion test

The Los Angeles Test method combines the actions of abrasion and impact by rotating a drum filled with steel spheres and aggregates. Trays are placed on the inside of the drum to pick up these steel balls and after making a half-turn, the free-falling balls generate an impact-crushing effect.

The evaluation of this test is also a coefficient called Los Angeles Coefficient (EN 1097-2:2020) or Los Angeles Abrasion Loss (ASTM C535-16:2016; ASTM C131/C131M-20:2020) which is the percentage of the initial sample reduced to a size smaller than 1.6 mm (EN) or 1.7 mm (ASTM) due to rolling. In the case of the Los Angeles Test, the required mass of the sample and abrasion charge, the number of revolutions depends on the fraction size of the specimen just like in the micro-Deval Test. The main difference between the two tests is the apparatus. The size of the drum is larger, the condition of the specimen is primarily dry, and the aggregate-steel ball mixture shall be rotated for 500 or 1000 revolutions. In the case of Los Angeles Abrasion Test the preparation of the test specimens and the test procedure depend on the fraction size of the aggregates, and it is different from standard to standard (see Table 2) like in the micro-Deval Abrasion Test.

It is important to note that while Eurocode distinguishes the range of the two test fractions according to the aggregate’s final application, ASTM defines different fractions regardless of use.

Table 2 Properties of the Los Angeles Abrasion Test method prescribed by various standards (ASTM C535-16:2016; ASTM C131/C131M-20:2020; EN 1097-2:2020)
Full size table

Impact test

The European Standard (EN 1097-2:2020) recommends an alternative test method for determining resistance to fragmentation, called Impact Test. The main aspect of this type of test is to define the fragmentation caused by a falling hammer from a specified height. The amount of fragmentation is measured by five specified test sieves. The impact value (SZ) is the average percentage of the masses passing each of the five sieves (sum of passing divided by five).

The EN standard has extended the Impact Test for railway ballast such as the Los Angeles Test. The main difference in the fraction size of the test specimen is the lifting height of the hammer, which is 420 mm instead of 370 mm and the number of blows, which is 20 instead of 10. In addition to these differences, the Aggregate Impact Value is expressed as the percentage of material passing through the 8 mm sieve and the initial mass of the test specimen.

The British standard Aggregate Impact Test method fundamentally differs from Eurocode recommendations (BS 812-112:1990). The test equipment size is smaller by scales in the case of the British standard; therefore, the impact work performed by the hammer is less than EN method (see Table 3). The expression of the results is the same as the Eurocode calculation specifies for railway ballast aggregates, with the difference that the size of the test sieve is 2.36 mm.

While the EN Impact Test is extended to railway ballast aggregates the aggregate sizes bigger than 14.0 mm is not appropriate to be tested according to BS.

Table 3 Properties of the Impact Test method prescribed by various standards (BS 812-112:1990; EN 1097-2:2020)
Full size table

Crushing test, Hummel test and Swiss compression test

The BS 812-110:1990 gives a relative measure of the resistance of an aggregate to crushing under a gradually applied compressive load. The cylinder shall be filled with the prepared specimens, then the samples shall be loaded with a uniform rate to reach 400 kN in 10 min. BS 812-111:1990 introduces a very similar aggregate testing method to the previous one.

The result of this method is the 10% Fines Value (TFV), which is also a measure of the resistance of an aggregate to crushing under a gradually applied compressive load with the difference that the upper limit of the loading force is not fixed during this test. The procedure is repeated with various loads to determine the maximum load, which generates the 10% Fines Value (TFV). The equipment used for the test is the same as the Aggregate Crushing Value (ACV) test (see Table 4) with the important addition that the loading limit of this apparatus is 500 kN. In the case of both test methods, after loading the sample shall be sieved over a 2.36 mm sieve and weighed. The degree of fragmentation shall be expressed by the following equation after both tests:

$$m=\frac{M_2}{M_1}\cdot100\;\left[\%\right]$$

where M1 is the initial mass of the test specimen in grams and M2 is the mass of the material passing the 2.36 mm test sieve in grams. In the case of the 10% Fines Test, if the result obtained is between 7.5% and 12.5%, the following calculation for 10% Fines Value, which is precisely the force to produce 10% fines, shall be used:

$$TFV=\frac{14 f}{m+4} \left[kN\right]$$

where f is the maximum force applied during the test, in kN and m is the percentage fines from the test.

