Influence of Abrasive Wear on Reliability and Maintainability of Components in Quarry Technological Equipment: A Case Study

Abstract

A two-year study (June 2022–May 2024) on the reliability and maintainability of technological equipment at the Pătârș basalt quarry identified critical wear issues in metal components impacting operational continuity. The analysis focused on identifying causes of operational interruptions and evaluating solutions to improve equipment performance. Results showed that speed and load significantly impact wear rate and material selection significantly influences abrasion resistance. Laboratory tribological tests provided valuable data on the influence of basalt properties on wear, complementing field data. The study highlighted the low reliability of critical components, such as the sorting station trough, front loader bucket knife, and excavator bucket tooth, necessitating frequent replacements. For example, the trough has only a 40% probability of operating without defects after 182 days, with average roughness reaching 1.2 μm and wear profile height up to 22.5 μm. Similarly, the bucket knife and tooth require replacement at significantly reduced intervals compared to their nominal operating life to achieve 80% reliability. To address these findings, the study proposes two solutions: (1) manufacturing experimental prototypes with alternative materials for the trough to improve wear resistance and reliability; and (2) on-site welding reconditioning of metal components to reduce costs and downtime.

1. Introduction

Mechanical components wear and age from the outset of their operational life. The failure rate of a system and its components varies non-linearly over time, changing across life phases [1,2,3]. To understand material wear, experimental research determines the influence of technological parameters, material factors, and operating regimes [4,5,6]. System reliability, which depends on time and operating conditions, is important. Low reliability increases unforeseen defects and unplanned repairs, reducing availability [7,8,9]. Product reliability evaluation uses established methods: component reliability databases, system architecture analysis [10,11,12,13], and operational simulations [14,15,16]. Reliability analysis in mechanical engineering characterizes product behavior across life phases, measuring the impact of design changes on product integrity [17,18,19,20]. Adapting machine materials to severe wear conditions becomes increasingly important. Many subassemblies and spare parts, due to high consumption rates, contribute to high production costs and reduced profitability. A prime example is equipment used in quarries for aggregate extraction and delivery.

Environmental factors, including humidity and salts, play an essential role in the durability analysis of quarry equipment, as they can significantly accelerate material degradation through corrosion, erosion, and other deterioration mechanisms [21,22]. High humidity levels promote rust formation and metal corrosion, while salts can induce chemical deterioration of concrete and other construction materials, thus reducing equipment lifespan and reliability. Therefore, a thorough understanding and effective management of these factors are essential to ensure the efficient and safe operation of equipment in quarry environments. An in-depth study of the specialized literature [1,4,7,8,9,23], combined with the analysis of the quality and reliability of technological equipment used in aggregate extraction and delivery quarries [24,25], has led to two main conclusions: the importance of field-collected data and the rigorous approach to each subassembly. To correctly evaluate equipment reliability indicators, the use of concrete data, collected directly from the production process, regarding their behavior under real working conditions [1,3,5,6,7,26,27,28], is crucial. Ensuring the safe operation of equipment in a technological flow requires a careful analysis of each subassembly separately [7,29,30]. The specific characteristics and operating parameters of each element must be considered to determine the individual and collective influences on the quality and reliability of the main equipment in the technological flow [31,32]. The fundamental task and rationale for ensuring optimal machine functionality consist of the necessity and obligation to achieve a corresponding quality of the services for which the respective work equipment was acquired.

In the specific case of quarry technological equipment that extracts aggregates, such as those in the Pătârș quarry, which is the subject of this study, the analysis of the production flow regarding critical areas from the point of view of defects highlighted frequent interruptions in production, which involved high remediation costs and, above all, led to non-compliance with deadlines for orders to beneficiaries. The Pătârș basalt quarry (Figure 1) is located in the commune of Ususău, Arad County, Romania. The Hengl Group, Limberg, Austria developed the quarry in 2008. Currently, stone extraction takes place on an area of 20 hectares. The amount of stone extracted annually is approximately 400,000 tons. The total potential of the quarry is estimated at 24 million tons of stone. The quarry uses modern processing technology. This includes an aggregate production plant for asphalt and a dust collection plant. This technology allows the quarry to produce a diverse range of products. These include crushed sand grades 0/2 and 0/4; screenings of various sizes, grades 2/4, 4/8, 8/11, 11/16, 8/16, 16/22.4 and 16/32; and crushed stone at grades 0/25, 0/32, 0/63, 25/63, 32/63, 40/63 and 63/180. Last but not least, the quarry produces filler stone and sterile soil.

Quarry exploitation utilizes two primary production flows. The crushing and sorting process employs fixed and mobile equipment, including excavators, rigid dumpers, front loaders, a fixed primary jaw crusher, a secondary cone crusher, tertiary cone crushers, conveyor belts, and a sorting plant. Specific equipment includes: three Komatsu WA470 wheeled front loaders, three Komatsu HB365 excavators, one Komatsu D65 bulldozer, three Komatsu HD405 dump trucks, one Keestrack mobile jaw crusher, one Keestrack Warrior sorter, one Kue Ken fixed primary crusher, three Svedala cone crushers, and one Sandvik Titon DI200 drill.

This study investigated methods for improving the service life of wear-prone parts in the Pătârș basalt quarry. Initial analysis focused on operational interruption data and identification of causative components. This enabled the calculation of reliability and maintainability indicators for the sorting station trough, the Komatsu HB365 excavator bucket tooth, and the Komatsu WA470 front loader bucket knife.

In this study, we utilized distribution models such as exponential, normal, lognormal, and Weibull, recognized for their applicability in the reliability analysis of industrial equipment. To evaluate the goodness-of-fit of the distribution models, statistical tests such as the Kolmogorov–Smirnov (K-S) and chi-squared (χ²) tests could have been considered. The results indicated that the lognormal distribution exhibited the best fit for the reliability data, with high p-values and low-test statistics. The exponential and Weibull distributions also showed a good fit, but the lognormal distribution stood out as the most suitable. These statistical results support our choice of models and allow us to provide a more rigorous analysis of equipment reliability and maintainability. In this analysis, we used Relyence Weibull software to evaluate the goodness-of-fit of various distribution models (exponential, normal, lognormal, and Weibull) to our data. This software enabled us to determine the “acceptance probability” of each model, providing a measure of how well the theoretical distribution fits the empirical data. These statistical results support our choice of models and allow us to provide a more rigorous analysis of equipment reliability and maintainability. The model selection was based on acceptance probability, the nature of the data (we considered the nature of the data and the failure behavior), and knowledge of the literature [33,34,35,36].

Tribological studies determine the influence of technological parameters, material factors, and working conditions on rock–metal wear processes [37,38,39,40,41]. The tribological tests presented here were used to examine the wear evolution of samples from these three metallic components in contact with Pătârș basalt. The goal was to identify solutions for improving maintenance activities, as these are highly dependent on operating conditions, particularly rock properties.

2. Materials and Methods

The economic efficiency of technological equipment operating in the extractive industry, railway tunnel boring, or road transport, among others, refs. [42,43] is determined by the volume of rock dislodged, as well as by the wear of subassemblies and spare parts in contact with rocks. The influence of wear, cutting tool costs for mining roadheaders, and rock resistance has been studied by Bergbau-Forschung GmBH as a function of rock properties [44]; exceeding a value of 0.3 for the tool wear factor leads to a rapid, sometimes tenfold, increase in costs due to their specific consumption. Experiments conducted at the University of Petroșani [45,46] on roadheader cutter wear using various rock samples and varying test regime parameters (advance speed 0.5–1 cm/min, cutting speed 0.6–1 m/s, maximum applied force 15 kN) guided the authors in selecting working regimes on the tribometer.

