Surface hardening of AISI 316L stainless steel using plasma carburizing

Surface hardening of AISI 316L stainless steel using plasma carburizing

rhino ELSEVIER Thin Solid Films 295 (1997) 185-192 Surface hardening of AISI 316L stainless steel using plasma carburizing Bong-Seok Suh, Won-Jong L...

599KB Sizes 4 Downloads 82 Views

rhino ELSEVIER

Thin Solid Films 295 (1997) 185-192

Surface hardening of AISI 316L stainless steel using plasma carburizing Bong-Seok Suh, Won-Jong Lee Department of Matertals Science and Engineering, Korea Advanced Instttute of Science and Technology, Taejon 305-701, South Korea Received 3 June 1996; accepted 22 August 1996

Abstract AISI 316L stainless steel was carbunzed using CH4/H2 plasma to increase surface hardness. Plasma carburizing conditions such as applied voltage, carburizing temperature, gas pressure, CH4/(CH4 + H2) ratio and process tlme were varied to achieve the desired surface hardness, effective hardening depth and carburizing uniformity. Plasma carburized specimens were analyzed with optical microscopy, microhardness tester, scanning electron microscopy and Auger electron spectroscopy. Effective removal of surface oxide which grows dunng plasma carbunzing is essential for successful carbunzing and good uniforrmty m AISI 316L stainless steel. Adding H2 to CH4 as well as increasing process pressure enhances carburizlng uniformity. Hardness obtained after plasma carburlzing has a linear relation w~th carbon concentration. By carrying out plasma carburizing over 90 min at 800 °C, 600 V, 2 Torr and 25% CH4/(CH4 q" H2) ratio, the back-ferrule type AISI 316L stainless steel specimen was uniformly carburized with a surface hardness above 735 HV50g and an effective hardening depth above 40 I~m. © 1997 Elsevier Sclence S.A. Keywords. Carbon; Glow discharge; Hardness; Heat treatment

1. Introduction

2. Experimental

Carburizing is a process by which carbon is added to the surface of steels. After carburizing, surface hardening is usually accomplished by rapid cooling. The steels treated by these processes have a hard and wear-resistant surface on a ductile core. Furthermore, the compressive stress generated in the hardened surface layer enhances fatigue resistance. Carburizing is widely used to produce various machine parts like automobile parts. It can be done in a gas atmosphere (gas carburizing), a salt bath (liquid carburizing) or a coke/charcoal (pack or solid carburizing). Recently plasma carburizing using glow discharge technology has been industrially applied. Plasma carburizing is carried out in the glow discharge of hydrocarbon gases to provide an ionized carbon species at the work surface [ 1-3 ]. Plasma carburizing is a clean vacuum process and can upgrade the product quahty [4,5]. It is also suitable to overcarburizing and deep carburizing because it can supply a large amount of carbon in a fast way. Moreover, it can be applied to stainless steels because the surface oxide layer can be removed by ion sputtenng [6,7]. In this study we applied plasma carburizing to austenitic stainless steel AISI 316L and investigated the variation of surface hardening properties with plasma carburizing conditions.

AISI 316L stainless steels (18Cr-10Ni) in the shape of 1/2 inch back ferrule were used as specimens. Fig. 1 shows the cross-sectional view of the back ferrule specimen. Each region of the specimen is named as tip, slant and side to indicate the analyzed region. The microhardness of the untreated specimens was 370 HV50g (HV50g indicates the Vickers hardness number when loading 50 g). Fig. 2 shows the schematic diagram of the plasma carburizing system. A glow discharge is generated by applying d.c. power between the specimen holder (cathode) and the chamber body (anode). The surface area of the specimen holder is 150 cm 2. The specimen holder is surrounded by graphite

0040-6090/97/$17.00 © 1997 Elsevier Science S A. All rights reserved PIIS0040-6090 (96) 09284-X

I/2 mc ~

~--"

Fig I Cross-section of the back ferrulespemmen.

186

B -S Suh, W.-3".Lee / Thm Solid Films 295 (1997) 185-192 Graphiteheater

T/C

\ IJ

.

Na (coolinggas)

~-leat stueld

/

Table 1 Plasmacarburizing conditions

I "v.'/A/J'///IIAP~A \

:

/

.

