an assessment of the effect of plasma nitriding pc borges,^ ae ... · n] contents vary between 15%...

An assessment of the effect of plasma nitriding

and nitrocarburizing parameters on the

corrosion resistance of sintered steel using the

response surface method

PC Borges, AE Martinelli, A.N. Klein, C.V. Franco

DAMEC-CEFET, Curitiba, PR, BrazilEmail: borges@damec, cefetpr.br

^UFRN-UFRG, Department of Physics, LagoaNova Campus,Natal, RN, Brazil

Email: [email protected]

Depto. de Engenharia Mecanica, UFSC, Florianopolis, SC> BrazilEmail: [email protected]

Email: franco@materiais. ufsc. br

Abstract

The Response Surface Method is an effective means of addressing the combinedeffect of plasma nitriding parameters on the corrosion resistance of materials. Inthis work, sintered steel was nitrided under different atmospheres: 75%N: -25%H2, and 25%N: - 75%Hi. In addition, a series of samples was alsonitrocarburized by adding CH4 to the nitrogen-enriched atmosphere. Thethickness and composition of the resulting compound layers were determined,along with the resulting microhardness profiles. Finally, the corrosion resistanceof the plasma-treated materials was evaluated by potentiodynamic polarizationtechnique. The results indicated that the addition of CH^ to the plasmaatmosphere increased the hardening depth and improved the corrosion rate of theresulting material.

1 Introduction

Powder metallurgy (PM) has been widely used in large scale production of steelcomponents, lowering the final cost of production for mechanical, automotive,and appliance components [1]. In addition, specific parts, such as metallic filters

Transactions on Engineering Sciences vol 25, © 1999 WIT Press, www.witpress.com, ISSN 1743-3533

182 Surface Treatment

and self-lubricating bearings can only be manufactured by powder metallurgy.Therefore, increasing efforts have been made towards improving theperformance of sintered steels. Residual porosity is particularly deleterious to themechanical and chemical behavior of powdered materials in applications thatdemand fatigue, wear and corrosion resistance. Although the recent developmentof stainless steel powdered alloys has provided a means of producing corrosion-resistant sintered materials [2-7].

The high price of raw materials cancels out the economic benefits of PMprocessing. Surface-modification techniques can be used to improve thecorrosion resistance of sintered steels. Surface-modifying methods include avariety of processes such as surface spraying, surface quenching, andnitrocarburizing, in addition to more sophisticated approaches where surfaceheating is achieved using laser beams, plasma, electron beams, or ion beams inbiased potentials [8-9]. The thickness and properties of the coating or modifiedlayer produced on the surface of the material vary widely as a function of boththe technique and corresponding parameters. For instance, the thickness ofcorrosion-protective barriers may vary from tens of nanometers, as in the case ofion implantation, to a few millimeters, as for laser-generated surface alloys[9-10]. Furthermore, from a mechanical standpoint, the hardness of coatinglayers range from around 200 Vickers (surface spraying) up to 10,000 Vickers(diamond deposition) [11]. Consequently, each method has its own advantagesand disadvantages associated, according to the final application intended for thematerial. This work addresses the use of plasma nitriding and nitrocarburizing tomodify the surface of sintered steel. The set of resulting properties includingmechanical (fatigue resistance and hardness), tribological (friction coefficient),and electrochemical (corrosion rate) are determined by the structure andcomposition of the compound layer that coat the metallic substrate. However, anequally important aspect is the distribution of nitride and carbonitride phases inthat layer, which is basically determined by both processing parameters such asplasma atmosphere, temperature, holding time, and cooling rate [12-19], and thenature and contents of alloying elements (especially carbon) in the base steel[20-22]. Although many aspects remain to be investigated, a few trends havebeen well established in previous studies. Plasma atmospheres containing lessthan 5% N] result in excessively thin compound layers. Thicker compound layershaving y'-Fe4N as their major phase are usually formed in atmospheres where theN] contents vary between 15% and 30% [16-18]. Atmospheres richer in N](more than 60%) and temperatures below 480°C drive the formation of s-Fe .sN.The relative concentration of the s-phase increases with the addition of carbon tothe nitriding atmosphere, resulting, in fact, in a nitrocarburizing process. Similarstructures can be formed from conventional methods such as salt bath and gasnitriding [13, 23]. Nevertheless, plasma is an environmentally safe process, inthe sense that no salt residues are exuded upon processing. However, recentreports on the superficial treatments, such as plasma nitriding, are not effectivein improving the corrosion resistance of materials with residual open porosity, asvolume remains exposed to the electrolyte [20-22]. The presence of even a smallnumber of uncoated pores may strengthen the effects of a corrosive action, as thepotential difference between substrate and nitride layer represent an effective