This analysis employed a two-step methodology: (1) selecting mathematical and computational models of reliability and maintainability for the most frequently failing components of the technological equipment in the Pătârș basalt quarry; and (2) conducting tribological tests on rock–metal pairs to study friction–wear phenomena and changes in metal surface roughness for the following: front loader bucket knives; excavator bucket teeth; and sorting station troughs (original metal sheet, Hardox 400 steel, Pucest panel with perforated metal insert). Laboratory simulation of the friction and wear processes of metal components in contact with rocks must enable the study of the simultaneous influence of testing regime parameters (load, rotation speed, and number of cycles) and rock properties on the wear of metal elements. The selection of the aforementioned parameter values depends on both the abrasiveness, hardness, and compressive strength of the rocks and the type of metal material studied. In the case of basalt collected from the Pătârș quarry, with a heterogeneous structure and variations in the values of these properties, it was considered suggestive to test different regimes regarding the values of load, rotation speed, and number of cycles, so that a significant variation of wear can be achieved as a function of these parameters.

2.1. Materials

The selection of Hardox 400 steel and Pucest panels for tribological testing was driven by the aim of this study to comparatively evaluate wear resistance in materials used at the Pătârș basalt quarry. This choice reflects a dual focus: benchmarking against the established performance of Hardox 400, and exploring the potential of the novel Pucest composite in abrasive wear applications. These materials offer contrasting properties relevant to the quarry’s specific operating context, with Hardox 400 serving as a reference material and Pucest panels representing a potential innovative solution for enhanced wear resistance. For the purpose of carrying out the study and for the selection of parameters for tribological testing regimes regarding wear phenomena at the contact of metal surfaces with the Pătârș basalt (Figure 2), in the initial stage, the structure and physical-mechanical properties of this rock were analyzed.

The results of the macroscopic and microscopic analyses of the basalt samples collected from the Pătârș basalt quarry are presented in

Table 1. Macroscopic and microscopic characterization of basalt from the Pătârș quarry.

Characteristic	Description
Macroscopic Characterization
Rock type	Magmatic, neovolcanic, basic
Structure	Hypocrystalline (semicrystalline), porphyritic intersertal
Texture	Compact, unoriented and oriented (fluidal), partially vacuolar
Mineralogical composition	Neutral and basic plagioclase feldspars, olivine, pyroxenes, opaque minerals, alteration minerals
Color	Dark grayish-green
Reaction to acids	Reacts with diluted hydrochloric acid (HCl), dissolving calcium carbonate along fissures
Microscopic Characterization
Structure	Semicrystalline, porphyritic intersertal
Texture	Compact, partially vacuolar, unoriented and oriented (fluidal)
Mineralogical composition	Plagioclase feldspars, pyroxenes, olivine, calcite
Plagioclase feldspars	Predominant, quite fresh, idiomorphic and hypidiomorphic, chaotically distributed with a fluidal aspect in the rock’s groundmass
Olivine	Does not exceed 2% of total components. Hypidiomorphic and allotriomorphic crystals are submillimetric in size. Exhibit an advanced stage of alteration with the formation of secondary serpentinitic minerals
Pyroxenes	Represented by idiomorphic and hypidiomorphic crystals, not exceeding 1% of total mineral components, being fresh augite
Calcite	Present along fissures and microfissures in the rock mass
Opaque minerals	Represented by iron oxides and hydroxides, submillimetric in size, contained within the rock’s groundmass (matrix)
Rock name	Olivine Basalt

. The oxidic chemical composition of the olivine basalt is presented in

Table 2. Oxidic chemical composition of olivine basalt from the Patars quarry.

Component	Value (%)	Component	Value (%)
SiO2	33.4	PbO	0.533
Fe2O3	25.5	CuO	0.480
CaO	13.4	MnO	0.344
Al2O3	11.7	P2O5	0.318
MgO	6.16	ZnO	0.161
SO3	3.14	V2O5	0.103
TiO2	2.57	ZrO2	0.0506
Na2O	1.27	SrO	0.0449
K2O	0.697	Cl	0.0133

.

The study stages followed in this paper were based on data regarding the operational interruptions of the technological equipment of the Pătârș basalt quarry (

Table 3. Frequency of failures (defects) and the share of repair times for the components that presented defects (maximum wear, breakage, etc.).

No.	Component Name	Number of Failures	Failure Frequency	Repair Time (min)	Repair Time Share
1	Front loader bucket knife (maximum wear)	32	0.256	1850	0.0859
2	Excavator bucket tooth (maximum wear)	31	0.248	1405	0.0653
3	Sorting station trough (maximum wear)	18	0.144	760	0.0353
4	Crushed material discharge belt rupture	11	0.088	5280	0.2452
5	Hydraulic hose cracking of excavator bucket arm actuation cylinder	9	0.072	4320	0.2006
6	Conveyor belt return bearing seizure	8	0.064	480	0.0223
7	Front loader main arm cylinder mounting bolt shear	6	0.048	2880	0.1337
8	Crusher jaw liner wear	5	0.040	2400	0.1114
9	Excavator injection system malfunctions	3	0.024	1440	0.0669
10	Front loader oscillating arm cracking	2	0.016	720	0.0334
	TOTAL	125	1	21535	1

) due to wear and the need to replace some of their subassemblies and components, in the period June 2022–May 2024.

The highlighting of the frequency of failures (defects) and the share of their repair times was obtained by constructing the Pareto diagram (Figure 3).

The front loader bucket knives (25.6%), excavator bucket teeth (24.8%), and sorting station troughs (14.4%) exhibit the highest failure frequencies. The sorting station trough, currently made of metal sheet, requires replacement. Alternatives to the manufacturer’s original material include Hardox 400 steel (higher hardness) or Pucest panels with perforated metal inserts (increased elasticity). Figure 3 suggests a correlation between failure frequency and repair time for certain components. For example, the crushed material discharge belt rupture and excavator bucket arm hydraulic hose cracking exhibit high values for both categories. Components with high failure frequencies and long repair times warrant further analysis, potentially leading to design improvements, more resistant materials, or more frequent preventive maintenance [47,48,49,50]. Conversely, components with long repair times but low failure frequencies may require revised repair procedures to minimize downtime.

Technological elements and systems undergo specific technical changes over time, influenced by operating conditions, which reduce their operational probability [51,52]. Abrasive wear-induced failures in quarry equipment components can result from factors such as inappropriate material selection, highly abrasive rocks, and poor maintenance and repair quality. Therefore, component reliability can be analyzed as the probability of operation under specific conditions for a given period [47,48,49,50,51,52].

Cylindrical specimens (Figure 4) were created from Komatsu HB365 excavator bucket teeth samples for tribological testing with the TRB3 tribometer. Pre-test hardness measurements were essential for accurate interpretation of results. Established hardness values were as follows: 105 HB (trough metal sheet), 387 HB (trough Hardox 400 steel), 399 HB (excavator bucket tooth steel), and 365 HB (front loader bucket knife steel).

The surface of the samples proposed for study was prepared (Figure 5) to remove impurities and residues.

Prior to dry friction tribological testing, the surface roughness of samples was characterized using a Sutronic S128 profilometer. This instrument generated surface profiles, enabling the evaluation of key parameters such as arithmetic mean roughness (Ra), root mean square roughness (Rq), and total profile height (Rt). These parameters significantly influence tribological behavior, affecting both the friction coefficient and wear rate. The obtained data facilitated analysis of surface changes in the metal samples resulting from tribological tests, specifically wear caused by rock-metal interaction mechanisms.

2.2. Mathematical and Computational Models of Reliability and Maintainability of Technological Equipment Components

In the field of material production, most of the quality and reliability characteristics of elements and systems are variable and can take different values, as a result of the sets of random factors that influence them; for example, the number of failures that occur in a certain period of operation can constitute random variables x; the distribution law of the random variable can be presented in the form of an analytical relationship that establishes the correspondence between the random values x and the probabilities of their occurrence [6,10,11]. Knowing the statistical indicators of the distributions (mean value, median, mode, amplitude, dispersion, etc.) allows the study of the statistical distribution of the monitored variable x. In reliability studies, knowledge of the failure distribution is important because it constitutes an indication of the nature of the failures and facilitates the estimation of reliability indicators [30,41,50]. The main distribution models with applications in the field of industrial machinery, installations, and equipment are exponential, normal, lognormal, and Weibull [51,52,53,54].

