EFFECT OF POST SINTERING THERMAL TREATMENTS ON

DIMENSIONAL PRECISION AND MECHANICAL PROPERTIES IN

SINTER-HARDENING PM STEELS

Bruce Lindsley and Thomas Murphy

Hoeganaes Corporation
Cinnaminson, NJ 08077 USA

ABSTRACT

Dimensional precision is a critical parameter in net shape processing of ferrous PM components. PM

parts producers continue to pursue larger parts, but absolute tolerances dictated by the end user generally
do not scale with part size. Therefore, in larger parts, the variation in percentage change in size, or
dimensional change, must be reduced. Beyond the dimensional changes associated with pressing and
sintering of typical low alloy PM steels, sinter-hardenable alloys present some unique challenges and
opportunities for PM part manufacturing. The ability to harden a part in the sintering furnace eliminates
the need for a secondary quenching operation. The resulting microstructure of untempered martensite is,
however, not ideal for dimensional stability and mechanical properties. Tempering hardened steels
results in improved mechanical properties and dimensional shrinkage, as the martensite converts to a
more stable ferrite and carbide microstructure of higher density. In addition, many sinter-hardening
grades contain high Cu and C contents that result in relatively high amounts of retained austenite.
Retained austenite can improve impact and ductility properties, but contributes to dimensional instability
as it can transform to lower density bainite and/or martensite with thermal fluctuations. Proper thermal
treatments of sinter-hardened steels are necessary to obtain the best combination of mechanical properties
and dimensional control. This paper reviews the effects of different post-sintering thermal treatments on
the dimensional, microstructural and mechanical property changes of sinter-hardened PM steels.

INTRODUCTION

Dimensional change in ferrous sintered compacts is influenced by several factors, including particle size,
density, composition (alloying method), sintering time and temperature, cooling rate and microstructure
[1-6]. Given a base iron and sintering conditions, factors such as admixed composition and cooling rate
play an important role on microstructure, dimensional change and mechanical properties. Previous
studies detail the effects of composition on dimensional change of sinter-hardenable steels [1,5]. Upon
cooling from the sintering temperature, the high temperature phase in steel, austenite, transforms to either
ferrite plus carbide (in the form of lamellar pearlite, divorced pearlite or bainite), martensite, or a
combination thereof. The type of martensite formed is dictated by the carbon content. At carbon
contents less than 0.6%, lath martensite forms. Above 0.8%, acicular (plate) martensite forms, and a
combination of both is found in alloys between 0.6 and 0.8%. The volume change of a martensite type is
also controlled by the carbon content. Sinter-hardening utilizes alloyed steels and/or accelerated cooling
rates to achieve fully hardened, martensitic structures directly out of the sintering furnace. In addition,
austenite is often present in the microstructure in heavily alloyed regions (retained austenite).

Austenite is the highest density phase in steel, followed by ferrite plus carbide, and finally martensite,
which is the lowest density phase. This decrease in density results in a corresponding increase in
compact size, and martensite formation leads to the largest growth resulting from microstructural
changes. Length changes in pore free materials can reach 1.4% upon transformation from austenite to
martensite [7]. Tempering of martensite leads to a ferrite and carbide structure, which has a higher
density. The post sintering operation of tempering should therefore decrease the dimensional change of a
martensitic sintered compact. However, retained austenite is not a stable phase at temperatures below the
eutectoid temperature. A reduction in temperature below room temperature will continue to drive the
martensite transformation, while an increase above ambient temperature will encourage bainite formation.
Transformation of retained austenite to either martensite or bainite will result in compact growth. The
purpose of this paper is to investigate the effects of post-sinter processing on the dimensional change and
microstructure of sinterhardened steels.

EXPERIMENTAL PROCEDURE

The alloys studied were FLC2-4808 (Ancorsteel® 737 + 2wt% Cu + graphite), FLNC-4408 (Ancorsteel
85HP + 2wt% Ni + 1.5wt% Cu + graphite), FLC-4608 (Ancorsteel 4600 + 2wt% Cu + graphite) and
Ancorsteel 4300 + graphite. The compositions are listed below in Table I. The mixes were compacted
into standard TRS, dogbone tensile, Charpy impact and dilatometry test specimens at a compaction
pressure of 690 MPa (50 tsi) at room temperature. The samples were sintered in a 90% nitrogen – 10%
hydrogen atmosphere at 1120 °C and accelerated cooling was used to achieve a cooling rate of 1.6 °C/sec
(2.8 °F/sec) between part temperatures of 650 °C (1200 °F) and 315 °C (600 °F). Tempering was carried
out in a nitrogen purged furnace at 205 °C (400 °F) for the majority of test conditions. After sintering,
samples were subjected to a variety of tempering (T) and liquid nitrogen quench (LNQ) operations: 1. as-
sintered, 2. LNQ, 3. T/1hr, 4. LNQ + T/1hr, 5. T/1hr, cool to room temperature + T/1hr, 6. T/1hr + LNQ,
7. T/1hr + T/1hr + LNQ, 8. T/4hr, 9. T at 225 °C/1hr (435° F/1hr). Dimensional change (DC) from die
size and mechanical properties were evaluated for all nine conditions.

