Surface and Coatings Technology74 75 (1995) 178-182
Adhesion and wear resistance of nitrided and TiN coated low alloy steels H.-J. Spies a, B. Larisch a, K. H6ck b, E. Broszeit b, H.-J.SchrOder c Institut ffir Werkstofftechnik, Technische Universitfit Bergakademie Freiberg, G. Zeuner Strafle 5, 09596 Freiberg, Germany b Vakuumtechnik Dresden GmbH, Bismarckstrafle 66, 01242 Dresden, Germany ° lnstitutfiir Werkstoffkunde, Technische Hochschule Darmstadt, Grafenstrafle 2, 64283 Darmstadt, Germany
Abstract Hardened and tempered low alloy steel grade 31CrMoV9 was gas and plasma nitrided to form a varied structure of the substrate for a subsequent TiN hardcoating. The process parameters of the nitriding were modified to form a sufficiently supporting diffusion layer with a thickness of more than 0.5 mm and a defined structure of the surface. Hardened cases without compound layer were formed using more step technologies in the nitriding process. The TiN coating was deposited by hollow cathode discharge evaporation. The composition and structure of the nitrided case, the mechanical pretreatment, the plasma assisted treatment directly before deposition as well as the deposition parameters influence the properties of the duplex treated steel. The adhesion can be improved essentially by an interlayer system Ti TiNx-TiN. Special consideration was given to the investigation of the adhesion and wear resistance of the composites. The wear tests include investigations of the sliding and abrasive wear. The results of testing rolling butt contact shows that a TiN layer on nitrided steel has no essential influence on the fatigue limit. Keywords: Nitriding; Titanium nitride; Wear resistance; Corrosion resistance; Fatigue
1. Introduction For more than seven decades, nitriding has been employed industrially under wide circumstances for the improvement of wear, corrosion resistance and fatigue limits of constructional parts. In comparison with other technologies it is distinguished by a great multiplicity of applications. However, in the case of very strong attack by wear and corrosion nitrided cases do not show sufficient resistance. By an additional protective layer, such as a hardcoating, the load bearing capacity can be further improved [ 1 ]. The sensitivity of nitrided parts towards tribological, chemical and electrochemical attack is controlled mainly by the structure of the compound layer, whereas their behaviour under cyclic, mechanical and thermal loads depends predominantly on the structure of the precipitation layer . The low fracture toughness of the porous compound layers is an important disadvantage especially under dynamic load and for the support o f a hardcoating. Nitrided cases of a depth greater than 0.3 m m and high surface hardness sufficiently support the hardcoating. The high hardness and internal compressive stresses of the case formed by nitriding lead to a high resistance to volume and rolling contact fatigue. The good tempering resistance of the nitrided case allows deposition up to nitriding temperature . 0257-8972/95/$09.50 © 1995 ElsevierScience S.A. All rights reserved SSDI 0257-8972(95 )08366-9
To attain a relevant increase in the fatigue limit and rolling fatigue limit the depth of the nitrided case must be approx. 0.4-0.5 mm, as is shown. For this higher depth of the nitrided case it is necessary to nitride the steels for longer times ( > 2 0 h). With regard to those process times it is advantageous to separate the nitriding process from the hardcoating process. The hardcoating, for instance TiN, after nitriding provides a very hard, wear, heat and chemical resistant outer layer. Thus properties obtained by the combination of nitriding with hardcoating allow function sharing between the core material, the hardened case and the surface, which is of special interest for application in complex stressed machine components. The purpose of the present investigations was to study the influence of the pretreatment of the substrates by controlled gas nitriding and plasma nitriding as well as the subsequent steps on the properties of duplex treated low alloy steels. Special consideration was given to the production of nitrided substrates with a defined structure, which fulfil the requirements for a sufficiently high contact fatigue limit support layer with a higher hardness depth (>0.3 mm) as well as a good adhesion of TiN.