.

x ...... i ...... /

IDC Power

Fig 2. Schematicdmgramof the plasmacarbunzmgsystem

heating elements. Reaction and cooling gases are supplied through the lines and holes inside the heating elements. Heat shields are placed between the graphite heater and the chamber wall which is cooled by cooling water. Treatment temperature is measured by a thermocouple which is covered with an alumina tube and located 1 cm above the specimen holder and the process pressure is measured by a McLeod gauge. Fig. 3 shows the schematic diagram of plasma carburizing process. After the chamber was evacuated, the specimen was heated to the treatment temperature (800 °C or 950 °C) at which the base pressure was about 200 mTorr due to the outgassing from the heat shield material (graphite soft felt). Even though the process was conducted in vacuum, the surface oxide layer grew due to the outgassing from the heat shield material, but at a much lower rate than in atmosphere. Specimen surface was sputter-cleaned by applying d.c. 600 V in a gas mixture of Ar (50 sccm) and Ha (50 sccm) for 30 mln. The plasma carburizing process was then carried out by flowing a mixture of CH4 and Ha as a reaction gas. The plasma carburizing conditions used in present experiments are summarized m Table 1. We tried plasma carbunzing at various apphed voltages and found that the specimen was successfully carburized at 600 V or above. In this study all experiments were carried out at 600 V. We also examined the influence of cooling rate and found that it had little effect

Temperature Sputter Cooh cleaning Carbtmzmg IHeatinl

Pumpmg

800-950'C

Room

latm

Temp :/t"

CHdl-h

1-2 7 Tort

\ 60rain 30rmn

10-160ram

0 2 Ton" ' Vacuum

Fig 3. Schematic diagram of the plasma carburizing process

Reaction gas CH4/(CI-L + H2) ratio Total gas flow rate Temperature Pressure Apphed d.c. voltage Current density ~

CH4, H2 100, 50, 25,10% t00 sccm 800, 950 °C 1 ~ 2.4 Tort 600 V 0.06 ~ 1.6 mA cm -2

a Current density was obtained by dividing the total current measured with a current-meter by the surface area of the specimen holder. Since plasma current density is higher at the sharp place such as tip region of the ferrule speclmen, it is not a real current density at the carburized place, but can be used as a reference to indicate the plasma condition.

on the hardness profile of the carburized specimen. This is probably because carburized stainless steels is hardened by precipitation hardening unlike low alloy steels which are hardened by martensite transformation during quenching. In order to examine the results of surface hardening, microhardness profiles at the cross-section of the specimens were obtained using a microVickers hardness tester (Mitutoyo MVK-H 1) with a load of 50 g, and from this hardness profile the effective hardening depth having 550 HV50g was measured. The microstructure of the carburized layer was observed with an optical microscopy and a scaning electron microscopy (SEM) after etching with a dilute marble etchant (HCI, 50 ml; distilled water, 450 ml; and CuSO4, 10 g). Carbon concentratlon profile was measured by Auger electron spectroscopy (AES) under the following conditions: electron energy, 5 keV; electron current, 300 nA; modulation voltage, 4 V; sputter ion voltage, 3 keV; and sputter ion current density, 400 IxA c m - 2.

3. Results

and discussion

Fig. 4 shows the typical cross-sectional micrographs of a plasma-carburized AISI 316L ferrule specimen after etching. From the optical view, the carburized layer can be divided into three zones: near-surface white zone, dark zone and core. It was found that the near-surface white zone had a carbon concentration above 4.0 wt.% and a hardness above 600 HV50g, whereas the dark zone had a carbon concentration of 1.5 ~4.0 wt.% and a hardness of 4 0 0 ~ 6 0 0 HV50g. From SEM views, elongated carbides and comparatively flat surface morphology were observed in the near-surface white zone, whereas relatively fine carbides and rough surface morphology were observed in the dark zone. The dark zone of the etched specimen appears dark under the optical microscopy examination because the light is scattered by the rough surface morphology. In the core, carbides were observed only at the grain boundary. Several workers [ 8-11 ] who studied the carburization of 316 stainless steel reported that (Fe,Cr)TC3 carbides formed predominantly when the carbon content was above 4 wt.% and the ratio of the (Fe,Cr)23C6

187

B-S. Suh, W.-J. L e e / T h i n Sohd Ftlms 295 (1997) 185-192

3~ ,[m

Fig. 4. Typlcal cross-sectional optical mlcrograph of the plasma-carbunzed AISI 316L specimen and SEM mlcrographs with higher magmfication of each zone. (1) near-surface white zone, (2) dark zone and (3) core. to (Fe,Cr)TC3 phases increased as the carbon concentration of the steel decreased. According to their results, it is believed that the near-surface white zone comprises predominantly

(Fe,Cr)TC3 and the dark zone comprises the mixed phase of (Fe,Cr) 7C3 and (Fe,Cr) 23C~. Plasma carburizing experiments were c a r d e d out with varying process time at 950 °C. Other carburizing conditions were as follows: reaction gas, CH4, 100 sccm; pressure, 2.4 Torr; and current density, 1.6 m A cm -2. Figs. 5 and 6 show the microstructures and hardness profiles after plasma carburizing for 10, 40 and 120 min. After carburizmg treatment the hardness of the steel core was lowered from 370 to 180 HV50g by the annealing effect of the high process temperature. W h e n the process time was only 10 min, hardness 1000

8O0

--Q-- 10 mm ---o-- 40mln

~,

12o 600

N

~

200t 0 !

