MATERIAL CHARACTERISATION

Microstructural investigation of SAE 1040 steel specimens

by ultrasonic measurements
C Hakan Gür and B Orkun Tuncer

The aim of this study is to characterise the microstructures attenuation and microstructure relationship in low carbon steels[6];
of the as-quenched and tempered SAE 1040 steel specimens and the effect of carbon content on the microstructure, sound
by sound velocity and attenuation measurements. An initial attenuation, and ductile-to-brittle transition temperature in steels of
heat treatment was applied to the specimens to obtain a approximately constant grain size[7] have been investigated. Sound
uniform starting microstructure. Following 850°C/0.5 h attenuation in fine pearlitic steels used for railway wheels has been
austenitisation and water-quenching of all specimens, a studied[8]. By investigating the influence of hardening, tempering,
tempering process was applied at 200oC and 600oC for 2 h. and annealing on sound velocity it has been concluded that among
Microstructures of the specimens were characterised first the other structural factors the phase composition of the alloy and
by metallography and hardness measurements. Then, the the distortion of its crystalline lattice have the strongest influence
velocities of both longitudinal and transversal waves were on the velocity of propagation of an ultrasonic wave[9]. Sound
measured. Backwall echo patterns were obtained to compare velocity and attenuation on as-cast, annealed, normalised, hardened
the attenuation behaviour. By using the specimens having and tempered conditions of steel castings have been studied[10].
the same thickness, and by eliminating the effects of grain Then, influence of deformation and thermal treatment on sound
size and texture, only the influence of microstructural phases velocity and attenuation in steel forgings has been determined[11].
on sound velocity and attenuation was investigated. Sound Isothermal annealing of a cold worked austenitic stainless steel was
velocity increased as a function of tempering temperature, characterised by sound velocity measurements[12], and variation
however no meaningful variation in sound attenuation was in sound velocity with ageing time was compared with hardness
observed. measurements[13].
The aim of this study is to determine the effect of individual
1. Introduction phases of steels on sound velocity and attenuation. By applying
careful heat treatments on SAE 1040 specimens, samples consisting
The objective of a tempering treatment is to relieve some internal of as-quenched martensite and tempered martensite were obtained.
stress and to reduce hardness, thereby creating higher ductility than By applying the same austenitisation and quenching procedure
is possible in as-quenched parts. Plain-carbon steels are tempered prior to individual treatments, the effect of austenitic grain size on
in the range of 175 to 700oC. Tempering modifies the structure of ultrasonic parameters was eliminated.
martensite, and this change can be used to adjust strength, hardness,
toughness, and other mechanical properties to specified levels.
In order to provide longer service life with higher performance
2. Experimental
of steel structures, quality control of the properties is important. Chemical composition of the steel used is given in Table 1. To
In many cases, non-destructive tests provide a good alternative to obtain fully martensitic structure necessitates high cooling rates,
traditional destructive tests. For instance, ultrasonic techniques and the 5 mm-thick specimens were cut perpendicular to the rolling
can be used to determine the elastic constants, microstructure and direction of the hot-rolled bar with 30 mm diameter. Following
morphological factors, and mechanical properties. Sound velocity the austenitisation process at 850°C for 0.5 h, the specimens were
and attenuation measurements are the key factors in ultrasonic non- quenched in water, and then a second group was tempered at 200oC
destructive material characterisation. Variations of velocity and for 2 h, and third group was tempered at 600oC for 2 h. Before
attenuation with respect to variations in the microstructure indicate measurements, scale was removed by grinding. This way, necessary
indirectly the changes in the mechanical properties. parallelity and the surface roughness of the specimen surfaces have
Various research studies have been performed on ultrasonic been obtained, and thicknesses of the specimens became identical.
material characterisation. Variations in the attenuation and velocity On each specimen, five Vickers hardness values were taken
of both longitudinal and transversal waves in hardened and tempered from the different regions, and then a mean hardness value for
steels as a function of austenitisation temperature and ultrasonic each group was determined. In the metallographic examination,
frequency have been studied[1]. Sound attenuation has been used after being cut into half, the specimens were ground, polished and
to determine the ferrite grain size in plain C-Mn steels[2], and it has etched. The through-thickness sections were examined under an
been concluded that chemical composition and sound attenuation optical microscope and a scanning electron microscope.
together could provide estimates of yield strength, tensile strength, Ultrasonic measurements have been performed using
and transition temperature that are nearly as precise as those longitudinal and transverse waves by the pulse echo technique.
provided by composition and grain size[3]. There are additional Signals were analysed by a Panametrics-5052UAX50 analyser
studies correlating grain size and sound attenuation[4,5]. Sound and a Philips-PM3365A oscilloscope. Longitudinal and transversal
waves were produced and detected by a 20 MHz probe and a
C Hakan Gür is with the Middle East Technical University, Metallurgical 5 MHz probe respectively. A constant force is applied to the probe
and Materials Eng. Dept., 06531 Ankara-Turkey. Tel: +90 (312) 2105916;
Table 1. Chemical composition of SAE 1040 steel
Fax: ++ (312) 2101267; E-mail: [email protected]
C Si Mn P S Cr Mo Ni Al Cu Sn
B Orkun Tuncer is with the Middle East Technical University, Welding
0.39 0.34 0.77 0.012 0.007 0.16 0.01 0.07 0.019 0.25 0.013
Techn. and NDT Research Center, 06531 Ankara-Turkey.

