CIRP Annals - Manufacturing Technology 58 (2009) 65–68
Contents lists available at ScienceDirect
CIRP Annals - Manufacturing Technology j o u r n a l h o m e p a g e : h t t p : / / e e s . e l s e v i e r . c o m / c i r p / d e f a ul ul t . a s p
CVD diamond coatings on geometrically complex cutting tools E. Uhlmann (1)* (1)*, J. Koenig Institute for Machine Tools and Factory Management, Chair of Machine Tools and Manufacturing Technology, Technical University Berlin, Berlin, Germany
A R TI C L E I N FO
A B S T R A C T
Keywords: Machining CVD diamond coating Wear analysis
The manufacturing of chemical vapour deposition (CVD) diamond coated shaft type cutting tools is demanding due to the complex design of the cutting edges and the cobalt content of the cemented carbide. The influencing parameters of substrate, pre-treatment and diamond film on the tool cutting performance are discussed. The optimised manufacturing route of CVD diamond coated thread milling drills is identified with the use of material and tribological tests. Following the optimised production of the tools, the thread milling drills are then applied in the machining of AlSi17Cu4Mg, whereby the tool performance is characterised with respect to their wear behaviour, the process forces and temperatures as well as the workpiece quality. 2009 CIRP.
1. CVD diamond diamond thin films on cutting tools
CVD dia diamon mond d thi thin n film filmss off offer er an app approa roach ch to com combin bine e dia diamo mond nd hardne har dness ss andwear res resist istanc ance e wit with h arb arbitr itrarytool arytool geo geomet metry.Stateof ry.Stateof the art in diamond tooling is the generation of diamond thin film systems on cemented carbide distinguished by crystallite size for indexable inserts and shaft tools (Fig. 1). Cemented carbide grades feasib fea sible le for CVDdiamo CVDdiamond nd dep deposi ositio tion n arelimit arelimited ed to a cob cobaltconte altcontent nt of 10 wt.% [1,2,3]. The manuf manufactur acturing ing chain of CVD diamond coated cemented carbide carbi de tools comm commences ences with the identification identification of a suitab suitable le substrate as well as the substrate pre-treatment to remove cobalt from the surface layer and to structure the tungsten carbide with underc und ercuts uts.. Thi Thiss is nec necess essaryto aryto pre preven ventt a cat cataly alyticreact ticreactionof ionof cob cobalt alt with wi th di diam amon ond d an and d to pr prov ovid ide e a me mech chan anic ical al bo bond nd be betw twee een n substr sub strate ate and dia diamon mond d film film.. The These se man manufa ufactu cturin ring g ste steps ps are follow fol lowed ed by cle cleani aning ng and dia diamon mond d see seedin ding g mea measur sures es bef before ore CVD diamond deposition is carried out [1]. The current state of research of diamond thin film technology on cutti cutting ng tool substrates substrates comp comprises rises diamond film depo depositio sition n adapted substrate development and treatment, CVD diamond film post-tre post -treatme atment nt and increased increased diam diamond ond film adhe adhesion sion by analysing analysing streng str ength th and res residu idual al str stress ess beh behavi aviour our.. Fin Fine e gra grain in cem cement ented ed carbides and partly partly silicon based ceramics ceramics are mainly employed employed as tool substrates. The residual stress profile of the diamond film and substrate interface depends on substrate type and pre-treatment as wel welll as CVDcondi CVDconditio tions ns andCVD dia diamo mond nd filmprop filmpropert erties[5–10 ies[5–10]. ].
