You are on page 1of 6

The Effect of medium strain rates on the mechanical properties of high

performance steels
S.Y. Kong & A.M. Remennikov
University of Wollongong, Wollongong, New South Wales, Australia
B. Uy
University of Western Sydney, Sydney, New South Wales, Australia

ABSTRACT: In this study an investigation into the effect that medium strain rates have on the mechanical
properties of cold drawn mild steel and 304L stainless steel was carried out using two 6mm diameter by
14mm long specimens machined from each grade of steel. The tests were performed using an Instron test ma-
chine under a quasi-static strain rate, and a high capacity instrumented falling weight axial impact (IFWI) test
rig developed at the University of Wollongong was used to conduct the medium strain rate tests. The axial
load time histories were measured by a load cell connected to each specimen and the deformation time history
by a strain gauge attached to each specimen. All the data was acquired by a high-speed data acquisition sys-
tem at a sampling frequency of 300 kHz. The load histories showed high frequency oscillations and were dig-
itally filtered by using a fourth order low pass Butterworth filter. The results showed that the yield strength of
both types of steels increased with an increased rate of strain.

1 INTRODUCTION types of austenitic and duplex stainless steel. An-


dersson et al. (2005) showed the dynamic response
Structures under impact and blast loading deform of four types of stainless steel and used the results to
rapidly, which induces a high strain rate in the mate- evaluate the parameters in the Cowper-Symonds,
rial. According to Harding et al. (1960), most metals Johnson-Cook and Jones models.
and alloys show a significant change in their me- Peixinho (2006) studied the response of H400
chanical response under increased rates of strain. stainless steel and dual-phase steel at strain rates in
The strain rate of structures under impact loading is the range of 0.001-1000 s-1. Both steels are strong
in the range of 0.1-200 s-1, but it is more than 200 s-1 candidates for use in crash-absorbing automotive
under blast loading. Instrumented falling weight im- components due to their high strength, ductility, and
pact (IFWI) testers are normally used to test material capacity to absorb energy. The results showed that
properties under medium strain rate (0.1-200 s-1) the strength of both steels increased with an increase
while a split Hopkinson pressure bar can be used to in the strain rates, and then the Cowper-Symonds
obtain material properties under high a strain rate. coefficients were derived. Uy et al. (2009) studied
The TM5-1300 outlines several important re- the dynamic behaviour of 304L stainless steel in the
sponses under high strain rate loading for mild steel range of 300-1000 s-1 using a tension split Hopkin-
compared to static loading conditions. The yield son bar.
stress increases significantly under a high strain rate A lot of studies have been conducted to investi-
while the ultimate tensile strength is not as sensitive gate the dynamic behaviour of stainless steel under
to the strain rate. The modulus of elasticity will gen- high strain rates, but only a few have been have been
erally remain insensitive to the rate of loading while carried out under medium strain rates. Based on this
elongation at rupture remains the same or is slightly research background, the dynamic response of stain-
reduced. less steel under medium strain rates will be investi-
Nicholas (1981) presented experimental results gated in this study using a high capacity IFWI test
for various grades of stainless steel including AISI rig developed at the University of Wollongong.
304, 321, 347, and 410 for strain rates up to 1000 s-1.
Zabotkin et al. (2003) showed the design concept of 2 COWPER-SYMONDS RELATION
an instrumented falling weight impact (IFWI) tester
and tested 316L stainless steel, titanium alloy, and The Cowper-Symonds relationship is an empirical
alloy C22 under medium strain rates. Cunat (2000) equation that describes material behaviour at differ-
presented Cowper-Symonds coefficients for various ent strain rates. The relationship between the dynam-
ic stress of a material to the strain rate is given in Wollongong. The drop hammer weighs 590 kg. The
Equation 1. height of the drop hammer can be raised up to 3 m

