A Metrologically-Grounded Analysis of Bridge-Type Bucket-Wheel Excavator Degradation under Variable Operating Conditions

Bridge-type bucket-wheel excavators (BWEs) are essential machinery in large-scale mining and material handling, where they operate under heavy mechanical stresses and environmental challenges. These machines, used primarily in the extraction of iron ore, face varying loads and operating conditions that lead to progressive wear and frequent failures. The critical components of BWEs, including the bucket wheel, rollers, and carriages, are subjected to abrasive contact, fluctuating loads, and humidity variations, which accelerate degradation and reduce operational reliability over time.

Previous research on BWEs has primarily focused on structural integrity assessments, failure analysis, and maintenance optimization for individual components. Studies such as the previous studies [1-4] have investigated the failure mechanisms of bucket wheel components, including fatigue and wear, while others have addressed specific damage modes such as fracture in pulley systems [5] and crack propagation in the boom [6]. These efforts have provided valuable insights into the damage mechanisms and operational challenges faced by BWEs, but most have been limited by deterministic assumptions, focusing on individual components or static load profiles. As a result, they often fail to capture the dynamic and interactive nature of degradation mechanisms that occur in real-world operational environments. For example, recent reliability analyses of mining machinery highlight that complex systems operating under varying loads exhibit highly stochastic failure behavior, underscoring the need for probabilistic degradation modeling rather than static assumptions [7].

Wear is a major degradation mechanism in BWEs and is influenced by operating load, contact conditions, and moisture-related service conditions [8-10]. Although previous studies have reported valuable results on structural failure, component damage, and maintenance issues in BWEs, these studies have generally treated wear analysis, reliability assessment, and operational modeling separately. As a result, the link between metrologically controlled wear observations and system-level degradation analysis under variable operating conditions remains insufficiently developed [11, 12]. The contribution of the present study is therefore not the proposal of a universal degradation law for all BWEs, but the development of a case-specific analytical framework in which calibrated wear measurements with stated uncertainty are combined with field failure data and discrete-event simulation to examine how operating conditions are associated with degradation behavior and reliability in one bridge-type BWE system. In this framework, load and moisture-related conditions are considered jointly at the level of scenario analysis and degradation interpretation, rather than through a fully estimated multivariate stochastic intensity model. The objective is to provide a measurement-grounded basis for reliability-informed interpretation of degradation trends under realistic industrial conditions [13, 14].

The remainder of the paper is organized as follows. Section 2 describes the investigated system and its operating context. Section 3 presents the degradation-modeling assumptions. Section 4 details the measurement methodology and metrological framework. Section 5 reports the statistical analysis of failure data. Section 6 presents the experimental wear study, and Section 7 reports the simulation-based production and reliability analysis. Section 8 discusses the results, limitations, and implications, and Section 9 concludes the paper.

A distinguishing feature of the present study is the integration of calibrated wear measurements with stated uncertainty into the degradation and reliability analysis of the investigated BWE system.

3.1 Wear model

To represent the progressive loss of material in the roller elements, the present study adopts the Archard wear framework. In this formulation, material removal depends on the contact load, the relative sliding path, and the resistance of the softer surface to deformation [10]. For the operating intervals considered here, the wear depth is treated as approximately linear with time. This simplification is appropriate for the normal service regime investigated in the experimental campaign and provides a practical basis for linking measured wear to system-level degradation behavior [12, 15-19].

3.2 Assumption on the impact of load and sliding velocity on wear

The abrasive wear formulation follows classical Archard-based tribological models [10, 15, 17].

It is assumed that the operating load applied to the production system significantly influences the wear rate. As the load increases, the extent of component wear also increases, which in turn raises the likelihood of failure. Under mechanical friction conditions involving two-body abrasive wear, the worn volume V (m³) is directly related to the normal load F_N (N) and the sliding distance L (m), and inversely related to the hardness H (N·m^-2) of the softer contacting surface. Accordingly, the abrasive wear volume can be expressed as follows [15, 16].