* Correspon Corresponding ding author. 0007-8506/$ – see front matter 2009 CIRP. doi:10.1016/j.cirp.2009.03.063
2. Anal Analysis ysis of the CVD diam diamond ond coated tool manufacturin manufacturing g chain 2.1. Research setup
In this research research manuf manufactur acturing ing of diam diamond ond coated tools by varying substr varying substrate, ate, prepre-trea treatmentand tmentand diam diamond ond film was focuse focused d on (Table 1). The design of experiment comprises the use of one ultra fine gra grain in car carbid bide e (su (subst bstrat rate e A) and two fine gra grain in car carbid bides es (subst (su bstrat rates es B andC) wit with h var variou iouss cob cobaltconte altcontentsfrom ntsfrom 6 to 10 wt. wt.%. %. The influence of polished and ground substrate surface roughness as wel welll as thr three ee dif differ ferent ent sub substr strate ate pr pre-t e-trea reatme tments nts (pt (pt)) for preparing prep aring the diam diamond ond deposition deposition on the tool wear behaviour behaviour was analysed. The prepre-treat treatments ments vary with respect to cobal cobaltt reduct red uction ion dep depth th in the sur surfac face e lay layer er and mec mechan hanica icall sur surfac face e structuring. The deposited diamond films are characterised by a variation varia tion of morp morpholog hology y (nano (nanocryst crystalline alline and multi multilayer layer)) and medium film thickness (d = 8 mm and d = 12 mm). Initially, characteristics of cemented carbides after each step of manufacturing were researched in cutting tool material analysis and tribological tests. Test samples with diameter d = 3.3 mm and length l = 38.0 mm were used. Afte Afterr measuring measuring surfa surface ce and cutting cutting edge roughness according to DIN EN ISO 4287 surface formation and cutting edge radius were determined. Transverse rupture strength according to DIN EN ISO 3327 was evalua eva luated ted in thr three ee poi point nt ben bendin ding g tes tests. ts. The tri tribol bologi ogical cal tes tests ts consis con sisted ted of osc oscill illati ating ng sli slidin ding g wea wearr tes tests ts of the cut cuttin ting g too tooll material in dry contact with AlSi17Cu4Mg and particle jet blasting tests of the diamond coated cemented carbides. The latter was carried out in a rotating setup. The cutting research was carried out with a two flute thread milling drill for manufacturing metric threads of 8 mm diameter (M8). The thread milling drill is a geometrically complex shaft tool applied for a combination cutting process of core hole drilling and countersinking for the thread entrance chamfer and a final thread
66
E. Uhlmann, J. Koenig/ CIRP Annals - Manufacturing Technology 58 (2009) 65–68 6.0
0.60
t R
eµm l i f o r p f 3.6 o t h 2.4 g i e h l 1.2 a t o T 0
a R
n
o µm t i a i v
e 0.36 d n
0.24 a e 8 µm multilayer diamond unc.
pt2
pt1
pt2
pt3
m . i r A
0.12 h t 0
Manufacturing state of substrate A
Fig. 1. Types of CVD diamond thin film modifications [4].
Fig. 2. Surface roughness of substrate A after each step of manufacturing.
milling process. It can be used for the manufacturing of various thread geometries. The significant challenges posed by this process are the tool geometry design as well as the diamond coating manufacturing steps. 2.2. Properties of cutting tool material and thread milling drills
All three cemented carbide surfaces exhibited similar surface roughness values following the same manufacturing step. The measured surface profile for substrate A in Fig. 2 showed the lowest roughness for the nontreated cemented carbide and an almost constant high roughness for the pre-treated and diamond coated samples. The high surface roughness of the diamond coating is causedby thesubstrate pre-treatment. In thecase of pretreatment 3 (pt3) a significantly lower value of total height of the profile Rt was obtained whereas no change of arithmetical mean deviation Ra can be observed. Even the ground substrate B with the highest surface roughness in the uncoated and nontreated state showed a similar surface roughness to the polished substrates after pre-treatment. It can therefore be seen that the surface roughness differences of polished and ground samples were almost eliminated by the pre-treatment process. So, the characteristics of the diamond coated cemented carbide are almost independent of the initial treatment and roughness state but mainly a function of the mechanical and chemical coating preparation. Thus, tool body grinding has low influence on diamond deposition and diamond film adhesion. Fig. 3 depicts the surface formation of substrate B in uncoated and untreated as well as in a pre-treated state (pt1). It can be seen that the grinding traces on the polished (Fig. 3a) and rough (Fig. 3b) samples were levelled after pre-treatment (Figs. 3c and d). These results were completed with measurements of the drill main cutting edge of the thread milling drills. The cutting edge roughness, also as known as shardness, increased on average by one fourth from Rt = 2.8 mm to Rt = 3.7 mm after diamond coating. The cutting edge roundness of CVD diamond coated tools doubled from r b = 10 mm in the uncoated state to r b = 21 mm, which correlates with previous research [5,6] The strength of samples representing the manufacturing steps of uncoated and untreated, pre-treated and diamond coated cemented carbide was analysed and compared with samples that had undergone only an annealing process. The transverse rupture strength, analysed in a three point bending test setup, is decreased by one third after pre-treatment, while the CVD diamond coating leads to a 70% strength increase. The samples which had been annealed only exhibited a similar strength to the untreated samples. The lowered Weibull module after each step of manufacturing compared to the initial Weibull module of the uncoated samples displays a higher distribution of the strength values (Fig. 4). It represents a diminished reliability of the diamond
Fig. 3. Surface formation of (a) polished and untreated, (b) rough and untreated, (c) polished and pre-treated pt1 as well as (d) rough and pre-treated pt1 substrate B.