0

~ o ,

r

100

i

m----_____ o_ i

i

200

i

300

400

Depth (gin)

Fig. 5. Cross-sectional micrographs of ferrule specimens after plasma carbufizmg for various process times at 950 °C, 600 V, 2.4 Torr and c g 4 100 sccm

Fig. 6. Hardness profiles at the tip regmn of ferrule specimens after plasma carbunzmg for various process times at 950 °C, 600 V, 2.4 Torr and C H 4 I00 sccm

188

B.-S. Suh, W -J Lee / Thin Sohd Flhns 295 (1997) 185-192

increased continuously but relatzvely slowly as it approached the surface, but it decreased at the very near surface because of the local heatzng of the surface by ion bombardment. Surface hardness and effective hardening depth increased with process time: after 120 min they became 700 HV50g and 150 ~m respectively. In these experiment conditions, carburlzing did not proceed uniformly through the ferrule specimen, rather it was concentrated on the tip region. The carburized tip region looked clean with a color of light gray but the other regions which were not carburized were covered with black soot (disordered carbon). Fig. 7(a) and 7(b) shows AES depth profiles at the carburized region and at the uncarburized region with soot respectively. At the carburized region, a very thin oxide film exists at the steel surface without soot and the carbon content is very high within steel. Whereas at the uncarburlzed region, thick soot covers the steel and a thick oxide layer exists between the soot and steel surface. It is believed that effective removal of surface oxide which grows during the high temperature plasma carburizing process is essential for successful carburizing, otherwise the surface oxide layer prevents the diffusion of carbon atoms and those carbon atoms aggregate on the stainless steel surface to form a soot. Since plasma current concentrates at the sharp place, the surface oxide of the tip reglon of the ferrule specimen is likely to be removed more effectively than other regions, resulting in the effective carburization at the tip region. Therefore in order to enhance carburizing unzformlty, one has to remove the surface oxide layer of all the specimen surface needed to be carburized. In the carburizing experiments at 950 °C, the surface was hardened to a high degree but the steel itself was softened

due to the high temperature of the carburizing process, and the carburizing uniformity was poor. Therefore in order to suppress the softening of the specimen and lower the surface oxide formation rate, carburizing at a lower temperature (800 °C) was carried out. Although the flow rate of C H 4 gas was same as that of 950 °C experiment, the pressure decreased to 1 Torr and a current density to 0.27 mA cm - 2 compared with the 950 °C experiment since the thermal decomposition of CH4 gas was greatly reduced at this temperature. Figs. 8 and 9 show the cross-sectional microstructures and hardness profiles after plasma carburizing for various times. The hardness of the steel core was 290 HVS0g which was a much higher

10

Fig 8 Cross-sectional rmcrographs of ferrule specimens after plasma carbunzmg for various process times at 800 °C, 600 V, I Torr and CH4 100

5

sccm

2

10oo~

1 00

(a)

2.4

4.8

7.2

800 ,',

Sputter time (min)

--E]u

10

10 rain

--z~-- 120 rain I 6oo E

4o0 &

2o0

[]

f

0

00 (b)

24

4.8

7.2

9.6

12,0 14.4 168

192

2].6 24.0

Sputter time (min)

Fig. 7. AES depth profiles (a) at the carbunzed region without soot and (b) at the uncarbunzed region with soot

;

5

i

i

100

i

t

150

i

t

200

i

250

Depth (l~m)

Fig. 9 Hardness profile at the slant region of specimens after plasma carbunzmg for various process times at 800 °C, 600 V, I Tort and CH4 100 sccm

B.-S. Suh, W -J. L e e / T h i n Sohd Fdms 295 (1997) 185-192

189

2 Torr

1 Torr 0.27 r n A / c m 2 I

9

I

'i

Arc

0.06 rnA/cm z

0.6 rmi/crn2 ,i

i .