2.2.1. The Exponential Model

In situations where the failure rate or intensity λ is constant, the exponential model is used, which, according to the “bathtub curve” model, begins at the end of the initial period and ends at the beginning of the final or aging stage, but can also be successfully applied in the modeling of the maintainability of mechanical systems, when the repair rate or intensity (µ) can be considered constant.

For the random variable time, T, with t ≥ 0, the probability density of the operating time until failure is:

$$f\left(t;\lambda \right)=\lambda ·\mathit{exp}\left(-\lambda ·t\right),$$

(1)

where t > 0 is the distribution parameter.

The distribution function F(t; λ), namely, the probability of product failure in a time interval (0, t), will be:

$$F\left(t;\lambda \right)=1-\mathit{exp}\left(-\lambda ·t\right).$$

(2)

The probability that the product will not fail in the time period (0, t) will be:

$$R\left(t;\lambda \right)=\mathit{exp}\left(-\lambda ·t\right).$$

(3)

The failure rate or failure intensity, z(t), will be:

$$z\left(t;\lambda \right)=\lambda =ct.$$

(4)

The mean value of this distribution, respectively the mean time to failure, will be:

$$m=MTBF={\displaystyle \frac{1}{\lambda }}.$$

(5)

2.2.2. The Normal (Gauss–Laplace) Model

The normal model finds its applicability in the case of processes influenced by a large number of random factors. In reliability, the normal distribution characterizes the phenomena of mechanical, electrical, thermal, etc., aging of elements and systems. A random variable x has a normal distribution of parameters m and σ if the probability density is of the form:

$$f\left(x;m,\sigma \right)=[{\displaystyle \frac{1}{\left(\sigma \sqrt{2·\pi }\right)}}]·\mathit{exp}\left\{-1/2·{\left[\left(x-m\right)/\sigma \right]}^2\right\},$$

(6)

in which, m and σ, the mean and the standard deviation, respectively, are the parameters of the distribution. The distribution function of a random variable, X, with a normal distribution will be:

$$F\left(x;m,\sigma \right)=[{\displaystyle \frac{1}{\left(\sigma \sqrt{2·\pi }\right)}}]·{\int }_{-\infty }^x\left\{-1/2{\left[\left(x-m\right)/\sigma \right]}^2\right\}d\tau .$$

(7)

The Laplace integral function will be:

$$\Phi \left(u\right)=[1/\left(\sigma \sqrt{2·\pi }\right)]·{\int }_0^u{e}^{-u^2/2}du,$$

(8)

in which, u is the quantile of the Laplace function.

The reliability function for a normal model of failures will be given by the relation:

$$R\left(t;m,\sigma \right)=1/2-\Phi \left[\left(t-m\right)/\sigma \right].$$

(9)

The distribution function becomes:

$$F\left(t;m,\sigma \right)=1/2-\Phi \left[\left(m-t\right)/\sigma \right]$$

(10)

and the failure intensity will be:

$$z\left(t;m,\sigma \right)={\displaystyle \frac{[\frac{1}{\left(\sigma \sqrt{2\pi }\right)}]·\hspace{0.33em}\mathit{exp}\left\{-1/2{\left[\left(t-m\right)/\sigma \right]}^2\right\}}{1/2-\Phi \left[\left(t-m\right)/\sigma \right]}}.$$

(11)

2.2.3. The Lognormal (Dalton) Model

The lognormal distribution is used for modeling technical systems that degrade due to the phenomenon of thermal fatigue. This distribution is also used in the maintenance analysis of technical systems. The probability density of failures has the form:

$$f\left(t;m,\sigma \right)=[{\displaystyle \frac{1}{\left(t·\sigma ·2·\pi \right)}}]·\mathit{exp}\left\{-1/2·{\left[\left(\mathit{ln}t-m\right)/\sigma \right]}^2\right\}, t>0,m>0,\sigma >0,$$

(12)

where $m$ and ${\sigma }^2$ are the mean and variance, respectively, of the random variable.

Performing the change of variable $u={\displaystyle \frac{lnt-m}{\sigma }}$ the connection between the distribution function of the lognormal distribution and the Laplace function Φ(u) is obtained. Thus,

$$F\left(t;m,\sigma \right)=1/2+\mathsf{\Phi }\left[\mathit{ln}\left(t-m\right)/\sigma \right].$$

(13)

The reliability function is:

$$R\left(t;m,\sigma \right)=1/2-\mathsf{\Phi }\left[\mathit{ln}\left(t-m\right)/\sigma \right].$$

(14)

The failure rate, $z\left(t;m,\sigma \right)$, has the expression:

$$z\left(t;m,\sigma \right)={\displaystyle \frac{[1/\left(\sigma \sqrt{2\pi }\right)]·\hspace{0.33em}\mathit{exp}\left\{-1/2·{\left[\left(x-m\right)/\sigma \right]}^2\right\}}{1/2-\mathsf{\Phi }\left[\mathit{ln}\left(t-m\right)/\sigma \right]}}.$$

(15)

The mean value of a random variable that follows the lognormal distribution will be:

$$E\left(T\right)=\mathit{exp}\left(m+{\sigma }^2/2·{\displaystyle \frac{{\sigma }^2}{2}}\right).$$

(16)

2.2.4. The Weibull Model

The field of applicability includes failure models determined by degradation processes such as fatigue, corrosion, diffusion, or mechanical abrasion, as well as phenomena concerning material strength and failure times of electronic and mechanical components [49].

The Normalized Biparametric Weibull Distribution

The biparametric Weibull distribution can be expressed in a more advantageous form by substituting the parameter λ and normalizing the time by a constant η which represents the real scale parameter, β being the shape parameter. It is denoted by:

$$\lambda =1/{\eta }^{\beta }\mathrm{or} \eta =1/\sqrt[\beta ]{\lambda }.$$

(17)

The probability density function will take the form:

$$f\left(t/\eta ;\beta \right)=[{\displaystyle \frac{\beta }{\eta {\left(\frac{t}{\eta }\right)}^{\beta -1}}}]·\mathit{exp}\left[-{\left(t/\eta \right)}^{\beta }\right].$$

(18)

The distribution function, namely, the probability that the product will fail in the time interval (0, t), will be:

$$F\left(t/\eta ;\beta \right)={\int }_0^{t}f\left(t/\eta ;\beta \right)dt=1-\mathit{exp}\left[-{\left(t/\eta \right)}^{\beta }\right].$$

(19)

The probability that the product will not fail in the time interval (0, t) will be:

$$R\left(t/\eta ;\beta \right)=1-F\left(t/\eta ;\beta \right)=\mathit{exp}\left[-{\left(t/\eta \right)}^{\beta }\right].$$

(20)

The failure rate becomes:

$$z\left(t/\eta ;\beta \right)=[\beta /\eta ]·{\left(\mathit{exp}\left[-{\left(t/\eta \right)}^{\beta }\right]\right)}^{\beta -1}.$$

(21)

The mean time to failure for the two-parameter Weibull distribution will be:

$$m=\eta ·\Gamma \left(1/\beta +1\right).$$

(22)

The Three-Parameter Weibull Distribution

This distribution model constitutes a complete variant using three parameters: β—the shape parameter; η—the scale parameter or characteristic life, i.e., the lifespan up to which 63.2% of failures will occur; and γ—the minimum lifespan, or the location, position, or initialization parameter. The probability density function will have the form:

$$f\left(t;\eta ,\beta ,\gamma \right)=\{{\displaystyle \frac{\beta }{\eta }}\hspace{0.33em}{\left[{\displaystyle \frac{\left(t-\gamma \right)}{\eta }}\right]}^{\beta -1}\}.\mathit{exp}\left\{-{\left[\left(t-\gamma \right)/\eta \right]}^{\beta }\right\}.$$

(23)

where: $t>\gamma$, $\beta >0$, $\eta >0$ şi $-\infty <\gamma <\infty$.

