Improvement of the Wear Resistance of Circular Saws Used in the First Transformation of Wood through the Utilization of Variable Engineered Micro-Geometry Performed on PVD-Coated WC-Co Tips

Abstract

Reduced performance of circular saws due to premature chipping of their teeth has been a critical issue in woodcutting industry for many years. This research examined the impact of surface coating and variable engineered micro-geometry of the cutting edges of carbide teeth (tips) on the wear resistance of circular saws used in primary wood processing. CrN/CrCN/DLC, CrN/AlTiN, CrN/CrCN, and CrCN/TiSiCN were deposited on tungsten carbide-cobalt (WC-Co) substrates using the cathodic arc evaporation technique. The CrN/CrCN coating proved to be the one with highest wear resistance and adhesion among those studied. No sign of delamination was observed around the indentation of the CrN/CrCN coating after the adhesion test. Furthermore, no abrasion, delamination or crack was observed on the surface of the CrN/CrCN coating after the three-body abrasion wear test. The results of the dry-sliding wear test revealed that CrN/CrCN coating significantly decreased the wear rate of WC-Co substrates by 74%, 66% and 77% at sliding speeds of 50, 100 and 250 mm/s, respectively. Afterwards, a CrC/CrCN coating was deposited on the teeth of conventional circular saws. Next, the cutting edges of teeth were modified through variable engineered micro-geometry. Tests were conducted at a sawmill with three series of saws: 1-coated and edge-modified, 2-coated and conventional edge geometry, and 3-uncoated and edge-modified. Wood processing was performed during two shifts of 480 min each. The width of the wear land was the criterion used as the wear index. The results of industrial tests showed that saws with edge-modified teeth had significantly less chipping and no breakage at their corners compared to the saw without edge modification (conventional saw). After 480 min of sawing, the wear rate of the coated saw with edge modification decreased by 46% and 16%, compared to the coated saw without edge modification and the uncoated saw with edge-modified teeth, respectively. Those values reached 73 % and 41%, respectively, after 960 min of sawing. The study shows that by optimizing the surface chemistry and the geometry of the cutting edge of WC-Co tips, tool life can be significantly increased therefore reducing downtime due to saw replacement and resharpening, thus significantly increasing productivity in the first transformation of wood.

1. Introduction

Wear of cutting tools is critical in the wood processing industry as the tools are subjected to high mechanical and thermal loads. Worn tools exert greater forces on the workpiece, thus degrading the surface finish of the final products [1]. As more stress is put on the wood sawing industry to maximize utilization of the wood fiber, it becomes critical to develop cutting tools that are significantly more resistant to wear in order to minimize sawing variation and generation of waste. Circular saws are commonly used in the first and second transformations of wood. Their performances are primarily determined by the wear resistance of their cutting edges made of WC-Co inserts brazed onto the steel body of the saw. Many factors influence the wear of saws including tool geometry, cutting parameters (cutting speed, feed speed, etc.), and workpiece conditions (wood species, moisture content, temperature, wood density, knottiness, etc.) [2]. Circular saws require periodic resharpening and/or replacement of carbide tips throughout their useful life. The cost of these operations can be minimized by improving the wear resistance of the carbide tips [3]. In addition, as tool wear increases, sawing variations increase as well, requiring that a larger volume of wood be removed in subsequent sawing operations such as planing. The microgeometry of the cutting edge significantly impacts temperature distribution, the surface quality of the workpiece as well as tool life [4]. According to Endres et al. [5], sharp-edged tools mostly present a high level of wear, especially near their sharp corners. As a result, optimization of the cutting edge geometry is essential to mitigate the chipping and fracture issues. It refers to the design of a controlled geometry of the cutting edge to reduce the impact of initial edge defects and micro-chipping that are inherent in all cutting tools [6].

Engineered modification of cutting edges has become more frequent in the metal cutting industry but received close to no attention in the wood processing sector. Proper edge preparation has the potential to improve the performance of cutting tools by delaying the onset of chipping and abrasive wear [7]. The microgeometry of the cutting edge is defined by the form factor K (K = Sγ/Sα), as presented in Figure 1 [8]. The K = 1 defines the radius-honed profile, while K > 1 and K < 1 describe the waterfall and reverse waterfall hone profiles, respectively [8].

Surface treatment is another solution to improve tool life. Hard and adherent coatings obtained by physical vapor deposition (PVD) have been recently used in wood machining. However, up to now, they have seen limited application for circular saws in primary wood processing. Coatings for wood cutting tools are typically characterized as having a high level of hardness and most importantly, a low coefficient of friction that lead to reduced wear rate [9]. In addition, due to the visco-elastic properties of wood and its high coefficient of friction, good adhesion to the substrate is another crucial feature of an adequate coating [10].

TiN and CrN are the most popular coatings for wood cutting tools. However, Cr-based coatings have attracted more attention in recent decades [11,12]. By adding a third element such as Vanadium (V), Aluminum (Al), Molybdenum (Mo), Silicon (Si), etc. in the system, nitride coatings show a good potential for increasing hardness and wear resistance of carbide-tipped circular saws. For instance, additions of Si increase the coating’s mechanical properties and abrasive wear resistance. Incorporating Al into a nitride-based coating results in the formation of a structure with high oxidation resistance and thermal stability [13].

Moreover, the development of multilayer coatings is advantageous to enhance the comprehensive properties of the overall coating by forming a layered microstructure that hinders crack propagation [14]. Pinheiro et al. [15] conducted an industrial cutting test with a CrN/CrWN multilayer coating deposited on a WC-Co cutting tool. The test resulted in an important increase of the total length of machined wood compared to the case where the coating was applied as a single-layer. The latter result was attributed to the interfacial bonding between layers restraining microcrack propagation in the coating, thus preventing premature failure of the cutting edge. Adding carbon to chromium nitride (CrN) coatings was also shown to improve mechanical properties, fatigue strength and tribological properties [16]. As reported by Warcholinski and Gilewicz [10], CrCN/CrN coatings had a wear rate ten times lower than that observed for a TiAlN/TiN coating. The reason was explained by the higher value of ion potential of CrO₂ (7.3) compared to TiO₂ (5.4). Indeed, when the ionic potential is higher, the cation in that oxide is more masked by the surrounding anions and the oxide easily shears accordingly [17].

