Effects of deep cryogenic treatment on the wear development of H13A tungsten carbide inserts when machining AISI 1045 steel

In this study, the effects of deep cryogenic treatment (93 K) on the surface and sub-surface wear development of H13A cobalt-bonded tungsten carbide cutting inserts during the wet machining of AISI 1045 steel were investigated. Cutting inserts were subjected to short periods (171–553 s) of turning at cutting speeds of 50–140 m/min, during which time mass measurements were taken and the worn edges were imaged and scanned, by optical microscopy and light interferometry, at regular intervals. Sections were taken following machining so that sub-surface features could be observed by scanning electron microscopy. It was determined that cryogenic treatment resulted in a 9.2 % increase in hardness and an increase in abrasive wear resistance, although microstructural changes and sub-surface behaviours suggested a corresponding decrease in toughness may have occurred.


Cryogenic processing
Cryogenic processing, which is alternatively known as cryogenic treatment or 'cryotreatment', involves the use of low temperatures (\193 K, [1]) to cause metastable or stable microstructural changes in materials and components. Cryotreating using forced convection of gaseous nitrogen in a sealed chamber, rather than direct immersion in liquid nitrogen, prevents the introduction of high thermal stresses into the material [2]. It also has the advantage of creating a dry atmosphere that will prevent damage to the surface of components [3].
Cryogenic treatment encompasses two commonly recognised classes of sub-zero heat treatments. Cryogenic treatments can be classed as 'shallow' and 'deep'; referring to temperature ranges of around 193-113 and 113-77 K respectively [1].
The past decade has seen an expansion of research into cryogenic treatments for cutting tool materials, such as cobaltbonded tungsten carbides. These studies have demonstrated a vast range of performance changes due to cryogenic treatment; among those already highlighted, increases in wear resistance or tool-life ranging from 0 to 1,257 % in tool steels and 0-36 % in tungsten carbides are reported. While material composition, treatment parameters and test conditions clearly have a significant impact on the effectiveness of cryogenic treatments, they undoubtedly offer significant benefits to component manufacturers and end users.

Wear of cutting tools in conventional machining
Conventional machining processes place numerous demands on cutting tool materials, which must be strong, hard, wear resistant and thermally resilient. Depending on the application they are also required to have a low chemical affinity and tendency for dissolution with the workpiece, and good corrosion resistance.
The combination of conditions experienced by cutting tools in conventional machining requires that they withstand high stresses (2 GPa), significant rubbing with the workpiece material and high temperatures (873 K) [29]. Commonly measured wear features include flank wear and crater wear, although cutting edge rounding and chipping may have a significant impact on cutting forces and surface finish [30,31]. The contributions made by these mechanisms to tool wear are significantly affected by temperature. For instance, diffusion overtakes abrasion as the dominant mechanism of crater wear at higher cutting speeds (and therefore temperatures), in tungsten carbide tools [32].
Complete understanding of the performance of a cutting tool requires the determination of its material properties (hardness, flexural strength, fracture toughness and thermal conductivity), tribological performance (wear resistance and friction characteristics) and an in-depth characterisation of its microstructure. However, useful understanding can be obtained through a combination of wear measurements (mass changes, surface wear and sub-surface wear analyses) to compare the behaviour (tough or brittle) of a tool material, and its resistance to various wear mechanisms (abrasion, adhesion and diffusion).

