Effect of Tooth Wear on the Technological Parameters of the Milling Process of Asphalt Concrete Layers

Abstract

This article discusses the findings of an experimental study designed to investigate the cutting forces encountered during the milling of asphalt pavement, considering the influence of cutter teeth wear. Experimental research was carried out for different values of wear, considered as a change in the shape of the active part of the tooth and a reduction in its height. The aspects studied continue the previous research of the authors regarding the study of cutting forces when milling asphalt pavement, using new milling teeth (without wear). Through this new study, the authors want to highlight how the phenomenon of wear influences the mechanical conditions of the chipping process and the efficiency of asphalt pavement processing. The experimental research was performed using an original stand, designed by the authors of the article, equipped with instruments for recording the values of the cutting force in the direction of advance. The experimental part is completed by the numerical modeling using the discrete element method (DEM). Research has shown that the increase in cutting forces is more pronounced at low advanced speed. Increasing the advanced speed leads to a reduction in differences between the cutting forces corresponding to the use or not of the milling tooth wear compensation. The study’s findings offer valuable insights into how milling parameters influence cutting forces, providing a basis for optimizing asphalt pavement milling processes.

1. Introduction

Asphalt milling is an essential operation in road maintenance, allowing for the removal of deteriorated pavement layers to prepare for resurfacing. The process is heavily dependent on the performance of milling teeth, which are exposed to abrasive forces that lead to wear. This wear not only affects the efficiency of material removal but also influences energy consumption, surface texture quality, and overall project economics. Understanding the effect of tooth wear on the technological parameters of asphalt milling is very important for optimizing tool performance and ensuring sustainable road rehabilitation practices.

The wear of cutting teeth in asphalt milling is influenced by a range of operational and material factors. Dumitru et al. [1] emphasize the importance of technological parameters, such as cutting speed and milling depth, in determining tool wear rates. Their work demonstrates how parameter optimization can reduce wear and improve performance. Similarly, Guo et al. [2] explore the effects of tool geometry, revealing that modifications to the rake angle and edge design significantly reduce wear progression during milling.

Several studies have used numerical modeling techniques to study the interaction between cutting tool and asphalt. Chen et al. [3] developed a micromechanical DEM approach to simulate the interaction between cutting teeth and asphalt layers. The simulations accurately predicted fracture patterns at varying temperatures and highlighted the effects of air void content and aggregate fraction on cracking behavior. Building on this, Wu et al. [4] simulated the milling process for aged asphalt mixtures, revealing that a cutting speed of 0.5 m/s and a cutting angle of 45° reduce aggregate damage, while a 40° angle with a 25 mm cutting depth at 1 m/s should be avoided to prevent large asphalt pieces. Similarly, in [5] a discrete element model, calibrated using uniaxial compression tests, was developed to analyze the milling process. Simulations identified optimal conditions—a cutting angle of 42°, speed of 0.5 m/s, and depth of 20 mm—that minimize tool wear achieve an 85.3% material utilization rate and reduce energy consumption by 33.53%. Parameter optimization improved material utilization by 17.4%.

You et al. [6] contributed to the field by creating three-dimensional DEM models to analyze dynamic modulus through stress–strain response under compressive loads. Using 2D images of aggregates and mastic, 3D models were reconstructed to simulate loading cycles and predict stiffness. The 3D model provided more accurate modulus predictions than 2D models, aligning well with experimental data across temperatures and frequencies. The effects of air void content and aggregate stiffness on mixture moduli were also evaluated. Extending this line of research, Xue et al. [7] reviewed advancements in DEM applications, providing a comprehensive framework for analyzing granular materials like asphalt during milling operations.

Tool wear monitoring is another essential aspect of the milling process. Yang et al. [8] explored fine-grained machine learning models for tool wear assessment, demonstrating their ability to predict wear states and guide real-time parameter adjustments. Similarly, Nooraie et al. [9] applied finite element methods to estimate tool wear, emphasizing the importance of monitoring systems in prolonging tool life and reducing operational costs.