Carbon-based coatings like diamond-like carbon (DLC) are characterized with having a lower coefficient of friction, high hardness, and high wear resistance. It is made of amorphous carbon, which is a disordered network of carbon atoms connected by a mixture of sp³ and sp² coordinated bonds [18]. The formation of an easy-shear-tribofilm on the surface of the counter body contributes to a lower friction coefficient [19]. Faga and Settineri [20] showed that the tools coated with DLC and multilayer CrN coatings performed much better compared to that of the uncoated tools when machining spruce wood. Review of the pertinent literature clearly shows that there is a significant advantage brought about by the deposition of ceramic coatings on the cutting edges of carbide-tipped circular saws.

The objective of this study is to quantify the efficiency of variable engineered micro-geometry modification of the cutting edges of circular saws used in the primary transformation of wood in combination with PVD-coatings of their WC-Co tips. Following a thorough literature review on the research work most pertinent to this study, it was decided to use the multilayer CrN-based coating. Moreover, in order to improve the tribo-mechanical properties of the coating such as hardness and wear resistance, the addition of another elements (Al, C and Si) was considered. It was also decided to use DLC as a top layer as a result of its excellent potential to reduce the coefficient of friction.

2. Materials and Methods

2.1. Coating Deposition Technique

Firstly, CrN/CrCN/DLC, CrN/AlTiN, CrN/CrCN, and CrCN/TiSiCN multilayer coatings were deposited by the cathodic arc evaporation method on the cemented carbide substrates (grade TMK-45, 88% WC and 12% Co, TechMet, Hickory, NC, USA) as well as on P-type Si (100) wafers having a thickness of 520 µm. This approach was selected to characterize mechanical and tribological properties of the coatings as well as surface morphology, and achievable coating thickness. The PVD-coatings were performed at Aurora Scientific Corp., Richmond, BC, Canada. Prior to deposition, the cemented carbide substrates were ground with a diamond grinding wheel having a grit size of 100, followed by ultrasonic cleaning. Finally, they were polished using 6 μm and 1 μm diamond suspensions. The obtained surface roughness (Ra) was approximately 0.03 μm. The cleaning process of the WC-Co tips consisted of sequence of operations that included degreasing with a high-alkaline solvent-free degreaser. This was followed by wet blasting using a solution of water and 3.9 micrometer quartz particles (1:5 ratio). The last step involved degreasing in an ultrasonic bath, air blowing, drying, and heating in a furnace at 120 °C for 30 min. Subsequently, all substrates were heated up to about 300 °C. They were then exposed to an Ar⁺ bombardment at a bias voltage of 800 V for 20 min and at a gas pressure of 0.66 Pa to remove the surface oxide layer and impurities.

Afterwards, a chromium sub-layer with a thickness of 0.1 μm was deposited to improve the adhesive strength between the substrates and coatings. The target’s current was 70 A. The deposition time was 5 min. For CrN/CrCN and CrCN/TiSiCN, C₂H₂ was added with a flow rate of 10 standard cubic centimeter per minute (sccm). For CrN/CrCN/DLC coating, the flow rate of C₂H₂ was 160 sccm. The pressure of nitrogen was 20 mtorr for CrN/CrCN, CrN/CrCN/DLC, and CrCN/TiSiCN coatings. For CrN/AlTiN, the nitrogen pressures were 21 and 25 mtorr for CrN and AlTiN layers, respectively. The distance between the target and substrate was 150 mm in all coating recipes. The list of coatings, with the temperature deposition, bilayer periods and target materials are shown in

Table 1. Target material, temperature deposition and bilayer periods of the coatings deposited at Aurora Co.

Coating	Target	Temperature Deposition	Bilayer Period
CrN/AlTiN	Cr and TiSi80/20	250 °C	1:1
CrN/CrCN	Cr	250 °C	1:1
CrN/CrCN/DLC	Cr	250 °C	1:1
CrCN/TiSiCN	Cr and AlTi67/33	250 °C	1:1

The purity of all targets was 99.999 wt.%. The diameter of all targets was 150 mm. The DLC layer is a-C: H, dopped with Cr element.

.

2.2. Coating Characterization Methods

Surface morphology and coating thickness were characterized in Scanning Electron Microscopy (SEM) FEI Inspect F50. The chemical compositions were determined using an energy-dispersive X-ray spectrometer (EDS) Ametek, Octane Super-A. The coating adhesion on the carbide substrates was evaluated following the Daimler–Benz test method with a conical Rockwell C intender (120° diamond cone, 200 µm radius) and a load of 1471 N [21]. The damages caused by the indentation on the layer adjacent to edges of the impressions were evaluated in scanning electron microscopy. The assessment of the coating adhesion was based on six-mode scales. HF1 to HF4 indicate sufficient adhesion, whereas HF5 and HF6 refer to poor adhesion to the substrate. Microhardness was measured using a Vickers microhardness indenter with 40gf (MMT-X7A, MATSUZAWA CO., Yokosuka, Japan) according to ASTM E384. The tribological behavior of the coated carbide specimens was evaluated through the dry sliding wear test and dry sand-rubber wheel abrasion testing (DSRW). Five coated carbide samples were tested using both wear tests.

The dry sliding wear test was carried out using a pin-on-disk instrument (Tribometer TRB3, Anton-Paar, Graz, Austria) according to ASTM standard G99. A ball made of Al₂O₃ and having a diameter of 6 mm was used as a counter body. The tests were conducted using a constant normal load of 10 N and a sliding distance of 10⁶ mm. The selected sliding speeds were 50 mm/s, 100 mm/s and 250 mm/s. Tests were conducted at 24 °C and 65% of relative humidity. A Surtronic S-100 Series surface profiler was used to measure the depth and width of the wear tracks at ten different locations in order to calculate wear rates. The morphology and composition of the worn surfaces, i.e., coated samples and alumina balls were characterized in Scanning Electron Microscopy equipped with an Energy-Dispersive X-ray Spectrometer. Three-body abrasion tests were performed according to ASTM standard G65 (procedure C), which is defined specifically for the use on thin coatings. The abrasive body was silica sand AFS 50/70. The sand flow rate, the applied force, and the wheel revolution were 330 g/min, 130 N, and 100, respectively. The latter parameters resulted in a lineal abrasion of 71,800 mm. The wear rate of each tested specimen was calculated as shown in Equation (1) [22].