Cryogenic treatment of tungsten carbides
Various studies investigating the effects of cryogenic treatment on tungsten carbides have been conducted in recent years. Investigations by Seah et al. [33] and Thakur et al. [34] demonstrated that cryogenic treatment had no significant effect on the hardness of tungsten carbides with a 7 % cobalt content. This finding was corroborated by SreeramaReddy et al. [35] when testing ISO P-30 (WC-17Co-1.4TiC-1.4TaC) inserts. However, it was determined that the cryogenically treated tungsten carbides had significantly improved hot hardness when compared with untreated materials; suffering only a 17 % reduction in hardness, compared to 29 % for untreated inserts, when tested at 873 K. The study of Gill et al. [36] is currently the only one to report modest improvements (*5 %) in the room temperature hardness of ISO P-25 (WC-6Co-1.4TiC) inserts following cryogenic treatment. While hardness changes in WC-MC-Co materials may be dependent on the presence of cubic carbides (M = Ti, Ta, Nb, V, Hf etc.), the lack of detailed processing routes and parameters available makes it impossible to gauge their significance.
A number of studies have used methodologies based on ISO 3685 [37]. In studies testing ISO P-25 (WC-6Co-1.4TiC) inserts, Gill et al. [38,39] determined that cryogenic treatment could change the life of uncoated tools by ?13 to ?36 %, and TiAlN coated tools by -3.9 to ?34 %; suggesting that deep cryogenic treatment weakened the adhesive bonding of the coating to the tungsten carbide substrate. In comparison with untreated inserts, cryogenically treated inserts displayed less evidence of abrasive wear and plastic deformation.
In contrast to the findings of Gill et al., when testing ISO P-30 (WC-17Co-1.4TiC-1.4TaC) inserts with a multilayer CVD coating (TiN, TiCN, Al 2 O 3 , TiN), SreeramaReddy et al. [35] observed no deterioration in their performance, when machining C45 steel, after cryogenic treatment. Their results suggested that cryogenic treatment gave an optimum increase in tool-life at cutting speeds of 250-300 m/ min, of 22 %, compared with 7.7 and 15 % at 200 and 350 m/min respectively, with a feed rate of 0.22 mm/rev and depth of cut of 1.0 mm.
Cryogenic treatment has been reported to result in three significant changes in cobalt-bonded tungsten carbides: (1) an increase in density of the cobalt binder (b-phase) [34]; (2) enlarged WC grains (a-phase) [35] and; (3) finer g-phase particles which are also stabilised by cryogenic treatment [40]. These changes could be expected to improve the corrosion resistance of the cobalt-binder and the thermal conductivity of the material due to a more contiguous WC grain structure. Although hard g-phase particles are often avoided in tungsten carbides to prevent premature fracture occurring, they may improve abrasive wear resistance and, if sufficiently refined and uniformly distributed, increase the toughness of cobalt-bonded tungsten carbides.
While a number of studies have highlighted the potential benefits of cryogenically treated tungsten carbides and begun to reveal the microstructural changes responsible, further studies are required to detail the surface and subsurface wear development of cryogenically treated tungsten carbide tools. It is important that their wear behaviour is understood for a wide variety of machining conditions, to optimise their use and avoid premature tool failures. In this study the wear development of untreated and cryogenically treated Sandvik Hard Materials H13A (SCMT 120408-KM) turning inserts when turning AISI 1045 (EN8) steel is presented with supporting sub-surface and microstructural observations. Uncoated inserts were selected to remove the uncertainty arising from the conflicting findings of previous investigators on the effects of deep cryogenic treatment on the strength of coating-substrate interfaces.
2 Tool material and test methodology

Preparation of cutting inserts
20 Sandvik Hard Materials H13A (SCMT 120408-KM) turning inserts were randomly numbered and labelled N1-10 (untreated or 'non-cryotreated') and C1-10 (cryotreated), with the latter set being subjected to deep cryogenic treatment by Cryogenic Treatment Services Ltd. (Nottinghamshire, UK) prior to testing. Inserts were cooled in a nitrogen atmosphere at 1-2 K/min, held at 93 K for 24 h, and then returned to ambient temperature at 1-2 K/min. The inserts were subsequently tempered at 453 K for 2 h, before being returned to ambient temperature prior to use. A Sandvik Coromant CoroTurn 107 tool holder (SSBCR 2020K 12) was used, giving the inserts a lead angle of 75°and a clearance angle of 7°. While these inserts (being ISO K-type cutting tools) are not optimised for machining of AISI 1045 steel, discussions with industrial sources confirmed that the tool-material pairing was not uncommon and would provide for a suitable comparative study.