Insight Vol 47 No 7 July 2005 421

against the specimen surface so as to have a constant thickness of the specimen tempered at 600oC. The hardness was decreased due
couplant layer at the interface between the specimen surface and the to a loss of carbon from solid solution and/or decrease in dislocation
probe. As couplant, machine oil was used for longitudinal waves density during tempering.
and honey was used for transversal waves. The sound velocity was
calculated by dividing twice the specimen thickness by the time-
of-flight between zero-crossing of the subsequent backwall echoes.
Backwall echo patterns of the longitudinal waves were obtained
using pulse echo technique (contact coupling), 10 MHz probe
and USD 15 testing equipment. The variation in beam divergence
was eliminated by using specimens having the same thickness. A
measurement condition that would bring the screen height of the
first backwall echo to 80% was maintained by adjusting the gain
and damping for each specimen.

3. Results and discussion

Microstructure and hardness
Fully martensitic structure was obtained in the quenched specimens Figure 2. Hardness values of as-quenched and tempered SAE
(Figure 1(a)). It is known that ferrous martensite has a very high 1040 specimens
dislocation density; and the morphology of 90% of martensite in
plain carbon steels having 0.4% carbon is lath-like[14]. Tempering
at 200°C did not produce a big difference in the microstructure. Sound velocity
However, martensitic microstructure was considerably changed The propagation rate of ultrasonic waves in a polycrystalline
after tempering at 600°C (Figure 1(b)). Tempering of martensite material is controlled mainly by elastic modulus and density of
at high temperatures leads to a marked reduction in the dislocation the bulk material. It is influenced by the microstructure through
density; martensite loses its tetragonality, cementite nucleates, changes in the elastic moduli of the individual grains, texture,
grows and spheroidises[14]. Results of the hardness measurements second phases, and energy absorption mechanisms. Generally, a
given in Figure 2 show that the highest hardness value belongs to substructure that strains the lattice or interrupts the continuity of the
the as-quenched specimen, and the lowest hardness is measured on matrix lowers the elastic modulus and propagation rate of the wave.
Wave velocity varies from grain to grain due to misorientation of
grains, which is related to variation in the elastic constant in the
same direction.
The parameters playing a role in variation of sound velocity in
steels include grain size, grain boundaries, the relative amounts
of phases, and lattice distortions. In this study, the influencing
parameters were minimised to correlate the ultrasonic properties
and microstructural phases. When cooling rate shows significant
variations as going from surface to the interior, differences in the
phase content through the thickness may occur. In such cases, the
hardness and the microstructure at the surface may not represent
the whole structure. The thickness of the specimens was chosen
considering the hardenability so that fully martensitic specimens
were obtained. On the other hand, since the specimens are thin,
the thermal stresses which may also be responsible for variations
in sound velocity were not high. Since the amount of the retained
austenite is very small, its contribution to the variations in sound
velocity is assumed to be negligible. No texture variations due
(a) to in grain shape and banding due to hot rolling was observed.
Previous studies stated that sound velocity is affected by both
grain size and microstructure. For instance, an increase in grain
boundary area, which means decrease in grain size, results in large
scattering of ultrasonic waves which causes ultrasonic waves to
take a longer path to cover the material thickness, which decreases
the sound velocity drastically[13]. To eliminate the effect of grain
size on sound velocity, all specimens have been austenitised and
quenched identically, and a constant tempering time was chosen.
Thus, the average size of prior austenite grains of all specimens
was maintained almost the same, which means that the parameter
affecting the ultrasonic measurement is solely the difference in the
phase content of the microstructure.
As seen in Figure 3, sound velocity measurements for both
longitudinal wave and transverse waves gave the same tendency.
The microstructure having the lowest sound velocity is the as-
quenched martensite. This can be explained by the high amount
(b) of lattice distortion. Martensite is the most random phase that
Figure 1. Microstructures of SAE 1040 specimens. a) as- causes an increase in the elastic anisotropy of the prior grain
quenched, b) quenched, then tempered at 600oC volume. Each austenite grain transforms to long and thin laths of