0 06 mA/crn~'

0 33 mA/cm2

0 27 mA/cmg

Fig 10. Cross-sectional rmcrographs of ferrule specimens after plasma carburizmg for 40 min at 800 °C, 600 V and various CH4/(CH4 + H2) ratios and pressures.

value comparing with the case of 950 °C treatment, as expected. Moreover, carburizing uniformity was improved: not only at the tip region but also the slant region was carburized. That is because it becomes easier to remove the surface oxide layer due to the reduced surface oxide formation rate at a lower process temperature. As the process temperature was lowered to 800 °C, the carburizing depth was greatly reduced; however, the surface hardness was even enhanced. In the previous experiments at 800 °C using C H 4 gas, the carburizing uniformity was improved, but carburizing still did not arise at the side region. In the following experiments, we used a mixture of C H 4 and Hz as reaction gases and examined the influence of the methane ratio C H 4 / (CH4 + Ha) and the process pressure on the carburizing uniformity. The decrease of the methane ratio will decrease the carbon mass flow by reducing CH 4 partial pressure and the H2 plasma itself will have an effect of removing surface oxide and soot by reduction. It is also well known that increasing pressure enhances carburizing uniformity because the plasma glow seam surrounds the specimen uniformly so the supply

of ion becomes uniform [12]. But increasing pressure imparts a difficulty in preventing an arc forming. With varying methane ratio, C H 4 / ( e l l 4 -b I--I2) , from 100% to 10% and process pressure from 1 Tort to 2 Tort, plasma carburizing was carried out for 40 rain at 800 °C. While the total gas flow rate was maintained at 100 sccm, the pressure was varied by controlling the conductance of pumping line. For the same conductance, strictly speaking, pressure increased somewhat with the methane ratio due to the dissociation of methane gas. The current density also increased with the methane ratio as well as the pressure. Fig. 10 shows the cross-sectional micrographs after plasma carburizing at various methane ratios and pressures. For the experiment conducted at 1 Torr, the addition of Ha did not enhance the carburizing uniformity much, rather the carburizing itself was greatly suppressed due to low current density. At 2 Torr, carburizing was successfully carried out even at low methane ratios and the uniformity was improved. Especially at 25% methane ratio, all the surface was uniformly carburized and no soot was observed, indicating that surface oxide was effec-

I90

B -S Suh, W -J Lee / Thm Solid Films 295 (1997) 185-192 1000

1000

66°_

(a) I--o--Tip

I--o--S'aotl

e00 ~

I+Side

I

80O r~

[

60O

460 O N ~ O C ~

(a)

--o-- Tip ----e-- Slant

4O0 200

0[

,

i

,

i

r

800F

,

r

,

I

0

,

8OO

(b)

00° 400

600

~0~

400 200

200 0

t

8O0

I

I

6

t

(c)

800

(e)

600

600

O 400 200

O

i 00

I i '1(30 150 Depth (lain)

i 200

250

Fig 11. Hardness profiles at various regaons of ferrule speclmens after plasma carburizing for 40 nun at 800 °C, 600 V, I Torr and various Ct-L ratios: (a) Ct-I4 100 sccm; (b) CH4 50 sccm, H2 50 sccm; (c) CH4 25 sccm, H2 75 sccm.

tively removed. Figs. 11 and 12 show hardness profiles of tip, slant and side regions after plasma carburizlng at 1 Ton" and 2 Ton- respectively. An increase of pressure and addition of H2 enhanced carburizing uniformity. However, addition of too much hydrogen ( i.e. too low methane ratio) suppressed carburizing due to too low carbon supply. To examine if the increase of uniformity at low methane ratio at 2 Ton- is simply due to the reduced carbon supply or not, a comparative experiment was carried out with CH4 25 sccm only at 800 °C and 1.7 Torr. In this case the carburizing was much less uniform comparing with the previous result using the gas mixture of CH4 25 sccm and Ha 75 sccm. Therefore it is concluded that the reduction effect of hydrogen plasma contributes to the improvement of carburizing uniformity. Fig. 13 shows cross-sectional micrographs after plasma carburizing for various times with the gas mixture of CH4 25 sccm and H2 75 sccm which was considered to be the best

0 0

50

100 150 2130 250 Depth (p.rn) Fig 12. Hardness profiles at various regions of ferrule specimens after plasma carburizing for 40 rain at 800 °C, 600 V, 2 Tort and various CH 4 rauos: (a) CH4 50 sccm, Ha 50 sccm; (b) CH4 25 sccm, H2 75 sccm; (c)

Ct-I410 sccm,H290 sccm.