Table I. Nominal compositions (in wt.%) of the alloys studied, balance Fe.

Alloy # Designation Ni Mo Mn Cu Cr Si C

1 FLC2-4808 1.4 1.2 0.4 2 - - 0.8

2 FLNC-4408 2.0 0.8 0.1 1.5 - - 0.8
3 FLC-4608 1.8 0.5 0.1 2 - - 0.8
4 Ancorsteel 4300 1.0 0.8 0.1 - 1.0 0.6 0.6

®
Ancorsteel is a registered trademark of Hoeganaes Corporation
RESULTS

The as-sintered microstructure of the four sinter-hardenable alloys is shown in Figure 1. Alloy 1 (FLC2-
4808) is fully martensitic, while the other three alloys are predominately martensitic with a small
percentage of bainitic regions. Nickel-rich regions can be found in the FLNC-4408 (Alloy 2)
microstructure. Alloy 1 has the highest hardenability of the four alloys and can be cooled at a relatively
slow rate while still maintaining a fully martensitic microstructure. The other three alloys have lower
hardenability, and cooling rate of the sintered compact will play an important role in the final
microstructure. One should be aware that the amount of martensite will play a role in the final
dimensions of the part and how it responds to tempering. Therefore cooling rate must be well controlled
to develop reproducible microstructures and dimensions.

Alloy 1 : FLC2-4408 Alloy 2 : FLNC-4408

Alloy 3 : FLC-4608 Alloy 4 : Ancorsteel 4300

Figure 1. As-sintered microstructures of sinter-hardened Alloys 1-4 (2% Nital/4%Picral Etch).

Alloys high in carbon and copper with martensitic microstructures, such as Alloys 1-3, often contain a
large fraction of retained austenite. Figure 2a shows the martensitic microstructure of Alloy 1 using a
typical etchant, whereas Figure 2b shows the same field etched to reveal retained austenite (white). The
majority of the retained austenite is directly associated with copper rich regions. Notice the retained
austenite is present surrounding the pore network where copper concentrations are highest. Earlier work
on this alloy has shown retained austenite levels as high as 10% [1]. Given that the retained austenite is
not uniformly distributed, local concentrations of austenite near the pore network will be much higher
than the average value reported in the earlier work. Dimensional control of these sinter-hardened
structures will be governed by not only tempering of martensite but also the transformation of retained
austenite.

(a) 2% Nital / 4% Picral (b) as (a), then Na2S2O5 in Water

Figure 2. Microstructure of the sinter-hardened Alloy 1 (FLC2-4808) at 1.6 °C/sec etched to reveal (a)
martensite and (b) retained austenite (white) in Cu-rich regions (same field).

The early stages of tempering involve 3 phenomena: 1) rapid stress-relieve and diffusion of C to
dislocations at relatively low temperature, 2) transformation of retained austenite to bainite, and 3) further
C diffusion out of the martensite associated with the formation of carbides above 200 °C (400 °F) [7].
The first and third phenomena involve a reduction in length, as the body center tetragonal (bct) structure
of martensite changes to bcc ferrite and carbide. The second produces a length increase, as the high
density austenite phase transforms to the lower density bainite. To demonstrate these three stages of
tempering, dilatometry was run on sinter-hardened Alloy 1. Figure 3 shows the change in length with
temperature and it can be seen that the slope of line upon heating is not constant. Between 125 and
175 °C (260 and 350 °F), the slope is greatly reduced as a result of the first stage of tempering. This
rapid change does not require a long time at temperature, as it occurs during heating prior to reaching the
final temperature. At the 1 hour hold temperature of 205 °C, a small amount of growth can be seen (see
inset box) due to retained austenite transformation. Upon cooling to room temperature, the reduction in
sample length is observed (-0.09%). To better illustrate the tempering effects at temperature, two
samples were heated to 225 °C (435 °F) to accelerate the dimensional changes. The first sample was
quenched in liquid nitrogen prior to tempering in order to transform retained austenite to martensite. A
greater length decrease (-0.14%) is observed as the sample is heated and held at temperature (Figure 4a).
This length decrease during the hold is a result of the third stage of tempering. No growth is observed as
the retained austenite was removed prior to the dilatometric test. The sample that was not quenched in
liquid nitrogen (Figure 4b) shows a different behavior. Growth is observed during the one hour hold at
temperature. This growth is a combination of shrinkage as the martensite is tempered and a larger
positive growth as the retained austenite is transformed to bainite. The growth at 225 °C is greater than
that at 205 °C, indicating that more retained austenite transforms at the higher temperature. Overall, the
sample length decreased by only -0.07%. By understanding these phenomena, the effects of post
sintering treatments on dimensional change and stability can be better understood in PM parts.

