ISIJ International, Vol. 33 (1993), No, 4, pp.

508-512

Elastic Moduli and Internal Friction of Low Carbon and Stainless

Steels as a Function ot Temperature

Mikio FUKUHARA
and Asao SANPEl
Technical Research Laboratory, Toshiba Tungaloy Co.. Ltd.. Tsukagoshi, Saiwai-ku. Kawasaki. Kanagawa-ken, 21 OJapan.

tReceived on October 14. 1992, accepted in final form on February 25. 1993)

Eiastic (Young, shear and bulk moduli. Poisson's ratio and Lam6parameter), Iongitudinal and transverse
internal friction values for low carbon steel and stainless steel were simultaneously measured over a
temperature range 300-1 500 K, by an ulttasonic pulse sing-around method. These elastic moduli decrease
and Poisson's ratio increases with increasing temperature, suggesting activation of shear modein a high
temperature region. Longitudina] and transverse internal frictions are sensitive to recrystalization, and to
oc(ferritic) /y (austenitic) phasetransition and solution of precipitated carbide phases into the austenitic matrix,
respectively. A relaxation peak with an apparent activation energy of 0.97 eV was observed at around 61 OK
for the carbon steel.

KEYWORDS:
Iow carbon steel; stainless steel; Young's modulus; shear modulus; bulk modulus; Lam6
parameter; Poisson's ratio; internal friction, relaxation peak.

1. Introduction
It recognized
is that
constantselastic
temperature range are important material parameters
for evaluation and design of ferrous alloys, whereas
no research work has carried out on simultaneous
measurementof various elastic parameters (Young, shear
and bulk moduli, compressibility and Lam6parameter)
from room temperature to high temperature up to
l 500 K. If these parameters are available up to high
temperature, one could precisely design the alloys for
high-temperature structural application with hlgh reli-
ability.
Internal technique provides a microscopic
friction
mechanical properties such as atomic diffusion, disloca-
tion bowing and grain boundary softening. Since the
pioneering works of internal friction in polycrystalline
metals by Zenerl) and K~,2) the internal friction studies
have been reported by various research groups3~5) up
to the elevated temperature. However, high-temperature
simultaneous measurementof internal friction by both
dilational and shear vibration modes has not been
reported as long as we know.
In previous papers,6 ~ 8) Young, shear and bulk moduli,
Poisson's ratio and Lam~parameter, Iongitudinal and
transverse internal friction values and acoustic wave
velocity anisotropy factors for four kinds of polycrys-
talline compounds, c(-Al203, (Y)TZP
tetragonal zirconia polycrystal) and p'-sialon, (Si, Al)3
(N, O)4 and ce-SiC, and two kinds of metallic compounds,
Inconel 718 and Ti-6AI-4V were simultaneously mea-
sured over a temperature range 295-1 818 K, and 298
l 385 K, respectively, by an ultrasonic pulse sing-around
method.
In this study, we report temperature dependence of
in elevated the elastic moduli, Poisson's ratio, Iogitudinal and
transverse internal frictions for low carbon steel, SIOC
and austenitic stainless steel, SUS304. The former is a
representative low carbon steel for machine structural
use, and the latter is the most commonlyused austenitic
stainless steel for acid-proof and heat-resisting steels.
Since these steels are isotropic polycrystal, the usual
elastic moduli can be calculated from two independent
moduli, i,e., Youngand shear ones, using the speed of
10ngitudinal and transverse waves.6) Therefore presenta-
tion of other elastic parameters is not generally necessary.
Since, however, it was very difficult for these ferrous
materials to measure simultaneously the speed of both
waves in high temperature range, the bulk modulus,
Poisson's ratio and Lam~parameter at elevated tem-
perature have not been reported in contrast with the
Youngand shear moduli. The bulk modulusis a measure
of the stiffness of the crystal or of the energy required
to produce a given deformation.9) Poisson's ratio is the
most important parameter for material design at high
temperature. In addition, Lam6parameter seems to be
sensitive to physical behaviors connected with shear; for
an example, the parameter of (Y)TZP, which is a ceramic
like steel, is fairly sensitive to relief of strain such as.
superplasticity and to relaxation damping.6) Thus we
(yttria-stabilized present all elastic moduli as a function of temperature,
- and discuss their acoustic properties in view of phase
transformation, softening of grain boundaries and
deformation mode, because high-temperature strength
effectively depends on material structureio) and residual
stress I I )
.

C 1993 ISIJ 508

 ISIJ International, Vol. 33 (1993), No. 4
Table l. Chemical composition and hardness of two kinds of steels used in this experiment.