 
high to achieve the desired impact energy and veloc-
 d   s 1   ' / D 
1/ q
(1) ity. During the test the drop hammer is released from
a predetermined height to impact on top of the steel
where  d = dynamic stress or strength;  s = static frame. The steel frame then accelerates downwards,
stress or strength;  ' = strain rate; D and q are the pulling the specimen until it fractures. Obviously,
Cowper-Symonds coefficients. Jones (1989) pre- varying the drop height means that the time needed
sented Cowper-Symonds coefficients for 304 stain- to fracture the specimen varies, which means that
less steel as 100 ( D ) and 10 (q) respectively. The different strain rates can be obtained.
main focus of this study was to investigate the valid- Before the test rig was fabricated, preliminary FE
ity of these coefficients under medium strain rates. analyses were carried out using the general purpose
finite element programme ABAQUS to evaluate the
3 DESIGN OF IFWI TEST RIG design concept. The actual geometry of the test rig
and a cylindrical steel specimen were modelled. The
Figure 1 shows the front view of the high capacity drop hammer was simplified into a cylinder striker
IFWI test rig developed at the University of Wol- with a mass of 590 kg. The performance of the test
longong. It consists of: 1) an impact transmitter rig was evaluated by subjecting it to different impact
frame, 2) two guide rails, 3) a load cell, 4) a nut, 5) velocities. The FE results showed that the test rig
a steel base plate, 6) a load cell holder, 7) a cylindri- could achieve various strain rates, depending on the
cal specimen, 8) four supporting columns, 9) a syn- impact velocity. The size of every component was
thetic rubber. The 50 mm thick steel base plate is adequate since no plastic deformation was predicted
bolted to the floor with two M25 high tensile bolts. for these components. Figure 2 shows the von Mises
An Interface 1210 Precision model load cell was stress contour plot for the test rig subjected to an
used for this test rig because it is accurate to 0.04% impact velocity of 5 m/s. It shows that the specimen
and has a maximum capacity of 50 kN. It is connect- reached a maximum tensile strength of 320 MPa and
ed to the load cell holder with eight M6 high tensile fractured whereas the stress distribution in the other
M6 screws. Four supporting columns connect the components was low. A test rig was then fabricated
load cell holder to the base plate and transfer the im- and a full experimental set-up is shown in Figure 3.
pact force to the ground. The specimen is screwed
into the load cell and then attached to the impact
transmitter frame with a nut. The nut was designed
to allow for easy installation and removal of the
specimen before and after the tests. The impact
transmitter frame is made from 50 mm thick high
strength steel to eliminate any deformation under re-

Drop hammer

Figure 2: Von Mises stress contour plot for the test rig un-
der impact.

Figure 1: Front view of the instrumented falling


weight impact (IFWI) test rig.
IFWI

peated impact.
The IFWI test rig is incorporated into the existing
small capacity drop hammer rig at the University of
Figure 3: Full experimental set-up for
medium strain rate tests.
A high capacity instrumented falling weight im-
pact (IFWI) test rig developed at the University of
Wollongong was used to perform tensile tests under
medium strain rates. Three tests were conducted for
mild and stainless steel under various drop heights,
namely 100 mm, 500 mm, and 1250 mm. The force
and the strain time histories in the specimen were
measured by the load cell and the strain gauge re-
spectively. The strain gauges used in the dynamic
tests were the same as those used in the quasi-static
test. A high speed data acquisition system was used
to record the load and strain time histories at a sam-
pling frequency of 300 kHz.

5 EXPERIMENTAL RESULT S
4 EXPERIMENTAL PROCEDURE
A mild and a stainless steel specimen were tested
Two grades of steel investigated in this study were under quasi-static loading conditions using the In-
cold drawn mild steel and 304L stainless steel. Their stron machine. The strain reading for the specimens
chemical composition is listed in Table 1. They were were constructed from both strain gauge and LVDT
machined into a 6mm diameter by 14mm long cy- records. For strain up to 0.15, records from the strain
lindrical specimens, as shown in Figure 4. Threads gauges were used but the strain gauges started to fail
were turned screwed into the testing equipment for after that, so the strain records from the Instron were
the quasi-static and medium strain rate tests. used. The composite strain records were used to plot
the engineering stress-strain curves for mild and
Table 1: Chemical composition of mild and stainless steel. stainless steel, as shown in Figure 5. It shows that
Cold drawn mild steel 304L Stainless steel the yield stress and ultimate tensile strength for the
Element % by mass Element % by mass mild steel are 407 MPa and 657 MPa. The stainless
C 0.20 C 0.059 steel has a higher yield stress and ultimate tensile
Si 0.22 Si 0.78 strength of 539 MPa and 766 MPa, respectively. The
Mn 0.45 Mn 2.00 fracture strain for stainless steel was 0.66 and 0.2 for
P 0.017 P 0.02 mild steel.
S 0.11 Ni 8.30 For the dynamic tests, three mild steel and three
Cr 18.37 stainless steel specimens were tested under various
N 0.051 drop heights using the IFWI test rig. All the strain