2. Experimental Hardened and tempered specimens of the nitriding steel 31CrMoV9 (360 HV10) with approx. 2.5% Cr were
H.-J. Spies et al./Surface and Coatings Technology 74-75 (1995) 178 182
ground up to a roughness R z = 0 . 6 - 0 . 8 jam (Ra<0.1 jam). The nitriding of the specimens was performed in industrial plants by controlled gas nitriding (Table 1) and pulse plasma nitriding to form nitrided cases with a nitriding depth greater than 0.5 m m and surface structures with thin and without compound layers. For plasma nitriding it was not possible to produce bright nitrided layers with hardness depths greater than 0.35 mm, so a nitrided layer with a thin compound layer were used for the further investigations (Table 2). The bright nitriding was realized by a two stage technology: (1) activation of the surface for nitrogen adsorption (nucleation of iron nitrides on the surface); (2) nitriding with a lower nitriding potential and lower nitrogen content in the plasma to inhibit the growth of a compound layer in the outer case. In some cases the nitrided samples were ground or polished before hardcoating to remove the nitride cover layer, the porous zone or the whole compound layer before deposition. The TiN was deposited by ion plating (hollow cathode discharge evaporation at 350 °C) with a coating thickness of approx. 2 jam (Table 2). A sputter cleaning by argon ion b o m b a r d m e n t was carried out before hardcoating. In the first stage of the deposition (5 min) a thin intermediate titanium layer was produced. The influence of the sputter cleaning in argon on the surface morphology of the different nitrided substrates before hardcoating is described in Ref. . The structure was investigated using optical microscopy, scanning electron microscopy (SEM), X-ray diffraction (XRD) and glow discharge optical spectroscopy (GDOS). Hardness-depth profiles and surface Table 1 Process parameters of the gas nitriding Temperature (°C} Nitrided with compound layer Bright nitridedb
Nitriding time (h) 32 (20') 44
" Ratio of the partial pressures p(NH3)/p(H 3/2) for gas nitriding. b Bright nitriding using a two-stage technology.
hardness were measured by Vickers microhardness testing and nanohardness measurement (50-1000 m N load) with registration of the indentation depth. The scratch test was chosen to evaluate the coating adhesion. The wear resistance of nitrided and duplex treated samples was tested using a plane-on-roll testing machine. First results of the behaviour under rolling contact were obtained by means of a double-disk testing machine.
3. Results and discussion 3.1. Structure and adhesion
The growth and structure of nitrided layers is influenced by the nitriding temperature, gas mixture for nitrogen adsorption and the alloying elements of the steel. As a result of the investigations of the surface morphology and the scratch tests of duplex treated low alloy steels it can be established that the structure, and especially the adhesion, of the TiN hardcoating depends on the structure of the nitrided case. Defined reproducible nitrided layers with a varied structure of the hardened case can be produced by plasma and controlled gas nitriding (Table 3). At the beginning of the nitriding process nuclei of iron nitrides are formed at the surface after the saturation of the :~-Fe with nitrogen. During nitriding they preferentially grow in a lateral direction to form a closed top layer. Recent investigations of the adhesion of this top layer by scratch tests and hardness indentations on gas nitrided surfaces have shown an extensive flaking-off of this cover layer from the orignal surface or the compound layer [-4-6]. The nitride top layer on plasma nitrided steels with compound layer is not so strongly developed as on gas nitrided samples. It has better adhesion as the cover layer on gas nitrided steel. Essentially smaller areas of failure or no delaminations were observed. In scratch test or sometimes even during deposition spalling of the hardcoating occurs because of the bad adhesion of the nitride cover layer to the original steel surface. Subsequently hardcoating of the top layers results in an inhomogeneous growth of the TiN and a coarse columnar structure of the hardcoating.