125
100 Heating
Cooling
75
Dilation ( µ m )

50
25
0
-25
-50
-75
0 50 100 150o 200 250
Temperature ( C )
Figure 3. Tempering of Alloy 1 (FLC2-4808) in the dilatometer at 205 °C (400 °F) for 1 hour.

125 125
100 Heating 100 Heating
Cooling Cooling
75 75
Dilation ( µ m )

Dilation ( µ m )

50 50
25 25
0 0
-25 -25
-50 -50
-75 -75
0 50 100 150o 200 250 0 50 100 150o 200 250
Temperature ( C ) Temperature ( C )

a b
Figure 4. Tempering of Alloy 1 at 225 °C (435 °F) a) after a liquid nitrogen quench and b) as-sintered.

The dimensional change of Alloys 1 - 4 for the various thermal treatments is given in Table II and Figure
5. The results have been normalized to an as-sintered dimensional change of zero. Alloys 1 – 3 have
significant levels of retained austenite, resulting in significant growth (+0.13% to +0.14%) with the LNQ.
Tempering of the martensite results in shrinkage, as the bct structure of martensite converts to a bcc
ferrite and carbide structure. Shrinkage results are seen in Table II, where the pre and post temper
condition of the as-sintered and LNQ samples are compared. More shrinkage occurs during tempering of
the LNQ sample, as there is no growth contribution from conversion of retained austenite. It is
interesting to note that the growth of Alloys 1-3 upon LNQ was similar, but the shrinkage during
tempering is consistently lower for Alloy 2. The double tempered samples and the samples tempered for
4 hours (conditions 5 and 8, respectively) show growth relative to the single tempered samples as the
retained austenite transforms to bainite. The growth of samples that undergo LNQ is greatly reduced if
they have been tempered previously, conditions 6 & 7. Samples only grew between 0.05% and 0.08%
relative to the tempered condition, whereas the LNQ samples grew between +0.13% and +0.14%. This
is consistent with a portion of the retained austenite transforming during the temper, and thereby less is
present to transform to martensite during the LNQ. The T + T + LNQ samples have nominally the same
DC as the T + LNQ samples. Given that the T + T operation resulted in more growth relative to the
single T operation and thereby had less retained austenite after tempering, it is expected that the T + T +
LNQ will have more bainite and less martensite than the T + LNQ. The 225 °C temper for 1 hour
resulted in both tempering of the martensite and transformation of the austenite, and is expected to be the
most dimensionally stable sample of the single step operations. All of the results are consistent with the
tempering stages developed for wrought alloys. Finally, the tempering difference between alloys 1 & 3
and alloy 2 is likely a result of the admixed Ni in Alloy 2. The retained austenite and martensite in the
Ni-rich regions is expected to behave differently from the rest of the microstructure.

Table II. Dimensional change (%) of post sinter operations normalized to the as-sintered length.
Alloy 1 is FLC2-4808, Alloy 2 is FLNC-4408, Alloy 3 is FLC-4608 and Alloy 4 is Ancorsteel 4300 with
0.6% graphite.
1. As- 4. LNQ 5. T + 6. T + 7. T + T 8. 9. T/
Alloy # 2. LNQ 3. T
Sinter +T T LNQ + LNQ T/4hr 225 °C
1 0 +0.13 -0.10 +0.01 -0.08 -0.05 -0.05 -0.07 -0.06
2 0 +0.13 -0.06 +0.04 -0.04 +0.01 +0.01 -0.03 -
3 0 +0.14 -0.10 +0.01 -0.07 -0.04 -0.03 -0.04 -0.07
4 0 +0.06 -0.03 +0.02 -0.03 -0.03 -0.03 -0.01 -0.02

