You are on page 1of 9

Trans. Indian Vo\.

55, NoA,

Inst. Met. August 2002, pp.225-263

HNS ' 02 TP 12

MICROSTRUCTURE AND CHEMICAL CHARACTERIZATION OF HIGH TEMPERATURE NITRIDED 12%Cr STAINLESS STEELS
Carlos Mario Garzon, Alejandro Toro, Andre Paulo Tschiptschin
Metallurgical and Materials Engineering Department, University of Sao Paulo, Av. Prof. Mello Moraes 2463, CEP 05508-900, Sao Paulo, Brazil.

ABSTRACT
AISI 410 (11.5 %Cr - 0.13 %C) and AISI 41 OS (13.0 % Cr - 0.07 % C) martensitic stainless steels were gas nitrided in Nz atmosphere at 1273-1473 K, 0.02 - 0.38 MPa and 0.9 - 172.8 ks. Nitrogen gradients were determined by chemical analysis through optical spectrometry and WDS microanalysis. The average nitrogen content of long-term nitrided thin specimens was measured by optical spectroscopy and fusion under inert gas. Precipitate extraction was performed by dissolution of the matrix, the precipitates being analyzed by fusion under inert gas and X-ray diffi'action. The results showed that by increasing nitriding time and pressure, as well as by decreasing temperature, both the nitrogen content at the surface of the steels and the tendency to form precipitates increased. Thermocalc@ calculations for the Fe-Cr-N-C. system allowed predicting nitrogen contents, as well as microstructures of the nitrided alloys. Good agreement between calculated and experimental values was observed. XRD of extracted precipitates confirmed Thermocalc calculations concerning the stability of nitrides. In specimens containing precipitated nitrides, the nitrogen content dissolved in martensite increased with the distance trom the surface, in the region where precipitation occurs. Beyond this region the nitrogen content decreased towards the core.

1. INTRODUCTION
Nitrogen has been added as an alloying element to stainless steels to improve mechanical properties and corrosion resistance!. Nitrogen can be introduced in stainless steels by a high-temperature thermochemical treatrnenr,3, which allows introducing nitrogen at temperatures"aboveAc! under nitrogen partial pressures that typically vary in the range 0.03-0.40 MPa. After 3.6-86.4ks treatments, a high nitrogen case is formed4. Underthese conditions nitrogen remains mainly in solid solution,without forming a white layer usually found in low temperature nitriding processes. The gas used (N2+Ar)is neither explosive nor toxic, and no gas flux is required. Dependingon the chemical composition of the steel and the phases in equilibrium at the nitriding temperature, eitherhigh strength austenitic or hard martensitic cases can be obtained. When applied to martensitic and martensitic-ferritic stainless steels, this process allows combining high hardness at the surface with good

impact resistance of the core. The resultant microstructure after hig!). tempt:rature nitriding treatment depends on the material and on nitriding times, temperatures and N2 partial pressures. Several authors4.5consider the existence of stable equilibrium between the steel surface and the gas, which implies that the phases present"-. t the surface as well as their a chemical composition can be read in a phase diagram including metal-gas equilibrium. According to Frisk6 the metal-gas equilibrium during nitriding of steel powders can be predicted by drawing in a phase diagram, the nitrogen activity for specific N2 partial pressures. In a similar approach, Tschiptschin5 proposes that metal-gas equilibrium during hightemperature nitriding may be predicted by using a metal-gas equilibrium diagram, constructed for a specific N2partial pressure. There is no systematic study of microstructural changes at the surface, as a function of time during hightemperature nitriding. Tschiptschin5 mentioned that under some specific nitriding conditions, the assumption

255

,~

TRANS. INDIAN INST. MET., VOL. 55, NO. 4, AUGUST 2002

Table 1
CHEMICAL COMPOSITION OF THE BASE MATERIALS (wt"A.)

Steel AI81 4108 AI81410

Cr 13.0 11.5

C 0.07 0.13

Ni 0.10 0.21

Mn 0.4 0.6

. Si ro:-25 0.36

of metal-gas stable equilibrium approximation.

is not

good

The microstructure of a region below the surface can be predicted in a phase diagram, if the mean nitrogen content of that region is known. Furthermore, the ni~ogen distribution in the nitrided case can be estimated by solving the Fick's second law for diffusion of nitrogen from the surface towards the core. The aim of the present work was to study the effects of nitriding temperature, Nz partial pressure and time, on the microstructure of high temperature nitrided and direct-quenched AISI 410 and 410S martensitic stainless steels.