Table 2 Parameters of the plasma nitriding and TiN hardcoating
Nitriding (48 h} TiN depositiona (40 mini
Substrate temperature CC)
Effective current (A)
" Hollow cathode discharge evaporation with Ti or Ti TiNx interlayer. b Pulsed bias voltage -50 V.
Gas mixture (%) Ar
Pulse on/off (its)
H.-J. Spies et al./Suffaee and Coatings Technology 74 75 (1995) 178-182
Table 3 Properties of the nitrided substrates Treatment
Depth of nitrided case (mm)
Maximum case hardness HV 0.3
Structure of the compound layer (thickness)
Gas nitrided with compound layer Bright nitrided Plasma nitrided
0.53 0.51 0.46
830 710 840
,,,' (5 gm) none ;,' (4 pm)
The problem of the decomposition of the outer part of the compound layer into soft c~-Fe by ion bombardment especially during the sputter cleaning was described in earlier works (Fig. 1). In this case the supporting effect of the compound layer for the hardcoating is lost . Therefore in further investigations bright nitrided substrates and substrates with a pre-polished surface were used. The sputter cleaning by pure argon ion bombardment leads to an incomplete removal of the nitride cover layer. This results in a macroscopic inhomogeneous surface of the hardcoating. Scratch tests on TiN deposited bright nitrided 31CrMoV9 without mechanical pretreatment show higher critical loads for adhesive failure than on polished samples (Fig. 2). This may be an effect of the morphological influence of the iron nitride cover layer on the TiN growth, i.e. its reaction in the scratch test. The scratch test results on hardcoated samples with compound layer show better adhesion after polishing of the nitrided surface. Further increase of the adhesion of
i 120 I ' 98 100 "~
6o 40 20 2 0
Fig. 2. Results of the scratch test on coated samples with Ti/TiN (Lcl, first adhesive failure at the edge of the scratch trace). Substrate variation: A, plasma nitrided; A1, plasma nitrided and polished; B, gas nitrided with compound layer; B1, gas nitrided and polished; B2, gas nitrided and ground; C, gas bright nitrided; C1, gas bright nitrided and polished.
TiN is found after the removal of the compound layer by grinding (Fig. 2). The highest critical loads in the scratch test are attained on bright gas nitrided steels, which have a very fine (small nuclei) and dense nitride cover layer structure. An essential improvement of the adhesion, more than by a single Ti intermediate layer (Fig. 2), is attained by a gradient Ti-TiNx interlayer (Fig. 3). The deposition of this Ti-TiN x intermediate layer reduces the hardness gradient and residual stress
10(I 77 Z~ 8O
Z 6o "~
Fig. 1. Scanning electron micrograph of duplex treated layer with a defect area in the compound layer-Ti interlayer as a result of denitriding during sputter cleaning.
Fig. 3. Critcal loads of the scratch test of TiN coated samples with a gradient Ti TiN~ interlayer (variations as in Fig. 2).
H.-J. Spies et aL/Surface and Coatings Technology 74-75 (1995) 178-182
gradient between the substrate and the stoichiometric TiN. 3.2. Wear behaviour
The sliding wear behaviour of various treated samples was charactarized in a plate-on-roll test. The nitrided and hardcoated samples (plates) were tested under dry sliding conditions in linear contact with a rotating roll of hardend and tempered X155CrMoV121 (hardness 58 HRC). Fig. 4 exhibits the decrease of the sliding wear by deposition of TiN on the nitrided-steel in comparison with nitrided-only steels and a SiCN coated-sample on the same pretreated steel. The higher linear wear on the nitrided 31CrMoV9 with compound layer is caused by the porous zone of the compound layer (wear depth < 2 rtm). The wear on non-coated surfaces was intensified by the formation of iron oxide. The presence of iron oxides was shown by XRD. This tribochemical reaction takes place between test piece and the roll. It results in an increase of the wear by the abrasive effect of the iron oxide. Oxidation on hardcoated surfaces only occurs if the hardcoating is locally removed in the wear trace (Fig. 5). By measuring of the tangential force during the sliding contact a considerable decrease of the friction on duplex treated samples in comparision to the non-treated and nitrided-only steel was proved. The fatigue limit under rolling contact, tested in a double disk machine, is obviously little influenced by the hardcoating (Fig. 6). Maximum stresses under rolling contact are formed below the surface• The thin layer of titanium nitride has no essential influence on the fatigue limit. The higher fatigue limit of nitrided steels is a result
Fig. 5. Tribochemical reaction with forming of iron nitride in the wear trace: local increase of removal by abrasion.