The dimensional change results indicate that higher carbon alloys 1-3 have greater amounts of retained
austenite and highly stressed plate or acicular martensite, due to the large growth resulting from the LNQ
and the large shrinkage due to the tempering, respectively. Alloy 4 is relatively insensitive to post
sintering thermal treatments. This is a result of the lower carbon content (0.53% sintered C) and no
admixed copper. The improved hardenability of Alloy 4 via a combination of Cr, Ni, Mo and Si allows
this lower carbon content to be used in sinter-hardening. The lower carbon content results in lath
martensite formation, which is less stressed than the higher carbon acicular martensite. Shrinkage upon
tempering is therefore reduced. The lower carbon and lack of copper greatly reduce the amount of
retained austenite, thereby reducing the sintered compact growth upon LNQ. Alloy 4 had a much
smaller growth (+0.06%) resulting from the LNQ quench as compared to Alloys 1-3 that had growths
more than double this amount. Alloy 4 is therefore more dimensionally stable after sintering.
 0.15
FLC2-4808
FLNC-4408
0.1 FLC-4608
Ancorsteel 4300
Dimensional Change (%)

0.05
T+ T +T + T/ T/
T T+T
LNQ LNQ 4hr 225C
0

LNQ LNQ
+T
-0.05

-0.1

-0.15

Figure 5. Dimensional change from as-sintered size due to different post sintering thermal treatments.

Given that mechanical properties are governed by the structure of the part, it is reasonable to assume that
the mechancial properties would follow these changes in dimensions. Table III shows the mechanical
properties of Alloy 1 in the 9 different thermal treatment conditions. The as-sintered hardness is quite
high in this alloy, while the other as-sintered properties are relatively low. As a reference, 60 HRA is
equivalent to 20 HRC and 70 HRA is equivalent to 39 HRC. The highly brittle martensite caused the
tensile samples to fail prior to yielding, resulting in extremely low tensile strengths in this alloy. One
property that provides reasonable performance in the as-sintered condition is impact energy, which is
aided by the high amount of tough retained austenite. LNQ causes all of the mechanical properties to
decrease, except for hardness. The increase in hardness is consistent with retained austenite
transformation to martensite.

Tempering of this alloy results in a dramatic improvement in properties. While the stress-relief during
tempering reduces hardness, TRS doubles and tensile strengths and elongations more than double. It is
interesting to note that once the martensitic structure has been tempered, little change in mechancial
properties occurs with further post-sintering thermal treatments. Continued tempering of martensite or
transformation of retained austenite by either LNQ or prolonged, elevated or repeated tempering has little
effect on mechanical properties. It appears that the stress-relief that was observed in the dilatometer
between 125 °C and 175 °C is the dominate stage of tempering for the development of good mechanical
properties. The results also suggest that tempering conditions can be modified to fine tune the
dimensions of a part with little effect on mechanical properties.
 Table III. Mechanical properties of Alloy 1 with different post sintering thermal treatments.
Sinter
Condition Density Hardness TRS 0.2% YS UTS Elong. Impact
(g/cm³) (HRA) (MPa) (MPa) (MPa) (%) (J)
As-sintered 7.03 75 979 - 389 0.4 14

LNQ 6.99 77 786 - 294 0.3 8

T 7.05 71 1869 755 1052 1.5 22

LNQ + T 7.02 72 1621 998 1086 1.0 16

T+T 7.05 71 1807 814 1053 1.6 20

T + LNQ 7.04 72 1766 857 1025 1.3 18

T + T + LNQ 7.05 70 1717 824 1002 1.1 20

T for 4 hrs 7.03 71 1745 806 1065 1.4 19

T at 225 C 7.05 70 1655 823 1046 1.3 20

The other 3 alloys were also mechancially tested in all 9 conditions. The hardness and TRS for the four
alloys is displayed in Table IV and Figure 6. Alloy 2 and 3 followed similar trends as Alloy 1. The as-
sintered and the LNQ samples had low strengths, while tempering essentially doubled TRS and increased
tensile and impact properties. Again, further thermal treatments after a single temper had little effect.
Alloy 1 had the best combination of hardness and strength for the first 3 alloys.