$$\mathrm{Volume}\mathrm{loss},{\mathrm{mm}}^3=\frac{\mathrm{mass}\mathrm{loss}\left(\mathrm{g}\right)}{\mathrm{density}\left(\mathrm{g}/{\mathrm{cm}}^3\right)}\times 1000$$

(1)

The adjusted volume loss (AVL) value was calculated according to the Equation (2) as the rubber wheel decreased in diameter [22].

$$\mathrm{AVL}=\mathrm{measured}\mathrm{volume}\mathrm{loss}\times \frac{228.6\mathrm{mm}\left(9\mathrm{in}.\right)}{\mathrm{wheel}\mathrm{diameter}\mathrm{after}\mathrm{use}}$$

(2)

2.3. Coating Deposition on Circular Saws

Based on the results of the characterization of the coatings (tribo-mechanical), the most suitable coating in terms of adhesion strength and wear resistance was selected. Four new conventional saws were sent to Aurora Scientific Corp. for coating deposition. The saw specifications are reported in

Table 2. Circular saw specifications.

Parameters	Value
Saw body material	AISI 8670
Tip material	88% WC and 12% Co
Number of tips	42
Saw diameter (mm)	610
Saw thickness (mm)	2.79
kerf width (mm)	3.91
Wedge angle	56.3°
Rake angle	24.7°
Top clearance angle	9°
Radial clearance angle (left and right)	1°
Left side clearance (mm)	0.56
Right side clearance (mm)	0.58
Supplier of the saws: Haskin Industrial Co., North Bay, ON, Canada.

.

The disposition of the saws in the chamber of the PVD coating system and the deposition area are presented in Figure 2. The distance between the saws was 75 mm. Two masking plates were mounted on both sides of each saw to cover the steel bodies. The target area for coating deposition was a 25.4 mm ring extending from the tip of the carbide inserts toward the center of the saw. Therefore, the inserts and the outside rim of the saw along with the gullet area were coated.

2.4. Variable Engineered Micro-Geometry of the Cutting Edges of Saws

For the conventional circular saw, following the brazing operations, the carbide tips of each saw were ground (Vollmer) at the manufacturer. The polycrystalline diamond grit sizes of the grinding wheels were 280 and 180 for the rake face and sides of the tips, respectively. The clearance face was ground with a polycrystalline diamond wheel with a dual grit combination 150/320. Therefore, after grinding all faces, the cutting edge of the tips is defined as having an up-sharp profile. It should be noted that for the conventional saws with up-sharp tips, coating was deposited on the tips after the grinding of their cutting edges.

Two coated and two uncoated conventional saws had the geometry of their carbide tips modified. The purpose was to change the profile of the cutting edges of all their carbide tips. The new profile was achieved by honing the cutting edge of each tip using brushes made of abrasive bristles of nylon loaded with micrometric natural diamond particles. The geometry of the modified cutting edges followed a waterfall hone profile (K = 2) with Sγ = 20 μm and Sα = 10 μm, as exemplified in Figure 3. The microgeometry of this profile was chosen based on previous work carried out internally at Conicity Technology, Turtle Creek, PA, USA [23].

It is well established that the thickness of the uncut chip changes from a maximum, which is equal to the feed per tooth, to a negligible amount on the tool’s corner radius according to Figure 4 [5]. In other words, the ratio of uncut chip thickness to the edge radius varies along the corner radius. In the cutting process, the ultimate goal is to keep the material removal process as efficient as possible [5]. As a result, the size of edge preparation must change along the corner radius.

To maintain a specific ratio of edge preparation size to the thickness of the uncut chip, edge preparation was distributed along the corner radius of the tip. Therefore, the edge preparation on the section of the cutting edge that transitions around the corner radius varied in size. The modification varied between 20 μm on the primary cutting edge to 4 μm on its flank side edges. It is well known that uneven kerf width or side clearance of the circular saw tips results in saw vibration [24]. Therefore, all the tips of circular saws were characterized using a video tooth inspector apparatus (developed by FPInnovations, Quebec City, QC, Canada) before performing the initial test. The average kerf width and average side clearances of the tips were measured to assure that these values were in accordance with the specifications reported by the saw manufacturer. The average kerf and average side clearances were 3.9100 mm (±0.0002) and 0.5588 mm (±0.0004), respectively for all saws.

2.5. Industrial Tests

To perform wood cutting tests in an industrial environment, six new circular saws were sent to a partnering sawmill located in the Province of Quebec, Canada. The tests were done for sawing cants of SPF (spruce/jack pine/balsam fir). Identification of the saw is reported in

Table 3. Saw identification for industrial tests.

Saw Identification	Coating	Edge Modification
#1, #2	Yes	Yes
#3, #4	Yes	No
#5, #6	No	Yes

.

The industrial wood cutting tests lasted for three periods of 480 min. The first phase involved utilizing saws 1, 3, and 5 under industrial conditions for 480 min of sawing. The saws were mounted on the vertical arbor of the bull-edger system (USNR VSS), as shown in Figure 5. The rotational speed of the saws and the resulting cutting speed were 2600 rpm and 83 × 10³ mm/s, respectively. The feed speed was 3050 mm/s for studs with a depth of 101.6 and 76.6 mm. For studs with a depth of 152.4 mm, the average feed speed was 2417 mm/s. Therefore, the feed per tooth was 1.67 and 1.33 mm, respectively.

Upon completing the test, tips from each saw were randomly selected and extracted to characterize their wear in scanning electron microscopy. The primary cutting edge, rake and clearance faces of each tip were analyzed. The second phase of the test included using saws 2, 4, and 6 for a sawing period of 480 min. Upon completion, the wear of the tips was characterized using a Reichert Jung 580 optical microscope. The images acquired were analyzed using the Image J 1.53e image analysis software. The analysis included determining the wear of the tips according to the parameters shown in Figure 6 [25]. The wear indexes were the change in recession on the rake and clearance faces and width of the wear land.

Once the wear characterization was completed, saws 2, 4, and 6 were again installed in the bull-edger system for the third phase of the study. They were used during another work shift of 480 min. The wear of the tips was then analyzed in optical and scanning electron microscope.