The Hummel test can be found in the MSZ 18287-3:1983. The national standards in Hungary were designated as MSZ before the Eurocode took over their role in 2002. MSZ 18287-3:1983 has not been officially revoked yet. The essence of the test is that the aggregates are crushed under static load force and the extent of crushing is measured. The fraction of the specimen shall be 16/40 mm, but 48 mm is the maximum allowed size. Before testing, the initial particle size distribution shall be determined by using 0.5 mm, 1.0 mm, 2.0 mm, 4.0 mm, 8.0 mm, 16.0 mm and 32 mm sieves. The sample shall be loaded with a uniform rate to reach 400 kN in 1.5 min. After loading, the particle size distribution shall be determined again by using 0.5 mm, 1.0 mm, 2.0 mm, 4.0 mm, 8.0 mm and 16.0 mm sieves. The result of a single test is the Hummel crushing value is given by the following formula:

$${H}_{48}={m}_{H1}-{m}_{H2}$$

where mH1 and mH2 are the logarithmic fineness modulus of the tested aggregate sample.

The Swiss Federal Laboratories for Material Sciences and Technology (EMPA) developed a testing method for aggregates in 1969 (Bach 2013). This procedure is quite similar to the previously presented procedure. The dimensions of the testing cylinder are over both in case of the ACV and Hummel test but the mode of the loading corresponds to the Hummel test. This method has been developed specifically for the railway faction. The degree of fragmentation is also defined in this test by measuring the change in particle distribution, similar to the Hummel test.

Table 4 Properties of the Crushing/Hummel/Swiss Compression Test methods (MSZ 18287-3:1983; BS 812-110:1990; Bach 2013)
Full size table

Summary of different types of mechanical degradation/attrition depending on the different types of degradation test methods

The aim of Table 5 is to summarize the different aggregate test methods and compare what type of stresses were indicated and what type of breakdown of particles occurred during them. It also contains the expression of the test results and the most common existing standards.

Table 5 Summary of different types of mechanical degradation/attrition depending on different types of degradation test methods
Full size table

Relationships between different aggregate degradation properties

Relationships between Aggregate Crushing Value (ACV) and aggregate impact value (AIV)

Turk and Dearman (1988) in their research studied the influence of aggregate size on the mechanical properties (AIV, ACV and 10% fines) of aggregates. Dolerite, granite, and limestone aggregates were tested with standard and non-standard sizes according to British Standard (BS 812-3:1975). They found the test results (AIV, ACV and 10% fines) depend on the aggregate size which size effect is rock strength related. The correlation between ACV and AIV were defined for different rock types and sizes.

The tested rock type of Irfan (1994) was granite with different grain sizes (fine-, medium- and coarse-grained granite). Like Turk and Dearman, Irfan defined a linear relationship between ACV and AIV on the standard size (BS 812-3:1975) quarry-produced aggregates also. He also found the bigger the grain size of the rock, the more the mean value of ACV or AIV of the quarry-produced aggregates.

Al-Harthi and Abo Saada (1997) classified the examined aggregates into three rock groups, basalt, granite and schist. However, correlations between different mechanical properties have been established for the mixed aggregate and the size of the tested fraction is not named in the paper. Regression analysis was performed on the data of the mixed natural aggregate samples by which a direct linear relationship was found between ACV and AIV.

One of the authors, Al-Harthi had another research on the same topic in 2001. The aim of his research was the determination of crushed aggregate strength characteristics with the use of field index tests such as the Schmidt hammer and point load tests. He used a wide selection of rock types, ages and degrees of weathering. In addition to the rock strength tests, the Crushing Impact and Los Angeles Abrasion Test were conducted on the aggregates. The ACV was determined on samples of 10/14 mm, while the Impact Test was undertaken on aggregates in size range of 9.5 to 12.5 mm.

The correlation between Aggregate Impact Value (AIV) and crushing values (ACV) was direct and linear like the previous relationships. Los Angeles Abrasion Value (LAAV), ACV and AIV tests were undertaken on various (20 types of crushed limestone, marble, and granite) aggregate samples by Palassi and Danesh (2016). ACV and the AIV were determined on 10/14 mm size aggregates. Both linear (1) and exponential (2) as well as logarithmic (3) fitting were performed during the regression. The linear curve fitting provided the best correlation between ACV and AIV with R2 of 0.88 which fact fits well with the results of previous studies. One of the main conclusions of their research was that LAAV, ACV and AIV tests are closely correlated with each other.