Table IV. Hardness and TRS for all four alloys with the different thermal treatments.
Alloy 1 - FLC2-4808 Alloy 2 - FLNC-4408 Alloy 3 - FLC-4608 Alloy 4 - 4300
Condition Hardness TRS Hardness TRS Hardness TRS Hardness TRS
(HRA) (MPa) (HRA) (MPa) (HRA) (MPa) (HRA) (MPa)
As-sintered 75 979 69 959 74 676 71 1690

LNQ 77 786 70 828 75 600 72 1593

T 71 1869 65 1517 69 1521 69 2062

LNQ + T 72 1621 68 1586 72 1362 70 1903

T+T 71 1807 66 1614 68 1528 69 2007

T + LNQ 72 1766 67 1524 69 1390 69 1986

T + T + LNQ 70 1717 66 1621 70 1472 70 1972

T for 4 hrs 71 1745 65 1572 68 1424 69 1903

T at 225 C 70 1655 64 1572 70 1472 69 1917

Alloy 4 exhibited a different behavior. The as-sintered properties were the best of all alloys tested and
the LNQ had little effect on properties. Tempering improved TRS and tensile properties by roughly
25%, and while significant for developing the best properties, tempering was much less important in this
alloy system compared to Alloys 1-3. Again, this is a function of the martensite type that is present in
the lower carbon steel. As-sintered hardness was lower than Alloys 1 and 3, but after tempering, Alloy 4
had similar hardness. The transverse rupture strength of Alloy 4 is equivalent or better than Alloy 1
(Table IV) and the tensile properties are similar. It should be noted that the sintered density of Alloy 4
was 0.05 to 0.10 g/cm3 higher than the other alloys due to a combination of improved compressibility and
less growth during sintering. Given the ability to properly sinter a Cr-containing PM alloy, Alloy 4
provides a more dimensionally stable sinter-hardening system as compared with more traditional sinter-
hardening alloys.

2500

2000
TRS (MPa)

1500 Alloy 1
Alloy 2
Alloy 3
1000 Alloy 4

500

0
T
Q

Q
T

s
ed

C
hr
LN

LN

LN
+

5
er

22
Q

r4
T
nt

+
LN
si

fo

at
T

T
-
As

T
+

T
T

Figure 6. TRS strength of the 4 alloys with different thermal treatments.

SUMMARY

The dimensional and microstructural changes of different sinter-hardening grades were investigated with
different post-sintering thermal treatments. Tempering of martensite reduces the growth associated with
its formation and produces a less brittle structure. Tempering is critical to mechanical properties in
sinter-hardened steels. It also improves the dimensional stability of the material. The presence of
retained austenite in high carbon alloys can lead to dimensional instability of a PM component. If cooled
to low temperature, the austenite will transform to untempered martensite and result in large growth. If
heated to a high enough tempering temperature, the austenite will transform to bainite. The lower carbon
martensite produced in Alloy 4 (Ancorsteel 4300) resulted in little dimensional change during the various
post sintering thermal treatments and provided equivalent or better mechanical properties than the more
traditional sinter-hardening alloy systems. The most dimensionally stable structure will contain little to
no retained austenite and well tempered martensite.

REFERENCES

1. B. Lindsley, G. Fillari and T. Murphy, “Effect of composition and cooling rate on physical properties
and microstructure of prealloyed P/M steels”, Advances in Powder Metallurgy & Particulate Materials,
compiled by C. Ruas and T. A. Tomlin, Metal Powder Industries Federation, Princeton, NJ, 2005, part 10,
p. 10-353.

2. N. Dautzenberg and H. J. Dorweiler, “Dimensional behavior of copper-carbon sintered steels”,

Powder Metallurgy International, Vol. 17, No. 6, 1985, p. 279.

3. R. L. Lawcock and T. J. Davies, “Effect of carbon on dimensional and microstructural characteristics

of Fe-Cu compacts during sintering”, Powder Metallurgy, Vol. 33, No. 2, 1990, p. 147.

4. F. Chagnon and M. Gagne, “Dimensional control of sinter hardened P/M components”, Advances in
Powder Metallurgy & Particulate Materials, compiled by W. G. Eisen and S. Kassam, Metal Powder
Industries Federation, Princeton, NJ, 2001, part 5, p. 5-31.

5. B. Lindsley and T. Murphy, “ Dimensional Control in Cu-Ni Containing Ferrous PM Alloys ”,

Advances in Powder Metallurgy & Particulate Materials, compiled by W. R. Gasbarre and J. W. von
Arx, Metal Powder Industries Federation, Princeton, NJ, 2006, part 10, p. 140-153.

6. T. Singh and T. F. Stephenson, “Ni-Cu-Mo interactions in sinter-hardening steels”, Euro PM2005,

EPMA, vol. 1, p. 261

7. R. E. Reed-Hill, Physical Metallurgy Principles – 2nd Edition, PWS-Kent Publishing, Boston, MA,
1973, p. 701-748.