Surface Treatment 183

galvanic pair [24]. The optimization of plasma processing, as applied toimproving the corrosion behavior of sintered steels, is an important tool inbroadening the application range of metallurgical powdered components.

In this scenario, the aim of the present study is to compile basic informationas to the combined effect of plasma processing parameters on the microstructureof nitrided layers as well as on the mechanical and chemical behavior of thesurface-modified materials.

2 Materials and methods

2.1 Material sintering

The chemical composition of the powders used to produce sintered 3.8 wt.% Mn- 0.4 wt.% C - bal. Fe alloys is described on Table 1. The alloying elements wereadded to the ferrous matrix in the form of a FeMnC master alloy, whose particleswere all finer than 45 um. The final carbon content was adjusted adding 0.08wt.% graphite to the powder mixture. Finally, 0.6 wt.% zinc stearate was added,and the final composition was mixed in a Y-mixer for 90 min at 45 rpm.

Table 1: Composition of the starting powders (in wt.%)

PowderFe Ancorsteel 1000BMaster alloy

C<0.016.5

O0.09-

Si<0.010.22

Mn0.176.05

P0.0050.25

S0.0090.01

Febal.bal.

The powder mixture was then uniaxially pressed under 600 MPa into 9.5 mmpellets using a double acting press equipped with a moving die body.Subsequently, the green pellets were pre-sintered at 600°C in 90% H?+10 %CH4. The furnace was heated up to the pre-sintering temperature at a rate of30°C/min, and the holding time was set to 1 hour. The specimens were thenpressed once more under 600 MPa using a 10-mm die, and the resulting pelletswere sintered at 1220°C under the same atmosphere employed in the pre-sintering stage. Finally, the material was either nitrided or nitrocarburized in aplasma reactor, as described in early report [25]. A series of samples were notplasma treated in order to establish grounds for a comparative analysis on theefficiency of the surface treatment.

2.2 Plasma nitriding

Prior to plasma surface treatment, the pellets were cleaned with acetone inultrasound bath for 2 minutes. The cleaning step was completed inside theplasma reactor, where the material was exposed to a discharge in % flow (2.0cnrYs) at 300°C for 15 minutes. At that point, the pressure in the chamber wasset to 200 Pa, and the applied potential ranged from 400 to 450 V. Table 2presents a full description of the processing parameters used in the final plasmatreatment.



Table 2: Set of plasma processing parameters

Temperature

Holding time

Atmosphere

FlowPressure

VoltageCooling atmospherePulse time on

Pulse time off

510°C(783K);2, 4, and 6 h

75% N: + 25%

4cm'/s(4x 10"400 Pa400 to 610V

No80 to 120 ms80 to 120 ms

540°C(813K), 570°C(84.

)%% + 1.25%CH4;(atm 1H? (atm 2)Ho (atm 3)"m/s)

3K)

)

2.3 Microstructural and mechanical characterization

Both optical and electronic microscopy were employed in the microstructuralanalysis of the compound layers. A Zeiss Neophot 30 optical microscope and aPhilips XL3O scanning electron microscope were used to inspect polished andetched cross sections of the plasma-treated surfaces. Phase identification tookplace by X-ray diffraction using Cu-kot radiation (k = 0.154060 nm). Diffractionpatterns were gathered in a Philips X'Pert equipment set to 30 mA and 40 kV.The angular range 30° < 28 < 100° was scanned at a goniometer speed ofo.o r/s.

Microhardness profiles were obtained from a Shimadzu equipment. Threeidentical samples were used for each set of plasma-processing parameters. Eachhardness value reported corresponded to an average of 5 indentations caused by aload of 0.025 kg applied onto the sample for 15 s. Marks were 30 fim apart.