The microstructure' of the specimens were examined by optical and scanning electron microscopy and by X-ray diffraction. The prior austenite grain size was determined according to the E 112 ASTM standard. The composition of the phases after nitriding was investigated by wavelength dispersive spectroscopy (WDS). Nitrogen gradients in the thick specimens were determined by chemical analysis through optical spectrometry (OS) and WDS microanalysis. The nitrogen content of long-term nitrided thin specimens was measured by OS and fusion under inert gas. Precipitate extraction was performed by the chemical dissolution of the matrix, and the precipitates were analyzed by fusion under inert gas and X-ray diffraction. For precipitate extraction, specimens weighing 1-2 g were dissolved in a Berzelius type solution at 300 K for 43.2 ks. The reagent was prepared by dissolving 320 g of CuClz2HzO,280 g of KCl, and 20 g of HOOC(CHOH]zCOOH in 1.85 I of distilled water and 150 mI of HCI. A magnetic wrist-action sbaker was used to agitate the specimens inside the reagent in a flask. The solution was suctioned through a 45 mm diameter PTFE filter with 0.45 nun maximum pore size. The residue was washed with 0.25 N hydrochloric acid and water and then. transferred to a watch glass and dried at 393 K for 24 h. Thermocalc@ was used to predict phase stability and chemical composition of the nitrided specimens. The gas-metal equilibrium was calculated for one specific Nz isobar allowing predicting the phases present and the maximum nitrogen content at the surface of the steel. The stability and the compositions of the phases in other regions inside the case were calculated assuming

2. EXPERIMENTAL
Two martensitic stainless steels were used: AISI 410 bars, 12.7 mm in diameter and AISI 410S sheets, 9 mm thick. Table I shows the chemical composition of these alloys. The AISI 410 and 410S steels were hightemperature nitrided in Nz atmosphere and direct quenched in water. Two sets of specimens were used: thick pieces 9 13 mm thick and thin pieces 0.3 - 0.7 mm thick. With the aim of obtaining steels with a high nitrogen case, the thick specimens were nitrided under 0.25 MPa Nz partial pressure at 1273 1473K for0.9 86.4 ks. The thin pieces were long-term (172.8 ks) nitrided under 0.02 - 0.38 MPa Nz partial pressure at 1273 1473 K, in order to obtain steels with bulk homogeneous nitrogen contents.

Metallographic preparation of the specimens consisted of grinding in emery paper up to 2400 grit, followed by 1J.1IIl diamond polishing.

256

GARZON et al : CHEMICAL CHARACTERIZATION OF NITRIDED 12% Cr STAINLESS STEELS

Table 2
THERMOCALC @ CALCULATED AND WDS MEASURED NITROGEN CONTENTS AT THE END OF THE MARTENSITIC LAYER

Temperature(K) 1473 1373 1353 1273

Calculated (wt%) 0.09 0.12 0.15 0.20

Measured (wt%) 0.11 :t 0.05 0.16:t 0.05 0.20 :t 0.05 0.22 :t 0.05

thermodynamicequilibrium as well. All thermodynamic calculationswere performed using the TCFE database.

Three zones with different microstructures were observed from the surface towards the center of the specimens: a martensitic surface layer, a transition zone, and a dual-phase martensitic-ferritic inner zone. The microconstituents in the martensitic surface layer were the same observed in the high-nitrogen case of AlSI 410 steel. The transitio!l zone was consisted of a martensitic matrix with isolated regions of untransformed ferrite, as shown in Fig. 2. No precipitates were observed under SEM. The inner zone showed a dual phase martensiticferritic microstructure containing 40-50% ferrite.

3. RESULTSANDDISCUSSION
3.1 Microstructure and Chemical Characterization of the Thick Specimens
Thick specimens of nitrided AISI 410 steel showed martensite both in the high- nitrogen case and in the low nitrogen core (Fig. 1). Different volume fractions of retained austenite, MX and M2X type precipitates were also observed in the case, depending on the nitriding conditions.

