Friction and wear of a nitrogen implanted austenitic stainless steel

Friction and wear of a nitrogen implanted austenitic stainless steel

216 Nuclear Instruments and Methods in Physics Research B19/20 (1987) 216-220 North-Holland, Amsterdam FRICTION AND WEAR OF A NITROGEN IMPLANTED S. ...

1MB Sizes 1 Downloads 12 Views

216

Nuclear Instruments and Methods in Physics Research B19/20 (1987) 216-220 North-Holland, Amsterdam

FRICTION AND WEAR OF A NITROGEN IMPLANTED S. F A Y E U L L E

AUSTENITIC STAINLESS STEEL

a n d D. T R E H E U X

Laboratoire de Metallurgie, UA CNRS 447, Ecole Centrale de Lyon, BP 163, 36, avenue Guy de Collongue, 69131 Ecully Cedex, France

Friction and wear behaviour of nitrogen implanted stainless steel 304 (18%Cr, 10%Ni) has been studied for different implantation conditions. Samples were implanted with 40 keV ions and fluences were between 1016 and 6 × 1017 ionscm -2. Friction and wear behaviour depends strongly on the implantation conditions and it is shown that only with some fluences (1017, 2 × 1017 ionscm -2) noticeable improvements are achieved. In order to understand the different mechanisms, TEM examination and grazing X-rays diffraction of the implanted layers are presented. The martensitic transformation of the matrix and the formation of nitrides are analysed. The formation of a hardened superficial layer responsible for good wear conditions is shown to depend on the kind of phases formed and on the microstructural morphology of these phases.

1. Introduction

Type 304 austenitic stainless steel, implanted with nitrogen ions, has been studied for many years as well for microstructures as for friction and wear properties. Superficial implanted layers are henceforth rather well characterized. Nitride formation and martensitic transformation have been analysed by several research groups [1-7]. Concerning wear resistance, a highly beneficial effect is generally achieved [8-10] but correlation with microstructures is still not clear. Two main hypotheses have been put forward: surface hardening by nitride precipitation [1,10], or stabilization of the austenitic phase towards the martensitic transformation [11,12]. In a previous paper [13], transmission electron microscopy (TEM) examination of implanted 304 steel was presented. This paper completes these results with grazing X-rays diffraction and tribological tests.

beam and the sample surface was varying from 0.5 ° to 0.9 ° . F o r each incident angle, the penetration depth can be calculated using the Fresnel formula [14]. These penetration depths are very small (from ¢ = 23 nm, when a = 0.5 ° to ¢ = 60 nm when ct = 0.9 °) because ~ae X-rays are highly absorbed. The iron fluorescence radiation was suppressed by using a silicon-lithium detector. Tribological tests were made using a tribometer which allows a cylinder-flat-pin contact. Friction coefficients and weight losses of the samples (with an uncertainty of 10 -4 gf) were measured. The tests were made in dodecan. The rotation speed of the cylinder (steel 42CD4: 0.4%C, 1%Cr, 0.2%Mo, hardness = 38 HRc) was 0.035 m s - i and the translation speed of the flat pin (304 steel) was 3 × 10-3 m s-1. The normal load was 50 N (contact pressure = 102 MPa).

3. Results 3.1. Microstructural characterization

2. Experimental procedure 3.1.1. T E M

Samples of type 304 austenitic stainless steel were implanted on one side after being electrolytically polished with 40 kev nitrogen ions by using the isotope separator of the Institut de Physique Nucleaire de Lyon. The ion fluences were between 1016 and 6 × 1017 ions cm -2 and the current density ranged from 10 to 20/~Acm -2. The temperature of the samples was maintained during implantation at room temperature using a water circulating device. Samples for TEM examination were prepared as previously described [13]. Grazing incidence X-ray diffraction technique (X-radiation Cu K~ wavelength ?~= 0.154 tam) was also used to analyse implanted samples. The incident angle a between the 0168-583X/87/$03.50 © Elsevier Science Publishers B.V. (North-Holland Physics Publishing Division)

The results of TEM examination of samples implanted with different fluences have already been presented [13]. The main results are the following: - for lower fluences (1016 and 5 × 1016 ionscm-2), the main effect of implantation is the appearance of c martensite; - n i t r i d e s (hexagonal and orthorhombic types) are formed beyond 1017 ions c m - 2; - for 2 x 1017 ions cm-2 implantation, a very compact layer is formed. It is mainly composed of oriented nltrides with a small amount of a'-martensite; - f o r higher fluence, the superficial layers are more defective.

S. Fayeulle, D. Tr~heux / N implanted austenitic stainless steel

217

"',:,, a =0.9 =

0.9

=oo, tA ~

J,,,A

at = 0.7

11o~

Ilo~ lol

at = 0.5

,

a0

is

z2

za

ao

26

z2

le

,

,

,

ao

,

~6

,

,

22

,

,

le

O (degree) Fig. 1. X-rays diffraction spectra of implanted samples at different glancing angles a = 0.5 °, 0.7 ° and 0.9 °. (a) 10 z7 ionscm -2, (b) 2 × 1017 ionscm -2, (c) 6 × 1017 ions cm -2.