3. Results and Discussion

3.1. General Description of Deposited Coatings

Figure 7 presents the typical surface morphology of the coatings studied. Spherical and shapeless microdroplets and macrodroplets can be observed on all the coatings, which are common features of cathodic arc vapor deposition [26,27]. Moreover, the voids can be seen on the coated surfaces correspond to sites where droplets were pulled out. The presence of such features and their origin is also reported in [28,29]. The number of droplets per unit area (n/mm²) was analyzed by the software Image J 1.53e [30]. For CrN/CrCN/DLC, CrN/AlTiN, CrN/CrCN and CrCN/TiSiCN coating, the number of droplets per unit area were 53,105, 115,105, 68,421, and 113,895, respectively. Therefore, the CrN/CrCN/DLC and CrN/CrCN coatings presented smoother surfaces compared to the CrCN/TiSiCN and CrN/AlTiN coatings. It was reported that the arithmetic roughness of the surface plays an important role on tribological properties [14,31].

The qualitative elemental analysis of as-deposited coatings and their characteristics are presented in

Table 4. Elemental composition and characteristics of coatings.

Coating	Atomic %						Thickness (µm)	Hardness (GPa)
	Cr	N	C	Al	Ti	Si
CrN/AlTiN	7	69	-	17	7	-	1.9	34
CrN/CrCN	19	57	24	-	-	-	1.5	24
CrN/CrCN/DLC	11	32	57	-	-	-	2.2	22
CrCN/TiSiCN	5	59	25	-	9	2	1.8	31

. The result of EDS characterization is an indication of the chemical compositions.

Typical SEM micrographs showing the cross-section views of the different coatings are presented in Figure 8. All multilayer coatings show a dense and homogeneous columnar structure without visible cracks or pores. The top layer of all coatings was thicker than the other sub-layers. For CrN/CrCN/DLC, CrN/AlTiN, CrN/CrCN, and CrCN/TiSiCN coatings, the top layer is DLC, AlTiN, CrCN, and TiSiCN, respectively.

3.2. Adhesion Strength

Figure 9 presents SEM micrographs of the coated carbide specimens. Following the Daimler-Benz adhesion test, the CrN/CrCN/DLC coating falls in the class HF4 group testifying acceptable adhesion to the substrate. Delamination is seen in Figure 9a. It could be the result of intrinsic stress owing to the cumulative impact of failures that appeared internally in the film during the deposition process [32]. Furthermore, few radial micro cracks were present around the imprint.

CrN/AlTiN and CrN/CrCN coatings are classified in the HF1 adhesion level denoting high adhesion strength. A few radial micro cracks were observed in the vicinity of the imprints of both coatings as a typical failure, demonstrating the relaxation of stress at boundaries between layers (Figure 9b,c). The cracks most likely stemmed from tensile stresses associated with the material piling up during the test, which relates to hard and brittle coatings [33].

CrCN/TiSiCN coating is in the HF3 category indicating a good adhesion strength. Radial micro cracks around the indentation and delamination are observed in Figure 9d.

SEM micrographs at higher magnification of the specimens used to carry out the Daimler–Benz adhesion tests clearly show the presence of radial micro cracks around the imprint of the tested coatings (Figure 10).

3.3. Tribological Properties

Three-body abrasion test: The results of the measurements of the volume loss during the DSRW abrasion test is presented in Figure 11. The best three-body abrasion resistance was obtained with the CrN/AlTiN coating followed by CrN/CrCN. The latter coatings are characterized with having better adhesive and cohesive properties as highlighted in Figure 9. CrN/CrCN/DLC gave the lowest level of protection against abrasive wear. Although CrN/CrCN had a high potential for reducing wear rate, applying DLC as a top layer decreased its resistance to abrasive wear. CrCN/TiSiCN coating showed medium resistance to abrasion when compared to the other three coatings studied.

The typical wear patterns of each coating are shown in Figure 12. It is seen that the CrN/AlTiN and CrN/CrCN coatings are firmly attached to the substrate (Figure 12b,c). No coating failure was observed for these two coatings. On the other hand, the CrN/CrCN/DLC coating showed severe wear. Fatigue wear occurred, as seen in Figure 12a owing to the presence of cracks, which are perpendicular to the sliding direction. Furthermore, the coating showed obvious signs of delamination, which were most likely caused by the lower adhesion strength of CrN/CrCN/DLC to the substrate compared to other coatings. Therefore, abrasive sand particles could promote coating detachment. Dalibón et al. [34] showed that DLC had lower abrasive wear resistance compared to TiN on steel substrate in DSRW tests. The behavior was attributed to the lower strength at the interface between the substrate–DLC film against shear stresses.

Delamination within the coating layers and abrasion occurred on the surface coated with CrCN/TiSiCN due to the abrasive sand particles, as seen in Figure 12d.

Dry sliding wear test: Figure 13 shows the wear rate of the four coatings studied along with the results obtained with uncoated substrates for different sliding speeds, i.e., 50, 100, and 250 mm/s and a constant normal load of 10 N. For CrN/CrCN and CrN/CrCN/DLC coatings, the wear rate as a function of sliding speed followed a similar behavior. As the sliding speed increased from 50 to 100 mm/s, the wear rate increased to its maximum value and remained constant with further increasing sliding speed. The wear rate of the uncoated sample, CrCN/TiSiCN, and CrN/TiAlN coatings increased by increasing the sliding speed. CrN/CrCN coating showed a lower wear rate compared to others.

It is well established that the addition of carbon to the CrN system reduces the residual stress and friction coefficient. It also increases wear resistance as well as the adhesion of the coating to the substrate [35]. Moreover, CrN/TiAlN coating increased the wear resistance of the WC-Co substrate. As stated by Gao et al. [36], higher wear resistance of hard films is associated with higher hardness. According to the results presented in Figure 13, the CrN/CrCN coating significantly decreased the wear rate of WC-Co substrates by 74%, 66% and 77% at sliding speeds of 50, 100 and 250 mm/s, respectively.