Field investigations have further validated these findings. Sivilevičius and Martišius [10] investigated the wear dynamics of milling picks used in reclaimed asphalt pavement (RAP) recycling. Their field experiments revealed that the length, carbide tip diameter, and steel body diameter of picks decreased proportionally to the milled surface. The statistical analysis revealed notable differences in wear intensity between picks from two manufacturers, offering essential data for choosing picks based on their wear performance. In parallel, Furmanov et al. [11,12] examined the wear mechanisms of cutting teeth, finding that variations in material composition and process parameters significantly influence wear rates.

Zaumanis et al. [13] conducted a full-scale milling experiment at four jobsites to study the impact of milling parameters (speed, depth, and drum rotational speed) on reclaimed asphalt pavement (RAP) properties. Their results showed that increasing the milling depth and machine moving speed led to larger RAP chunks, while milling in larger chunks produced more filler. Additionally, they found that high milling temperatures influenced the RAP properties, highlighting the importance of milling conditions in determining the performance of recycled asphalt mixture.

Experimental studies, such as those by Liu et al. [14] and Alagarsamy et al. [15], have focused on optimizing cutting parameters to reduce tool wear and enhance surface finish.

Finally, research [16] on wear-resistant coatings for cutting tools explored the use of PVD HiPIMS TiAlVN coatings in milling INCONEL^® 718. The study found that milling parameters such as cutting speed, feed per tooth, and cutting length significantly affected wear mechanisms, including adhesion, abrasion, and delamination. Higher cutting speeds increased wear, while higher milling parameters led to rougher surfaces, with delamination due to poor coating adhesion and crack propagation resistance. These advancements offer promising solutions for extending tool life, though economic feasibility remains a key consideration.

Despite the wealth of research on tool wear and milling parameters, many studies focus on isolated factors or rely on controlled laboratory conditions that do not fully replicate real-world scenarios. Furthermore, the interplay between material properties, tool geometry, and operational parameters under field conditions is not well understood. The present study seeks to highlight the role of the milling parameters (namely, milling depth and advanced speed) in the evolution of cutting force values when milling asphalt pavements under the conditions of the manifestation of specific wear phenomena. It is also aimed at establishing the effectiveness of tooth wear compensation by adjusting the cutting depth during milling. The findings aim to inform strategies for optimizing tool performance and operational efficiency in the processes of asphalt milling.

2. Materials and Methods

2.1. Experimental Stand

To assess the consequences of wear on the asphalt cutter teeth, the research was continued in two additional stages, which involved the use of worn cutter teeth as follows. The research was conducted in two phases: initially, experimental investigations assessed how cutting regime parameters influence horizontal cutting forces; subsequently, discrete element method (DEM) modeling simulated the interaction between worn cutter teeth and asphalt concrete. In order to make a comparison, both the experimental part and the DEM modeling were conducted under the same conditions and using the same conditions as in the case of the unworn tooth analyzed by the authors in paper [17].

The experimental procedure was conducted using the device shown in Figure 1. For this purpose, five used milling teeth were selected from a total of 50 teeth obtained from the machines of Strabenbau Logistic SRL Blejoi (Blejoi, Romania). These selected teeth shared similar characteristics, specifically a height reduction of 4 mm, as shown in Figure 2.

To determine the values of the angles at which tooth wear occurs (the angle formed by the wear facet with the axis of the tooth), the 50 teeth were measured on the STARRETT HF600 (Horizontal Floor Standing Optical Comparator) microscope (Figure 3). The values of the wear angles were around 42 degrees, which proves a uniform operation of the milling equipment. It was aimed that the five teeth had a configuration that was as similar as possible to ensure the same experimental conditions. With the wear angle being the same, for the teeth to have the same reduction in the height of the active part was also sought. This was important when studying the effect of wear compensation.

The teeth mounted on the drum of the milling machine have the possibility to rotate around their own axis. This favors the exposure of the teeth over the entire surface of the active part (the conical surface of the body together with the tip of the tooth). However, during milling, fine dust particles resulting from the degradation of asphaltic aggregates in the presence of chipping fluids cause teeth locking. The result is the impossibility of rotating the tooth. Consequently, the active part of the tooth will expose, in contact with the chipped material, only one generator of the conical surface. Over time, the wear will show unevenly and cause a facet to be produced, as seen in Figure 2.