422 Insight Vol 47 No 7 July 2005

 Sound attenuation
Backwall echo patterns were utilised to observe the attenuation
behaviour (Figure 4). The factors contributing to attenuation are
mainly ultrasonic beam divergence, absorption, and scattering. In
steels, the main factors of scattering appear to be grain structure,
morphology and dislocation density. Absorption is caused mainly
by internal friction at the grain boundaries, and elastic hysteresis
due to imperfect elasticity. Effects of heat conduction and thermal
relaxation are negligible. Any inhomogeneity scatters ultrasonic
waves depending upon the differences between the elastic constants
and the densities of the inhomogeneity and the matrix, and upon the
kind of wave. When the wavelength becomes comparable to the
grain size by increasing the probe frequency, scattering appears.
It depends on the elastic anisotropy of the individual grains, on
precipitates and multiple phases within the grains and in grain
(a) boundaries, and on the texture of the metal[10].

Figure 4. Backwall echo patterns of the as-quenched and

tempered SAE 1040 specimens
(b)
Figure 3. Sound velocity of the as-quenched and tempered SAE The analysis of scattering becomes difficult when different
1040 specimens. a) Longitudinal wave, b) Transversal wave
phases are present in the microstructure. The orientation of grains
with respect to the propagation direction of the wave will also affect
martensite suddenly by diffusionless lattice shear. Since these laths the degree of attenuation in the multi-phase structure. Since the
are highly dislocated, martensite contains very large dislocation same probe has been used, and the thickness of the specimens is the
density that is comparable to a cold-worked structure. It has been same, the effects of the measurement system and beam divergence
previously reported that sound velocity decreases with the increase can be ignored. All specimens consist of a single phase, and the
in the degree of deformation on the material, and this decrease is average prior austenite grain size of the specimens is almost the
only due to the increase in dislocation density[11]. It has also been same since the same austenitisation procedure was applied.
reported that sound velocity changes by the changes at the lattice Transformation of austenite grains to martensite introduces
distortions and dislocations[12]. Increase in sound velocity has been internal stresses resulting from lattice deformation, being
explained by the reduction in the distortion of the lattice caused by accompanied by an increase in dislocation density, and this causes
the annihilation of point defects and dislocations[13]. a greater absorption of sound energy[7,11]. It is known that for a
The sound velocity in the specimen tempered at 200°C is slightly given cooling procedure, the higher the austenitising temperature,
higher than that in the as-quenched specimen. Above 100°C, ie, the larger the prior austenite grains, the higher the sound
normally the dissolved carbon precipitates into carbides, which attenuation. The effect of cooling severity on attenuation from a
is associated with a decrease of the specific volume of martensite given austenitisation temperature, ie, varied phase content in the
and by this volume change micro-residual stresses completely microstructure, however, is not so clear. Since the effect of prior
relieve. Due to the decrease of macro-residual stresses, sound austenite grain size was eliminated, these curves represent only
velocity increases. Partial loss of tetragonality, and a very small the effect of phases. The backwall echo patterns show that the
drop of dislocation density may also be considered for this slight attenuation behaviour of as-quenched and tempered specimens is
increase. However, tempering of martensite at 600°C produces very similar (Figure 4). However, it can be said that tempering has
some definite changes in the martensite: carbide particles form a slightly lowering effect on sound attenuation. These results show
within and between the remanent martensite laths which are almost that compared to the effect of the average grain size, the effect of
dislocation-free. The marked reduction of dislocation density, thus microstructural phases on attenuation is very slight.
residual stresses, combined with the changes in the morphology,
results in a less distorted lattice and, therefore, an increase of sound
4. Conclusion
velocity compared to the as-quenched martensite. These results
support those of the limited number of studies: sound velocity Microstructures of the as-quenched and tempered SAE 1040
of oil-quenched SAE 4150 specimens increases on tempering at specimens were investigated, and hardness, longitudinal and
650oC primarily due to differences in the elastic moduli[13], whereas transversal wave velocities and attenuation measurements were
in another study it was concluded that the phase composition of the done. Application of an identical austenitisation and quench
alloy and the distortion of its crystalline lattice have the strongest procedures eliminated the influence of the prior-austenite grain
influence on the sound velocity[12]. size and texture on ultrasonic properties. The lowest sound velocity