Figure 4: Geometry of a cylindrical specimen


.
An Instron tensile testing machine was used to
perform material property tests on mild steel and
stainless steel specimens under a quasi-static loading
rate of 0.0005 s-1. The length of each specimen was
too small to use extensometers and strain values rec- Figure 5: Engineering stress strain curves for mild and
orded by LVDT in the linear elastic range were too stainless steel obtained from quasi-static tests.
high because its measurement included compliance
of the entire load path (Zabotkin et al., 2003). There- gauges attached to the specimens recorded strain up
fore, a fine post yield strain gauge was attached to to 0.05 before the adhesive used to bond the strain
the specimen to measure the strain. The strain gauge gauges to the specimens failed for unknown reasons.
was designed to measure large strain up to 15-20%. The strain time histories showed that the strain in-
The backing length and width are 12 mm and 4 mm, creased exponentially with time until it reached 0.02,
respectively. after which it increased almost linearly. This means
that the velocity of the impact transmitter frame in- mens in Figure 7. The yield stress for each specimen
creased from zero to almost constant when the strain was determined from these initial elastic-plastic en-
in the specimen reached 0.02.
To obtain a complete strain time history, the gra-
dient of the strain curve before the adhesive failed
was first determined. Then using this gradient, a
straight line was extruded from the strain gauge rec-
ords to the end of the impact duration. The fracture
strain predicted by this straight line was compared to
the fracture strain recorded for each specimen after
the tests. Generally, the fracture strain predicted by
this method was close to the experimental fracture
strain. In order to match the experimental fracture
strain, the gradient of the straight line was adjusted
slightly. The composite strain time history for the
mild steel specimen under 500 mm drop height is
shown in Figure 6. The experimental fracture strain
for every mild steel and stainless steel specimens are
Figure 7: Initial elastic-plastic engineering stress-strain
presented in Table 2.
curve for mild steel specimen.
The average strain rate in this study was taken as
gineering stress-strain curves, and the results are
the fracture strain of the specimen over the loading
presented in Table 2. To obtain real load time histo-
duration. The fracture strain for each specimen was
ries, the load signal up to the yield strength was re-
measured after the tests while the loading duration
tained and the high frequency oscillations were fil-
was determined from the corresponding load time
tered using a fourth order low pass Butterworth
history. The loading duration is the time the load
digital filter with a cut-off frequency of 2.5 kHz.
starts to increase up to the point where the load
The processed load time history for the mild steel
drops to zero, which indicates that the specimen had
specimen under 500 mm drop height is illustrated in
fractured. With increased height of the drop ham-
Figure 6. The processed load time histories were
mer, the loading duration decreased, and the average
then plotted against their corresponding strain time
strain rate increased. The loading duration and aver-
histories for the mild steel and stainless steel speci-
age strain rate for every specimen is shown in Table
mens.
2.
Engineering stress-strain curves for mild steel and
stainless steels under various strain rates are shown
The raw load time histories contain high frequen-
in Figures 8 and 9, respectively. The yield stress and
cy oscillations due to a hard impact between the
ultimate tensile strength for both steels are summa-
drop hammer and transmitter frame, as shown in
rised in Table 2. Based on the experimental results,
Figure 6. The load time histories up to the initial
the yield stress and tensile strength for both steels
increased with an increasing strain rate. The yield
stress of cold drawn mild steel increased up to 61%
and the ultimate tensile strength increased up to 24%
when the strain rate increased to almost 450 s-1. The
yield stress and ultimate tensile strength of stainless
steel showed a 52% and a 12 % increase when the
strain rate increased to about 450 s-1. The fracture
strain of mild steel increased from 0.2 for quasi-
static test to approximately 0.25 for dynamic tests.
However the fracture strain of stainless steel de-
creased from 0.66 to approximately 0.5 with an in-
creased strain rate. The enhancement of the yield
stress and ultimate tensile strength for both steels
with increased strain rate are plotted in Figure 10.
Figure 6: Processed load and strain time histories.