The components of the experimental stand are detailed in

Table 1. The components of the experimental stand.

No.	Component	Functional Role
1	Asphalt part	The test material that simulates the asphalt layer subjected to stripping.
2	Milling tooth	The component that comes into direct contact with the asphalt, performing the stripping.
3	Tooth holder	The mechanism that holds the milling tooth in position (provided with a small pitch thread, on the left, to avoid jamming during operation).
4	Spring	Elastic component that allows shock absorption.
5	Force transducer	Sensor that measures the forces applied to the tooth during the milling process.
6	Machine table	The platform on which the entire system is mounted, ensuring the necessary stability and support.
7	Display for maximum recorded force values	Screen where the force values measured by the transducer are displayed.
8	Base plate	The structure on which all the components of the stand are mounted.
9	Linear bearings	Mechanisms that allow smooth and control the movement of the support table (platform).

.

Tests on worn teeth (Figure 4) were performed in two cases:

-

Without tooth wear compensation. In this case the milling device was positioned identically to the situation where the new tooth was used. The consequence was that the depth of cut decreased (naturally) with the amount of wear—4 mm;

-

With tooth wear compensation. In this case, the milling device was close to the asphalt concrete sample, so that the chipping depth was the same as in the case of using a new tooth.

Similarly to previous work of the authors of [17], the workpiece was composed of asphalt concrete blocks constructed from layers of road asphalt. The study focused on evaluating the milling process by examining the interaction between a single cutter tooth and the asphalt concrete block. This methodology facilitated the measurement of cutting reaction forces and enabled the adjustment of various cutting parameters, such as milling depth, rotational speed, feed rate, and advanced speed, across a wide range. The experimental asphalt concrete samples were prepared using the same composition as employed in road construction. These samples originated from a batch produced at the Strabenbau Logistic asphalt plant, subsequently cast into molds with dimensions of 350 × 240 × 100 mm and compacted. After casting, the samples underwent curing under ambient conditions for 240 days to replicate the natural consistency of asphalt commonly found in road surfaces. The asphalt concrete material used in the experiment had the following composition: 22% mineral aggregate, sized 4–8 mm; 43% mineral aggregate, sized 8–16 mm; 19% sand, 10% filler, and 6% bitumen.

2.2. DEM Simulation

Numerical modeling was performed with the discrete element method (DEM) using Rocky 2022 R2 software. In the present analysis, a hysteretic linear spring model was adopted for the normal contact forces, which accounts for energy dissipation and ensures a realistic representation of particle interactions. For tangential contact forces, a linear spring model Coulomb limit was employed to simulate the frictional behavior between particles. Additionally, adhesion forces were modeled using a constant approach to reflect the cohesive interactions present in the asphalt mixture [1]. In order to make the simulation more efficient—in the sense of the efficient use of hardware resources and obtaining reasonable runtimes (of the order of hours)—the aggregate particles were modeled using spherical particles with a diameter of 8 mm. Similarly to previous work [18], the interaction between aggregates (particles), filler/bitumen, and the milling drum was further quantified using coefficients such as static friction, dynamic friction, rolling, and restitution coefficients. For instance, a restitution coefficient of 0.3 was adopted to reflect the elasticity of particle collisions, while the static friction coefficient was set at 0.6 (for stone–stone) and 0.4 (for stone–tooth) interaction, respectively, dynamic friction coefficient was set at 0.6 (for stone–stone) and 0.1 (for stone–tooth) interaction.

Space Claim was used to create the geometric models used in DEM analysis. All the details regarding the choice of parameters for DEM simulation are presented in the previous work of the authors of [17].

The performed investigation aimed to use DEM modeling to simulate the interaction of worn cutter teeth with the asphalt concrete structure. Firstly, the authors intended to reproduce the active part of a milling cutter tooth that presents wear. For this, the three-dimensional scanning of a real tooth, representative for this application, was used, using the SHINING 3D scanner, model EinScan SP (Hangzhou, China) (Figure 5) located in the Department of Mechanical Engineering within the Faculty of Mechanical and Electrical Engineering. The obtained results regarding the shape of worn teeth are shown in Figure 6.