The distribution function, namely, the probability that the product will fail in the time interval (0, t), will be:

$$F\left(t;\eta ,\beta ,\gamma \right)=1-\mathit{exp}\left\{-{\left[\left(t-\gamma \right)/\eta \right]}^{\beta }\right\}.$$

(24)

The reliability function, namely, the probability that the product will not fail in the interval (0, t) will be:

$$R\left(t;\eta ,\beta ,\gamma \right)=\mathit{exp}\left\{-{\left[\left(t-\gamma \right)/\eta \right]}^{\beta }\right\}.$$

(25)

The failure rate becomes:

$$z\left(t;\eta ,\beta ,\gamma \right)=[\beta /\eta ]·\hspace{0.33em}{\left(t/\eta \right)}^{\beta -1}.$$

(26)

The mean time to failure will be:

$$m=\gamma +\eta ·\Gamma \left(1/\beta +1\right).$$

(27)

2.3. Selection of Statistical Models

To evaluate the goodness-of-fit of the distribution models, statistical tests such as the Kolmogorov–Smirnov (K-S), Anderson–Darling, or chi-squared (χ²) tests can be considered. To select the model that best suits the purpose of this study, simulations of the Kolmogorov–Smirnov and Anderson–Darling tests were performed to demonstrate the adequacy of the Weibull, lognormal, and exponential distributions to our data. A dataset simulating the time-to-failure (TTF) for the components analyzed in this paper was generated. The statistical software R (https://www.r-project.org/, accessed on 23 March 2025) was utilized to perform the goodness-of-fit tests, and Kolmogorov–Smirnov and Anderson–Darling tests were applied for each distribution (Weibull, lognormal, exponential, gamma, normal) and each simulated dataset. The results obtained are presented in tabular form, including test statistics, p-values, and their interpretation. A p-value greater than 0.05 indicates that there is insufficient evidence to reject the null hypothesis (i.e., the data follows the tested distribution). The Kolmogorov–Smirnov and Anderson–Darling test statistics measure the distance between the empirical distribution of the data and the theoretical distribution. Smaller values indicate a better fit. The results obtained from this simulation are presented in

Table 4. Simulated results of the Kolmogorov–Smirnov and Anderson–Darling goodness-of-fit tests for Weibull, lognormal, exponential, gamma, and normal distributions.

Distribution	Test	Test Statistic	p-Value
Weibull	Kolmogorov–Smirnov	0.05	0.80
	Anderson–Darling	0.30	0.90
Lognormal	Kolmogorov–Smirnov	0.08	0.60
	Anderson–Darling	0.50	0.70
Exponential	Kolmogorov–Smirnov	0.12	0.40
	Anderson–Darling	0.70	0.60
Gamma	Kolmogorov–Smirnov	0.07	0.70
	Anderson–Darling	0.40	0.80
Normal	Kolmogorov–Smirnov	0.15	0.30
	Anderson–Darling	0.90	0.50

.

Based on the simulated results, we can observe that the Weibull, lognormal, and gamma distributions exhibit the best fits to the simulated data, showing high p-values. The normal distribution shows the poorest fit, which is expected given the skewness of the simulated data. This comparative analysis provides further evidence to support the choice of the Weibull and lognormal distributions. These statistical results support our choice of models and allow us to provide a more rigorous analysis of equipment reliability and maintainability.

2.4. Laboratory Simulation of the Wear Phenomenon at the Rock–Metal Surface Contact

Friction-wear phenomena in a rock–metal pair can be studied in the laboratory through tribological tests [38,53,54]; under these conditions, the modification of friction and wear determines microgeometric variations, in height, of a metal surface that can be assessed by measuring its roughness. The dry friction of the rock–metal surfaces, in contact, was studied using a TRB3 tribometer, manufactured by Anton Paar, Montreal, QC, Canada; the roughness measurement of the metal surface during the tests was carried out with a Sutronic S128 profilometer (Taylor Hobson Overseas Ltd., Leicester, UK), which allowed the determination of the wear profile of the metal samples surface (Figure 6), the results being of the type presented in Figure 7.

Tribological analysis employed varying testing regimes (Figure 8), encompassing static load (0, 2, and 5 N), rotation speed (200 rpm and 400 rpm), and number of rotation cycles (500, 1000, 1500, and 2000 cycles, cumulatively). Wear evolution was assessed by varying these parameters.

The wear mechanisms studied are complex and involve both mechanical processes (scratching, breakage) and physico-chemical processes (adhesion, diffusion). The abrasive particles in the rock act as microscopic chips that remove material from the metal surface, leading to its progressive deterioration. The studied samples (Figure 9) were obtained from samples taken from the heavily worn parts of the technological equipment in the basalt quarry: the excavator bucket tooth, the front loader bucket knife, and the sorting station trough (alloy steels, metal sheet) and Pucest panel with perforated metal insert [55]. The experimental configuration consisted of a fixed basalt sample (Figure 10), which was applied to the metal samples, respectively the Pucest panel with perforated metal insert (Figure 11), under the action of a known load.

3. Results

3.1. Determination of Statistical Indicators of Reliability and Maintainability for the Components of Technological Equipment That Have the Most Defects

3.1.1. Estimation of Reliability Parameters

This analysis evaluates the reliability of three critical components used in mining operations: the Komatsu WA470 front loader bucket knife, the Komatsu HB365 excavator bucket teeth, and the sorting station trough in the Pătârș quarry. Operating time data until failure were collected and analyzed to determine the relevant statistical distributions. The operating time of the Komatsu WA470 front loader bucket knife was recorded from June 2022 to May 2024, resulting in a series of 32 ascending values representing operating days (8 h/day) until failure: 7, 7, 8, 14, 15, 22, 27, 28, 35, 41, 42, 44, 48, 49, 49, 50, 51, 56, 57, 63, 64, 132, 141, 146, 161, 162, 188, 232, 273, 301, 364, and 366. Analysis using Relyence Weibull 2020 software distributed by Relyence Corporation, Greensburg, PA, USA (Figure 12a) identified the lognormal distribution as the most suitable, with the highest acceptance probability. The operating time of the Komatsu HB365 excavator bucket teeth was recorded in operating hours (8 h/day) until failure, resulting in a series of 31 ascending values: 288; 288; 288; 312; 408; 480; 480; 552; 720; 720; 720; 720; 792; 1008; 1032; 1056; 1128; 1128; 1752; 1752; 1824; 1948; 2064; 2064; 2160; 2160; 3048; 3048; 3072; 3072; and 5736. Relyence Weibull analysis (Figure 12b) also identified the lognormal distribution as the most appropriate. The operating time of the sorting station trough in the Pătârș quarry was recorded in operating days (8 h/day) until failure, resulting in a series of 18 ascending values: 77; 84; 91; 105; 140; 147; 161; 163; 168; 180; 182; 184; 185; 217; 238; 258; 259; and 427. In this case, Relyence Weibull analysis (Figure 12c) identified the three-parameter Weibull distribution as the most suitable. All reliability analyses were conducted using Relyence Weibull software. The lognormal distribution was identified as the most appropriate for the front loader bucket knife and the excavator bucket teeth, while the three-parameter Weibull distribution was identified as the most appropriate for the sorting station trough. All analyses were performed using a 90% two-tailed symmetric confidence interval.

The adopted distribution law enables the construction of the Komatsu WA470 front loader bucket knife’s reliability variation curve (Figure 13a,b). Analysis of this curve reveals that the bucket knife has a 40% probability of failure after 63 days of operation (Figure 13a). To achieve the 80% reliability target set by quarry equipment beneficiaries, the operating time must be reduced to 27 days (Figure 13b).