coatedcementedcarbide anda higherrisk of shaft type cutting tool failure. An influence of cemented carbide cobalt content was evaluated. The lower the cobalt content, the lower the strength after pretreatment but the higher the strength after diamond coating. Reason for this is that the cobalt content exhibits a higher toughness which supports the strength even if the surface layer is weakened by cobalt etching. After diamond coating the residual stress in the interface of substrate and diamond film is determined by theamount of cobalt which causes thermallyinduced stress due to its high thermal expansion in contradiction to tungsten carbide and diamond. 2.3. Tribological behaviour of cutting tool material
Fig. 5 demonstrates the influence of roughness Rt against the friction coefficient m600s after t = 600 s friction time of the tool substrates to AlSi17Cu4Mg in oscillating sliding tests. The uncoated and untreated as well as the pre-treated cemented carbides showed the highest friction coefficient, independent of their surface roughness. The influencing factor is the thermochemical affinity of cobalt in the case of the untreated cemented carbide, which leads to micro-weldings between the cobalt and workpiece material. The pre-treated cemented carbide causes micro-chipping of the aluminium alloy due to its surface topography of tungsten carbide particle edges. The influence of surface roughness of the pre-treated cemented carbides and pretreatment type on friction coefficient is observed to be less significant. The CVD diamond films have a broad range of surface roughness. Almost independent of their roughness state, their
Table 1 Researched cemented carbides and variation of process steps.
Grade
Carbide grain size ( mm)
Co content (wt.%)
Surface roughness
Pre-treatment
Diamond morphology
Diamond film thickness ( mm)
A B C
0.2–0.5 0.5–0.8 0.5–0.8
9 6 10
Polished Polished, ground Polished
pt1, pt2, pt3 pt1, pt2, pt3 pt1, pt2,
Nanocrystalline, multilayer Nanocrystalline, multilayer Multilayer
8, 12 8, 12 8
E. Uhlmann, J. Koenig / CIRP Annals - Manufacturing Technology 58 (2009) 65–68
Fig. 4. Transverse rupture strength of substrate A after each step of manufacturing.
friction coefficient was found to be approximately half that of the uncoated cemented carbide. Reasons are the low chemical affinity of diamond to aluminium silicon and its high hardness which suppresses mechanical interactions with the counter body. Pretreatment pt3 is superior for supporting the tribological behaviour of the diamond film. A rotating setup for particle jet blasting test was used to model the load on varied diamond coatings on cemented carbides at changing load angle, simulating respectively the load direction during drilling and milling. The poorest performance was obtained by substrate C and partly substrate A, both with high cobalt content. Best results could be detected in the case of the low cobalt content containing substrate B with high diamond film thickness. These tests were aborted following 3000 s blasting time (Fig. 6). In thefixed setup,multilayer films with 12 mm onsubstrate B had the highest wear resistance. 2.4. Application of CVD diamond coated thread milling drills
During cutting tests with AlSi17Cu4Mg, diamond coated thread milling drills based on substrate A showed the same ability to reach the tool life criteria of N VB0.3 = 400 threads with diameter M8 as uncoated tools. This result was independent of whether the entire cutting tool or only the drill edges were diamond coated. Thread milling drills of substrate B couldoutperform the respective uncoated tools when the drill edges were coated. The diamond coating of the entire cutting tool led to total tool fracture however.
Fig. 5. Friction coefficient as function of total height of the profile for the cemented carbides after each step of manufacturing.
67
Fig. 6. Average jet blasting time in rotating setup for selected CVD diamond coated cemented carbides.
The pre-treatment was shown to be the strongest influencing factor on tool life quantity of the diamond coated tools. Total tool fracture could only be completely avoided by pre-treatment pt3 (Fig. 7). The other pre-treatment types induced surface layer damage at the fracture relevant zone between milling part and countersink, thus leading to early fracture of the tool. Temperature as well as axial feed force and cutting torque were measured during the cutting process. The theoretical process temperature was detected by indirect thermographic measurements of core hole drilling and subsequent analysis of the thermodynamic behaviour of the workpiece material. The maximum temperature at the cutting edge was estimated to be T = 240 C and therefore lies far from a temperature which could induce wear of the diamond [1]. Cutting torque M c was measured for the first and last thread to be cut with uncoated as well as pt3 pre-treated and diamond coated thread milling drills (Fig. 8). Despite higher surface roughness and cutting edge radius of the diamond coated cutting tool compared to the uncoated thread milling drills, the mechanical process loads are often lower or in the same range. During drilling no difference of cutting torquevalueswere observed dueto the modification of the thread milling drill. During countersinking, maximum cutting torque occurs due to the combined drilling when four cutting edges are engaged. The CVD diamond coated tools often showed similar or lower cutting torque than the uncoated tools. Similar results were obtained for thread milling. Usually, cutting torque while using diamond coated tools is increased due to their higher cutting edge radius and surface roughness [7,8]. Reason for the observed behaviour is only partly 8
Fig. 7. Tool life quantity of uncoated and variedly diamond coated thread milling drills M8 based on substrate A and B.