The aim of the study of Ahmad et al. (2017) was to evaluate the quality of four types of igneous (basalt, granite, gabbro and diorite), three types of sedimentary (sandstone, limestone and dolomite) and four types of metamorphic (quartzite, schist, phyllite and amphibolite) aggregates with several physico-mechanical tests. Single and multiple regressions were performed. For comparability, the correlation equation between ACV and AIV of the best fitting linear curve is reported in this paper. The size range of the examined aggregates was not detailed in their study.

The subject of the research of Agyeman et al. (2019) was to understand the properties and qualities of the dominant quarry product (granite and its modifications) in Ghana. During the test program 10/14 mm aggregates were used in both cases of ACV and AIV. Although the correlation is discussed with the results of other previous studies (see Table 6), the regression curve showed completely different, therefore not included in the graph summarizing the correlations. The reason for the difference might be that while the Crushing Test was performed according to the British Standard, similarly to other research, the Impact Test was performed according to the specifications of the EN 1097-2:1997. As discussed earlier, the two different test methods and equipment differ greatly. This fact presupposes the viability of research that would create interoperability between the results of the two test methods. Based on the correlation relationships determined by previous studies, it can be concluded that there is a strong linear relationship between ACV and AIV values. In addition to the regression curves discussed, Fig. 2 also contains the measurement results of studies in which the two abrasion values were reported, but no correlation was established.

Fig. 2
figure 2

Aggregate degradation test results and established relationships between different aggregate degradation parameters based on previous studies. The single data points represent the measured values, the lines represent the correlations. Microsoft Excel was used to display the dataset

Full size image

Kazi and Al-Mansour (1980) investigated the abrasion characteristics of aggregates of igneous origins. They found in their study that grain size and pore volume have a significant role in the resistance to wear.

Engidasew and Barbieri (2014) tested 10/14 mm basalt aggregates according to BS 812. They evaluated the results focusing on petrology, petrography, and fabric. They interpreted the variety of ACV and AIV values with the influence of aggregate particle shape, bulk composition, grain size, texture and weathering (Fookes 1991).

Nigerian granitic rocks were evaluated by Afolagboye et al. (2016, 2017) from the perspective of road construction aggregate use. Extensive petrographical and physical-mechanical research was conducted. For crushing and impact testing, the BS 812 was followed. The results showed that the different textural properties affect the physical and mechanical properties of the rocks. Both the resistance to impact and crushing increase with decreasing grain size. This finding agrees with Kazi and Al-Mansour (1980) and Engidasew and Barbieri (2014). The main purpose of the study of Kamani and Ajalloeian (2019) was to find correlations with which the prediction of aggregate degradation is possible from rock strength parameters such as unconfined compressive strength (UCS) and point load index (PLI). For this reason, the regression analysis was undertaken between one of the UCS and PLI, and an aggregate degradation indicator value. The crushing and Impact Tests were carried out in the range of 9.52/12.7 mm on chemical sedimentary rocks (limestones, travertines, dolomites).

Table 6 Comparison of the correlations between Aggregate Crushing Value (ACV) and Aggregate Impact Value (AIV) developed in previous studies
Full size table

Relationships between Aggregate Impact Value (AIV) and Los Angeles abrasion value (LAAV)

One of the most widely used tests to evaluate the quality of aggregates is the Los Angeles Abrasion Test. In the case of roads, surface wear is a very important issue (Irfan 1994; Al-Harthi 2001). The benefit of this test is that the standard exists both in America and Europe. While AIV and ACV tests as a part of BS 812 are defined in a national standard and it is mostly used and known in UK and in areas with an Anglo-Saxon past, while Los Angeles Test is prescribed in the most up to date standards such as EN 1097-2, ASTM C131 and ASTM C535. There is a large overlap in the source of the correlations discussed in Table 7 with the studies discussed above. The reason for this could be the fact these research were associated with the aggregate toughness/abrasion resistance and ACV, AIV and LAAV tests are suitable for measuring this resistance. Therefore, mostly these tests were carried out to study the abrasion behavior of different aggregates. In most studies, the AIV is the British version, not the Impact Test defined in the EN 1097-2.