2.4 Electrochemical analysis

Electrochemical analyses were carried out at ambient temperature (22 ± 2°C)using a EG&G-Princeton Applied Research Potentiostat/Galvanostat 273 Aconnected to a personal computer. The electrochemical cell was formed by athree electrode cell system. The working electrode consisted of the sample itselfhaving a total area of 0.283 cm". Saturated Calomel Electrode coupled to aLuggin capillary functioned as the reference electrode. Finally, the counter-electrode consisted of two inert graphite bars. The measured voltage was laternormalized to a Hydrogen Normal Electrode (HNE) [26]. The electrolyteconsisted in a KNOg 0.5 M solution (pH = 6) prepared from highly pure KNO]and pure water, obtained from a MILL1PORE system. Each scan lasted 3600 s.The bias potential of the working electrode ranged from -250 mV x EQC to 1.6 V,and the scan rate was set to 0.8 mV/s.



2.6 Statistical analysis

The influence of raw materials and sample preparation, as possible sources ofdiscrepancies, was ruled out as the starting powders came from single batches,and the preparation methodology was rigorously the same for all samples. Thestatistical analysis was designed based on a factorial model with a central point,such as

y = ao

The method of minimum squares was then used to plot the response surfacefor the thickness of the compound layer and corrosion rate [27].

Automated variance analysis and surface fitting was carried out using anappropriate software.

3 Results and discussion

3.1 Thickness of the compound layer and hardening depth

The response surfaces corresponding to the thickness of the compound layer(Figure 1) are inclined with respect to their reference axes. Under CH^freeatmospheres (Figure Ib and c), increasing the nitriding holding time from 2 to 6hours at 570°C did not result in significantly thicker compound layers.Conversely, a similar increase of the holding time at 510°C, resulted inreasonably thicker layers for both atmospheres. The saturation of the growth rateof the compound layer for shorter times at higher temperatures is possibly relatedto an increase of the diffusion coefficient of nitrogen in iron, as the temperatureincreases. Further evidence confirming this hypothesis has been reported in theliterature [14,25]. In addition, Figure 1 also shows that thick compound layerscan alternatively be obtained increasing the relative amount of No in theatmosphere. Other than that, all response surfaces depicted in Figure 1 are quitesimilar.

An analysis of the response surface obtained from nitrocarburized specimens(Figure la) indicated the growth of the compound layer as both temperature andholding time increased. Figure la and b also suggested that the addition ofCH*to the atmosphere prompted the formation of reasonably thick compound layers.In addition, the presence of CFU in the atmosphere changes the holding time intoan important process parameter, significantly affecting the final characteristics ofthe nitrocarburized surface. This can be rationalized considering that thediffusion coefficient of carbon in iron is higher that that of nitrogen for thetemperature range investigated Therefore, the diffusion of nitrogen was enhancedadding carbon to the atmosphere, via a mechanism of mutual diffusion.



847

(a) 2LUQ.

783

CZD 0,067CZ3 0,192CZ3 0,317CZ3 0,443CZ3 0,568F~~71 0,693CZ3 0,818D3 0,943ma 1,068

1,194

TEMPO (h)

(b)813

783

847

(c)

783

| 1 2,103| 1 2,262I 1 2,420I 1 2,579I 1 2,737["""I 2,896[Z] 3,054CO 3,213^ 3,372ggg 3,530

TEMPO (h)

Figure 1: Response surface for thickness of the compound layer as a function ofthe nitriding temperature and holding time, (a) atm 1: 75%N2-23.75%H?-1.25%(%; (b) atm 2: 75%N2-25%Hi; (c) atm 3: 25%N]-75%H2.



3.2 Microhardness profile

Figure 2 depicts microhardness profiles that illustrate the effects of theprocessing parameters on such property. Increasing the nitriding holding timeincreased both the absolute hardness and the hardening depth of the material(Figure 2a). This was expected, since longer times imply in higher contents ofnitrogen in the diffusion zone. On the other hand, increasing the nitridingtemperature softened the material (Figure 2b). As the temperature decreased, sodid the diffusion coefficient of nitrogen in iron, and, consequently, theconcentration of nitrogen in the diffusion zone. Nevertheless, the average size ofnitride precipitates also decreased, at the same time that the number of sitesincreased. As a result, the precipitate particles packed together, thus inhibitingthe movement of dislocations which, in turn, hardened the material. Adding CFUto the plasma atmosphere also increased both the hardness and hardening depthof the material. Finally, the effect of increasing the relative amount of nitrogen inthe atmosphere was also towards hardening the material, as a result of anincrease in concentration and chemical potential of nitrogen.