In a first step, DEM modeling of the milling process was performed using the 3D representation of the tooth, obtained by scanning. Since the available hardware resources proved to be insufficient—taking over the model and starting the simulation process taking more than 10 days—this approach was abandoned and the use of a new tooth model whose active part was modified was resorted to, respecting the values of the real geometry by sectioning with a plane inclined at 42 degrees to the axis of the tooth, as seen in Figure 7. The tooth thus obtained has a configuration very close to that of a real worn tooth.

The simulation of the tooth interaction with asphalt concrete was carried out on the same principles as in the case of a new tooth, as in the previous work of the authors of [17]. The principle of machining by milling using a worn tooth is shown in Figure 8. Although Figure 8b illustrates particles with a uniform size for computational efficiency, the particle size distribution in the model was calibrated as in [17,18] to correspond with the actual gradation of the asphalt mixture. This approach ensures that critical mechanical responses (contact forces), align with those observed in real materials.

2.3. DOE Analysis

In this study, Minitab 19 software and full-factorial design method were used to investigate the influence of milling parameters (milling depth and advanced speed) on horizontal cutting forces. For each parameter involved in statistical analysis, three levels were considered.

The working parameters used in the investigation are as follows: milling depth a_p was set at 15 mm, 30 mm, and 50 mm. The milling drum rotation speed was fixed at 75 rpm. The advanced speed v_f was tested at 190 mm/min, 235 mm/min, and 375 mm/min.

3. Results and Discussion

3.1. Experimental Results

After performing the tests, the recorded values of horizontal cutting forces (calculated as the average of three values obtained under identical chipping conditions) were centralized in

Table 2. The experimental values of horizontal force during milling.

Milling Depth, mm	Advanced Speed, mm/min	Horiontal Force Fz, N
		New Tooth	Worn Tooth, (Without Wear Compensation)	Worn Tooth, (With Wear Compensation)
15	190	605	269	701
15	235	518	532	863
15	375	616	828	863
30	190	791	675	1301
30	235	965	919	1452
30	375	1724	1532	1561
50	190	1452	954	2210
50	235	2706	1555	2892
50	375	2234	2337	2460

. During the experimental tests, the rotation speed was kept constant at 75 rpm. The values shown in Table 2 represent an average of three records. It was considered necessary to make several determinations for each individual case, considering the anisotropy of the chipped material.

The graphical representation of the experimentally obtained results can be found in Figure 9 and Figure 10.

Analyzing the results presented in Table 2 and Figure 9 and Figure 10, the following conclusions can be drawn:

-

The cutting forces increase with advanced speed and the milling depth;

-

As it resulted from the previously presented statistical analyses, the determining factor in the evolution horizontal cutting force value is the milling depth;

-

As the tooth wears, the value of cutting force drops below the value measured for the unworn (new) tooth. This is due to the reduction in the contact surface between the tooth and the chipped material (which also causes the chip thickness to decrease);

-

In the conditions of tooth wear compensation, by penetrating the milling cutter into the chipped material to an additional depth of value equal to the wear value, the value of the chipping force increases above the value measured for a new tooth. It can be concluded that the tooth geometry damage makes the chipping process difficult. This tendency is manifested in small values of the milling depth. When the value of milling depth increases (in the conducted study, over 40 mm), the chipping forces for the new tooth and for the used tooth have close values. This can be explained by the fact that the tooth, at depths exceeding the height value of the tungsten tip, will enter the chip with the whole body, and the geometry of the active part will influence the chipping process less significantly.

It is observed that, at high values of advanced speed (in the analyzed case, 375 mm/min), the cutting forces no longer present large differences (in the sense of increasing the value) between the new tooth, the worn tooth without wear compensation, and the worn tooth with wear compensation, respectively. Under these conditions, it can be said that the wear evolution, at high advanced speeds, is difficult to notice. It is recommended to periodically view the condition of the cutter teeth and, if possible, to compensate for wear to increase the milling process efficiency.