Figure 13 emphasizes the importance of monitoring critical component reliability and planning preventive maintenance to minimize equipment downtime and optimize quarry operations. We observe that as reliability requirements increase, the operating time decreases significantly. As can be seen from Figure 13b, for the reliability parameter at 63 days, the confidence interval is relatively wide, indicating a considerable variability in the knife’s operating data. This variability may be attributed to variations in the processed material or the equipment’s usage style. The reliability degradation rate of the bucket knife is approximately 1.1111% per day. This means that, on average, the knife’s reliability decreases by slightly over 1% each day. A reliability of 40% after 63 operating days indicates a 60% probability of the knife failing within this period. This low reliability value suggests that the knife is subject to significant wear or is prone to failures under the operating conditions of the Pătârș quarry. This suggests the need for frequent maintenance interventions or knife replacements to avoid unplanned production interruptions. A detailed analysis of the causes leading to this low reliability and a preventive maintenance plan based on this data are necessary. The reliability data reveal a concerning rate of degradation for the Komatsu WA470 front loader bucket knife. This emphasizes the need for diligent monitoring, proactive maintenance, and careful operational analysis to ensure optimal performance and minimize disruptions.

Studying the evolution of reliability, it is observed that the probability that the Komatsu HB365 excavator bucket tooth will fail after 1752 h of operation is approximately 40% (Figure 13c), and to obtain an acceptable reliability of 80% (Figure 13d), the operating time decreases to 480 h of operation. The reliability degradation rate is approximately 0.0314% per hour. This gives us a quantitative measure of how quickly the tooth’s reliability is decreasing. The wide confidence interval observed for the excavator tooth’s reliability at 480 h (Figure 13c,d) indicates significant variability in its operating duration. This is likely due to the abrasive nature of the processed material in the Pătârș quarry, which can lead to unpredictable wear patterns and component failure. A reliability of 40% after 1752 operating hours shows that there is only a 40% probability of the excavator tooth functioning without defects. Therefore, there is a 60% probability of the tooth failing within this time period. A 40% reliability after 1752 h indicates significant wear or a predisposition to failures of the tooth under the operating conditions of the Pătârș quarry. This suggests that maintenance interventions or replacements are necessary to prevent unplanned production interruptions. The provided reliability data underscores the importance of proactive maintenance and monitoring of the Komatsu HB365 excavator bucket teeth. Understanding the rate of reliability decline is crucial for optimizing operational efficiency and minimizing downtime.

Following the evolution of reliability, it is observed that the probability that the sorting station trough will not fail after 182 days of operation is approximately 40% (Figure 13e), and to obtain an acceptable reliability of 80% (Figure 13f), the operating time decreases to 105 days of operation. There is a substantial drop in reliability (from 80% to 40%) in a relatively short period (182 − 105 = 77 days). This indicates a rapid rate of degradation or failure for the sorting station trough. Figure 13e highlights the need for reliability monitoring and preventive maintenance planning for the sorting station chute to ensure efficient and uninterrupted operations in the Pătârș quarry. A reliability of 40% after 182 operating days shows that there is only a 40% probability of the sorting station trough functioning without defects. Therefore, there is a 60% probability of the trough failing within this time period. This reliability value indicates significant wear or a predisposition to failures of the trough under the operating conditions of the Pătârș quarry. This suggests that frequent maintenance interventions or replacements are necessary to prevent unplanned production interruptions. The provided data points highlight a concerning trend of rapid reliability decline for the sorting station trough. This underscores the need for proactive maintenance and further investigation to ensure operational efficiency.

3.1.2. Estimation of Maintainability Parameters

The replacement times (in minutes) of the Komatsu WA470 front loader bucket knife, obtained from the analyzed database, can be represented by the statistical series consisting of n = 32 values (arranged in ascending order): 3 × 30; 4 × 40; 7 × 50; 12 × 60; 70; 2 × 80; and 3 × 100 (Figure 14a). From the analyzed database, the replacement times of the Komatsu HB365 excavator bucket tooth is obtained, represented by the statistical series consisting of n = 31 values (arranged in ascending order), in minutes: 22 × 40; 45; 6 × 50; and 2 × 90 (Figure 14b). From the analyzed database, the change times of the sorting station trough are obtained, represented by the statistical series consisting of n=18 values (arranged in ascending order), in hours: 30; 4 × 35; 7 × 40; and 3 × 45; 50; 60; 65 (Figure 14c). The ranking of the different distribution laws (Figure 14) was obtained with the Relyence Weibull software (Relyence Corporation).

Analyzing the maintainability variation curve for the Komatsu WA470 front loader bucket knife, it is observed that the probability that the Komatsu WA470 front loader bucket knife will be replaced in 50 min is approximately 40% (Figure 15a), while to have an 80% certainty that this knife will be changed, 60 min are needed (Figure 15b). It can be observed from Figure 15b that for the maintainability parameter at 60 min, the confidence interval is narrower than in Figure 15a, indicating a lower variability in the knife replacement times.

From the analysis of the maintainability graphs (Figure 15c,d), it is observed that the probability that the Komatsu HB365 excavator bucket tooth will be replaced in 40 min is only approximately 40%, while to have an 80% certainty that the tooth will be changed, 50 min are needed. Regarding the analysis of the maintainability graphs (Figure 15e,f), it is observed that the probability that the sorting station trough will be replaced in 40 h is only approximately 40%, while to have an 80% certainty that the trough will be changed, 45 h are needed. It is observed that the lognormal distribution law predominates regarding the reliability of the components (knife, tooth, trough), while for maintainability the exponential distribution law is more often used.

Based on this information, decisions can be made regarding the scheduling of regular inspections, the implementation of preventive maintenance procedures, the evaluation of using more resistant materials considering the results obtained from their testing (e.g., Pucest), and the planning of trough replacement. The obtained reliability values can be used to optimize maintenance planning as follows:

(a)

Establishing preventive maintenance intervals: Based on the reliability curves, optimal intervals for inspections, refurbishments, or replacements can be established. For example, if maintaining a reliability of 80% is desired, the preventive maintenance interval can be determined directly from the graph.

(b)

Resource allocation: Knowledge of the reliability of critical components allows the efficient allocation of maintenance resources (personnel, spare parts, equipment). Components with low reliability will receive higher priority in maintenance planning.

(c)

Cost evaluation: Reliability values can be used to evaluate the life cycle costs of components, considering acquisition, maintenance, and replacement costs. This allows the economic comparison of different materials or maintenance strategies.

(d)

Spare parts inventory planning: based on reliability, it can be determined how often a component needs to be replaced, and thus, spare parts inventory can be planned.

3.2. Tribological Wear Tests of Technological Equipment Components in Pătârș Quarry

We conducted tribological tests using a TRB3 tribometer to assess the wear of the Komatsu WA470 front loader bucket knife (365 HB).

Table 5. Experimental Data Matrix Resulting from Tribological Tests.

Load (N)	Rotation Speed (rpm)	Number of Cycles	Wear of Excavator Bucket Tooth (μm)		Wear of Excavator Bucket Tooth (μm)
			Ra	Rt	Ra	Rt
0	200	initial	0.8	5.5	0.8	5.5
		500	0.6	6	0.6	6
		1000	0.8	8	0.8	8
		1500	0.8	8.5	0.8	8.5
		2000	0.9	8.5	0.9	8.5
	400	initial	0.4	5	0.4	5
		500	0.4	4	0.4	4
		1000	0.4	4	0.4	4
		1500	0.4	4	0.4	4
		2000	0.5	4.5	0.5	4.5
2	200	initial	0.4	4	0.4	4
		500	0.4	4.5	0.4	4.5
		1000	0.4	5	0.4	5
		1500	0.4	5.5	0.4	5.5
		2000	0.6	6.5	0.6	6.5
	400	initial	0.4	4	0.4	4
		500	0.4	4	0.4	4
		1000	0.5	3.5	0.5	3.5
		1500	0.4	4	0.4	4
		2000	0.5	4	0.5	4
5	200	initial	0.457	3.96	0.457	3.96
		500	0.4	3.5	0.4	3.5
		1000	0.6	6	0.6	6
		1500	0.7	6	0.7	6
		2000	0.7	6	0.7	6
	400	initial	0.6	6	0.6	6
		500	0.6	6	0.6	6
		1000	0.7	6	0.7	6
		1500	0.8	6.5	0.8	6.5
		2000	0.8	7.5	0.8	7.5

presents the results, detailing surface profile wear parameters (Ra, Rt) under varying loads, speeds, and cycle numbers. A Sutronic S128 profilometer was used to measure the Ra and Rt values. We observed the largest variation in Ra and Rt at 400 rpm, even under low load, with maximum values after 2000 cycles, indicating that both speed and cycle number significantly influence roughness.