(a)

(c)

PAPB

Distancia (urn)

(d)

100 150 200Distanaa (\.<m)

^800-

(b) I

100 200Distancia (pm)

Figure 2: Microhardness profiles showing the effect of (a) nitriding time; (b)nitriding temperature, (c) addition of CH^ to the nitriding atmosphere, and (d)NziHz ratio.



3.3 Composition of the compound layer

The composition of the compound layer was assessed by X-ray diffraction(XRD). Prior to surface treatment, the pattern corresponding to the sinteredsubstrate depicted only the presence of the BCC Fe-a structure. Although areasresembling cementite (8-FegC) had been previously observed by opticalmicroscopy, the presence of that phase was not confirmed by X-ray diffraction.Figure 3 illustrates the XRD patterns that followed plasma treatments. Inaddition to Fe-a, resulting from the interaction of the incident beam with themetallic substrate, peaks corresponding to nitride phases of composition y'-Feand G-Fe2_]N were also identified. Samples nitrided under ami 1 (containing CHLJrevealed the presence of a mixture of s-Fe2_3N and y'-Fe N (Figure 3a), exceptfor the sample treated at 570°C/2 h (PD), whose compound phase consistedexclusively of 6-Fe2_]N. The relative intensity of the diffraction linescorresponding to E-Fe2_]N and y'-Fe N revealed that the former was the majorphase resulted from the nitrocarburizing process. Samples labeled PF to PH,nitrided under atm 2, at relatively low temperatures (510°C and 540°C), alsodepicted high contents of c-Fe2_]N (Figure 3b). However, in this case, increasingthe nitriding temperature to 570°C enhanced the precipitation of y'- Fe^N. Thisresult is in good agreement with previous studies that pointed out an increase ofthe relative amount of y' as the nitriding temperature increased [13-19]. Figure3c shows the XRD patterns obtained from samples nitrided under atm 3 (labeledPK to PO). Reducing the relative amount of nitrogen in the atmosphere stabilizedy'- Fe N. As a matter of fact, y'- Fe4N was the major phase observed for allsamples nitrided under atm 3, except sample PL (510°C/6 h). In that case, thehigher content of G-Fe^-sN was most likely related to the excessively longnitriding time, responsible for depleting the substrate of carbon, which diffusedtowards the compound layer (Figure not shown here).

3.4 Corrosion resistance

In addition to the corrosion rate, estimated at the Tafel region of the curve,potentiodynamic plots also offered information on active dissolution, active-passive transition, passivation, and transpassivation.

Econ vs. time scans revealed that the plasma treatment shifted Econ- to noblervalues. Nevertheless, the rapid decrease of the potential as the nitriding holdingtime increased is an indicative of the beginning of the corrosion processs and thetendency of the material not to passivate. Samples nitrocarburized in thepresence of CH^ displayed a rather distinct behavior, with a slightly increase onthe potential, such as it was the case of sample PA.

Potentiodynamic polarization scans were performed on both sintered andsintered/plasma-treated specimens (Figure 4). Although nitrocarburizing reducedthe anodic dissolution current of the material (Figure 4a), an increase in thepassivation current of the plasma-treated material with respect to the plainsintered material could also be noticed. This indicated that, in the range of highpotentials, the oxide layer was more resistant to corrosion than the compound



layer. At the vicinity of the active-passive region, the performance of thecompound layer improved. A reasonable explanation to this fact is thebreakdown of the thinnest areas of the compound layer, resulting in theformation of a galvanic cell between substrate and nitride layer. The resultsobtained for the material nitrided under atm 2 and 3 are shown in Figure 4b andc, respectively. As a result of the nitriding process, the passivation regionbetween 0.7 and 1.6 V could no longer be observed. However, the region ofanodic dissolution of the nitrided material extended all the way to thetranspassivation region. Moreover, the current values were lower than thoseobtained for the passivation region of the sintered material (samples labeled P).