double disk contact -t-.test
. 4.34/ 4.418 ~!i ,~Ioo '.
plane-on-roll test 2280 \\\\\
F= 100 N v= 0,04 m/s
X155CrMoV121 no lubricant
"\\'\" ,~\\,X ,\\\, *
\\\\, \\\,,\ \\\\"
with with compound compound layer layer & TiN
\\\\\ \\\\" •~
_ .' _ // bright bright nitrided nitrided & TiN
Fig. 6. Results of testing rolling butt contact.
bbb:?: ~--2~." , : , .:... ~.~ a
Fig. 4. Sliding wear of nitrided and duplex treated 31CrMoV9 in the plane-on-roll test: a, non-nitrided; b, nitrided with compound layer; c, bright nitrided; d, bright nitrided/TiN; e, bright nitrided/polished/TiN; f, bright nitrided/polished/SiCN.
of the increase of the hardness and the formation of compressive residual stresses in the hardened case. Pitting, indicating the test piece failure, starts in the subsurface. The bright nitrided 31CrMoV9 shows a lower fatigue limit than the samples with compound layer. The long nitriding time, necessary to obtain the hardness depth of approx• 0.5 mm, leads to a decrease in surface hardness and maximum compressive stress because of aging effects.
H.-~L Spies et al./Surfiwe and Coatings Technology 74 75 (1995) 178 182
The duplex surface engineering (nitriding+hardcoating) of steel grade 31CrMoV9 leads to advantages important for its use on complex stressed machine components. Nitriding results in a considerable increase of hardness and the formation of compressive residual stress. The nitrided case supports the hardcoating sufficiently. It reduces the hardness and stress gradient between substrate and TiN, too. However, it must be taken into consideration that the adhesion of TiN is influenced by the structure of the outer part of the nitrided layer and the formation of a nitride cover layer. Bright nitriding or the removal of the nitride cover layer and of the porous zone of the compound layer should be carried out, to guarantee good adhesion of the hardcoating. The adhesion can be further improved by deposition of a gradient Ti-TiNx intermediate layer. Hardcoating of nitrided steel reduces the friction and sliding wear essentially. Tribochemical reactions are reduced. The fatigue limit in rolling contact is increased by nitriding. Hardcoating has no essential influence on the wear behaviour in rolling contacts.
The financial support for this investigation by the Deutsche Forschungsgemeinschaft (DFG Nr. Sp 376/7-1) is gratefully acknowledged.
References  Y. Sun and T. Bell, Mater. Sci. Eng., A140 (1991) 419-434.  H.-J. Spies, H.-P. Winkler and B. Langenhahn, Hiirterei-Techn. Mitt., 44 (1989) 75 82.  H.-J. Spies, K. H6ck, E. Broszeit, B. Matthes and W.Herr, Surf. Coat. Technol., 60 (1993) 441 445.  N. Van Stappen, C. Quaeyhaegens, L. Stals, J.R. Roos and J.P. Ceils, Proc. IPAT '91, Brussels, May 1991, CEP Consultants, Edinburgh, 1991, pp. 208 213.  A.Yang, Y. Yufei, G. Xiang and L. Peng Xing, Proc. Heat Treat. Surf. Eng. A S M Int. (1988) 43-48.  M. Zlatanovic and T. Gredic, Mater. Sci. Forum, 102-104 (1992) 655-666.