3.2. DEM Results

Within the DEM analysis, the cutting forces were evaluated along the Oz axis that aligns with the direction of advancement; one example is presented in Figure 11. In order to determine the cutting forces, the peak values (derived from the summation of forces recorded during the milling process simulation) were identified from the diagrams shown in Figure 11.

The simulation results for different values of the parameters of the milling cutting regime are shown in

Table 3. The values of the horizontal force during milling, obtained by DEM, for rotation speed of 75 rpm and different parameters of the cutting regime.

Millind Depth, mm	Advanced Speed, mm/min	Horizontal Force Fz, N
		New Tooth	Worn Tooth, (Without Wear Compensation)	Worn Tooth, (with Wear Compensation)
15	190	1980	1584	2158
15	235	1881	1802	2244
15	375	2178	2125	2178
30	190	2191	1914	2462
30	235	2673	2310	2970
30	375	2838	2726	2858
50	190	2772	2792	3102
50	235	3036	2356	3036
50	375	2970	2422	3076

and Figure 12 and Figure 13.

The graphs from Figure 12 reveal distinct trends when comparing the cutting forces associated with new and worn teeth under various milling conditions. For a new tooth, the cutting force increases steadily with the advanced speed across all examined milling depths. This scenario demonstrates the influence of wear compensation on cutting efficiency. For a worn tooth without wear compensation, the resulting chips have a reduced cross-sectional area, leading to lower milling forces. However, when wear is compensated, the chip cross-section becomes comparable to that produced by a sharp tooth. The primary difference lies in the increased frictional surface between the worn tooth and the chipped material, resulting in higher frictional forces and an overall increase in cutting force. The increase in cutting forces is more pronounced at low advanced speeds. In contrast, worn teeth with wear compensation exhibit consistently higher cutting forces due to greater friction and reduced cutting efficiency. For the worn tooth without wear compensation, the cutting force curves are notably reduced compared to those with compensation, illustrating the benefits of this adjustment.

At the deepest examined depth of 50 mm (Figure 12c), the cutting force reaches its maximum due to the larger material volume being processed. In this scenario, the differences between worn teeth with and without compensation are even more pronounced. Specifically, as the advanced speed increases, the observed reduction in cutting force differences (with and without wear compensation) becomes evident. Effectively, the conclusion noted based on the analysis of the experimental values presented in Table 2 is maintained.

Analyzing the plots from Figure 13, it can be seen that for a new tooth, the cutting force increases with the milling depth, regardless of the advanced speed. This trend confirms efficient cutting with minimal resistance, as the force consistently remains lower than that seen with worn teeth. In contrast, worn teeth with wear compensation experience higher cutting forces at all milling depths.

At a speed of 190 mm/min (Figure 13a), the cutting force increases steadily as milling depth rises. At a 235 mm/min advanced speed (Figure 13b), the pattern of increasing force with depth persists, but the rate of increase is steeper compared to the lower speed. This highlights the greater challenge posed by intermediate speeds. Additionally, the force gap between a worn tooth with compensation and a new tooth becomes more significant, indicating an amplified effect of tool wear. When the advanced speed reaches 375 mm/min (Figure 13c), the cutting force is higher across all conditions due to the faster speed and the wear impact becomes even more pronounced.

3.3. Statistical Analysis

The Pareto charts in Figure 14 highlight that milling depth is the main factor influencing cutting force, as demonstrated for the new tooth in the previous work of the authors of [17,18]. This finding aligns with research [19,20] which highlighted the essential influence of cutting depth on the cutting force during the milling process.

The chart in Figure 14a reveals that, based on the experimental results, in the case of a worn tooth without wear compensation, even though the primary factor is milling depth, the factor B (advanced speed) also has an important role because the horizontal bar passes over the red line (threshold for statistical significance 2.776), indicating that the factor associated with that bar is statistically significant. This can be correlated with the p-value of 0.015 determined with ANOVA.