Hardness measurements
Low-force hardness tests were performed on the surfaces of inserts prior to testing using a Vickers micro-hardness tester (Mitutoyo HM-101), with an applied load of 0.5 kgf (4.9 N) and a duration of 15 s. Ten measurements were taken from each of three untreated (N2-4) and cryotreated (C2-4) inserts, with the mean values and coefficients of variation (C v = standard deviation/mean) reported.

Tool wear testing
Tool wear testing was conducted using a methodology based on ISO 3685 [37]. As recording the initial development of wear was the primary objective of this study, the full range of tests required to determine tool-life was not conducted. A standard bench lathe (Colchester 600 Series) was used for testing, which did not allow for continuously variable spindle speeds. To maximise the use of the limited workpiece material available, the cutting speeds selected (50, 95, 140 m/min) were achieved to within ±10 %. Cutting fluid was applied to the top of the workpiece 90°a head of the tool to ensure that the cutting zone was consistently flooded throughout each cut.
Three untreated and three cryotreated inserts were used to machine AISI 1045 (EN8) steel bars at each cutting speed, using a fixed feed rate of 0.2 mm/rev and depth of cut of 0.5 mm. Two bars (with initial diameters of 69 and 71 mm) were used for this study. Each bar was slotted to provide four separate cutting lengths of between 95 and 97 mm. After cutting one of these lengths (equivalent to 34-112 s of machining depending on cutting condition and spindle speed of 260, 470 or 840 rpm), the inserts were removed, cleaned and weighed, before being optically imaged and scanned by light interferometry as described in Sect. 2.4. Before each pass of the workpiece, the inserts were clean and at ambient temperature.
The order in which the inserts were used to make passes on the steel bars was systematically randomised, to minimise the effects of any inhomogeneity in the material and the variation in material removed by each insert. Figure 1 illustrates the cutting lengths (1-4) created on the machined bars, the effect of successive passes by the inserts tested and their geometry.

Wear measurements and characterisation
Before testing, the flank and rake faces of one untreated (N1) and one cryotreated (C1) insert were scanned using a light interferometer to establish initial roughness values, and imaged using an optical microscope to identify notable surface features. Figure 2 shows these images taken from an untreated insert (N1), highlighting the areas on both the rake and flank faces where five area roughness measurements were made, from which to determine mean values. The initial mass of every insert was measured using precision digital scales (Sartorius Analytical Balance BP210D) with a readability of ±10 lg.
Both untreated and cryotreated inserts were found to have similar surface roughnesses of approximately 0.64 and 0.47 lm (R q ) on their rake and flank faces respectively. Their initial masses were randomly distributed about a mean of 7.925 g with a negligible standard deviation.
After each pass of the workpiece, each insert was removed and cleaned with acetone in an ultrasonic bath to remove loose debris and cutting fluid residue, without removing any builtup edge (BUE). Following weighing, the flank and rake faces of each insert were imaged using an optical microscope (Carl Zeiss AxioImager, darkfield filter, 109 and 59 objectives respectively) and the rake faces were scanned, by light interferometry (Bruker ContourGT, 109 objective, narrowband green light), to measure the extent of crater wear. Following testing, one untreated and one cryotreated insert, tested at cutting speeds of 50 (N3 and C3), 95 (N7 and C7) and 140 m/min (N9 and C8), along with the two unused inserts (N1 and C1), were mounted in phenolic resin, before being ground with diamond grinding paper to reveal a cross-section through the cutting edge and crater wear zone on the rake face of the tool. Samples were then polished with diamond suspensions. Sub-surface behaviours were imaged using a scanning electron microscope (FEI Inspect F) at up to 5,0009 magnification.

Microstructural analysis
Microstructures were imaged with the specimens in their polished condition (without chemical etching) so that any change in the porosity of the material due to cryogenic treatment could be observed. A scanning electron microscope (FEI Inspect F) was used to generate micrographs at magnifications of up to 10,0009.