$V=\frac{K_a F_N L}{H_s}=\frac{G}{\gamma}$   (1)

where, G (kg) represents the weight loss, γ (kg/m³) denotes the material density, and V/L (m³/m) corresponds to the volumetric wear rate. In tribological systems characterized by frictional contact between a hard surface and a significantly softer counter-surface, wear mainly affects the softer material. Nevertheless, the resulting worn volume cannot be predicted reliably by deterministic formulations alone.

The wear coefficient formulation is commonly adopted in tribology studies [12, 15].

In real-world conditions, the wear factor K (m²/N) is a practical approach to describing wear and is more widely used than the K_a/H_s ratio, which has less applicability across varying operational scenarios. The relationship is therefore expressed as:

$K=\frac{K_a F_N L}{H_s}=\frac{V}{F_N L}$   (2)

This assumption is commonly used for plastic contact conditions in tribological systems [11, 15].

Under conditions of plastic deformation, the normal contact pressure may be assumed to be approximately equal to the hardness H_s of the worn material. On this basis, the total real contact area can be written as follows:

$A_s=\frac{F_N}{H_s}$   (3)

By combining Eqs. (1) and (3), the abrasive wear volume of the material can be expressed as follows:

$V=K_a A_r L$   (4)

The sliding distance may be expressed as L = v × t, where v (m/s) is the sliding velocity and t (s) is the duration of the wear process. Similarly, the normal load F_N on the contact interface is given by F_N = p × A, where p (N·m^-2) denotes the average contact pressure and A (m²) refers to the nominal contact area.

Substituting these expressions into Eq. (1) yields:

$V=\frac{K_a F_N L}{H_s}=\frac{K_a(v t) F_N}{H_s}=\frac{K_a(v t) p A}{H_s}=\frac{K_a(p A) t v}{H_a}$   (5)

The thickness h (m) of the linear wear quantity can be expressed as follows:

$h=\frac{V}{A}=\frac{K_a(p v) t}{H_s}=\frac{G}{A \gamma}$   (6)

The linear wear rate, or wear intensity, is defined by the ratio h/L (mm^-1). When the wear depth h is used as the reference quantity for characterizing wear evolution, the time-dependent wear rate may be written as $\lambda=d h / d t$ [16, 18]. For most engineering calculations, a linear relation between wear depth h and operating time t can be assumed. This means that, during the normal service period, the wear rate remains approximately constant. Under this assumption, the wear evolution with time may be simplified as $\lambda=h / t$. Therefore, the wear rate as a function of time can be expressed as follows:

$\lambda=\frac{h}{t}=\frac{K_a(p v)}{H_s}$   (7)

In this study, h denotes the linear wear depth and λ denotes the corresponding wear rate expressed as wear-depth change over time. Under the assumption of approximately steady operating conditions over short intervals, λ is interpreted as a local average wear rate used to compare degradation behavior under different operating conditions.

3.3 Stochastic representation of degradation

The Archard-based formulation presented above is used in this study as a mechanistic basis for interpreting wear progression under different operating conditions. However, the present manuscript does not estimate a fully parameterized stochastic degradation process directly from the wear data. Instead, degradation variability is represented more cautiously through the combined interpretation of measured wear, grouped failure records, and simulation-based operating scenarios.

Within this framework, load and moisture-related conditions are introduced as operating variables associated with changes in wear behavior and failure occurrence. Their role is therefore interpretive and comparative rather than that of formally estimated covariates in a validated stochastic intensity function. Accordingly, the model should be understood as a measurement-grounded degradation-assessment framework rather than as a fully specified stochastic process model with identified distributional form and parameter-estimation procedure.

This clarification is important for the interpretation of the results. The manuscript supports the conclusion that varying operating conditions are associated with different degradation trends in the investigated BWE system, but it does not claim to establish a universally parameterized stochastic wear law for broader direct transfer without further model identification and validation.

5.1 Data origin and aggregation

The failure data used in this study were obtained from industrial maintenance and operational logs of the investigated bridge-type BWE under varying production-load conditions. Each recorded event corresponds to an actual maintenance intervention documented during operation. Because failure events at the individual-component level were relatively infrequent and unevenly distributed over time, the raw observations were grouped by operating-load range and aggregated over extended operating periods. The resulting grouped dataset was used for comparative statistical analysis. The statistical analysis was then conducted using analysis of variance (ANOVA) and Spearman’s rank correlation to examine the association between operating load and failure occurrence. These analyses should therefore be interpreted as based on aggregated industrial field data rather than on a large set of independent raw failure events.