Fig.2 :

Micrographs of the AISI 410S steel nitrided at 1473 K for 3 h at 0.25 MPa. (a) Optical micrograph showing the high nitrogen martensitic surface layer (MSL), the transition zone (TZ) and the low nitrogen core (LNC); (b) Scanning electron micrograph showing ferrite (cx) and marten site (cx') in the transition zone.

Table 2 shows the Thermocalc@ calculation and WDS measured values of nitrogen content at the end of the martensitic layer (beginning of the transition zone) for the AIS 4IOS specimens, nitrided at 0.25 MPa for 21.6 ks. A good agreement between the calculated and the measured values were observed.

Fig.I :

Optical micrograph of the AISI 410 steel nitrided at 1473 K for 6 h at 0.25 MPa, showing the high nitrogen case (HNC) and the low nitrogen core (LNC).

Thick specimens of nitrided AISI 4IOS steel showed a martensitic case and a martensitie-ferritic core (Fig. 2).

257

I
TRANS. INDIAN INST. MET., VOL. 55, NO. 4, AUGUST 2002
,
I

Table 3 shows the microconstituents at the surface of XRD and WDS techniques. Lowering the temperature the thick specimens nitrided at 0.25 MPa, identified by Table 3
MICROCONSTITUENTS AT THE SURFACE OF THICK SPECIMENS NITRIDED AT 0.25 MPa

T(K) 1473 1473 1373 1373 1353 1273 1273

AISI 410 steel Micro constituents Time (ks) 0.9/1.8/3.6 a' 7.2/10.8/86.4 a' + 'Yretained 3.6 a' 6.3/10.8/21.6/86.4 3.6/10.8/21.6/86.4 1.8/3.6/63 10.8/21.6/86.4 a'+MX a'+MX a' + M2X a'+MX

AISI 410 S steel Micro constituents Time (ks) 0.9 a' 1.8/3.6/10.8/86.4 a' + 'Yretained 3.6/6.3 10.8/21.6/86.4 3.6 6.3/10.8/21.6/43.2 3.6 6.3/10.8/21.6/86.4 a' a' + M2X a' + M2X a'+MX a' + M2X a'+MX

Table 4
PRECIPITATES IN THE CASE OF SPECIMENS NITRIDED AT 0.25 MPa

Identified Precipitates

T(K) 1473 1373 1353 1273

AISI 410 None MX MX MX

AISI410S None M2X MX + M2X MX+M2x

enhanced nitride precipitation; the volume fraction increased with increasing nitriding time, as shown in Figs. 3 and 4 for the AISI 410 stet:! (nitrided at 1273 K, 0.25 MPa). In the precipitate-free specimens (nitrided at 1473 K) grain growth occurred reaching a size of 2-4 ASTM. On the other hand, in specimens with intense precipitation (nitrided at 1273 and 1353 K) grain growth was inhibited, and grain sizes of 5-6 ASTM were obtained. Fig. 5 shows the effect of nitriding time and temperature on prior austenite grain size for the AISI 410 steel nitrided at 0.25 MPa. Table 4 shows the precipitates formed inside the entire case of specimens nitrided at 0.25 MPa for 21.6 and

4 tl

I
Fig.3 : Scanning electron micrographs at the surface of the AISI 410 steel nitrided at 1273 K, 0.25 Mpa for (a) 10.8 ks and (b) 43.2 ks.

\
I

free cases were obtained in both the steels. For the other nitriding temperatures, MX and M2X precipitates were identified in AISI 410S steel, while only MX type precipitates were found in AISI 410 steel.

258

GARZON et al : CHEMICAL CHARACTERIZATION OF NITRIDED 12% Cr STAINLESS STEELS

Figure 6 shows the relation between nitriding time and nitrogen content at the surface of the specimens nitrided
0.15 c 0 200

1473 K

.&. 150

U0.10
.... Co)

I!

!
0) u i! c

;: ..

c 100 'i
50 , 0 0 10 20 30 40 50 Q

>

~ 0.05 ;:, '0

0 0 20 40 60 80 100

Nltrldlng time, (ks)

1353 K o 1273 K

Nitriding time, (ks)

Fig.4 :

Effect of nitriding time on the amount of precipitates at the surface of the AISI 410 steel nitrided at 1273 K, 0.25 Mpa.

Fig. 5 : Effect ofnitriding temperature and time on austenite grain size in the case of the AISI 410 steel nitrided at 0.25 MPa.