The values of the corrosion rate were interpreted from a statistical standpoint,resulting in the response surface depicted in Figure 5. As it can be seen, allsurfaces are at an angle with respect to their reference axes. The lowest corrosionrates obtained for each atmosphere investigated corresponded to samples labeledPD, PI, and PL, respectively (Table 3). The corrosion rates for samples PD andPI showed an improvement with respect to the plain sintered material.Comparing the response surface for the corrosion rate (Figure 5) to those for thethickness of the compound layer (Figure 1), it can be suggested that below acritical value, plasma nitriding actually deteriorates the corrosion behavior of thematerial. Apparently, the composition of the compound layer also determines thecorrosion rate of nitrided materials. The improvement observed for the corrosionresistance of nitrocarburized specimens was generally linked to the thickening ofthe compound layer. Nevertheless, the best corrosion rates were obtained forsamples nitrocarburized at 570°C/2 h (label PD), although the correspondingcompound zones were thinner than those of samples PE. This can be explainedbased on the XRD patterns, which revealed that whereas the compound layer ofsample PD consisted exclusively of G-Fe^-gN, that of sample PE contained amixture of c-Fe?_3N and y'-Fe N. Another important aspect is the substantialpresence of pores in the compound layer of samples PE. Porosity is generally aresult of the decomposition of E-Fei-sC^N into FegC and nitrogen [28-29].Comparing Figure 6a and b indicated that the effective action of the corrosionbarrier is related to a uniform coating rather than the thickness of the compoundlayer itself. Consequently, lower corrosion rates were obtained from samplesnitrided under atm 3, regardless of their relatively thinner compound layers. Thevariance analysis of the overall results suggested that for CH^-free atmospheresnone of the processing parameters or their combination were determinant on thefinal characteristics of the compound layer. On the other hand, adding CH^ to theatmosphere drastically changed the relative importance of all parametersinvestigated (Figure 5), and they all became meaningful. Furthermore, theatmosphere also affected the curvature of the response surface, suggesting thatthe response surface for nitrocarburized specimens would be better fitted by aquadratic model, such as

y = bo + b]X] + b]X2 + bi ix -+- b2]X] + b^XiX] (2)



However, the increase in the number of constants to be adjusted, as comparedto equation (1) would require a great number of experiments, which wouldprevent the conclusion of the present study in a timely fashion.

Table 3: Results from potentiodynamic polarization and material ranking

Material

PPAPBPCPDPEPFPGPHPIPJPKPLPMPNPO

Ecoi-r(mV)-0.50-0.45-0.42-0.44-039-0.40-0.52-0.46-0.39-0.40-0.44-029-0.41-0.43-0.28-035

E(i=o)(mV)-0.67-0.68-0.53-0.46-0.57-0.53-0.66-0.66-0.54-0.63-0.64-0.45-0.52-0.61-0.46-0.52

*corr(uA/cm*)41.5195.7014.5716.745.5610.8767.5085206200503077.0918.3615.1044.9717.5334.49

Corrosion rate(mm/year)

0.48 + 0.081.11 +0.150.18 + 0.050.20 +0.040.06 + 0.030.13+0.020.78 + 0.350.99 + 0.240.72 + 0.080.58 + 0.170.89 + 0.260.21 +0.240.18 + 0.060.52 + 0.450.20 + 0.080.40 + 0.34

Rank

GoodPoorGoodGood

ExcellentGoodRegularRegularRegularRegularRegularGoodGoodGoodGoodGood

' average + 1.96 x sample standard deviation

Table 3 also summarizes the average values of Ic E(i=0), andcorrosion rate. Along with a qualitative rank on the performance of the differentmaterials in the corrosive environment investigated (KNO] 0.5 M solution). Theresults showed that, under specific plasma conditions it was possible to improvethe corrosion resistance of the original sintered material. The best results wereobtained from nitrocarburized specimens, as the presence of CH^ in the plasmaatmosphere generally resulted in uniform and relatively thick compound layers,that extended to the innermost regions of open pores (Figure 6), thus improvedeven further the corrosion resistance of the material.