Since the coefficient of friction (COF) of any given surface plays an important role on its wear resistance, the latter was characterized. Figure 14 presents the variation of the coefficient of friction of the different coatings studied as a function of sliding distance at sliding speeds of 50, 100, and 250 mm/s. All samples experienced an abrupt increase of its COF at the beginning of sliding. The initial formation of wear debris and its further involvement in the sliding contact is responsible for such an increase in friction coefficient [37]. The values of the COF of the CrN/CrCN and CrN/CrCN/DLC coatings are very similar, and they provided the lowest COF values among the coatings studied. After the initial sharp increase, the COF of the CrN/CrCN coating reached a stable state at all sliding speeds. The COF of CrN/CrCN/DLC increased gradually, decreased afterward, and finally stabilized for all sliding speeds. The COF of CrN/CrCN/DLC coating decreased as the sliding speed increased. This observation is most likely due to a higher contact temperature between the pin material and the coating, which is said to accelerate the trend of graphitization of the DLC layer during the wear test [19].

The friction coefficient of the WC-Co sample was quite similar to those of the CrN/CrCN and CrN/CrCN/DLC coatings at the sliding speed of 50 mm/s. However, a sudden increase in COF occurred for the uncoated specimen at a higher sliding speed. This behavior is related to the worn surface of the uncoated sample.

CrN/AlTiN has the highest COF followed by the CrCN/TiSiCN coating. The latter had a gradual increase of its COF followed by an abrupt increase to finally reach a stable regime at higher sliding speeds. The lower COF of the CrCN/TiSiCN coating at the beginning of the test is caused by the accumulation of carbon particles alongside the wear track. These particles serve as a solid lubricant, lowering the friction coefficient [38]. Moreover, the following increase of the COF of the CrCN/TiSiCN coating is due to materials being removed from the worn surface. The higher friction coefficient of the CrN/AlTiN coating is probably due to the presence of a large number of droplets that formed at the surface of the specimen during the evaporation of the AlTi cathode [14].

Figure 15 shows SEM micrographs of typical worn surfaces of coatings and uncoated WC-Co substrate for the tests performed at sliding speed of 50 mm/s. Tribofilms are observed within the wear tracks of the CrN/CrCN/DLC, CrN/AlTiN, and CrCN/TiSiCN coatings. EDS analyses of zones A-C (in Figure 15a,b,d) show the presence of oxygen (Figure 16a–c). This indicates the occurrence of tribo-oxidation, thus the formation of an oxide layer during the test. The worn surfaces were covered with a tribofilm having fish scale-like pattern. The fish-scale morphology shows that the tribofilm underwent shear deformation during the test [39]. The tribofilms for the CrN/AlTiN coating were found to be present on almost the entire wear track, except for its edges. They were highly dense and bonded well, thus protecting the surface and leading to a lower wear rate.

Traces of Al in the tribofilms characterized by EDS analysis confirms the transfer of material from the ball to the CrN/CrCN/DLC and CrCN/TiSiCN coatings (Figure 16a,c), which is the sign of adhesive wear. Parallel grooves indicating abrasive wear were observed on the wear surfaces of all coatings. These grooves were caused by the wear debris detached from the coating, leading to three-body abrasion. According to Wang et al. [40], the primary wear mechanisms of cemented carbide coated with CrN/AlTiN during the friction test were abrasive and adhesive wear. The presence of W identified in the EDS spectra of Figure 16a, c, reveals that the CrN/CrCN/DLC and CrCN/TiSiCN coatings were detached from the substrates in some areas, which was the main reason for the higher wear rates of these coatings as reported in Figure 13. As claimed by Wang et al. [41], the dominant wear mechanisms of a TiSiCN coating under a slow sliding speed are abrasive and adhesive wear. However, according to Zhou et al. [42], abrasive wear was the only main failure mode for this coating. The reason for the latter observation was attributed to the accumulation of the highest proportion of amorphous Si₃N₄ hard phase in the microstructure [42].

The CrN/CrCN coating was worn in a slight abrasive manner in the centre of the wear track, whereas at the edges, the coating experienced mild abrasion. The wear debris or weakly bounded droplets were drawn from the surface of the sliding track to its edges by the counterpart body, leading to mild abrasion in those areas. Furthermore, the presence of aluminum from the counter body within the wear track of CrN/CrCN coating confirmed the activation of adhesive wear. Gilewicz and Warcholinski [33] reported the same wear mechanisms.

The tribofilm was observed on the worn surface of the uncoated WC-Co, as shown in Figure 15e. The detection of Al via EDS analysis (Figure 16d) indicates the transfer of material from the alumina counter body to the surface of the WC-Co substrate during the test. Additionally, cobalt binder extrusion took place on the worn surface (Figure 15f). This condition leads to the relaxation of the internal compressive stresses in WC grains, thus initiating crack formation [43]. Transgranular cracks occurred within most WC grains, as seen in Figure 15f.

Figure 17 shows SEM micrographs of typical worn surfaces of different coatings and uncoated WC-Co substrates at the sliding speed of 100 mm/s. The dominant wear mechanisms for CrN/CrCN/DLC coating were delamination and abrasive wear. As seen in Figure 17a, the coating is rubbed off from the surface, mostly in the middle section of the wear track. The detached coating particles may also serve as abrasive particles increasing the wear rate through the development of three-body abrasion mechanism. Mild adhesive wear was also observed. Oxide layers formed during the test were present within the wear track (Figure 17a).

The analysis of the wear track of CrN/AlTiN coating (Figure 17b) indicates abrasive wear as a principal wear mechanism. The accumulation of wear debris at the edge of the wear track was responsible for constituting the grooves at those locations. Those particles which were detached from the coating via a brittle failure mechanism, played the role of a third body in the tribological system. A similar finding was reported by Hsieh et al. [44]. The CrN/CrCN coating presents a smooth worn surface with shallow grooves indicating slight abrasive wear (Figure 17c). The formation of the smooth surface is related to the easy removal of debris particles from the contact surface [45]. Gilewicz and Warcholinski [33] found the same wear mechanism because of the good tribo-mechanical properties of CrN/CrCN coatings. As shown in Figure 17d, the wear track of the CrCN/TiSiCN coating is covered with oxide films and adhesive layers. The elements detected by EDS analysis of these layers (zone 1 in Figure 17d) are shown in

Table 5. Qualitative EDS analysis of zone 1 covering with oxide films and adhesive layers in Figure 17d.