Hardness
Hardness measurements indicated a slight increase in the hardness of the tool inserts due to cryogenic treatment (22,600 MPa compared to 20,700 MPa; an increase of 9.2 %), although it was noted that this increase lay within the C v values determined from the untreated and cryotreated inserts, of 15.5 and 15.6 % respectively.

Tool wear testing
At each of the cutting speeds at which tool inserts were tested, distinct wear mechanisms and behaviours were observed. At cutting speeds of 95 and 140 m/min, flank wear and a BUE developed on each insert, whereas at 50 m/min they did not. While crater wear developed at every cutting speed, the intrusion of a built-up edge into the crater wear zone at cutting speeds of 95 and 140 m/min restricted its development and prevented its subsequent measurement.
Inserts were subjected to differing durations of machining depending on the selected cutting speed, however all inserts made five 'passes' of the workpiece after which measurements were taken. Table 1 relates the number of passes made by the cutting inserts with the corresponding cumulative machining durations and volumes of material removed.
The change in mass of each set of inserts gave an indication of the dominant wear mechanisms at work. Abrasive wear on the flank and rake faces, or dissolution of the tool into the chip led to mass losses, whereas adhesion of the workpiece material to the tool and BUE led to mass gains. From mass measurements taken ( Table 2) it was clear that the greatest adhesive behaviour was displayed by cryotreated inserts at 95 m/min, while untreated inserts at 140 m/min demonstrated the most consistent mass loss driven by abrasive wear, which was likely to be as a result of the coolant not working effectively as a lubricant at higher cutting speeds due to less entrainment in the contact as a result of greater rotational speeds.

50 m/min tests
While initially, there was some evidence of adhesion on cryotreated inserts tested at 50 m/min, both sets of inserts predominantly suffered mass loss ( Table 2) due to crater wear until the latter stages of testing when a BUE formed in isolated cases (N4 after 5 passes and C2 after 4 passes) and material began to adhere to the tool in the crater wear zone (as seen on inserts N4 and C2 in Fig. 3).
No flank wear was visible on any of the inserts as the cutting fluid had effectively lubricated the tool-workpiece flank interface under these conditions. The cutting edges of untreated inserts developed slight abrasive wear features, while only minimal edge wear and rounding was observed on cryotreated inserts after machining at 50 m/min.
Crater wear measurements following machining tests at 50 m/min (Fig. 4) indicated no significant change due to cryogenic treatment with both sets of inserts bearing craters with mean depths of approximately 26 lm. The apparent reduction in the rate of growth of crater wear after 3 passes ([7 min of machining) correlates with the observations that the workpiece material began adhering to the inserts.

95 m/min tests
The greatest adhesion of material and development of a BUE was observed from tests at 95 m/min (inserts N5 and C6 in Fig. 3). A large BUE rapidly developed, leading to a significant increase in mass for both untreated and cryotreated inserts ( Table 2). The mass of both untreated and cryotreated inserts fluctuated during testing as the BUE repeatedly developed and broke off, ultimately resulting in almost no overall mass change after 5 min of machining.
Tools tested at 95 m/min were observed to have evenly distributed flank wear and a significant BUE (Fig. 5). The extent of flank wear on cryotreated tools had a mean value 6 % less than that observed on untreated tools (Fig. 6), with measurements indicating that cryotreated tools supported a BUE with a mean height 19 % higher than that of untreated tools, although with considerable variation due to the tendency for these edges to build up and break off during testing. As this BUE was often observed to protrude ahead of the cutting edge and with a natural correlation existing between BUE height and protrusion over the cutting edge, it was thought to be at least partly responsible for the reduced flank wear observed on cryotreated tools. It should be noted, however, that continuing BUE formation and renewal would result in relative sliding motion with the tool face, and therefore abrasion would still continue (although probably under substantially altered contact conditions).