1017 tons.cm -2 '7

~

g t~ 0.5,

~

o

o-

250W

0

//

°"

30

°

60

°-

9'o

1~,o

2 X 1017 lons.cm-2

FRICTION TIME (MINUTE)

Fig. 2. Weight loss of 304 steel as a function of friction time for nonimplanted (e) and implanted (v 5 × 1016; © 1017; O 2 × 1017; A 6 × 1017 ionscm -2) samples.

2gOU

,oF-"

6 X 1017 tons.cm-2

3.1.2. Grazing incidence X-ray diffraction Fig. 1 gives the diffraction spectra obtained for different fluences (1017, 2 × 1017 and 6 x 1017 i o n s c m -2) a n d three incident angles (0.5 ° , 0.7 ° , and 0.9°). After a fluence of 1017 ions c m - 2 , the main phenomena - b r o a d e n i n g of ),-phase peaks (in comparison with the n o n i m p l a n t e d samples) indicating variations in the composition of the austenJte. This is mainly visible for a = 0.5 ° a n d a = 0.7°;

250U

,oF-"

a r c :

Fig. 3. Profiles of wear tracks after wear in implanted 304 samples. II. METALS

218

S. Fayeulle, D. Tr$heux / N implanted austenitic stainless steel

- formation of a ' maxtensite detected only in the near surface region for a = 0.5 ° ( , = 23 rim). On spectra achieved with samples implanted with 2 × 1017 ions c m - 2, several phases are identified: - a ' martensite, essentially for a = 0.5°; - n i t r i d e s ( d = 0 . 2 3 8 rim, d = 0 . 2 2 rim, d = 0 . 2 1 2 rim) hexagonal or orthorhombic (ASTM data of ~'Fe2N, Cr 2 (C, N) and (Cr, Fe) 2N1 - ~); - 3' austenlte in very small quantity, for a = 0.9 ° (seen by 200 reflexion). After a fluence of 6 x 1017 ionscm -2, a ' martensite is detected in a great amount. A very intense reflexion occurred for d = 0.234-0.236 nm, which can be fitted with cFe2N 110 reflexion. 3.2. Tribological tests

....

=

Fig. 4. Optical micrographs of worn surfaces. (a) 1017 ionscm-2; (b) 2 x 1017 ionscm-2; (c) 6 × 1017 ionscm -2.

/ -T E

•/

/

10

N o significant change in the friction coefficient was detected. F o r all tiuences, it was about 0.2. Fig. 2 gives wear results obtained for 304 steel implanted at various fluences. N o effect is seen for the lower (5 x 1016 i o n s c m - 2 ) and the higher fluence (6 × 1017 ionscm -2) values. Wear resistance is most increased after implantation of 2 × 1017 ions c m - 2. Profilometer traces of,wear tracks (fig. 3) confirm the better performance of the 2 × 1017 N cm -2 implanted samples. The very good surface state of these samples is due to the formation of a superficial film as seen on fig. 4 (micrographic observation of the surface after two hours of friction). Fig. 5 shows the evolutions for longer tests (t = 100 h). The effectiveness of nitrogen implantation occurred essentially during the first twenty hours. Nevertheless, the wear rates of 1017 and 2 × 1017 implanted samples remained slightly lower after twenty hours of friction. SEM observations of the worn surfaces clearly show that at the b e # n n i n g of the friction test (fig. 6a), abrasive wear and plastic flow of metal occurred for nonimplanted samples when only small plastic deformation was seen on the 2 × 1017 ions cm-2 implanted surface. F o r longer tests (fig. 6b), SEM observations on the nonimplanted samples show the formation of a strain hardened layer that decreased the wear rate.

E

3

4.

Discussion

I-

4.1. Microstructure

0.5

1

5

10

50

100

FRICTION TIME (HOUR)

Fig. 5. Weight loss of 304 steel as a function of friction time for nonimplanted (O) and implanted (© 1017; 12 2 × 1017; ,x 6 x 1017 N cm -2) samples.

F o r fluences of 1017 and 2 × 1017 N cm -2, results obtained by T E M and by X-ray diffraction axe coherent. Nitrides which are detected by TEM as soon as a fluence of 1017 ionscm -2 is reached, are surely too small to be detected by X-ray diffraction. This fact has already been noticed by Moncoffre et al. At a fluence of 2 × 1017 ionscm -2, the 002 reflexion of nitride is not

S. Fayeulle, D. Tr~heux / N implanted austenitic stainless steel

Fig. 6.

