Floor Scrubber Brush Wear Mechanism and Service Life Prediction: A Tribological Study of PA6, PP, and PBT Brush Materials

Abstract

This study investigates the wear mechanisms and service life of three commonly used brush materials — polyamide 6 (PA6), polybutylene terephthalate (PBT), and polypropylene (PP) — in industrial floor scrubber applications. Pin-on-disc tribometer tests were conducted at three normal loads (20N, 40N, 80N) to establish the Archard wear coefficients for each material. Scanning electron microscopy (SEM) analysis identified abrasive wear as the dominant mechanism on concrete surfaces and adhesive-fatigue wear on epoxy surfaces. A service life prediction model was developed and validated against 2000 hours of field data from 12 operating facilities. Results show PBT exhibits the highest wear resistance (k = 2.8×10⁻⁵, 95% CI [2.5, 3.1]×10⁻⁵), followed by PA6 (k = 4.1×10⁻⁵, 95% CI [3.8, 4.5]×10⁻⁵) and PP (k = 5.7×10⁻⁵, 95% CI [5.2, 6.3]×10⁻⁵). The wear coefficient increased by 2.3× on concrete versus epoxy surfaces for all materials. The proposed model predicts brush replacement intervals with a mean absolute error of 12.4% when validated against field data.

1 Introduction

Floor scrubber brush discs are the primary consumable component in industrial cleaning operations, directly influencing both cleaning quality and operational cost. Despite their critical role, quantitative data on brush wear mechanisms and service life prediction remains scarce in the published literature. Most replacement decisions are based on visual inspection alone, leading to either premature replacement (unnecessary cost) or delayed replacement (reduced cleaning quality).

Previous investigations of brush materials have focused on cleaning performance rather than wear durability. Studies by Schmitz et al. (2021) examined bristle stiffness effects on soil removal, while Park and Lee (2022) compared the fatigue life of thermoplastic bristles under cyclic bending. However, no published work has systematically characterized the abrasive wear behavior of scrubber brush materials using the Archard wear framework, nor provided validated service life prediction models for field use. This study addresses that gap through a combination of controlled tribometer testing, surface morphology analysis, and multi-site field validation.

2 Theoretical Framework

2.1 Archard Wear Model

The Archard wear equation describes the relationship between wear volume and operating parameters:

where V = wear volume (mm³), k = dimensionless wear coefficient, F = normal load (N), s = sliding distance (m), and H = material hardness (N/mm²). For rotary brush applications, the sliding distance per hour of operation is:

For the standard 510mm disc brush rotating at 180 rpm, the peripheral sliding distance is approximately 28.9 km per hour of continuous operation. The Archard wear coefficient k represents the probability that a contacting asperity will produce a wear particle and is a material-specific parameter determined experimentally.

2.2 Wear Mechanism Classification

Brush bristle wear in floor scrubbing applications involves three primary mechanisms: (i) abrasive wear — caused by hard particles (dirt, grime, floor debris) cutting or plowing the softer bristle surface; (ii) adhesive wear — localized material transfer between the bristle tip and the floor surface; and (iii) fatigue wear — cyclic stress accumulation leading to subsurface crack propagation and delamination. The dominant mechanism depends on floor surface type (epoxy vs. concrete), soil load, and contact pressure.

3 Materials and Methods

3.1 Brush Material Specifications

3.2 Tribometer Testing

Pin-on-disc tribometer tests (ASTM G99 standard) were conducted using brush fiber specimens mounted as stationary pins against rotating counterfaces of epoxy-resin and concrete tiles. Three normal loads (20N, 40N, 80N) were applied at a sliding speed of 0.5 m/s, representing the range of contact pressures experienced by scrubber brush discs. Each test condition was replicated 5 times (3 materials × 3 loads × 2 surfaces × 5 replicates = 90 total tests). Wear volume was measured by mass loss using a precision balance (0.1 mg resolution), converted to volumetric wear using material densities.

3.3 SEM Analysis

Post-test bristle surfaces and cross-sections were examined using a scanning electron microscope (SEM) at 50×, 200×, and 500× magnification to identify wear mechanisms. Energy-dispersive X-ray spectroscopy (EDS) was performed on selected samples to detect foreign particle embedment.

3.4 Field Validation Protocol

Twelve facilities were recruited for field validation: six with epoxy floors and six with sealed concrete floors. Each facility operated a BIOCCE BC500 walk-behind scrubber at standardized parameters (40N brush pressure, 3.0 km/h forward speed, 1.0 L/min flow rate) for 2000 hours total. Brush length reduction was measured at 200-hour intervals using a digital caliper (±0.01mm). Replacement was triggered when bristle length reached 50% of original (8mm reduced to 4mm).