Table 7 Comparison of the correlations between Aggregate Impact Value (AIV) and Los Angeles Abrasion Value (LAAV) developed in previous studies
Full size table

Irfan (1994) performed the Los Angeles Test on 20 mm nominal size aggregates. He found a linear relationship between AIV and LAAV. Like in the case of AIV and ACV, the lower the Los Angeles Abrasion Value is, the finer-grained the granite is.

The regression analysis of Al-Harthi and Abo Saada (1997) led to the result that AIV is directly related to LAAV outside of ACV. The authors did not elaborate on the particulars of the Los Angeles Test only the standard is known (see Table 7).

In the study of Al-Harthi (2001) it is also not mentioned in what size range the Los Angeles Test was implemented. He found a relatively strong linear relationship between AIV and LAAV. As Al-Harthi compared the results with earlier studies, it was noted similarly. In this paper, two correlations were taken from the study of Al-Harthi (2001) directly, namely the regression equation of Chang (1973), Kazi and Al-Molki (1982).

Koukis et al. (2007) compared LAAV and AIV with each other and with the petrographic composition of the rock material. Mostly limestone and sandstone gravels were tested. The aim of their study was to inquire into the properties of optional useable lower-quality aggregates in infrastructure construction. During the Los Angeles Abrasion Test, grading A was used, which means 9.5/37.5 mm fraction size range, while during the Impact Test the standard, 10/14 mm size was applied. After evaluating the results of the mechanical damage tests, it was concluded that both LAAV and AIV are influenced by the same geological and clastic features.

The research of Naeem et al. (2014) was specified in Margala Hill Limestone aggregates (in Pakistan). In the context of the physical characterization study, the Los Angeles Abrasion Values and Aggregate Impact Values of these limestones were measured. LAAV and AIV showed a significant positive correlation (R = 0.80) with each other. The size of aggregates was not detailed.

Palassi and Danesh (2016) expanded their research on Los Angeles Test analysis also. They used aggregates with grading B (9.5/19.0 mm) to define the LAAV. The exponential curve fitting gave the best result with R2 of 0.92. They found inconsistencies between their study and the previous ones. They mentioned two possible explanations for the discrepancy, the rock types or the size distribution of the aggregates. This conclusion also sheds light on the need to carefully compare different research findings (see Fig. 2).

Ahmad et al. (2017) did not publish the size range of the Los Angeles Test either. In addition, the graph summarizing the correlations between AIV and LAAV also shows measurement results for which no regression was performed.

The goal of the paper of Rigopoulos et al. (2013) was a better understanding of the process of aggregate strength properties changing in service. To determine the degradation of the aggregates, Los Angeles, micro-Deval and Impact Tests were applied. This research was one of the few studies that have extended the investigation to the size range of the ballast aggregates used in railway construction. However, Eurocode was not used consistently in the use of standards. Basically, during the Los Angeles Test the grading A (9.5/37.5 mm) of ASTM was applied. The Impact Test was carried out according to the specification of the BS 812.

The studies of Engidasew and Barbieri (2014), Afolagboye et al. (2016, 2017) and Kamani and Ajalloeian (2019) were discussed earlier. In both research, Los Angeles Abrasion Value was defined, but the size range (9.5/19.0 mm) of the tested aggregates is mentioned only in the study paper of Kamani and Ajalloeian (2019). Afolagboye et al. (2016, 2017) found that the fine-grained rocks produced aggregates had lower LAAV, which means the fine-grained materials are more resistant to degradation, to the combined effect of abrasion and impact. The previously established regressions can generally be said that the authors have found in most cases direct, linear correlation in the international literature. The smaller the magnitude of the degradation (LAAV and/or AIV) is, the less their value is affected by the type of the rock (features like mineralogy, texture, etc.).

Relationships between aggregate crushing value (ACV) and Los Angeles abrasion value (LAAV)

Palassi and Danesh (2016) found the exponential curve fit the best out of linear, exponential, and logarithmic forms with R2 of 0.88. Ahmad et al. (2017) established a linear correlation between ACV and LAAV with R2 of 0.73.