Element	% Mass	% Atomic
C	7.83	13.34
N	12.79	18.70
O	41.11	52.60
Al	1.07	0.81
Si	0.95	0.69
W	2.93	0.33
Ti	12.50	5.34
Cr	20.81	8.19

. The presence of W reveals that the coating was removed from the surface of the substrate.

In addition, slight abrasive wear at the edges of the wear track of CrCN/TiSiCN coating was observed. Abedi et al. [38] confirmed the evidence of ploughing due to abrasion damage of loose wear debris on TiSiCN coating at a medium sliding speed of 150 mm/s. Additionally, the presence of microcracks perpendicular to the sliding direction attests that fatigue wear is taking place on the worn surface of the CrCN/TiSiCN coating. The dominant wear mechanisms for the uncoated substrate were transgranular cracks in the WC grains and their brittle fracture, which was attributed to the cyclic loading during the sliding wear test (Figure 17e).

Figure 18 shows the SEM micrographs of typical worn surfaces of different coatings and uncoated WC-Co substrates at the sliding speed of 250 mm/s. As presented in Figure 18a, CrN/CrCN/DLC coating is delaminated from the substrate within the wear track. The surface is worn due to the severe consequences of the ploughing mechanism. The worn surface is covered with wear debris. Entrapment of debris particles in the wear track led to high abrasion stress, causing the wear mechanism to shift from two-body to three-body abrasion and ultimately producing deep grooves. Severe delamination, crack and abrasive wear accounted for the high wear rate of this sample, as shown in Figure 13. It is seen that the CrN/CrCN coating was worn due to microcutting (Figure 18c). Wear tracks of the CrN/CrCN and CrN/AlTiN coatings were relatively shallow, and smoother compared to other samples. The wear mechanism of these coatings did not change by increasing the sliding speed. Fatigue is responsible for the formation of cracks in the CrCN/TiSiCN coating, which resulted in the coating chipping away from the surface (Figure 18d). Stresses in a coating may cause crack propagation and plastic deformation via local coating fracture, thus increasing the wear rate [46]. Slight abrasive wear was observed due to the presence of wear debris within the track (Figure 18d). Moreover, tribofilms formed on the surfaces coated with CrN/AlTiN and CrCN/TiSiCN. For the uncoated sample, cracks occurred in most WC grains, which were predominantly transgranular, followed by grain crushing. Increasing sliding speed resulted in increased cobalt binder extrusion and widespread grain crushing, which led to a higher wear rate of this sample compared to what was observed at lower sliding speeds (Figure 13). The fragmented grains served as three-body abrasive hard particles promoting the abrasive wear mechanism (Figure 18e).

Figure 19 shows the worn surface of the alumina counterparts at the sliding speed of 50 mm/s for the series of specimens studied. Grooves and tribofilms can be seen on the balls sliding against the coated substrates. The grooves are due to the three-body abrasion taking place during the friction test. The balls that were in contact with the CrN/CrCN/DLC and CrN/CrCN coated samples showed smoother surface. The characterization of these balls showed the presence of chromium as well as a large amount of carbon from the coatings. The latter was responsible for lowering COF as it constitutes an easy-to-shear tribofilm on the surface of the balls. It was reported that the transfer layer facilitates sliding by lessening the contact between the counterparts [47]. Elements such as Ti, Cr, and C were detected on the balls that were used against the CrCN/TiSiCN and CrN/AlTiN coatings. Their presence confirmed that adhesive wear was active during the test. Furthermore, the EDS spectrum of the counter body used to characterize the wear resistance of the CrCN/TiSiCN coating shows the presence of significant weight fraction of W (16.62 wt%) along with Co (Figure 19f), which proves that the coating was removed during the sliding test. Examination of the ceramic counter body used to test the uncoated sample showed the presence of W (12.86 wt%) and Co, which is evidence of adhesive wear (Figure 19g).

Figure 20 shows the wear scar area of Al₂O₃ counterparts for different coatings and the uncoated specimen in terms of sliding speed. For the coated samples, the wear scar area became larger as the sliding speed increased from 50 to 100 mm/s and then got smaller in size at the speed of 250 mm/s. However, the wear scar area on the balls against the uncoated substrates increased in size with increasing sliding speed and it was found to be proportional to the wear rate.

The results of the characterization performed on the different coatings in terms of wear resistance and COF are summarized in

Table 6. Assessment of coating characterization following wear tests.

Test		CrN/CrCN/DLC	CrN/AlTiN	CrN/CrCN	CrCN/TiSiCN
Sliding wear test	Abrasive wear	×	×	×	×
	Adhesive wear	×	×	×	×
	Cracking	×			×
	Delamination	×			×
DSRW test	Abrasive wear	×			×
	Cracking	×
	Delamination	×			×
Friction coefficient at 250 mm/s		0.30	0.65	0.31	0.56
Wear rate at 250 mm/s (mm3 N−1·m−1)		3.01 × 10−6	4.67 × 10−7	3.18 × 10−7	3.63 × 10−6
Volume loss after DSRW test (mm3)		11.9 × 10−1	2.14 × 10−1	3.61 × 10−1	8.13 × 10−1

. As a result of its higher wear resistance and lower COF compared to the other coatings studied, CrN/CrCN showed the best results, and it was selected for deposition on the tips of commercial circular saws to perform the industrial test at the sawmill.

3.4. Results of Industrial Tests

Wear Characterization of Tips

The recession of the rake and clearance faces and the width of the wear land of all saws that were measured before the industrial tests (0 min) are presented in

Table 7. Initial measurements at 0 min on the clearance face, rake face, and the width of the wear land of the saws.

Saw Identification	Modified Saw	Up-Sharp Saw
Recession on clearance face (µm)	10	3
Recession on rake face (µm)	20	3
Width of the wear land (µm)	17	2

. Note that for the modified saws, the initial values at time 0 min presented in Table 7 is due to the modification of their cutting edges according to Figure 3b.

The comparison of wear between the three types of circular saws listed in Table 3 as a function of sawing time is shown in Figure 21. The recession on the rake and clearance faces as well as the width of the wear land are presented as wear indexes for the wear comparison between all saws. As observed in Figure 21a–c, the saw with cutting edge modification and coating (#2) performed much better than the other saws (#4 and 6). The edge-modified saw without coating (#6) showed the second-best behavior in terms of wear resistance in an industrial setting while the up-sharp and coated saw (#4) showed the worst wear resistance.