'#

1.2
-0(0) AISI410S -1473 K -1473K
410S

1
25 2.5 AlSI 410 1473 K AISI410S /

1.0
0.8 0.6

4>-(bIAISI410
-Q(1;1 AIS!

1!

~2
!

~
c fit
:t:::

- 12731<

fS
1.0
05 Z

&

l
0

'

/
/

"/

12.0 1:1 1.5

/'///

004 0.2
0 0 0.5 1.0 1.5 2.0 2.5

1.0

I
"
. 10

1~ K
............... 30 40

.."

.20 MItrkIInu

, so

Z
05 10 20 M_lno 30 time. (D)

time, (~.)

Distance from the surface, (mm)


Fig. 7:
Nitrogen content in steels, measured by OS, as a function of the distance from the surface: (a) AISI 410S and (b) AISI 410 precipitate-free steels nitrided at 1473 K, 0.25 Mpa for 21.6 ks. (c) AISI 410S steel with high amounts of precipitates, nitrided at 1273 K, 0.25 MP~ for 21.6 ks.

Fig. 6 : Effect of nitriding time and temperature on nitrogen content at the surface of the specimens nitrided at 0.25 MPa.

at 0.25 MPa, 1273 - 1473 K. The nitrogen content increasedwith nitriding time up to a saturation value, whichdecreased with increasing nitriding temperature. The nitrogen content decreases towards the core, as shownin Fig. 7 for three representative specimens. In specimens containing precipitated nitrides, both nitrogenand chromium contents in martensite increased withthe distance from the surface in the region where intense precipitation occurs. Beyond this region the martensitenitrogen content decreased towards the core

while the chromium content remained almost constant, as shown in Fig 8 for the AISI 410S steel nitrided at 1273 K, 0.25 MPa for 21.6 ks. At the surface of the specimens where nitrides precipitated, the martensite nitrogen content increased with nitriding time up to a point where the volume fraction of precipitates is still small and the chromium depletion of austenite is negligible. From this region, it decreases with increasing nitriding time reaching a limit value. In addition, the chromium content in martensite

259

1\ TRANS. INDIAN INST. MET., VOL. 55, NO. 4, AUGUST 2002

decreaseswith nitriding time reaching a limit value. Fig. 9 shows the relation between nitriding time and
15

nitrogen and chromium contents in martensite, at the

:\ ~\ II I

.. CJ 13 'i!-

14

12 11 10

z i!-

0'6

0.4 0.2 0

Distance

~ ~
0 0.1

0.06

0.2

0.3

0.4

0.5 (mm)

0 0 0.1 0.2 0.3 0." (mm) Distance from the sutface,

from the surface,

Fig. 8 :

(a) Nitrogen and chromium contents in martensite and (b) amount of precipitates, as a function of distance trom the surface for the AISI 410S steel with intense precipitate formation, nitrided at 1273 K, 0.25 MPa for 21.6 ks.

~ I .

::~~ i:[ ~
~

0.'
0.2
06~ 0

10

20

30

40

. '-I 50

Nltrlding time, (ks) Fig.9 :


Notrogen and chromium contents in martensite at surface as a function of nitriding time for the AISI 410 steel containing large amount of precipitates, nitrided at 1273 K under 0.25 MPa Nz..

martensite as a function of total nitrogen content at the surface of the AISI 410 steel, nitrided at 1273 K, 0.25 MPa. The experimental points were obtained by correlating the total nitrogen content at the surface, for different nitriding times, with the chromium and nitrogen contents of austenite shown in Figs. 6 and 9. Each point was taken for the nitriding times indicated in the graph. The Thermocalc curves were calculated assuming that stable equilibrium was reached for given nitrogen contents at the surface. The nitrogen content in martensite at the surface increases with the total nitrogen content up to a maximum value (0.54 wtOlo N for the mentioned steel), and then decreases reaching an equilibrium limit value. The chromium content in martensite remains almost constant for low nitrogen contents at the surface. As the nitrogen content increases with nitriding time the chromium content starts to decrease, reaching an equilibrium limit value, as a consequence of the precipitation of nitrides.

. I I I

surface of the AISI 410 steel nitrided at 1273 K,-0.25 MPa.