140 m/min tests
During testing at 140 m/min the mass loss of tool inserts was relatively constant ( Table 2), indicating that the loss of tool material due to flank and crater wear was more  significant than adhesion of the workpiece material. The BUE observed was substantially smaller (38 % lower mean height) and the extent of flank wear greater than for inserts tested at 50 and 95 m/min (Fig. 6). At 140 m/min the mean extent of flank wear on cryotreated tools was 5 % greater than that of the untreated tools, with the mean height of BUE on cryotreated tools 12 % less. Significant notch wear was also observed on both sets of inserts, with the cryotreated inserts again performing marginally worse (Fig. 6). Considering the postulation already made regarding a link between flank wear extent and BUE, it is suggested that this apparent change in BUE behaviour plays a significant role in the wear of cryotreated cutting tools.

Sub-surface features and microstructures
Electron micrographs taken through sections of the crater wear zones in inserts tested at 50 m/min (Fig. 7) revealed more significant damage in untreated inserts, with evidence of the removal and separation of tungsten carbide grains (bright) from the cobalt binder (dark) near the surface. In cryotreated inserts a compaction of the grain structure up to 3-5 lm from the surface was observed. External observations suggested that the crater wear zones of inserts tested at 95 and 140 m/min suggested they may be protected by the 'tail' of the BUE formed on the cutting edge. However, sub-surface observations (Fig. 8) revealed that this tail entrained tungsten carbide grains removed from the tool surface, likely resulting in significant abrasion of the tool rake face and crater wear zones. As the BUE tail broke down, it was observed to form a tribofilm of approximately 10 lm across the insert surface. Different behaviours were seen at the nose of inserts, the point of separation between the adhered material and rake face, and underneath the tail of the BUE. In untreated inserts at 95 m/min (Fig. 8a) the BUE appeared to detach from the insert and break up more readily than in cryotreated inserts (Fig. 8b). In cryotreated inserts the BUE was not only significantly larger but the point of separation marked the start of a crack propagating towards the cutting edge.
After testing at 140 m/min, a more cohesive BUE was again seen on the cryotreated insert (Fig. 8d), which did not break up upon separating from the tool material. The onset of catastrophic failure was observed in the untreated insert (Fig. 8c) instead, with a clear fracture of the tool nose being hidden from external observation by the BUE. By contrast the cryotreated inserts showed only minimal signs of sub-surface damage.
Observations of the microstructures of unused inserts (Fig. 9) revealed the cobalt phase to occupy a greater portion of the visible sections following cryogenic treatment but, unlike other reported studies [35], there was no visible swelling of the tungsten carbide grains. While it was not possible to quantitatively verify these observations based on the samples available, it is suggested the changes observed would have resulted in an increase in the abrasive wear resistance of the material, but a reduction in toughness as a result of a less contiguous microstructure.

Conclusions
From this study, it was determined that subjecting H13A tungsten carbide inserts to deep cryogenic treatment resulted in: • An increase in the hardness of the material of 9.2 % (based on a sample of 30 measurements per treatment), although it was noted to lie within one standard deviation of the mean hardness of both the untreated and cryotreated inserts. • Mean reductions in flank wear of 6 % when machining at 95 m/min, which was primarily abrasive in nature. This was suggested to be strongly influenced by the development of built-up edges on the inserts at this cutting speed.
• No significant change in the depth of crater wear, although sub-surface analyses indicated that cryotreated inserts suffered less pull-out and separation of carbide grains, resulting in a smoother worn zone. • Greater adhesion between the steel workpiece material and the tungsten carbide inserts, resulting in larger BUE that protected inserts from abrasive wear but increased the stress they were subjected to, as suggested by subsurface cracking. • Significant changes in the cobalt binder phase, suggested to have resulted in an increase in abrasive wear resistance but a likely reduction in toughness.
In summary, deep cryogenic treatment was found to offer an improvement in the wear resistance of SHM H13A turning inserts, although care must be taken when choosing machining parameters as key wear behaviours were observed to change when these were varied.