The wear rates of the three types of saws are presented in

Table 8. Rate of wear of saws 2, 4 and 6 as a function of sawing time at the sawmill.

Saw Identification	First Phase of the Test (between 0 min and 480 min of Sawing)			Second Phase of the (between 480 min and 960 min of Sawing)
	Rate of Wear on the Clearance Face (µm/h)	Rate of Wear on the Rake Face (µm/h)	Rate of Increase of Wear Land (µm/h)	Rate of Wear on the Clearance Face (µm/h)	Rate of Wear on the Rake Face (µm/h)	Rate of Increase of Wear Land (µm/h)
#2	11.1	7.6	8.7	3.7	3.7	2.9
#4	18.0	15.6	16.2	15.9	6.3	10.9
#6	14.1	8.5	10.4	5.0	5.7	4.9

Saw #2: Coated and edge-modified saw. Saw #4: Coated and up-sharp saw. Saw #6: Uncoated and edge-modified saw.

. For all saws, the wear rate was higher at the beginning of the test corresponding to the break-in period.

Taking all wear indexes into account (recession on the rake and clearance faces and the width of the wear land), the percentages of decrease in wear rate when using the coated and edge-modified saw are presented in

Table 9. Decrease in wear rate (percentage) when using coated and edge-modified saw.

Saw Identification	After 480 min of Utilization (between 0 min and 480 min of Sawing)			After 480 min of Utilization (between 480 min and 960 min of Sawing)
	Clearance Face	Rake Face	Wear Land	Clearance Face	Rake Face	Wear Land
Decrease in wear rate of saw #2 compared to saw #4	38%	51%	46%	77%	41%	73%
Decrease in wear rate of saw #2 compared to saw #6	21%	11%	16%	26%	53%	41%

Saw #2: Coated and edge-modified saw. Saw #4: Coated and up-sharp saw. Saw #6: Uncoated and edge-modified saw.

. As seen, the above-mentioned saw significantly decreased wear rate compared to other saws especially in the second leg of the test (between 480 and 960 min). The results revealed the value of edge modification combined with the deposition of a CrN/CrCN coating in improving the wear resistance of circular saws used in primary wood processing.

The wear rate of the edge-modified saws was lower compared to the up-sharp saw. For the former, the size of the edge preparation on the rake face (20 µm) is large enough to withstand the shock of interrupted cutting. Furthermore, the smaller edge preparation (10 µm) on the clearance face prevents the saws from rubbing the workpiece, reducing frictional heat. On top of that, the variable configuration of edge preparation provided by cutting edge modification reduces the ploughing action and frictional heat during cutting. In variable edge modification, the edge radius changes along the corner of the tips. Therefore, the ratio of the uncut chip thickness to edge radius remains constant. In this case, the cutting edge should experience a lower temperature gradient. Sima et al. [48] stated that variable edge microgeometry can potentially reduce heat build-up during the turning of a titanium alloy. In the present study, the edge preparation evenly tapered from the full size (20 µm) on the primary cutting edge to 4 µm on the flank side edges. The latter mainly remained sharp to cut the spring back wood and reduced the frictional heat developed from side rubbing. This led to a lower wear rate of the edge-modified saws compared to the up-sharp one. In addition, the combination of cutting edge modification and coating treatment significantly decreased the wear rate, as seen in Figure 21 and Table 9. CrN/CrCN multilayer coating was efficient in reducing frictional heat on the rake and clearance faces as well as sides of the carbide tips. Its high adhesion and low coefficient of friction assisted in significantly increasing wear resistance.

As highlighted by Djouadi et al. [49], high adhesion and low COF of coatings play a crucial role in the performance of wood cutting tools by reducing the cutting forces and increasing the stability of the cutting process. Typical results ensuing from SEM characterization of the rake face of tips taken from each saw are presented in Figure 22a,b,d,e. The latter show the absence of crumbling, delamination, or cracks in the coating after 480 and 960 min of sawing. The saws without coating (#6) show evidence of cobalt binder removal, transgranular cracks in the WC grains, and grain pull out on the rake face (Figure 22c,f). As stated by Cristóvão et al. [50] and Guo et al. [51], the synergetic interaction between mechanical wear and heat generation, particularly frictional heat, is the main cause for partial extrusion of the cobalt binder, followed by carbide grain removal. These results highlight the importance of depositing multilayer coating on the saw tips in primary wood processing.

SEM characterization of the tip corners of the saws tested after 480 and 960 min of sawing are presented in Figure 23. The micrographs were taken from the rake face view of the tips. The coating did not provide advantages for cutting edge protection of saw #4 as it was peeled off from the sharp edges and corners of the tips (Figure 23b,e). The corner of the up-sharp tip (Figure 23e) sustained significant local breakage, highlighting the weakness of the tip in this area. However, no apparent chipping or breakage could be seen for saws #2 and 6 (Figure 23a,c,d,f). The reason is attributed to the increased corner strength brought about by the honing process. Furthermore, after 480 min of sawing, the rake face of some tips of saw #4 experienced cratering, mainly located at the corners (Figure 23b). The latter observation is attributed to the synergetic interaction of the abrasive action of chips over the rake face and the frictional heat generated at the wood-tip interface as the primary cutting edge was chipped.

Representative SEM micrographs of the cutting edge after 480 min of sawing are presented in Figure 24. The main wear mechanism involved for saw #4 was chipping (Figure 24b,e). It is well established that the stress concentration on the cutting edge of the tips with an up-sharp geometry is high [52], which leads to chipping and early breakdown of the cutting edge [53]. On the contrary, most tips of saws that have undergone edge modification (saws #2 and #6) appeared to be intact with only a few tips experiencing microchipping (Figure 24a,d,c,f). According to Ventura et al. [7], edge preparation improved the performance of cutting tools during metal machining by reducing the rate of chipping. It is reported that cutting edge preparation has the potential to alleviate the issues of edge chipping by eliminating typical edge defects such as micro fractures and irregularities [53].