Chemical composition (masso/*)

Material Hardness
U.S. designation
(JIS) C Si
Mn P S Cr Ni
(HB)

SIOC O.
11 0.24 0.65 0.017 0,010 O. 12 83 AISISAE1010
SUS304 0.06 0.48 O,
IO 0.35 0.024 I8.25 8,09 156 SAE30304

Although high-ternperature dynamic measurementsof Yaung's Medulus

Young and shear moduli have mainly carried out by
flexural resonance4) or torsional2,3,s) method, these
~
O 200
Shoar Modulus
Bulk Moduius
methods have many disadvantages such as uncertalnty 'o Lam6Parameter
of dimensionality, high frequency fatigue under applied s
stressl2) jn flexural :'
resonance, adjustment of suspension 1'
position,13) exfoliation or deterioration of coated elec- o
trical film on the sample. In contrast, the pulse method
E 1OO
is suitable for high temperatures wheninternal damping
~
makesthe resonant method very difficult.14) Especially,
[i
the ultrasonic pulse sing-around method is still more
suitable for accurate measurement.4)
o
300 600 900 1200 1500
TemperatUre (K)
2. Experimental Procedure
Fig. 1 Young, shear, bulk moduli and Lam~parameter of
Chemical compositions and hardness of two ferrous low carbon steel. SIOCas a function of temperature.
steels used in this study are listed in Table
I with the
corresponding U.S. designations. The former is as cold
work and the latter is annealed. A11 elastic moduli and
internal friction values were accurately measuredin argon ~ 200
O
of ambient pressure by means of the ultrasonic pulse
sing-around method6,15) with zero cross time detection
and multiple delay circuits.
transverse velocities were determined by
Both longitudinal
meansof mode
and
~
o
conversion in one run. The specimen was in the form of
Ei 1Oo
a long rod (20 mmin length) united by buffer rod
:F
(waveguide) with threads of pitch I .5mm, to eliminate
the generation of spurious signals by modeconversion Ei
at the sides.16) The experimental procedure is minutely O
described in the previous paper.6) 300 6OO 900 1200 1500
In order to avoid propagation loss due to high fre- TemperatUre (K)
quency of ultrasonic, the frequencies of 5 and 2.25 Fig. 2 Young, shear, bulk moduli and Lam6parameter of
MHZWere used for the low carbon and the stainless austenitic stainless steel, SUS304 as a function of
steels, respectively. An apparent activation energy for a temperature.
relaxation peak was determined using the frequencies of
2.25, 5and 10 MHz.The specimens were heated at a rate sibility that the change is due to oc(ferritic)/y(austenitic)
of O.25 K/s to 1500 K, where transverse wavedisappears. phase transition. The larger modulus-temperature slop~s
Determination of all elastic moduli in the high tem- in ferritic phase range show higher deformability of the
perature range is limited by this disappearance of the ferritic phase comparedwith the austenitic phase one at
transverse wave. high temperature.
In comparsonwith other frequency resonance datal7)
3. Results of the normalized steel with O. 12 masso/o up to 873 K, C
Youngand shear moduli in Fig, are somewhathigher I
3.1. Temperature Dependenceof Elastic Moduli
K
(206.6 GPaat 293 and 158.7 GPaat 873 in Young's K
The four elastic moduli (Young, shear and bulk modulus, 80.7GPaat 293 K and 60.7GPaat 873 K in
moduli, Lam6parameter) of the low carbon and the shear modulus) over whole temperature range. Static
stainless steels as a function of temperature are shown Young's modulus curve reported by Hara et al.18) in
in Figs.
I and 2, respectively. The four moduli of the elevated temperature range up to 1222 K is similar to
carbon steel decrease substantially as the temperature the curve of Young's one in Fig. I but their data (96 GPa)
,
increases. The modulus-temperature slope of Young, at 1222 K is lower than that of the latter, Fourthermore,
shear and bulk moduli and Lam~parameter change to their data are subject to large variations. There is a
the more sluggish one at around I OOO K. There is a pos- difference of 7o/* in room temperature value of bulk

509 C 1993 ISIJ

 ISIJ International, Vol. 33 (1993). No. 4
0.38

O
0'36
o
1-
E, Loflg~Jdin~d Wave

~ e sloc >( 2.0 C Transve,~5e Wave

,O
C:o
0.34 O
CO c:
,O 0.32 2
o 1
.O
O
CL
1'
dE
0.30
c5
CL
0.28
o 0.0
30O 600 900 1200 1500 ~
C:
600 900 1200 1500
300
,