As before in case of ACV and LAAV relationships, those results were plotted as well on the summary graph (Fig. 2) in which case no correlation was determined (Irfan 1994; Al-Harthi and Abo Saada 1997; Engidasew and Barbieri 2014; Kamani and Ajalloeian 2019). The established correlation of Ahmad et al. (2017) is consistent with the test results of these studies, while a set of test results of Palassi and Danesh (2016) and the fitted curves are separated from the other studies (see Table 8 and Fig. 2).

Table 8 Comparison of the correlations between Aggregate Crushing Value (ACV) and Los Angeles Abrasion Value (LAAV) developed in previous studies
Full size table

Relationships between 10% fines value (TFV) and aggregate crushing value (ACV)/ aggregate impact value (AIV)  

Carrying out the 10% fines tests on different aggregate sizes Turk and Dearman (1988) found that the larger the aggregate particle size is, the smaller load is needed to get the 10% fines. They also noted that the 10% fines test result is rock type-dependent regarding the fact that the weaker the rock is, the less the results are affected by the aggregate size. Correlation analysis showed a high correlation coefficient for both AIV and ACV (see Table 9).

Table 9 Comparison of the correlations between Aggregate Crushing Value (ACV) and 10% Fines Value (TFV), and Aggregate Impact Value (AIV) and 10% Fines Value (TFV) developed in previous studies
Full size table

The study of Uche and Muhammad (2011) focused on the relationship between TFV and ACV. The aim of the established mathematical correlation was to save time and money. They found the 10% fines test a non-effective method to determine the resilience of aggregates to crushing. With a good correlation equation, the unknown variable can be well estimated from the known variable. They found an inverse relationship between ACV and TFV (both tests were carried out on 10/14 mm aggregates), which means that the higher the ACV the lower load is needed to produce 10% fines (see Table 9). In both studies, the results led to a strong relationship, this may be since the same apparatus is required for both tests and each test is used to determine the crushing resistance.

The comprehensive research program of Irfan (1994) included the 10% fines test method also. He found that fine-grained granites have the highest TFV compared with coarser-grained granites. Two linear regression lines were fitted to the TFV and AIV results, one for laboratory crushed aggregates and one for quarry-produced aggregates, although Irfan himself points out that the number of measurements on laboratory crushed aggregates was low. Therefore, the established correlation is more reliable based on the quarry-produced aggregates test results. Between TFV and ACV no interrelation was identified, only the measurement data points were plotted on the graph containing the correlations (see Fig. 2).

In their research, Al-Harthi and Abo Saada (1997) found that, as the inverse relationship between TFV and ACV, there is an inverse proportionality between TFV and AIV (see Table 9). This finding is consistent with the studies of Turk and Dearman (1988), and Irfan (1994).

Over ACV and AIV the study of Engidasew and Barbieri (2014) as well as Afolagboye et al. (2016, 2017) was comprised of TFV also. Afolagboye et al. (2016, 2017) found that the texture of the aggregate rock influences the TFV. They measured the highest values on the fine-grained types of the examined Precambrian basement rocks.

Relationships between micro-Deval coefficient (MDE) and Los Angeles abrasion value (LAAV)

The micro-Deval Test is a rotary drum aggregate degradation test method similar to the Los Angeles Test. The micro-Deval Test was developed to quantify rounding resistance, while the Los Angeles Test was developed to measure the tendency to fragmentation. The other important difference is that while the micro-Deval test is recommended to be performed in wet conditions, but it also allows for a dry test. Meanwhile, the Los Angeles test can only be performed on dry test specimens.

Erichsen et al. (2011) carried out both abrasion tests on fraction range of 10/14 mm. EN standard tests were implemented by the extended sieving method to get more accurate grading curves. In some cases, the fraction size was escalated to the railway ballast fraction size, 31.5/50 mm. They found the distribution well-graded after Los Angeles Test which fact indicates the degradation due to fragmentation. The difference in the shape of the grain size distribution curves led to the conclusion that the degradation depends on the quality of the rock. Based on the coarse size fraction test results, the authors assumed that the reason for the steeper falling of the micro-Deval curve was caused by more abrasive degradation. They also found that Los Angeles Test was characterized by fragmentation and the more and less similar distribution curves concluded that the two types of wearing test methods express the same type of degradation.