68
E. Uhlmann, J. Koenig/ CIRP Annals - Manufacturing Technology 58 (2009) 65–68
Fig. 10. Accuracy to thread gauge of manufactured threads M8.
Fig. 8. Cutting torque offirst and last thread tobe machinedwithuncoated andCVD diamond coated thread milling drills over tool life quantity.
After evaluating the cutting tool wear and process behaviour the thread quality was determined using a thread gauge. Without milling path compensation the threads machined with cutting tools of substrateB exhibitedthe longest lasting accuracy to gauge, with a mean of 280 threads. The uncoated thread milling drills exhibited only limited accuracy to thread gauge (Fig. 10). This is probably due to higher tool deflection induced by increased mechanical loads and stronger elastic workpiece material deformation caused by increased friction related temperatures at the interface of the workpiece and the uncoated thread milling drill. 3. Summary
Fig. 9. Worn cutting edges of pt3 pre-treated and d = 12 mm CVD diamond coated thread milling drills: (a–c) multilayer film on substrate A, (d) nanocrystalline film on substrate B.
the minimised roughness of the pt3 pre-treated and diamond coated tools. The core reason is the pre-treatment dependent suppressed tool fracture which leads to an artificial wear development of the diamond coating. Small film volumes from diamond coating delaminate directly at the drill and thread mill cutting edges. Due to the good film adhesion of the diamond film, cutting edge displacement occurs but the load resisting diamond film further protects the cutting edge. It appears in combination with a highly sharp structureof the diamond fracture at the cutting edge even in a worn state. The geometry of cutting edges is altered by mechanical wear, especially at the drill minor cutting edges, which also work as milling teeth during thread milling (Fig. 9a). The diamond film at the thread milling teeth is also mostly delaminated. Remaining diamond film parts resist further cutting edge deviation (Fig. 9b). Partly cohesive film delamination in the thin film itself was observed for multilayer diamond coatings on substrate A (Fig. 9c) as well as on substrate B. This wear behaviour was also seen on the nanocrystalline diamond films on substrate B (Fig. 9d). Nanocrystalline diamond deposited on substrate A delaminated directly from the substrate and could be classified as adhesive thin film failure.
The deposition of CVD diamond thin films on geometrically complex cutting tools is possible even on ultrafine tungsten carbide grain size cemented carbide substrates with higher cobalt content. The strength of CVD diamond coated tools is mainly influenced by the initial strength and substrate pre-treatment. The cobalt content of the substrate determines the strength behaviour of the CVD diamond coated tool. Diamond thin films with low roughness value Rt , which is mainly a function of substrate pre-treatment, demonstrated the best performance in sliding wear tests. The substrate cobalt content and diamond film thickness significantly influences fatiguewear resistance.Lowertungsten carbidegrainsize supports the extension of this target parameter. At the cutting edge of the diamond coated tool film delamination occurs during machining. Due to high film adhesion the diamond film serves as a tribological partner with highly sharp cutting edges. References
[1] Dischler B, Wild C, (Eds.) (1998), Low-Pressure Synthetic Diamond. Manufacturing and Applications. Springer-Verlag, Berlin. [2] Field JE,(Ed.) (1992), The Propertiesof Natural and SyntheticDiamond. Academic Press, London, pp. 473–513. [3] Uhlmann E, Kott R, Hu¨bert C (2004) Wear Mechanisms of Mono- and Multilayer CVD Diamond Films,. Production Engineering 11(2):83–86. [4] Lemmer O, Cremer R, Breidt D, Frank M, Mu¨ ller J (2005) CVD-Diamantbeschichtung nach dem Hot-Filament-Verfahren. in Weinert K, (Ed.) Spanende Fertigung . Vulkan-Verlag, Essen, pp. 189–201. [5] Grams, J (2004) Untersuchungen zum Fra¨ sen mit CVD-diamantbeschichteten Werkzeugen, Dissertation, RWTH Aachen. [6] Kott R (2007) Eigenschaften und Einsatzverhalten CVD-diamantbeschichteter Hartmetallwerkzeuge. Dissertation, TU Berlin. [7] VDI 2841 Blatt 1: 2008 (Public Draft) CVD Diamond Tools. Categorisation, Production and Characterisation, Beuth-Verlag, Berlin. [8] Uhlmann E, Kott R (2002) Investigation on the Wear Behaviour of Diamond Coated Taps. Proceedings of the 3rd International Conference on THE Coatings , Salonika, 149–156. [9] Settineri L, Bucciotti F, Cesano F, Faga MG (2007) Surface Properties of Diamond Coatings for Cutting Tools. Annals of the CIRP 56(1):573–576. [10] Byrne G, Dornfeld D, Denkena B (2003) Advancing Cutting Technology. Annals of the CIRP 52(2):483–550.