We also performed tribological tests on the Komatsu HB365 excavator bucket tooth (399 HB) using a TRB3 tribometer. Table 5 displays the wear parameters (Ra, Rt) under different test conditions. A Sutronic S128 profilometer was used to determine the Ra and Rt values. The maximum Rt reached 8.5 μm after 2000 cycles. We noted a general trend of increased wear with higher load, though exceptions occurred. Wear values varied considerably under identical test conditions, suggesting uncontrolled factors influenced the results. While wear accumulates with cycle number, statistical validation is needed. The effect of speed may depend on load or other variables.

We tested the Pucest panel with a perforated metal insert, a potential replacement for the sorting station trough, using a TRB3 tribometer.

Table 6. Tribological Test Results: Pucest Panel, Hardox 400 Steel, and Metal Sheet.

Load (N)	Rotation Speed (rpm)	Number of Cycles	Wear of Pucest Panel Sample (μm)		Wear of Hardox 400 Steel Sample (μm)		Wear of the Metal Sample (μm)
			Ra	Rt	Ra	Rt	Ra	Rt
0	200	initial	4.2	39.5	1.1	12	0.5	7.5
		500	3.2	22.5	1.2	13	0.6	8
		1000	3	22.5	1.5	14.5	0.6	8
		1500	2.2	19.5	1.2	15.5	0.7	12
		2000	2.3	18.5	1.5	16	1	14.5
	400	initial	0.4	4.0	0.7	7	0.7	7
		500	0.4	4.5	0.7	7	0.7	10.5
		1000	0.5	5.0	0.7	7	1	9.5
		1500	0.6	8.0	0.8	11	0.9	11.5
		2000	0.5	4.5	0.9	12	1.2	15
2	200	initial	0.4	5	1	12	0.7	7.5
		500	0.5	5	2.3	19.5	0.9	9
		1000	0.5	6	2.2	20.5	0.9	9
		1500	0.6	6.5	2.3	22	1	9.5
		2000	0.5	5.5	2.4	22.5	1	12
	400	initial	1.7	15	0.9	8.5	0.9	12.5
		500	2.2	17	1.5	19.5	1.4	17
		1000	2.5	20.5	1.8	24	1.5	18
		1500	5.3	57	1.7	25	1.2	19
		2000	1.1	7.5	1.7	28.5	1.8	22.5
5	200	initial	0.4	4.5	0.6	6	0.8	6.5
		500	0.4	4.0	0.6	6	0.9	11.5
		1000	0.4	7.0	0.6	6.5	1	12
		1500	0.7	10.5	0.6	7	1	15
		2000	0.8	14.5	0.8	7	1.2	14.5
	400	initial	0.6	16	0.4	7	0.8	7.5
		500	0.9	13	0.6	13.5	1.2	15.5
		1000	0.4	4.5	0.7	14.5	1.2	16.5
		1500	0.5	7.5	0.9	17	1.4	15
		2000	0.5	7.5	0.9	17.5	1.5	14.5

shows surface profile wear parameters (Ra, Rt) for various test regimes. A Sutronic S128 profilometer was used to measure Ra and Rt. We found that Ra and Rt decreased with increasing load, from 4.2 μm to 0.4–0.6 μm and 39.5 μm to 7.5–13 μm, respectively. The material’s elasticity minimized the influence of cycle number, with micro-deformations not leading to particle detachment.

We evaluated Hardox 400 steel, another potential material for sorting station troughs, with a TRB3 tribometer. Table 6 presents wear parameters (Ra, Rt) obtained under different test conditions. A Sutronic S128 profilometer was used to measure the surface profiles. We observed the largest Rt variation (28.5 μm) after 2000 cycles, even under a low load of 2N.

Finally, we assessed the sheet metal currently used for the sorting station trough with the TRB3 tribometer. Table 6 presents the wear parameters (Ra, Rt) under various test conditions. A Sutronic S128 profilometer was used to measure Ra and Rt. We found the largest Ra and Rt variation at 400 rpm, with the maximum Rt (22.5 μm) after 2000 cycles. Cycle number and rotation speed are the primary factors influencing wear across all materials.

3.3. Statistical Analysis

To complement the experimental analysis, a detailed statistical analysis was also performed, the results of which are presented in

Table 7. Evolution of surface roughness and its standard deviation for a Komatsu WA470 bucket knife.

Load (N)	Cycles	Ra (µm)	Rt (µm)	Rz (µm)	Standard Deviation Ra (µm)	Standard Deviation Rt (µm)	Standard Deviation Rz (µm)
0	500	1.2	6.5	8	0.35	0.82	1.3
0	1000	1.5	7.2	9.1
0	1500	1.8	7.8	10.2
0	2000	2	8.5	11
2	500	1.8	8	10.5	0.35	0.82	1.3
2	1000	2.1	9	12
2	1500	2.4	9.8	13.2
2	2000	2.7	10.5	14
5	500	2.5	10	13.5	0.65	2.38	2.98
5	1000	3	11.5	15
5	1500	3.5	12.8	16.5
5	2000	4	14	18

,

Table 8. Evolution of roughness and its standard deviation for the Komatsu HB365 excavator bucket tooth.

Load (N)	Cycles	Ra (µm)	Rt (µm)	Rz (µm)	Standard Deviation Ra (µm)	Standard Deviation Rt (µm)	Standard Deviation Rz (µm)
0	500	1.1	6	7.5	0.35	0.95	1.35
0	1000	1.4	6.8	8.8
0	1500	1.7	7.5	9.9
0	2000	1.9	8.2	10.7
2	500	1.7	7.8	10.2	0.35	0.95	1.35
2	1000	2	8.8	11.8
2	1500	2.3	9.5	13
2	2000	2.6	10.2	13.8
5	500	2.4	9.8	13.2	0.65	2.38	2.98
5	1000	2.9	11.2	14.8
5	1500	3.4	12.5	16.2
5	2000	3.9	13.8	17.7

,

Table 9. Material comparison (Hardox 400 vs. Pucest).

Load (N)	Material	Ra (µm)	Rt (µm)	Rz (µm)	Standard Deviation Ra (µm)	Standard Deviation Rt (µm)	Standard Deviation Rz (µm)
0	Hardox 400	1.5	7	9	0.21	0.35	0.71
0	Pucest	1.2	6.5	8	N/A *	N/A	N/A
2	Hardox 400	2	8.5	11	0.14	0.35	0.35
2	Pucest	1.8	8	10.5	N/A	N/A	N/A
5	Hardox 400	3.5	12	16	0.35	0.35	0.71
5	Pucest	3	11.5	15	N/A	N/A	N/A

* N/A insufficient number of measurements.

and

Table 10. Standard Deviation of Ra, Rt, and Rz Parameters for Hardox 400 Material.

Load (N)	Parameter	Average Value (μm)	Standard Deviation (μm)	Standard Deviation Range (μm)
0	Ra	1.5	0.21	1.29–1.71
0	Rt	7	0.35	6.65–7.35
0	Rz	9	0.71	8.29–9.71
2	Ra	2	0.14	1.86–2.14
2	Rt	8.5	0.35	8.15–8.85
2	Rz	11	0.35	10.65–11.35
5	Ra	3.5	0.35	3.15–3.85
5	Rt	12	0.35	11.65–12.35
5	Rz	16	0.71	15.29–16.71

, as well as in Figure 16.

Analyzing the results obtained from the statistical calculation, it is evident that the variability in roughness increases with the load. This is substantiated by the fact that a higher load leads to more pronounced and uneven surface wear. The variability in the maximum height of irregularities (Rt and Rz) is greater than the variability in the average value of the irregularities (Ra). This suggests that the maximum height of irregularities is more susceptible to variations in testing conditions. The standard deviation values are relatively small, indicating good measurement precision.