Rigopoulos et al. (2013) carried out the micro-Deval Tests according to EN 1097-1, while the Los Angeles Tests on small-size coarse aggregates by using ASTM C-131 (gradation A) and extended the tests on ballast aggregate according to EN 1097-2. In Fig. 2 the MDE values are compared to LAAV value tested on gradation A.

Tuncay et al. (2016) determined the mineralogical, petrographic, mechanical, physical, and chemical properties of different rock types in their study. They used Los Angeles and micro-Deval Test methods to determine the resistance to fragmentation and wear. A linear relationship was established between MDE and LAAV (see Table 10). This result contradicts the statement of Gálos and Kárpáti (2007) which is discussed in a former research. They found no clear connection between the two parameters. Therefore, they concluded that it is important to use both tests for the characterization of crushed stones. In everyday practice, the examination of aggregates accordingly covers both tests.

Table 10 Comparison of the correlations between micro-Deval Coefficient (MDE) and Los Angeles Abrasion Value (LAAV)developed in previous studies
Full size table

Sogut and Kocabey (2020) tested highway and railway construction aggregates with magmatic origin. The aim of the study was to evaluate the examined rock types for these uses. In this paper, the MDE and LAAV values are presented in Fig. 2. Both abrasion tests were carried out according to EN standards.

Conclusions

Theoretically, if an appropriate quantity of aggregate is not available to test, then with the help of a well-defined correlation between aggregate degradation properties, the missing mechanical properties can be predicted (Al-Harthi and Abo Saada 1997; Kamani and Ajalloeian 2019). Many correlations were defined between different aggregate degradation parameters in the previous studies. This paper aims to summarize these results and emphasizes the limits of the comparisons since, in the international literature, a wide range of standards, rock types, and aggregate fraction sizes were used and tested.

Many researchers have concluded that the degradation resistance of aggregates is highly dependent on the fraction size, which size effect is rock strength related. Palassi and Danesh (2016) discovered a difference in their research results compared to the results of previous studies, which difference was explained by the distinct type of rocks and particle size distribution. During the comparison of the degradation of fraction range of 10/14 mm and the railway ballast size fraction of 31.5/50 mm Erichsen et al. (2011) showed that the shape of particle size distribution curves after the Los Angeles and micro-Deval Tests differ from each other in case of the two different size fraction range. They also mentioned that behind the size, the quality of the rock is a significant influencing factor (Ugur et al. 2010; Ozcelik 2011; Kuna and Bögöly 2024).

The quality of the rock can be determined as the physical property of the rock and expressed as a rock strength parameter. Behind the rock type, it depends on many geological features as texture, mineral composition, weathering, origin, etc. Several research dealt with petrographic analysis of the tested rocks (Koukis et al. 2007; Engidasew and Barbieri 2014; Afolagboye et al. 2016, 2017; Tuncay et al. 2016; Kenmoe et al. 2020). Irfan (1994) found that the degradation resistance of granite aggregates increases with decreasing of grain size. Afolagboye et al. (2016, 2017) stated the same for granites. Kazi and Al-Mansour (1980) found that grain size and pore volume of igneous origin rocks both affect the resistance to wear. Kahraman and Fener (2007) found that the correlation coefficients increase comparing Uniaxial Compressive Strength and Los Angeles Abrasion Value by the classification of rocks according to porosity. It can be said in general that these findings agree.

In summary, the aim of this paper is to highlight that the degradation test results are highly affected by numerous properties such as the fraction size, the texture, mineral content, and other geological features of the rocks. Therefore, all these data are necessary to make a reliable comparison. This fact also justifies the need to consider other physical properties during the evaluation of the mechanical tests. From the overview of the literature, it can be concluded that there is no universal formula that could be used in practice to convert one mechanical degradation parameter from another. The difference in standards and the question of their comparability are well illustrated by the two types of Impact Tests, which are identical only in their names. The comparison of data sets from the literature could be even more complicated if some critical information were not discussed in detail.

In addition, there are examples in the international literature where the researchers partially deviated from the standard specifications and according to the results, they suggest new qualifier properties for crushed stones (Czinder and Török 2017; Czinder et al. 2021). The focus of the present paper was only the standard procedures, but a further research can also comprise the comparison of newly suggested methods.