Temperature '
(K)
TemperatWe (K)
Fig. 3. Poisson's ratio oftowcarbon steef;~roc andaustenitic friction for
Fig. 4. Temperature dependencc of intemal '

staintess steeL SUS304 as a tunction of temperature. longitudinal and transverse wavcsof low carbon
steel. ,

SIOC.
modulus compared with data (158.9GPa) by Date.17)
Lam6parameter decreases only slightly between room el 2
temperature and the oe/y phase transition temperature and o
r a Loogitudinal wave
then sluggishly increases up to 1290 K. Over 1290 K, it >(
' Traosverse wave
begins to decrease again.
Thetemperature dependenceof Young, shear and bulk O
moduli in the stainless steel shows monotony decrease 4
with increasing temperature. Roomtemperature value O
t: 1
~:
of Young's modulus is consistent with the dynamic data o
reported by Andrews,19) but the modulus-temperature L
LL
slope is smaller than the reported data (118.7GPa at ,u
,:
1270K). Shear modulus at room temperature is a little
*
lower than the other data20) (86.2 GPa), whereas degree ~o o300
E:
600 900 1200 1500
of decrease in elevated temperature range is also smaller
TemperatUre (K)
than the data (57 GPaat 922 K), as well as in Young's
modulus. Onthe other hand, Lam6parameter increases Fig. 5. Temperature dependence of internal friction for
only slightly betweenroomtemperature and 1421 K, and longitudinal and transverse waves of austenitic
then begins to decrease from around 1421K. The stainless steel. SUS304.
increase in temperature range up to 1421 resembles K
augmentation from the o(/y point in austentic carbon steel 10
of Fig. l.
Both Poisson's
stainless steels
ratios for the low carbon
Overlapping of
are presented in Fig. 3.
and the ~
E
two lines from around I OOO means that Kthe temprature
~
correspond to the a/y phase transition. Poisson's
may 05
ratios for the stainless steel over whole temperature range
are almost the sameas nickel-based alloy, Inconel 718
(-0.36),8) but are smaller than that of Ti-6Al~V ~:
~
(-0.45).8) In contrast, Poisson's ratios of the iron
group-metals are considerably large comparedwith those
of c(-SiC (-0.17)7) and p-Si3N4 (-0.28)6) coming from 1
its covalent nature in bonding, suggesting activation of 1 55 1.65 1.75
shear mode in deformation at elevated temperature 10~3 /T (1 IK.)

range. Fig. 6. Arrhenius plot of ultrasonic frequency for Iow

3.2. Temperature Dependenceof Internal Friction
temperature peak in low carbon steel, 10C. S
Internal friction curves of longitudinal and transverse
wavesfor the low carbon and stainless steels are presented frequence dependency was investigated. Since the
in Figs. 4
and 5, respectively, as a function of temperature. frequency Fvarieswith the peak temperature according
The longitudinal curve of the low carbon steel begins to to
I K
increase from around 080 which corresponds to A3
F=Fo exp(E/kT)
point, while the transverse one shows three peak at 610, ,

l OOO 1
and 238 K. Since the lowest temperature peak may the apparent activation energy found from the slope
is

be related to relaxation one in iron crystal lattice, the (Fig. 6) to

= (
be I .55 x lO~ 19 J 0.97 eV). The I OOO peak K
C 1993 ISIJ 51 o
 ISIJ International. Vol. 33 (1993), No. 4
Table 2. The modulustemperature slope of bulk modulus for two steels, SIOCand SUS304,cc-Al203, (Y)TZP, p'-sialon,
c(-SiC. Inconel 718 and Ti-6A1~1V.

Materials
SIOC SUS304 cc-Al203 (Y)TZP p'-sialon c(-SiC Inconel 718 Tl 6AI~IV

Modulus temperature slope (1/K) 22x 10 4 -1.4x l0~4 -3.9 x l0~5 -9.7x IO-s
_ 1.4xl0-4 - 4.2xl0~5 - 4.2x lO~ 4
-
4.3 x l0~5
Temperature range (K) 302-1 478 298-1 416 29 1-1 773 306 1773 292l 771 298-1 539 309--1 379 300-1 172