peak were plotted against the experimental strain


time histories, as exemplified by mild steel speci-
Table 2: Summary of experimental results for mild and stainless steel under various strain rates.
Mild steel Stainless steel
Quasi- 100 mm 500 mm 1250 mm Quasi- 100 mm 500 mm 1250 mm
static drop drop drop static drop drop drop
Fracture strain 0.20 0.25 0.25 0.25 0.66 0.49 0.48 0.48
Loading duration - 3.12 1.04 0.54 - 5.53 1.78 1.09
(ms)
Average strain 0.0005 78 236 469 0.0005 89 270 444
rate s-1
Yield stress 407 516 621 656 539 667 722 819
(MPa)
Ultimate tensile 657 766 772 815 766 826 845 857
strength (MPa)

Figure 8: Engineering stress strain curves for mild steel Figure 9: Engineering stress strain curves for stainless
under various strain rates. steel under various strain rates

showed that the yield stress and tensile strength for


both mild steel and stainless steel increased with an
increase in the strain rate, but the yield stress was
more sensitive to the strain rate than to the tensile
strength. More tests will be conducted in the future
and the experimental results will be used to derive
the Cowper-Symonds coefficients.

7 ACKNOWLEDGEMENT

The authors wish to acknowledge senior technical


officer, Mr. Alan Grant from the University of
Wollongong for his contribution in the design of
the IFWI test rig. This research was supported un-
der Australian Research Council's Discovery Pro-
Figure 10: Increment of the yield and tensile strength
jects funding scheme (project number
for mild and stainless steel under various strain rates.
DP0879733).

6 CONCLUSION 8 REFERENCES

Andersson, R., Syk, M., Powell, J. & Magnusson, C. 2005.


The dynamic response of cold drawn mild steel
Development of high strain rate equations for stainless
and 304L stainless steel under medium strain rates
steels. Journal of Material Engineering and Performance,
has been investigated using a high capacity instru-
Vol. 14 (5), pp. 553-562.
mented falling weight impact (IFWI) test rig de-
Cunat, P.J. 2000. Stainless steel properties for structural au-
veloped at the University of Wollongong. The
tomotive applications. Proceedings of Metal Bulletin In-
maximum average strain rate achieved in this study
ternational Automotive Materials Conference,
was about 500 s-1. The experimental results
Cologne.
Harding, J., Wood, E.O. & Campbell, J.D. 1960 Tensile test-
ing of material at impact rates of strain. Journal of Me-
chanical Science 2, pp. 88-96.
Jones, N. 1989. Structural Impact, Cambridge University
Press, pp. 333-383.
Nicholas, T. 1981. Tensile testing of material at high rates of
strain. Experimental Mechanics, Vol. 21 (5), pp. 177-185.
Peixinho N. 2006. Determination of crash-relevant material
properties of stainless steel alloy and constitutive equa-
tions. Journal of Physics IV France 134, pp. 243-249.
TM5-1300 Structures to Resist the Effects of Accidental Ex-
plosions, Departments of the Army, the Navy and the Air
Force, 1990
Uy, B., Tao, Z. & Chen, B. 2009. Dynamic behaviour of
stainless steel investigated with a tension split Hopkinson
bar. 8th International Conference on Shock & Impact
Loads on Structures. December 2-4, Adelaide, Australia.
Zabotkin K., O’Toole B. & Trabia M. 2003 Identification of
the dynamic tensile properties of metals under moderate
strain rates. 16th ASCE Engineering Mechanics Confer-
ence.

You might also like