The chromium content of the precipitates, measured through WDS technique, was between 50 and 70 wtOlo. 3.2 Thin Specimens Gas-metal Equilibrium The WDS-measured nitrogen contents of the MX and M2Xwere 22f1 wtOlo l3:tI wt%, respectively. and OS and fusion under inert gas measurements showed that the long-term nitrided thin specimens were Figure to shows the experimental and the Thermocalc@ chemically homogeneous. In addition, quantitative calculated nitrogen and chromium contents in metallography and XRD of the collected residues showed that the amount and type of precipitates at the 260

I I , I I I I I I I I 1 , I I I

GARZON et al : CHEMICAL CHARACTERIZATION OF NITRIDED 12% Cr STAINLESS STEELS

surface and inside the specimens were the same. conditions were reached. Thereafter, it was considered that the stable equilibrium Table 5
THERMOCALC@PREDICTEDAND EXPERIMENTAlLY IDENTIFIED PRECIPITATES SPECIMENS FOR THE LONG-TERM NITRIDED THIN

Nitriding temperature (K) and pressure (MPa) 1473 K-0.25 MPa 1373 K-0.I0 MPa 1373 K-0.25 MPa 1353 K-0.25 MPa 1273 K-0.25 MPa

Predicted precipitates None None None M2X MX

AISI410 Identified precipitates None None MX MX MX

Predicted precipitates None None M2X M2X MX

AISI 410 S Identified precipitates None None M2X MX MX

! !

0.6

i:- 12 3.6fr:

i 0.4 .!:

Ut if

r
1.0 2.0

101

-ir'o

?21.6

! i
j
0
Total nitrogen content

"

'j
8

\
1.0
c:ont8nt

f43.2 r6.4

02

4 0 2.0
(wt-%J

B 2

Total nitrog8n

Fig. 10:

Experimental and Thermocalc@ calculated nitrogen and chromium content in martensite as a function of the total nitrogen content at the surface of the AISI 410 steel, nitrided at 1273 K, 0.25 MPa. The values in the graph are the nitriding times.

Figure 11 shows the measured and Thermocalc@ calculated nitrogen contents at the surface of the thin specimens as a function of the nitriding temperature and N2 partial pressure. A good agreement between calculated and experimental values was observed. Table 5 shows the Thermocalc @ predicted and the XRD-identified precipitates in nitrided thin specimens. The predicted precipitates were read in an isobar line for each temperature studied, obtaining a rather good agreement with the experimental results.

The nitrogen contents analyzed by fusion under inert gas of the extracted precipitates were 19:1:2wt'1o nd a
11:1:2wt'1ofor the MX and M2X, respectively.

In Fig. 12 the (%Ns / %Neq)ratio (total nitrogen content / equilibrium nitrogen content ratio) at the surface is plotted as a function of the nitriding time for the AISI 410 steel nitrided at 0.25 MPa, 1273 - 1473 K. The time necessary for reaching the nitrogen content

261

TRANS. INDIAN INST. MET., VOL. 55, NO. 4, AUGUST 2002

corresponding to equilibrium metal-gas decreased with the temperature.

amount of precipitates in the microstructure, that assumption could not be valid for specimens containing nitrides. In precipitate-free specimens the nitrogen content at the surface increases very quickly during nitriding, given that after 3.6 ks the surface has already attained 80%of the nitrogen equilibrium content. On the other hand, in specimens with intense precipitation, the nitrogen content at the surface increases very quickly for lower nitriding times (when the amount of precipitates is small) and very slowly for longer times. As an example, when nitriding was carried ou,t at 1273 K the nitrogen content at me surface after 3.6 ks was about 20% of the equilibrium content, and a value of about 85% was reached only after 86.4 ks. The increase in the total nitrogen content at the surface with time is related to the nitrogen uptake reaction (1/2

301 1J s

!' 20 l. \ \ \\ \ ". 1:'


C"

AISI"OS . AlSI"O

0.8T

~
10,6 '!:
1

<i +-

AISI410S AlSI"O ~ <1 <1

'

g z

C '.0

\,;\1

~
8 C
3.

t,r

0',
'

P.,= 0,25 MP'

g.. /~>/ /
~,,'//' j;:./' 02 0,' T=1473K 0.6
0,1

t
'200

~t:~:..:l.c",cc.,.,., '400 Temperature, (K)

~ Z

0,2

O.