corresponds to changing point of the modulus-temper-
ature slope in Fig. 1, showing the o(/y phase transforma-
tion. When measurementwas repeated from room tem-
perature to 1500 K, the peak at 1238 in Fig. 4 disap- K modulustemperature slope as well as Young, shear and
peared. This shows that the 1238K peak is the so-
called annealing peak or due to lattice defects such
as dislocations and point defects, produced by cold-
working. The origin is not clear; it needs further in-
vestigation.
Figure5 shows the internal friction of the stainless
steel. Rapid increase from around 1200 K for longi-
tudinal wave could perhaps be due to recrystalization
of grains in the austenitic phase. For transverse wave,
besides manysmall peaks, internal friction showsgradual
increase from roorn temperature to around I 122 K and
then slow decrease; it indicates one broad peak at around
1122 K. This broad peak temperature corresponds to
solution limit of precipitated carbide phases into the
austenitic matrix, expected from pseudo-binary vertical
section between iron-solid solution (18 masso/o Cr,
masso/o Ni) and carbon.2 1) In contrast with the relatively
smoothinternal friction curve in ceramics,6,7) manysmall
peaks are observed at 449, 556, 661, 793, 888, 983, 055,
l I 1
122, 194, 308 K. These peaks are repoducible within
exprimental errors. As regards the origins of these peaks,
since there are so
manypossibilities, e,g., precipitation
of carbides, mechanical dampingby dislocation motion,
migration of intersticial atomes etc., we can not make
any assignment at present time. It needs further in-
vestigation.
4. Discussion
The modulus-temperature slope in Youngand shear
moduli obtained by the ultrasonic pulse sing-around
method is smaller than that by resonant frequency
method. Similar behaviors were observed in ceramics6)
such as c(-A1203, (Y)TZP and p'-sialon. This would be
due to fatigue deterioration in high frequency resonance
method.12) Furthermore, the resonant method prevents
measurementsabove recrystallization temperature, as
assumedfrom study by Bell. 13) Wecan determine elastic
parameters in higher temperature range beyond re-
crystalization temperature, by meansof the ultrasonic
pulse sing-around method.
The modulustemperature slopes of bulk modulus for
two steels are listed in Table 2together with the data6 ~ 8)
for other materials. The slopes for these two steels are
almost the sameas those of (Y)TZP, p'-sialon and nickel
based alloy, and are 10 times larger than those of
cc-alumina, c(-SiC and titanium based alloy. This could
be explained by binding strength amongatorns in these
crystals.
Although Lam~parameter shows generally negative
bulk moduli,6 -8) as deduced from definition for C12 in
the elastic constant matrix, the austenitic low carbon and
stainless steels show the positive modulus-temperature
one. Wecan not makeclear this
reason at present time.
As reported in studies of the (Y)TZP
previous
ceramic6) and Ti6AI~l,V alloy,8) abrupt increase in
internal friction for longitudinal wave corresponded to
superplasticity accompanied by relief of strain. Since
superplasticity has been explained by motion of dislo-
cation along grain boundaries,22,23) by analogy we in-
fer that this rapid increase of internal friction in Fig. 5
is connected with recrystallization where formation and
growth of visible strain-free grains occur.24)
Onthe other hand, the transverse waveis susceptible
to phase transformation. Many studies2s,26) have re-
8 ported on a relation between phase transformation and
lattice softening indicated by decrease in shear phonon
mode. Thus it is assumed that the o(/y phase trans-
I K
formation gives rise to the I OOO peak in the low car-
bon steel.
Arelaxation peak was observed at around 610K for
the carbon steel. Theapparent activation energy (0.97 eV)
of this peak is not coincident with that (O.57 or 0.46 eV)
of Hasiguti peak27) which is observed at 503 or 471 K,
respectively, under the frequency of I kHZin cold worked
ferritic Whena line in Fig. 6 is extrapolated to
iron.
l Hz, corresponding peak temperature is 329 K. This
the
temperature is not coincident with Snoek-Kbster peak
at around 543 K,28) but close to a peak temperature
(around 300 K) of Bordoni y peak which has found in
worked ferritic steel.29) Bordoni peak in body-centered
cubic metals has been explained by a thermal instability
of the screw dislocation segments.30) However, the
detailed study for the peak in high frequency range is
called for, because the frequency dependent damping,
which occurs from breakway of pinned dislocation loop
oscillating under the influence of an applied stress, is
found to have a maximum in the high megacyclerange, 31)
5. Summary
Elasticmoduli (Young, shear and bulk moduli,
Poisson's ratio and Lam6parameter) and longitudinal
and transverse internal friction values for low carbon
and stainless steels were simultaneously measured over
a temperature range 300-1 500 K, by an ultrasonic pulse
sing-around method. These elastic moduli decrease and

51 1
C 1993 ISIJ
 ISIJ International. Vol. 33 (1993), No. 4
Poisson's ratio increases with increasing temperature, l l) M. Fukuhara: J. Am. Ceram. Soc., 72 (1989), 236.
suggesting activation of shear modein elevated tem- l 2) D. E. Macdonald: Engineering Fatigue Mechanics, 8(1976), 17.