219

electron micrographs of worn surface: (al) nonimplanted sample t = 15 mn; (a2) 2 x 1017 N cm - 2 implanted sample t = 15 mn; (bl) nonimplanted sample t = 5 h; (b2) 2 x 1017 N cm- z implanted sample t = 5 h.

Scanniog

clearly visible (it is hidden at the beginning of the broad reflexion). This can be explained by a preferential orientation of the nitride phases similar to that observed by Moncoffre et al. [15,16] in low alloyed steel. Austenite reflexions are hardly visible, even for a = 0.9 °, which confirms the TEM observations of a very regular superficial layer composed only of nitrides and a small amount of martensite. The situation is more ambiguous after an implantation with a fluence of 6 × 1017 ions cm-2. TEM shows that the same kind of nitrides are formed (i.e. orthorhombic phase) than after implantation of 2 × 1017 ions cm-2, but now, the superficial layer is made of a mixture of austenite, martensite and apparently ntisoriented nitrides. The very intense reflexion ( d = 0.234-0.236 rim) does not reflect this defective aspect of the layer and on the contrary seems to indicate a strong preferred orientation of the nitrides. A more complete study of this point is in progress. 4.2. Friction and wear

The importance of the fluence on wear resistance, which has already been observed on 304 steel [9,10] or on other materials [17] can be explained by the microstructures after implantation.

The hardening phases achieved by nitrogen implantation are a' martensite and nitrides. The concentration of these phases is too small for the lower fluence (1017 ionscm-2). On the contrary, at a fluence of 2 × 1017 ionscm -2, a very regular layer of nitrides is obtained which prevents severe wear to occur immediately. The texture of the nitrides is probably also partly responsible for the good wear behaviour. The bad wear resistance of samples implanted with 6 × 1017 ions cm -2 shows that the kind of phases formed is not the only factor responsible for an improvement of wear resistance. The morphology of superficial layers is also very important, and a defective layer where austenite, martensite and nitrides (and maybe blisters [7,18]) are mixed is not beneficial. Correlation of wear versus time behaviour with phases appearing at different depths is difficult because the kind of the superficial layers (composition, structure, and soon) can be modified by friction. So, further interest has to concentrate on the evolutions of layers during wear. In a forthcoming paper TEM observations will be presented on this problem. The authors wish to thank M.A. Plantier for carrying out the implantations. This work was partly funded by the Ministere de la Recherche et de la Technologie. II. METALS

220

S. Fayeulle, D. Tr~heux / N implanted austenitic stainless steel

References

[1] F.G. Yost, S.T. Picraux, D.M. Follstaed, L.E. Pope and J.A. Knapp, Thin Solid Films 107 (1983) 287. [2] R.G. Vardiman, R.N. Bolster, I.L. Singer, in: Metastable Materials Formation by Ion Implantation, eds., S.T. Picraux and W.J. Choyke (Elsevier, New York, 1982) p. 269. [3] J.L. Whitton, G.T. Ewan, M.M. Ferguson, T. Laursen, I.V. Mitchell, H.H. Plattner, M.L. Swanson, A.V. Drigo, G. Celotti, M. Servicori and W.A. Grant, Mater. Sci. Eng. 69 (1985) 111. [4] I.L. Singer and J.S. Murday, J. Vac. Sci. Technol. 17 (1980) 327. [5] M. Baron, A.L. Chang, J. Schreurs and R. Kossowsky, Nucl. Instr. and Meth. 182/183 (1981) 531. [6] W.M. Bone, R.J. Colton, I.L. Singer and C.R. Gossett, J. Vac. Sci. Teclinol. A2 (1984) 788.

[7] D.C. Kothari, M.R. Nair, A.A. Rangwala, K.B. Lae, P.D. Prabhawalkar and P.M. Raole, Nucl. Instr. and Meth. B7/8 (1985) 235. [8] W.C. Oliver, R. Hutchings and LB. Pethica, Metall. Trans. A15 (1984) 2221. [9] P.D. Goode and I.J.R. Baumvol, Nucl. Instr. and Meth. 189 (1981) 161. [10] H. Dimigen, K. Kobs, R. Leutenecker, H. Ryssel and P. F_.ichinger, Mater. Sci. Eng. 69 (1985) 181. [11] I.L. Singer, Appl. Surf. Sci. 18 (1984) 28. [12] A. Cavalleri, L. Guzman, P.M. Ossi and I. Rossi, Scripta Metall. 20 (1986) 37. [13] S. FayeuUe, D. Tr~heux and C. Esnouf, Appl. Surf. Sci. 25 (1986) 288. [14] M. Brunel and F. de Bergevin, Acta Cryst., to be published. [15] N. Moncoffre, Th6se de doctorat, Lyon (1986). [16] N. Moncoffre, M. Brunel, P. Deydier and J. Tousset, Surface and interface analysis (to be published). [17] S. FayeuUe, Wear 107 (1986) 61. [18] I.L. Singer, Vacuum 34 (10/11) (1984) 853.