optimized condition. Fig. 14 shows the changes of hardness profile and carbon concentration profile with process time at the slant regions of ferrule specimens. After 10 min, the specimen was hardly carburized, however, as the carburizing proceeded, both maximum hardness and the effective hardenmg depth increased. Maximum hardness and effective hardening depth obtained were, respectively, 600 HV50g and 25 txm after 40 min, 735 HV50g and 40 ~m after 90 min, and 800 HV50g and 49 i~m after 160 min. Carbon concentration was highest at the surface, and saturated about 6.5 wt.%. It has been reported that the saturated carbon concentration of high Cr steel is related to Cr content and corresponds to carbide precipitation amounts [7,11]. Fig. 15 shows the relationship between the carbon concentration and the microhardness, which was obtained from the data of Fig. 14 except those of 10 rain experiment where carburizing had hardly proceeded. The microhardness had a linear relation with the

191

B -S. Suh, W - J L e e / T h m Sohd Ftlms 295 (1997) 185-192 1000

(a) 800

"~

b

--o-- 10 m=n ---o-- 40 rain --'~-- 90 mm + 160min

400

200 1 0

0

r 100

i 150

i 200

250

2

2

Depth(gm) ,

~'~'~ i ~

8

(b)

i 10

50

100

150

Depth(ram) Fig. 14. Hardness and carbon concentration depth profiles at the slant regmn of ferrule specimens for various process times at 600 V, 800 °C, 2 Torr and 25% CH4/(CH4+H2) ratio.

1000

Fig 13. Cross-sectmnal micrographs after plasma carbunzing for various process times at 800 °C, 600 V, 2 Torr and 25% CH4/(CH4'-F H2) ratio. 0

carbon concentration. Some deviation from the linear relation at high carbon concentrations was attributed to the surface annealing effect.

-r" v

800

0

600

~O

t-

£

.o

q~

0 O0

0

0

0 0

0.. 0 0

400 "~ oO0% 2O0

4. Conclusions

0 Plasma carburizing was applied to AISI 316L stainless steel and the following conclusions were drawn. 1. For plasma carburizing ofAISI 316L material, the surface oxide needed to be removed effectively. Otherwise soot was formed on the surface and carburizing became nonuniform. This problem can be solved by adjusting the C H 4 / ( C H 4 + H 2 ) ratio and increasing the process pressure. By carrying out plasma carburizing over 90 min at 800 °C, 600 V, 2 Tort and 25 % C H 4 / ( C H 4 -'k H2) ratio, we obtained uniform carburizing result with a surface hardness above 735 HV50g and an effective hardening depth above 40 ~m. 2. The carburized layer can be divided into three zones. The near-surface white zone, where elongated carbides were observed, had a carbon concentration above 4.0 wt.% and

4

6

Carbon Concentration (wt.%) F:g. 15. Relation between carbon concentration and rmcrohardness

a hardness above 600 HV50g. The dark zone, where fine carbides were observed, had a carbon concentration of 1.5 ~ 4.0 wt.% and a hardness of 400 ~ 600 HV50g. In the core, carbides were observed only at the grain boundary. Hardness of carburized layer had a linear relation with carbon concentration.

Acknowledgements This work is supported by the Korea Science and Engineering Foundation (KOSEF) as part of the UniversityIndustry Collaborative Research.

192

B.-S Suh, W.-J. Lee /Thin Sohd Ftlms 295 (1997) 185-192

References

[1] M Booth, T Farrell and R H Johnson, Heat Treat Metals, (February, 1983) 45-52. [2] W Rembges and J. Luhr, Ion Nitriding and Ion Carburizing, ASM Intemauonal, OH, 1989, pp. 235-243 [3] F. Schnatbanm and A. Melber, Heat Treat Metals, (February, 1994) 45 -51. [4] W. L Grube and J. G Gay, Metall. Trans. A, 9,4 (1978) 1421-1429. [5] M Booth, M Lees and A. M. Staines, Vac. Metall. (1983) Paper 23/123/6.

[6] K. Akutsu and M. Nakamura, Ion Nztriding andIon Carburizing, ASM International, OH, 1989, pp. 249-256. [ 7] T Klmura and K Namlki, J. Jpn. Soc. Heat Treat., 34 ( 1 ) 18-23. [8] J. Gwyther, M Hobdell and A. J. Hooper, Conf. Fuel Element Behavwur, Metals Somety, London, 1972. [9] S. Terauchl, H. Teraucba and K. Kamei, Heat Treatment '76, Metals Society, London, 1976, pp. 45-49. [10] J. P. Souchard, P. Jacquot and M. Burton, Mater. Sci. Eng., A140 ( 1991 ) 454460. [ 11 ] I Asano, T. Araki and Y. Ikawa, Mater. Sci. Eng., A140 ( 1991 ) 461468 [ 12] J. G. Conybear, Heat Treating (May, 1988) 24-27.