Furthermore, the main effect plots presented in Figure 15 allow establishing the optimal combination of milling parameters for minimizing cutting force. Notably, the results indicate that a milling depth of 15 mm combined with an advanced speed of 190 mm/min is the most effective setup for achieving reduced cutting forces, as obtained in [17] for the new tooth.

4. Conclusions

This article presents the methodology and findings of research conducted to study, in laboratory conditions, the milling process of asphalt concrete. A specialized experimental stand, designed by the authors, was used to measure cutting forces in the horizontal direction (feed direction of the milling machine). As detailed in the first part of the paper, the stand allowed for controlled modification of various parameters affecting the milling process, namely, milling depth, rotational speed, and advanced speed.

The experiments employed real Wirtgen-type teeth made of steel with tungsten carbide tips, which exhibited wear on their active surfaces (Figure 2). This type of tooth is commonly used in asphalt milling applications due to its durability and effectiveness. The experimental research focused on a frequently encountered practical situation in which a single tooth operates along the cutting line. The experiments were complemented by a simulation program that modeled the interaction between the milling tooth and asphalt concrete. The simulation aimed to identify efficient, less costly, and more easily operable methods for studying the physical phenomena specific to the milling process.

The main conclusions from the experimental program and numerical modeling are as follows:

-

The experimental results show a relative consistency in the values of the horizontal cutting force with respect to advanced speed provided the rotational speed remains constant;

-

Certain anomalies in the force values are attributed to the structural anisotropy of the sample, as asphalt concrete is characterized by a random arrangement of aggregate particles that vary significantly in geometric shape, size, and mechanical properties, as also found in [1,17,18];

-

Numerical simulation results indicate that as the rotational speed increases, the horizontal cutting force generally tends to increase. However, at a given rotational speed, increasing the advanced speed results in the horizontal cutting force remaining approximately constant, albeit with some random variations;

-

Discrepancies between the experimental results and the numerical (DEM) simulations can be attributed to the simplifying assumptions used in numerical modeling, where spherical particles—homogeneous in terms of material properties—were employed to reduce computation time;

-

For a given milling depth, the evolution of cutting force with advanced speed does not follow a consistent trend. Higher advanced speeds generally result in increased cutting forces, making the milling process more demanding;

-

For large milling depths—over 30 mm, as shown in Figure 10 and Figure 11, and according to [18], and especially over 100 mm—a significant increase in cutting force values is observed;

-

At constant cutting speeds, an increase in milling depth from 15 mm to 50 mm generally results in higher horizontal cutting forces, indicating that greater milling depths demand more substantial cutting forces;

-

In most cases, the numerical modeling results are significantly higher than the experimental values. This difference may stem from various factors such as the simplifications applied to the numerical model or the structural anisotropy of the materials used in the experiments. These differences highlight the challenges in accurately modeling and simulating the milling process of asphalt layers and, more broadly, heterogeneous materials;

-

Both experimental and DEM modeling results demonstrate that the wear of the milling tooth notably influences the cutting force. Tooth wear manifests as a decrease in cutting force, primarily due to reduced milling depth and, consequently, a smaller chip volume being removed. When the milling process is monitored and milling depth is adjusted to compensate for the axial wear of the milling tooth, ensuring consistent penetration into the material (asphalt concrete), a notable increase in cutting forces is observed. This is mainly due to changes in the geometry of the tooth’s active part and the increased use of the tooth body, which lacks cutting capabilities;

-

It is important to note that not all process parameters can be replicated in a laboratory setting using devices like the one presented here. Cutting speeds achievable on such equipment are significantly lower, which can lead to reduced cutting forces. Furthermore, refining DEM models to better replicate the heterogeneity and anisotropy of real asphalt materials, exploring the application of optimal milling parameters identified in current road construction and maintenance scenarios to validate laboratory results, and developing new models and materials for milling teeth remain areas for further investigation;

-

The experimental program described in this section makes a notable scientific contribution by introducing an innovative approach to determining the forces involved in asphalt milling and validating DEM modeling as a cost-effective alternative to physical experimentation. This research advances the understanding of the interplay between milling parameters and cutting forces, offering valuable insights for optimizing milling processes.