(a)

(b)

60 80 (28) 100

(c)

o-POT3•g PN ,.

J= ..;-}&

; A f /

.; 1 ; A i ^ y

40 80 (2U) 130

Figure 3: X-ray diffraction patterns of specimens treated under different timesand atmospheres, (a) atm 1: 75%N2-23,75%H2-1,25% CH4;, (b) atm 2:25%H:; and (c) atm 3: 25%N]-75%H2.



(a)

(b)

(c)

2,0 r

m 15X

tp

-0,5 i

<0;

2,0

1,5

w 1,0x —

,fpF5 ko

-0,5

- 6 - 5 - 4 - 3Log i(A/cm ) )

-1,06,5 -6,0 -5,5 -5,0 -4,5 -4,0 -3,5 -3,0 -2,5 -2,0 -1,5

Figure 4: Potentiodynamic polarization curves of plain sintered specimens (P),and samples nitrided under atm.l at 510°C/2 h (PA), atm. 2 at 570°C/2 h (PI),andatm. 3at510°C/2h(PK).



847

(a)Icc

813

(b)

783

847

813

TEMPO (h)

0.

783

847

I 1 0,577I 1 0,621CD 0,664CD 0,708g 1 0,752[ED 0,795[3 0,839

0,883HI 0,926« 0,970

TEMPO (h)

(c)

783

| 1 0,176| 1 0,203| j 0,230CD 0,257|—1 0,28501 0,312CD 0,339

0,3670,3940,421

CZD

TEMPO (h)

Figure 5: Response surface for the corrosion rate as a function of the temperatureand nitriding holding time, along with corresponding variance analyses, (a) atm.l;(b)atm. 2, (c)atm. 3.



(a) (b)Figure 6: Backscattered electron image of specimen nitrocarburized under (a)570°C/6 h.

4 Conclusions

• The thickness of the compound layer is generally determined by thenitriding temperature. The holding time becomes an important parameterwhen the plasma atmosphere contains CH^

• The hardening depth is significantly affected by the holding time,particularly at low temperatures. Temperature affected the hardening depthonly upon nitrocarburizing;

• Nitrocarburizing atmospheres are more efficient in increasing the thicknessof the compound layer and hardening depth;

• In N2-rich atmospheres, the relative amount of y'-Fe N increased with thenitriding temperature. Reducing the relative amount of N] assisted in theformation of y'-Fe N, whereas CH^ drove the precipitation of 8-Fe2-sN;

• Nitriding and nitrocarburizing resulted in a decrease of the anodicdissolution current which indicated an improvement of the corrosionresistance of the material in that potential range. At higher potentials, thecompound layers ruptured at their thinnest areas, thus exposing thesubstrate. Corrosion in the interface-to-substrate direction was observed;

• The presence of CH^ in the atmosphere resulted in uniform coatings thatworked as effective corrosion barriers on sintered substrates.

AcknowledgmentsThe authors wish to acknowledge the support given by our sponsors FINE? andCNPQ, as well as to all who directly or indirectly contributed to the outcome ofthis work.

References1) White, D. G., M/P in North America, The International Journal of Powder

Metallurgy, 32, No.3, pp 221-228, 1996.



2) E. Maahn, S. K. Jensen, R. M. Larsen, T. Mathiesen, Advances in PowderMetallurgy and Particulate Materials, 7, pp 253, 1994.

3) T. Mathiesen and E. Maahn, Advances in Powder Metallurgy and ParticipateMaterials, 3, pp 45, 1995.

4) L. Fedrizzi, F. Deflorian, A. Tiziani, I. Cristofolini, A. Molinari, Advancesin Powder Metallurgy and Paniculate Materials, 7, pp 273, 1994.

5) A. Tremblay and R. Angers, Advances in Powder Metallurgy andP articulate Materials, 7, pp 22 5, 1995.

6) E. Otero, A. Pardo, M. V. Utrilla, F. J. Perez and C. Merino, CorrosionSc/gMCg, 39,pp453, 1997.

7) Sobral, A . V. C.; Ristow, W. Jr. and Franco, C. V. 3rd NACE LatinAmerican Region Corrosion Congress, Compact Disk Paper # S17-S15.