perature region. The larger modulus-temperature slope l 3) J. B, Wachtman,Jr. W. E. Tefft: Rev. Sci. Instrum,, 29 (1958),
517.
of Young, shear and bulk moduli, and parabolic in-
l4) J. F. W. Bell: Phylos. Mag., 2 (1957), I 13.
crease in Poisson's ratio in ferritic phase range of the I5) N. P. Cedroneand D. R, Curran:J. Acoust. Soc. Am., 26 (1954),
carbon steel show higher deformability of the ferritic 963; R. L. Forgacs: J. Accousl. Soc. Am., 32 (1960), 1697.
phase comparedwith the austenitic phase one. Longi- i6) A. Sather: J. Acoust. Soc. Am., 43 (1968), 1291; H. J. McSkimin
tudinal and transverse internal frictions and E. S. Fisher: J. App/. Phys., 31 (1960), 1627.
are sensitive 17) Z, H, F. Date: J. Iron Stee! Inst., 207 (1969), 988.
to recrystallization accompaniedby relief of strain, and
to oc(ferritic)/y(austenitic) phase transition and solution
l 8) T. Hara, M. Takata, K, Ota and T. Karasudani: J. Iron and Steel,
49 (1963), 27.
of precipitated carbide phases into the austenitic ma- 19) A. Krisch: Stah/ Eisen, 73 (1953), 1349.
trix, respectively. A relaxation peak with the apparent 20) SAEHandbook, Society of Automotive Engineers,
(1971), 244.
NewYork,
activation energy of 0.97eV was observed at around
21) Acid-Proof Alloys (in Japanese), Japan Soc. Met. Sendai, (1971),
K
610 for the carbon steel. 42.
22) Y. Uematu: Private communication, Nisshin Steel.
Acknowledgements
23) R. C, Gifkins: Meta!! Trans, 7A (1976), 1225; A. Ball and M, M.
Wewould like to thank Dr. T. Matsuyamaof Toshiba Hutchinson: J. Met. Sci., 3 (1969), l.
Ceramics, Mr. Y. Uematsuof Nisshin Steel, Mr. M. 24) Y. Katsumura and M. Fukuhara: High Tech Ceramics, 6th Int,
Sakakibara of Nippon Steel and Mr. H. Uchida of Kobe Meeting on Modern Ceramic Technologies, Milan, Italy, June
Steel for useful discussions. 1986, ed. by P. Vincenzini, E]sevier, Amsterdam,(1987). 2735.
25) C. S, Barret: Structure of Metals, McGraw-Hill, NewYork,
(1943), 375.
REFERENCES 26) G, Shirane: Rev. Mod. Phys., 46 (1974), 437; J. Perkins: Scr.
l) C. Zener: Phys. Rev., 40 (1956), 906. Metall., 8 (1974), 31.
2) T. S, Ke: phys. Rev., 71 (1949), 533. 27) N, Nakanishi, Y. Murakamiand S. Kachi: Scr. Metal!., (1 971),
5
3) J. L. Snoek: Physica,
8 (1941), 71 l. 433; T. Suzuki and M, Wutting: Meta!1. Trans,, (1972), 155. 3
4) P. G. Bordoni: Ric. Sci, 16 (1949), 851. 28) R. R. Hasiguti, N. Igata and G. Kamoshita: Acta Metall., 10
(1 962), 442.
5) W. K6ster, L. Bangert and R. Hahn: Arch. Eisenhtittenwes., 25
(1954), 442. 29) L. B, Magalas, J. F. Dufresne and P. Moser: J. Phys. Colloq.,
6) M. Fukuhara and I. Yamauchi: contributing to J. Mater. Sci, Supp!. 10, 42 (1981). C5-127.
7) M. Fukuhara and Y. Abe: J. Mate,'. Sci. Lett., in press. 30) G. Fantozzi and I. G. Ritchie: J. Phys. Colloq., Supp!. 10, 42
8) M. Fukuhara and A. Sanpei: J. Mater. Sci. Lelt., in press. (1981), C5-4.
9) C. Kittel:Introduction to Solid State Physics, fourth ed.. John 31) P. Asti6. J. P. Peyrade and P, Groh: J. Phys. Colloq., Suppl. IO,

Wiley & Sons, NewYork, (1971), I lO. 42 (1981), C5-91; P. Groh and H. Schultz; J. Phys. Colloq., 42
(1981), C5-25.
l O) M. Fukuhara: "Super High Strength Ceramics," Super Fine
Ceramics, 19th Ceramic Lecture Meeting. JapaneseCeramic Soc., 32) A. Granato and K. Lticke: J. App/. Phys., 27 (1956) 583.
Tokyo. Sept. (1985), 27.

@1993 ISIJ 51 2