4 Results

4.1 Archard Wear Coefficients

The Archard model provided excellent fit (R² > 0.94) across all material-surface combinations, confirming its applicability for brush wear prediction. Concrete surfaces produced wear coefficients 2.2-2.3× higher than epoxy surfaces for all three materials, attributed to the higher abrasiveness of silica-based concrete aggregates. PBT demonstrated the lowest wear coefficient on both surfaces, approximately 32% lower than PA6 and 51% lower than PP.

4.2 Load Dependence of Wear Rate

The load exponent was close to 1.0 for all materials, confirming linearity between wear rate and applied load as predicted by the Archard model. The slight deviation in PP (1.08) suggests a small non-linear contribution from fatigue wear at higher loads.

4.3 SEM Surface Morphology Analysis

SEM examination of worn bristle surfaces revealed distinct wear mechanism signatures between floor types. On epoxy surfaces, PA6 bristles exhibited smooth, polished wear facets with fine abrasion grooves (width 2-5 µm) oriented parallel to the sliding direction, characteristic of two-body abrasive wear. Occasional localized melting zones (identified by recast polymer layers) were observed at 80N load, indicating frictional heating exceeding the polymer's softening point at high contact pressures.

Concrete surfaces produced significantly different wear morphology. All three materials showed deep plowing grooves (width 10-30 µm) with embedded silica particle fragments confirmed by EDS analysis. PP bristles additionally exhibited transverse microcracking (crack spacing 50-120 µm) perpendicular to the sliding direction, consistent with fatigue crack propagation under cyclic loading. PBT bristles on concrete showed the least surface damage, with predominantly mild abrasive wear and no evidence of fatigue cracking.

4.4 Field Validation Results

Field validation (n = 12 facilities, 2000 h each) showed good agreement with the Archard-based model, with a mean absolute error of 12.4%. The systematic pattern of under-prediction on concrete (model predicts shorter life than observed) and over-prediction on epoxy (model predicts longer life) suggests that the model's assumption of linear wear progression throughout the bristle's life may not fully capture the non-linear wear acceleration that occurs as bristle length decreases and contact stiffness increases.

5 Discussion

5.1 Wear Model Limitations and Refinements

While the Archard model provided satisfactory prediction accuracy, two refinements emerged from the field data. First, the wear rate is not constant throughout the brush life — analysis of sequential length measurements shows that wear accelerates by approximately 15-20% in the final third of brush life (p = 0.01). This is attributed to reduced bristle stiffness as length decreases, increasing the effective contact pressure at the bristle-floor interface. A refined model incorporating a wear acceleration term is proposed:

Second, the water absorption behavior of PA6 (1.6% by weight at 24h saturation) was found to reduce its effective hardness by approximately 8% in wet scrubbing conditions, increasing the effective wear coefficient. This partially explains PA6's higher-than-expected wear rate in field conditions where the brush operates continuously in a water-detergent solution. PBT and PP, with water absorption below 0.1%, are unaffected by this mechanism.

5.2 Practical Implications for Brush Selection

• Epoxy floors with light soil: PBT offers the longest service life (400-550 h at 40N) and is the most cost-effective option when labor cost for brush replacement is high.
• Concrete floors with heavy soil: PA6 provides the best balance of wear resistance and cleaning performance. While PBT has lower wear on concrete, its lower tensile strength (56 MPa vs. 78 MPa for PA6) makes it more susceptible to bristle fracture under high-impact conditions.
• High-pressure environments (>60N): PBT is recommended for high-pressure cleaning on both surface types, as its higher hardness (Shore D 82) resists deformation and maintains bristle geometry longer than PA6 or PP.
• Chemical-intensive environments: PP, despite its higher wear rate, offers superior chemical resistance to alkaline and acidic cleaning agents used in food processing facilities.

5.3 Economic Analysis

Based on a brush disc cost of $25-45 (depending on material) and a labor cost of $30/hour for brush replacement, the total brush operating cost per 1000 hours ranges from $38 (PBT on epoxy) to $138 (PP on concrete). Optimizing brush material selection and operating pressure can reduce annual brush costs by 40-60% for a single-scrubber operation running 8 hours daily.

6 Conclusion

This study presents the first systematic tribological analysis of floor scrubber brush materials combining controlled laboratory testing, microscopic wear mechanism analysis, and multi-site field validation. The Archard wear model provides a reliable framework for brush service life prediction (R² = 0.94-0.98 in lab testing, 12.4% mean error in field validation). PBT is identified as the most wear-resistant brush material, offering 30% longer life than PA6 and 50% longer life than PP under identical operating conditions. The load-dependence of brush wear confirms that operating at moderate brush pressure (30-50N) is critical not only for cleaning efficiency (as established in prior work) but also for maximizing brush service life and minimizing consumable costs.

These findings have direct practical implications for users of BIOCCE walk-behind scrubbers and ride-on scrubbers, enabling data-driven brush replacement scheduling that balances cleaning quality with operating cost.