In addition, a crack perpendicular to the cutting edge of saw #4 was observed, which is due to fatigue in combination with frictional heat. It is well known that heat is generated during wood cutting. Wood exhibits a low thermal conductivity, and therefore heat is either accumulated on the cutting edge and, to a much lesser extent, released into the surrounding air [9] Thus, the cutting edge experiences periodic temperature fluctuations. This situation combined with the cyclic nature of the cutting force during sawing led to crack initiation and propagation, [54] as shown in Figure 24e.

Different wear mechanisms on the cutting edges of all saws after 960 min of sawing are presented in Figure 25. The wear mechanisms are identical to those observed for the same types of tips after 480 min (Figure 24). Chipping was the principal wear mechanism responsible for the degradation of the cutting edges of saw #4 having up-sharp tips (Figure 25b,e). On the contrary, saws #2 and #6 with edge modification showed micro chippings and few chippings on their cutting edges, but to a much lesser extent than what was observed of saw #4, as shown in Figure 25a,c,d,f. Thus, variable engineered micro-geometry significantly reduced the tendency of the carbide tips to chip after 960 min of woodworking.

To precisely investigate the failures on the cutting edge, SEM micrographs of the worn cutting edge (from the rake face view) acquired at higher magnification are presented in Figure 26. The coated saw with up-sharp tips sustained transgranular crack in most WC grains after 480 and 960 min of sawing (Figure 26b,e). Apart from the crack in the cobalt binder, most WC grains lacked interconnecting cobalt. Furthermore, grain crushing, debonding, and removal of many WC grains occurred due to the continuous shock of intermittent cutting (Figure 26e). Ndlovu [43] pointed out that during wood machining, fluctuating cutting forces coupled with frictional heat generated on the rake and clearance faces causes oscillation of WC grains leading to cobalt extrusion followed by grains falling out from the matrix.

According to Hernández et al. [55], as the cutting edge is worn, the edge radius increases, causing the rake and clearance angles to decrease. This condition increases the ploughing action, creating frictional heat and inducing larger cutting forces on the cutting edge [56], thus leading to rapid tool wear [57]. However, there were a few cracks in the binder phase, reduced binder extrusion, and fewer transgranular cracks in WC grains for the saws having edge modification (Figure 26a,c,d,f). Most regions on the modified cutting edges remained intact. The reason is explained by the strengthening of the cutting edge, particularly at its corners. The variable configuration of edge preparation made the edge keener and reduced the ploughing area in front of the tool tip [58], thus minimizing edge degradation.

Typical SEM micrographs (from the clearance face view) of the corners of all saws used during 480 and 960 min of sawing are shown in Figure 27. Apart from breakage (Figure 27b), some of the tips of the coated saw with up-sharp cutting edge geometry suffered from abrasion wear (Figure 27e). The latter is due to the degradation of cutting edges, leading to larger contact length between the clearance faces and the wood. As a result, the level of frictional heat was concentrated in this area. Moreover, frictional heat developed from side rubbing on the spring back layer of the wood is another source of heat affecting the abrasion wear mechanism. As seen in Figure 27e, the coating was partially delaminated from the abraded area. No breakage or chipping was observed at the corners of edge-modified tips as a result of corner strengthening by the honing process (Figure 27a,d,c,f).

Abrasion took place on a small area at the corner of the tips of the uncoated saw that had undergone edge modification, as presented in Figure 27f. The corners of the tips of the coated saw with edge modification remained intact and no breakage and abrasion wear took place (Figure 27a,d). Coating on the sides of the saws had the potential to reduce the detrimental build-up of heat during sawing, thus helping to prevent abrasion on the clearance face. Reduction of wear on the corner of tips effectively increases the cutting capabilities of the saw. In a study conducted by Benlatreche [59], edge honing of WC-Co inserts improved tool wear resistance during machining MDF, especially when it was combined with the application of CrVN and CrAlN coatings.

SEM micrographs of the worn cutting edge from the clearance face view at higher magnification are presented in Figure 28. Transgranular cracks in a few WC grains and partial extrusion of the cobalt binder were observed on the cutting edge of the saws having variable edges preparation (Figure 28a,c,d). The grainy structure of the worn cutting edge of saw #6 (Figure 28f) showed that the latter lost most of its interconnecting cobalt binder after 960 min of sawing. The coated saw with up-sharp tips sustained a crack parallel to the cutting edge that is the product of impact wear, as seen in Figure 28b. The cutting edge could no longer support the cutting forces as efficiently, resulting in chipping. As explained earlier, this condition promotes an increase in the ploughing area in front of the tip. As a result, additional frictional heat is generated, assisting in binder removal, transgranular crack in WC grains in all directions, and fragmented grains (Figure 28b,e). It should be pointed out that the coating was fully removed from the up-sharp cutting edge. Therefore, the coating could no longer assist in the protection of the cutting edge from the aforementioned failures.

4. Conclusions

Premature chipping of circular saws due to the lower toughness of their up-sharp cutting edges has been a critical issue in the wood processing industry for many years.

This project characterized the benefits of applying variable engineered micro-geometry and surface coating technology on WC-Co carbide inserts mounted on circular saws. The objective was to improve the wear resistance of saws used in primary wood processing. The mechanical and tribological characterization of CrN/CrCN/DLC, CrN/AlTiN, CrN/CrCN and CrCN/TiSiCN multilayer coatings on WC-Co substrates showed that CrN/CrCN coating offered the best properties in terms of wear resistance and adhesion to the substrate. The results revealed that variable engineered micro-geometry combined with CrN/CrCN coating imparts significant improvements in the wear resistance of circular saws utilized in primary wood processing. Compared to the coated saw without edge modification and the uncoated saw with edge modification, the wear rate of the coated saw with edge modification decreased by 46% and 16%, respectively, after 480 min of sawing. Those values reached 73 % and 41%, respectively after 960 min of utilization in a mill. Furthermore, variable engineered micro-geometry paired with multilayer CrN/CrCN coating significantly reduced the tendency of the carbide tips to chip, to sustain breakage at their corners and to suffer from abrasive wear. This improvement has a major role in maximizing fiber recovery in the wood industry, contributing to significant cost savings, possible improvement of sawmill revenues, and process reliability.

The present research lays the groundwork for subsequent studies to optimize the size of edge preparation applied on the cutting edges of the saw teeth. Optimization of the coating parameters and chemical composition of the layers could likely become another interesting avenue for future work. Moreover, reduced or delayed wear will more than likely result in lower sawing variations.