(pressure]!, (MP.!)

Fig. 11:

Nitrogen content at the surface of thin specimens as a function of (a) nitriding temperature and (b) N2 partial pressure.

1.0T
o.a~
.J. 0.6 .,. .,.

r--

,..0'
0' .

.0",

cl

0.4 ! 0.2 0 0

r
20

1473,1<

135~ K.."""'.!
K c

/.,J)-"""'i27'

40

60

80

100

Nltriding time (ks) Fig. 12: Effect of nitriding time and temperature on the total nitrogen content/equilibrium nitrogeri content ratio (%Ns/%Neq) at the surface of the AISI 410 steel nitrided at 0.25 MPa.

N2(gas)-*[N]y) and to the nitride precipitation rate. In the specimens with intense precipitation, when the amount of precipitates is still small (shorter times), the controlling reaction is the absorption of nitrogen in austenite, which is a very fast reaction. For higher times, the controlling reaction is the precipitation of chromium nitrides, which strongly depends on the diffusion of chromium in austenite. This precipitation reaction strongly reduces the rate of nitrogen increase at the surface. During nitriding, chromium depletion of austenite occurs in the specimens containing precipitates. This depletion changes the ratio between total nitrogen content and nitrogen content in martensite (as shown in Fig. 10), thus explaining the time variation of nitrogen content at the surface shown in Figure 9. Moreover, the austenite nitrogen content / total nitrogen content ratio cannot be cal:;ulated by assuming stable equilibrium conditions, because the time necessary for nucleation and growth of precipitates is not negligible. This fact explains the differences between the experimental and calculated values in Figure 10.

3.3 Discussion
The results disclose two basic characteristics of the microstructure of high temperature nitrided martensitic stainless steels: (i) The microstructure at the surface reached the metalgas equilibrium condition after nitriding times of the order of hours. This time is much greater for nitriding conditions in which precipitates are formed. (ii) In precipitate-free specimens, the microstructure in a region of the nitrided case can be predicted by assuming equilibrium. Depending on the nitriding time and the

4. CONCLUSIONS
(1) The microstructure of the nitrided steels near the surface was mostly martensitic. Depending on the

262

GARZON et al : CHEMICAL CHARACTERIZATION OF NITRIDED 12% Cr STAINLESS STEELS

nitriding conditions, different volume fractions of retainedaustenite, MX and M2X precipitates were also observed. (2)The nitrogen content at the surface and the tendency to nitride precipitation were increased by increasing nitriding time and pressure and by decreasing temperature. (3)In the specimens containing precipitated nitrides, the martensitenitrogen content at the surface increases with the nitriding time up to a point where the volume ftactionof precipitates is still small; from this point on, it decreases reaching a limit value. In the same specimens, the martensite nitrogen content increased withthe distance from the surface in the region where intense precipitation occurs, decreasing beyond this regiontowards the core. (4) For the precipitate-free specimens the assumption that all components are in equilibrium is useful to predict the microstructure in a region of the nitrided case.
\

REFERENCES
1. 2.
3. 4.
Gavriljuk V G, ISIJ intl, 36 (1996) 738.

Bems H, and Siebert S, High Nitrogen Steels


Institute of Metal Physics, Kiev (1993) 566.

- HNS - 93.

Siebert S, Randaufsthicken nichtrostender Stahle, doctoral thesis, Ruhr-University Bochum, Germany (1994). Gavriljuk V G, and Bems H, High Nitrogen Steels: structure, properties, manufacture, aplications, Springer-Verlag, Berlin (1999) p 378.
Tschiptsc:hin A P , On the Behavior of Stainless Steels Exposed Habilitation to High Temperature Nitrogen Atmospheres Thesis for becoming Associate Professor, University of Sao Paul0, Brazil (2000) P 142 (in Portuguese).

5.

6.

Frisk 1(, Materials Science Forum, 318-320 (1999) 95.

Depending on the nitriding time and the amount of precipitates,this assumption is not valid for specimens containingnitrides. (5) Therrnocalc@calculations described fairly well the phase stability in high temperature nitrided martensitic
stainless steels.

ACKNOWLEDGEMENTS
The authors would like to acknowledge the support of the Fundacao de Amparo A Pesquisa do Estado de Sao Paulo(FAPESP), Brazil.

263