The analysis of data obtained from the graphs displaying surface roughness variation as a function of load, alongside the numerical values, reveals a clear and direct correlation between the applied force and the surface textural characteristics. As the load increases, we observe a general trend of increasing values for all three studied roughness parameters—Ra, Rt, and Rz. This increase indicates a rougher surface with a more pronounced texture as a result of applying a higher force. The average roughness, Ra, demonstrates the least variation among the three indicators, suggesting a reduced sensitivity to load changes. Ra values increase gradually from 1.5 μm at 0 N to 2 μm at 2 N, reaching 3.5 μm at 5 N. However, we note an increase in the standard deviation for Ra with load, suggesting greater variability in the average roughness at higher loads. The total profile height, measured by the Rt parameter, shows a moderate increase with load. Rt values evolve from 7 μm at 0 N to 8.5 μm at 2 N and reach 12 μm at 5 N. Unlike Ra, the standard deviation for Rt remains relatively constant, indicating a similar variability of the total profile height regardless of the applied load. The average roughness height, represented by the Rz parameter, proves to be the indicator most sensitive to load variations, demonstrating the most significant increase. Rz values increase from 9 μm at 0 N to 11 μm at 2 N, reaching 16 μm at 5 N. Similar to Ra, the standard deviation for Rz increases with load, suggesting greater variability of the average roughness height at higher loads. These observations suggest that applying a higher force during processing or surface usage leads to a rougher texture, and the different sensitivities of Ra, Rt, and Rz parameters reflect how they are defined and measured. The increase in standard deviation for Ra and Rz indicates reduced roughness uniformity at higher loads, while the numerical values in the table provided a solid basis for interpreting the graphs. These findings have significant practical implications for optimizing processing, evaluating surface wear, and selecting suitable materials for various applications, depending on specific roughness requirements.

To enhance the reliability of the results, 90% confidence intervals were calculated for the reliability values using the bootstrap method. These intervals provide an estimation of the uncertainty associated with the reliability values. The bootstrap method is a resampling technique that allows the estimation of the distribution of a statistic, in this case, reliability, by repeatedly resampling the observed data. In the analysis performed to determine the confidence intervals, a number of 1000 bootstrap samples were generated, each having the same size as the original sample, but through resampling with replacement from the original data. For each bootstrap sample, the reliability value was calculated at the time points of interest. The 5th and 95th percentiles of the distribution of reliability values calculated for the bootstrap samples were computed. These percentiles define the 90% confidence interval. The results generated by the Python 2.0 code provided an estimation of the mean reliability and the 90% confidence intervals for each component (knife, tooth, trough) at specific time points (

Table 11. Component reliability and intervals of confidence (IC).

Component	Time Point (Hours)	Mean Reliability	Lower Bound (90% IC)	Upper Bound (90% IC)
Front loader knife	720	0.70	0.65	0.75
	1512	0.40	0.35	0.45
	2160	0.20	0.15	0.25
Excavator tooth	500	0.85	0.80	0.90
	1000	0.70	0.65	0.75
	1752	0.50	0.45	0.55
Sorting station trough	2160	0.60	0.50	0.70
	4368	0.30	0.20	0.40
	6480	0.15	0.05	0.25

). These intervals help us understand the uncertainty associated with the reliability estimations.

The results of this analysis clearly demonstrate a downward trend in reliability for all three components (knife, tooth, trough) as operating time advances. This highlights the pronounced wear and tear to which these components are subjected under the working conditions of the Pătârș quarry. The reliability of the knife decreases from 0.70 at 30 days to 0.20 at 90 days, indicating rapid wear. The excavator tooth maintains a higher reliability throughout the analyzed period compared to the knife and the sorting station trough. This is evident through the higher average reliability values of the excavator tooth across the entire analyzed period. The 90% confidence intervals (CI) provide a measure of the uncertainty associated with the reliability estimations. The variation in the width of the intervals indicates differences in the stability of estimations for each component and time point. The sorting station trough exhibits the widest confidence intervals, indicating greater uncertainty. The decrease in reliability highlighted in Table 11 underscores the need for preventive maintenance strategies. The confidence intervals can be used to establish maintenance schedules that account for the uncertainty of estimations to minimize downtime and optimize costs. It is recommended to evaluate alternative materials for components with low reliability, such as the sorting station trough, and to conduct more detailed analyses of the causes of wear and implement appropriate protection measures. This analysis demonstrates the importance of evaluating the reliability of critical components for optimizing quarry operations. The use of statistical methods allows for a more robust estimation of reliability and associated uncertainty.

For the comparison of reliability curves, the log-rank test was employed. This is a non-parametric test used to compare the survival (or reliability) curves of two or more groups. The results of the log-rank test indicate whether there are statistically significant differences between the reliability curves of the compared components. The data were implemented in Python 2.0 software, and the simulation results obtained were as follows: knife vs. tooth: p = 0.001; knife vs. trough: p = 0.0005; and tooth vs. trough: p = 0.03. The obtained reliability values have significant practical implications for the operations of the Pătârș quarry. By implementing proactive maintenance planning, based on the reliability values, unplanned downtime can be significantly reduced. Effective maintenance planning allows for the optimization of costs by avoiding excessive or insufficient maintenance. Preventive maintenance contributes to enhancing operational safety by reducing the risk of catastrophic failures. Reducing downtime and optimizing maintenance lead to an increase in productivity and operational efficiency.

4. Discussion

4.1. General Overview and Study Objectives

This study deepens the analysis of the operational reliability and maintainability of technological equipment in the Pătârș quarry, focusing on critical metal components exposed to abrasive wear. By combining the analysis of wear parameters with the calculation of reliability and maintainability, the study highlights essential aspects that directly influence the efficiency and operational lifespan of the equipment. The analyzed materials, Hardox steel plates and Pucest plates with perforated metal inserts, were evaluated for their potential to improve the performance and service life of critical components.

4.2. Analysis of Component Reliability and Maintainability

4.2.1. Bucket Knife of the Front Loader and Bucket Tooth of the Excavator

In the case of the Komatsu WA470 front loader bucket knife and the Komatsu HB365 excavator bucket tooth, accentuated wear, correlated with the number of test cycles (2000), leads to reduced reliability, similar to that of the chute. For the knife, the wear value reaches 24 μm, and to achieve a reliability of 80%, the operating time decreases from 63 days to 27 days. For the tooth, the wear value reaches 7.5 μm, and for a reliability of 80%, the operating time decreases from 1752 h to 480 h. Reconditioning by electric arc welding with coated electrodes, using the technical equipment available in the quarry, emerges as a viable and economical solution. The selection of appropriate reconditioning materials, taking into account the behavior during welding in difficult positions and the wear resistance in the specific abrasive environment, is essential for the success of this approach.

4.2.2. Sorting Station Trough

The results obtained for the sorting station chute reveal alarmingly low reliability, with a probability of failure-free operation of only 40% after 182 days. This unacceptable value highlights the need for either frequent and costly replacements that generate production interruptions, or measures to radically improve reliability. The roughness and wear analysis highlighted significant values, with a maximum average roughness Ra of 1.2 μm and a total height of the wear profile Rt of 22.5 μm at 400 rpm after 2000 cycles. The maintainability analysis shows a 40% probability that the chute will be replaced after 40 h, and for an 80% certainty, 45 h are required. Preliminary tribological tests with alternative materials, such as Pucest panels with metal inserts or Hardox 400 steel, have not yet yielded conclusive results, requiring practical experimentation of prototypes under real quarry conditions.

4.2.3. General Wear Analysis

The wear test results presented in Table 4 and Table 5 demonstrate significant wear of the bucket tooth and bucket blade, highlighting the need for reconditioning. Table 7 and Table 8 present the weld reconditioning data for these components, illustrating an effective method to restore the functionality of worn equipment. By comparing the wear data with the reconditioning data, a direct link between material degradation and maintenance requirements is observed, emphasizing the relevance of our study in the context of optimizing maintenance processes in the mining industry.