Insight Vol 47 No 7 July 2005 423

of as-quenched specimen is due to high amount of tetragonal 6. R L Smith, W N Reynolds, H N G Wadley, Ultrasonic
lattice distortion and increased elastic anisotropy caused by Attenuation and Microstructure in Low Alloy Steels, Metal
austenite-martensite transformation. Tempering at 200°C results in Sci, Vol 15, pp 554-558, 1981.
a slight increase in the sound velocity because of partial loss of 7. R L Smith, K L Rusbridge, W N Reynolds, B Hudson,
tetragonality and a very small drop of the dislocation density. After Ultrasonic Attenuation, Microstructure, and Ductile to Brittle
tempering at 600°C, the dislocation density reduces remarkably and Transition Temperature in Fe-C Alloys, Mater Eval, Vol 41,
the morphology of the microstructure changes. This causes a less pp 218-22, 1982.
distorted lattice, and therefore, a detectable increase in the sound 8. B Kopec, V Hanak, Ultrasonic Attenuation in Quenched
velocity compared to the as-quenched structure. No meaningful Railway Steels, NDT Int, Vol 22, pp 21-25, 1989.
variation in the sound attenuation as a function of tempering was 9. V V Muravʼev, Interrelationship of the Velocity of an Ultrasonic
observed. Wave in Steels and Their Heat Treat Cycles, Plenum Publ.
Corp, pp 135-37, 1989.
References 10. R Prasad, S Kumar, An Investigation into the Ultrasonic
1. E P Papadakis, Ultrasonic Attenuation and Velocity in Behaviour of Cast and Heat-Treated Structures in Steel, Insight,
SAE 52100 Steel Quenched from Various Temperatures, Vol 33, pp 506-08, 1991.
Metallurgical Trans, pp 1053-57, 1970. 11. R Prasad, S Kumar, Study of the Influence of Deformation
2. R Klinman, G R Webster, F J Marsh, E T Stephenson, and Thermal Treatment on the Ultrasonic Behaviour of Steel,
Ultrasonic Prediction of Grain Size, Strength, and Toughness J Mater Processing Techn, No 42, pp 51-59, 1994.
in Plain Carbon Steel, Mater Eval, Vol 38, pp 26-32, 1980. 12. M Vasudevan, P Palanichamy, S Venkadesan, A Novel
3. R Klinman, E T Stephenson, Ultrasonic Prediction of Grain Technique for Characterising Annealing Behaviour, Scripta
Size, and Mechanical Properties in Plain Carbon Steel, Mater Metallurgica et Materialia, Vol 30, No 11, pp 1479-83, 1994.
Eval, Vol 39, pp 1116-20, 1981. 13. P Palanichamy, M Vasudevan, T Jayakumar, S Venugopal,
4. P Palanichamy, A Joseph, T Jayakumar, Grain Size B Raj, Ultrasonic Velocity Measurements for Characterising
Measurements in Austenitic Stainless Steel Using Ultrasonics, the Annealing Behaviour of Cold Worked Austenitic Stainless
Insight, Vol 36, pp 874-878, 1994. Steel, NDT&E Int, Vol 33, pp 253-59, 2000.
5. E P Papadakis, From Micrograph to Grain-Size Distribution 14. R W K Honeycombe, H K D H Bhadeshia, Steels: Microstructure
with Ultrasonic Applications, J Appl Phys, Vol 35, pp 1586- and Properties, 2nd Ed. Edward Arnold-London, 1995.
1594, 1965.

Enquiry No 507-05

424 Insight Vol 47 No 7 July 2005