8) Lampman, S., Introduction to Surface Hardening of Steels, pp. 259-267.9) Rickerby, D. S., and Matthews, A., Advanced Surface Coatings: a

Handbook of Surface Engineering, Chapman and Hall, 1991.10) Steen, W. M., Laser Material Processing, Springer - Verlag London

Limited, 1996.11) Chardin, H., Grillon, F., Jeandin, M. and Frainais, M., Superficial

Densification of P/M Parts Using Ultrasonic Shot-Peening, TheInternational Journal of Powder Metallurgy, Vol. 32, No.2, pp 245-254,1996.

12) Bas, J. A.,and Molins, C.,MM6/mg p /zW/o f/Mfar%, PM 86, Part 1, pp365-368, 1986.

13) Rie, K. T., Lampe, Th., Eisenberg, St.,Surface Tratment of PM-Material byPlasma Nitrocarburizing, PM 86, Part 1, pp 361-364, 1986.

14) Bocchini, G. F., Molizari, A., and Tesi, B., Ion Nitriding of Sintered Steels -?r g/?r zYw zoM aWfc/ cr/v MPR, Nov., pp 772-778, 1990.

15) Metin, E. and Inal, O. T., Formation and Growth of Iron Nitrides During/o/7-M/nWmg, Journal Of Material Science, 22, pp 2783-2788, 1987.

16) Kovacs, W., Russel, W., An Introduction to Ion Nitriding. What is it? Whereis used?, International Conference on Ion Nitriding, Clevelend, Ohio, pp.186-187, 1977.

17) Inal, O. T., and Ozbaysal, K., Mechanism of Iron Nitriding in Cr, Al and Cr+ Al Containing 1040 Steel, International conference on Ion Nitriding,Clefelend, Ohio, Usa, Sept., 1986, Proceedings edited by T. Spalvins, ASM,pp 31-36, 1987.

18) Lightfoot, B. J., and Jack D. H., Kinetics of Nitriding whih, and WithoutWhite-Layer Formation, Source Book on Nitriding, American Society forMetal, Ohio, pp. 248-254, 1977.

19) Edenhofer, B., Production lonitriding, Metal Progress, March, pp. 38-42,1976.

20) Nilton C. Pereira, Friedrich G. Mittelstadt, Almir Spinelli, Ana M. Maliska,Joel L. R. Muzart. Aloisio N. Klein and Cesar V. Franco, Journal ofM Wj c/ ce, 30 Number 19, pp 4817-4822, 1995.

21) Ana C. Boblitz Parente, Ana V.C. Sobral, Aloisio N. Klein, Joel L.R.Muzartand Cesar V. Franco, Advances in Powder Metallurgy & P articulate

, 4-13, pp 167-176, 1996.



22) Ana V.C. Sobral, Ana C.B. Parente, Joel L.R. Muzart, Cesar V. Franco,Surface and Coatings Technology, 92, pp 10-15, 1997.

23) Mridha, S., Jack, D. H., MzW Sconce, Vol. 16, Aug., pp 398-404, 1982.24) P. C. Borges, N. C. Pereira, C. V. Franco and A .N. Klein, Advances in

Powder Metallurgy & Paniculate Materials, Metal-Powder IndustriesFederation, 2, pp 61, 1994.

25) Borges, P.C., Martinelli, A.E., Muzart, J.L.R., Klein, A.N., and Franco, C.V.Materials Science Forum, - Advanced Powder Technology Ed. TranstechPublications. Switzerland, pp 286, 1999.

26) Friedrich G. Mittelstadt, Cesar V. Franco, Journal of Materials Science, 31,pp 431-435, 1995.

27) Montegomery, D. C., Design and Analysis of Experiments, John Wiley &Sons, New York, 1991.

28) Somers, M. A. J., Mittemeier, E. J., Surface Engineering, 3, no. 2, pp. 123-137, 1987.

29) Li, S., Manory, R. R., Hensler, J. H., Surface ans Coatings Technology, 71,pp 112-120, 1995.


an assessment of the effect of plasma nitriding pc borges,^ ae ... · n] contents vary between 15%...

Documents