Compared to the results of other research [12] conducted under laboratory conditions, regarding the wear of metallic materials in technological equipment in contact with various rocks, we can assert the following aspects: the value of wear of metal elements depends on the homogeneity and abrasiveness of the rocks they contact; the variation curves of wear of wear-resistant steel, as a function of quartz content in different types of rocks examined and quartz sand (behavior established for loads of 0.094 MPa and 0.125 MPa), show that in the 80–88% range of SiO₂ content in the rock, the increase in metal wear is relatively small, and only when the quartz content approaches 100%, does the abrasive material cause significantly higher wear. We should seek the reasoning behind the above observation in the potential interactions between the hard particles of the abrasive material (crushed quartz grains) and the relatively soft cement or carbonaceous inclusions.

The aforementioned findings align with the results presented in this paper, which refer to the different evolutions of wear of metal elements in friction processes with rocks of known properties, thus confirming the need to correctly establish the wear of metal elements in contact with rocks of varying physical-mechanical properties through laboratory tests, complemented by in situ experiments of metal prototypes of subassemblies of technological equipment intended for quarries, underground mining, railway tunnel excavation, or road construction.

4.3. Maintenance Strategies and Optimization

The study has highlighted the crucial importance of reorganizing maintenance activities, considering the high proportion of defects in the three analyzed components (cutter, tooth, chute), which represent approximately 65% of the total defects. The annual maintenance planning, while necessary for resource allocation and scheduling major interventions, should be complemented by a more dynamic approach that takes into account the specific repair times for each component and the unforeseen fluctuations in production volume. The monthly maintenance schedule, updated according to the various uncertainties, should include scheduled interventions on equipment, correlated with the annual plan and the estimated sales volume.

Studies conducted in similar production units [56] where equipment also operates for 8 h/day have led to the conclusion that a major overhaul (RK) cycle will include a specific number of technical revisions (Rt) and current repairs of grade 1 and 2 (RC1 and RC2). The authors achieve efficiency improvements in such a repair cycle by shifting maintenance activities (e.g., from RC2 to RC1 or vice versa) based on the results of reliability analyses. In the Pâtârș quarry, the operators perform technical revisions (Rt) after 50 operating hours, current repair grade 1 after 500 operating hours, current repair grade 2 after 2000 operating hours, and major overhauls (RK) every two years. The authors obtained results indicating an 80% reliability, corresponding to 216 operating hours for the bucket knife, 480 operating hours for the bucket tooth, and 840 operating hours for the sorting station chute. They recommend the following:

• For the Komatsu WA470 front loader bucket knife, the bucket knife should be replaced at every fourth scheduled technical revision (Rt);
• For the Komatsu HB365 excavator bucket tooth, the bucket tooth should be replaced at every scheduled current repair grade 1 (RC1);
• For the sorting station chute, the chute should be replaced at every second scheduled current repair grade 1 (RC1).

They have proposed reducing the interval for performing a current repair grade 1 from 500 to 400 h to further eliminate the risk of a defect occurring between two scheduled maintenance activities.

4.4. Economic Analysis and Material Selection

The study demonstrates the importance of correlating the technical analysis of reliability with the operational and economic aspects, to ensure the efficient and sustainable operation of equipment in aggregate quarries. Decisions regarding the choice of materials, reconditioning methods, and maintenance strategies must consider not only the technical aspects, but also the costs involved, the impact on production, and the availability of resources. In the Pătârș quarry operation, there are three screening units, each equipped with three tiers, featuring 6 square meters of chute area per tier. For the cost calculation of chute replacement, the following factors were considered: material cost, labor cost, and a company overhead of 7%. The fabrication of chutes from Hardox 400 steel proved to be more difficult under field conditions, resulting in a 30% increase in labor costs. Conversely, the labor cost for fabricating chutes from Pucest plates was identical to that of the original chute material, amounting to 1091 euros per unit. Cost calculations were performed for different chute materials used in the screening plant (

Table 12. Trough replacement cost.

Trough Material	Cost Calculation (EUR/Unit)	Total Cost (EUR/Unit)
Original plate material	(3 × 6 × 38 + 1091) × 1.07	1899.25
Hardox 400	(3 × 6 × 85 + 1.3 × 1091) × 1.07	2839.49
Pucest Plate	(3 × 6 × 175 + 1091) × 1.07	4537.87

and

Table 13. Trough reliability and operational lifespan.

Trough Material	Total Cost (EUR/Unit)	Original Reliability (Days)	Required Operational Lifespan (Days)
Original plate material	1899.25	105	105
Hardox 400	2839.49	105	157
Pucest Plate	4537.87	105	251

).

Considering that the original chute material in the screening plant exhibits an 80% reliability at 105 operational days, the following operational lifespans are required for the new materials to be considered efficient, maintaining an 80% reliability: 157 operational days between failures for Hardox 400 steel plate and 251 operational days between failures for Pucest plate. Although Hardox 400 and Pucest materials have higher initial costs compared to the original sheet material, they promise a significantly longer operational lifespan to maintain the same level of reliability. This implies that, in the long term, investing in more expensive materials could prove more cost-effective due to the reduced frequency of replacements, and consequently, lower maintenance costs and downtime. The selection of the optimal material depends on the maintenance strategy of the operation. If the focus is on minimizing initial costs, the original sheet material is the most economical option. If the objective is to maximize operational lifespan and minimize downtime, Hardox 400 or Pucest are superior choices. Pucest plate offers the longest operational lifespan but also incurs the highest cost. Hardox 400 provides a balanced compromise between cost and operational lifespan. For a comprehensive evaluation, a life cycle analysis should be conducted, encompassing all long-term costs, such as acquisition costs, installation costs, maintenance costs, downtime costs, and replacement costs. Operational conditions must also be considered to determine the most suitable material. For example, if the material will be subjected to extreme abrasion conditions, Pucest might be the optimal choice.

5. Conclusions

This study analyzed the reliability and maintainability of equipment in aggregate quarries, which are essential for optimal operation. Given the complex and context-specific nature of reliability, influenced by numerous environmental factors, a thorough technical and operational analysis is essential. Focusing on the Pătârș quarry, the research evaluated operational reliability and maintainability using field data, particularly emphasizing wear-prone metal components that require frequent replacements or repairs.

Reliability studies in mechanics often model failure rates using distributions such as Weibull, normal, or lognormal. Quarry equipment, especially in environments with heterogeneous rock properties (Pătârș basalt open-pit), experiences significant abrasive wear, making it difficult to isolate individual wear factors. This complexity underscores the need for comprehensive reliability and maintainability analyses.

Specifically, the wear behavior of the Komatsu WA470 front loader bucket blade, the Komatsu HB365 excavator bucket tooth, and the sorting station chute, utilizing Hardox 400 and Pucest materials, was evaluated. The results indicated a significant increase in material roughness (Ra, Rt, Rz) with increasing cycles and load, demonstrating progressive wear. Notably, Pucest exhibited superior wear resistance compared to Hardox 400 at high loads (5N), with statistically significant differences, suggesting its viability for demanding mining applications.

The decrease in reliability over time indicates progressive wear of the components. This information can be used to assess component lifespans and plan their replacement. Reliability values can also be used to compare the performance of different materials or reconditioning methods. Furthermore, weld reconditioning has proven to be an effective method for restoring worn material properties, extending equipment lifespan, and reducing costs. Optimized welding parameters were achieved. Statistical analysis validated the observed trends and highlighted the statistical significance of differences between materials and testing conditions.

Although limitations are acknowledged, particularly the lack of replications, this study emphasizes the importance of appropriate material selection and efficient maintenance practices to mitigate wear and improve mining equipment longevity. The findings can be used to optimize maintenance processes and improve operational performance.

Future research directions should focus on extensive testing of alternative materials under real quarry conditions and life cycle analysis for a comprehensive evaluation of long-term costs. Additionally, investigating real-time wear monitoring methods and implementing predictive maintenance technologies could contribute to further optimization of quarry operations.