5.2 Correlation analysis between load and failure: An approach using the analysis of variance method

The statistical relationship between operating load and failure occurrence was evaluated using correlation analysis, and the corresponding coefficients and significance levels are reported in Table 6. The load-dependent mean failure rates for the bucket wheel, trolley, bridge, and total system are summarized in Table 7.

To enhance the robustness of the statistical analysis, failure data were aggregated over extended operating periods and reorganized into a larger dataset of size N = 30, reflecting the observed failure trends under different load levels. This approach enables the application of ANOVA while preserving the physical consistency of the original industrial observations, with a sufficient sample size (> 30 by the empirical rule for approximate normality).

Accordingly, the statistical results reported in this section support load-related trends within the grouped industrial dataset used in this study, but they should not be interpreted as establishing fully general effect sizes beyond the investigated case.

Table 6. Correlation analysis between operating load and failure occurrence

Table 7. Load-dependent failure rates: Bucket wheel, trolley and bridge

A discrete-event simulation model was developed to analyze the dynamic behavior of the production line under degradation constraints. The logical structure of the simulation algorithm is summarized in Figure 18, which presents the flowchart of the proposed model, including machine states, accumulator levels, stochastic failure generation, and repair processes.

Figure 18. Flowchart of the new dynamic production line simulation algorithm

Figure 19. Machine operating condition with high flow rate and 20% failure rate

Figure 20. Machine operating condition with average flow rate and 10% failure rate

Figure 21. Machine operating condition with low flow rate and 5% failure rate

The simulation was used to evaluate system performance under three operating scenarios corresponding to low, medium, and high flow rates. The resulting machine operating conditions are illustrated in Figures 19-21, which show the evolution of machine states and failure occurrences over time for each scenario.

Table 8. Simulation table for the system with low flow rate

Note: NSLM = Number of Stops due to Lack of Material; DT = Downtime; MCT = Mission completion time

Table 9. Simulation table for the system with average flow rate

Note: NSLM = Number of Stops due to Lack of Material; DT = Downtime; MCT = Mission completion time

Table 10. Simulation table for the system with high flow rate

Note: NSLM = Number of Stops due to Lack of Material; DT = Downtime; MCT = Mission completion time

Key performance indicators derived from the simulation include downtime (DT), mission completion time (MCT), and the number of system-level maintenance events. These indicators are summarized in Tables 8-10, enabling a quantitative comparison of reliability and productivity across different operating policies.

This study highlights the importance of understanding the operating conditions associated with degradation in bridge-type BWEs used in iron-ore handling. The results indicate that higher operating loads and wetter service conditions are associated with faster degradation of critical components and with less favorable reliability outcomes.

The findings should therefore be interpreted as a case-study contribution that provides a measurement-grounded basis for reliability-informed maintenance assessment in the investigated excavator system, rather than as a broadly generalizable system model derived solely from the monitored roller sample.

Within the scope of the present case study, the simulation framework is intended to compare operating scenarios and their reliability implications. It should not be interpreted as a formal optimization algorithm for determining an optimal load policy. The main contribution of the study is therefore a measurement-grounded basis for reliability-informed maintenance assessment under variable operating conditions.

The analysis also highlights the influential roles of production rate and initial storage levels, which directly affect overall system performance. A detailed investigation into the interactions among these parameters is essential for designing resilient systems capable of meeting production targets while mitigating risk. Furthermore, seasonal fluctuations in humidity call for environment-specific adaptive maintenance techniques. Maintaining equipment performance and structural integrity requires the implementation of particular preventive measures in response to climatic fluctuations.

To improve bucket excavators' longevity and operational efficiency, it is highly advised that these various factors be incorporated into their design, monitoring, and maintenance. A proactive and multifaceted approach to managing degradation in industrial excavation systems is supported by the convergence of environmental awareness, data-driven optimization, and algorithmic load control.