International Journal of Plasticity 21 (2005) 651–670

www.elsevier.com/locate/ijplas

L€
uders bands propagation of 1045 steel under
multiaxial stress state
Jixi Zhang, Yanyao Jiang *

Department of Mechanical Engineering (312), University of Nevada, Reno, NV 89557, USA

Received in final revised form 14 April 2004
Available online 19 June 2004

Abstract

Inhomogeneous plastic deformation of 1045 steel under monotonic loading was experi-
mentally studied. Thin-walled tubular specimens were used in the experiments and custom-
made small strain gages were bonded on the specimen surface to characterize the local
deformation. Experiments were conducted under tension, torsion, and combined tension–
torsion. During the propagation of L€ uders bands, the local deformation experienced two-
stage deformation: an abrupt plastic deformation stage followed by a slower deformation
process. In some area of the gage section of the specimen, a small amount of initial plastic
deformation occurred before the L€ uders front reached. During the propagation of L€ uders
bands, multiple L€ uders fronts can be formed. Under tension, torsion, and combined ten-
sion–torsion with a constant axial load, the L€uders front was approximately parallel to the
material plane of maximum shear stress. When the combined axial-torsion followed a
proportional fashion, the stress–extensometer strain responses were dependent on the axial/
torsional loading ratio, and the L€
uders fronts were oriented differently and propagated along
the specimen axis at a different velocity. The local strain was inhomogeneous even at the
work-hardening stage. The relationships between the equivalent stress and the equivalent
plastic strain were found to be practically identical for all the loading cases studied.

2004 Elsevier Ltd. All rights reserved.

Keywords: Combined tension–torsion; L€

uders bands; Mild steel

*
Corresponding author. Tel.: +1-775-784-4510; fax: +1-775-784-1701.
E-mail address: [email protected] (Y. Jiang).

0749-6419/$ - see front matter
2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijplas.2004.05.001
652 J. Zhang, Y. Jiang / International Journal of Plasticity 21 (2005) 651–670

1. Introduction

A number of materials exhibit discontinuous yielding under monotonic tension. A

typical example is annealed mild steel. The monotonic stress–strain curve is not
smooth but shows marked irregularities, with negative slopes occurring at or near the
initial yield. Elastic deformation is terminated at a stress level known as the upper yield
stress. Deformation proceeds at a decreased stress level known as the lower yield stress
accompanied with inhomogeneous deformation. The specimen is divided into regions
where the strain is relatively high (L€uders strain) and regions which are still elastic. The
distinct plateau in the stress–strain curve is characterized by the propagation of L€ uders
bands. After the whole gage length has been strained by the amount of the L€ uders
strain, deformation becomes essentially homogeneous again (Hall, 1970).
The formation and propagation of L€ uders bands in mild steels under monotonic
tension are generally attributed to the strong interactions of interstitial atoms with
dislocations, known as Cottrell atmospheres (Cottrell and Bilby, 1949). In Cottrell
atmospheres, interstitial atoms segregate to dislocations and pin them in position to
lower the lattice distortion energy. At the upper yield stress, dislocations are un-
pinned from the Cottrell atmosphere and become free dislocations. These free dis-
locations can move at the lower yield stress and lead to multiplication of new
dislocations. Therefore, during the propagation of L€ uders bands, the elastic zones
are essentially free of dislocations whereas L€ uders bands have a high dislocation
density. The upper yield stress is regarded as the nucleation stress, and the lower
yield stress is the growth stress of the L€ uders bands.
The macroscopic and microscopic characteristics associated with the propagation
of L€uders bands under monotonic tension have been studied extensively. They in-
clude the angle and the propagation velocity of the L€ uders fronts, the in-plane
kinking of the strips, the influence of the number of L€ uders fronts on the yield stress,
and the strain profile at the L€ uders fronts. Macroscopic shear bands can be readily
observed at the front of the L€ uders bands, and very often the band front makes a 45
angle with the loading axis. That is, the band front follows the plane of maximum
shear stress. However, this is a first approximation. By using specimens with a
pentagon-shaped cross-section, Lomer (1952) showed that the angle between the
tensile axis and the band front was 48 ± 1.5. In the study of L€ uders bands in Co-
lumbium, Conrad and Stone (1964) observed the rotation of the band front. A kink
band usually exists in the front of shear band and the kink angle is in the order of 1
(Ananathan and Hall, 1989; Kyriakides and Miller, 2000). Ananathan and Hall
(1991) suggested that the grip effects and bending moment associated with the shear
kink influenced the deviation of the orientation angle of the band front.
The propagation of L€ uders bands is influenced by many factors including crystal
structure, grain size, composition and microstructure, shape and stiffness of the
testing sample, strain rate, and the type of loading (Lomer, 1952; Butler, 1962;
Ananathan and Hall, 1991). Grain size has a great influence on the L€ uders strain and
the morphology of the L€ uders bands, particularly in the case for mild steels (Hall,
1950, 1951; Morrison and Glenn, 1968). The L€ uders strain decreases significantly as
the grain size increases. When the grain size is less than 30 lm, sharp L€ uders bands
 J. Zhang, Y. Jiang / International Journal of Plasticity 21 (2005) 651–670 653

are commonly observed. When the grain size is larger than 40 lm, diffuse L€ uders
bands may occur. Lomer (1952) pointed out that the type of loading influenced the
orientation and morphology of the L€ uders front. For example, a tubular specimen
under torsion will result in a single L€
uders band with the planar band front normal
to the specimen axis. Complex multiple bands will be formed if the tubular specimen
is pulled in tension.
Different techniques have been developed to observe the propagation of L€ uders
bands and to evaluate the strain profile near the L€ uders front. Due to the conve-
nience and suitability for surface observations, a thin plate specimen was often used
in the investigation of L€
uders bands. Usually, surface scratches (Hall, 1951; Iricibar
and Mazza, 1975; Iricibar et al., 1977), surface grid (Prewo et al., 1972; Lloyd and
Morris, 1977), contrast interferometer (Iricibar et al., 1975, 1977; Verel and Sle-
eswyk, 1973), stresscoat (Fisher and Rogers, 1956), electroplating of the specimen
prior to testing (Liss, 1957), optical ‘‘T€opler Schlieren’’ method (Hall, 1951; Syl-
westrowicz and Hall, 1951), critical macro-illumination (Boxall and Hundy, 1955),
grid-reflection (Verel and Sleeswyk, 1973), clip-extensometer (Iricibar and Mazza,
1975; Moon, 1971), high speed camera (Miyazaki and Fujita, 1979), and infrared
camera (Louche and Chrysochoos, 2001) were used to observe the L€ uders bands and
to identify the local strains. However, these methods do not provide a direct and
accurate quantitative measurement of the local strain.
By using small distributive strain gages bonded on the specimen surface within the
gage section, Zhang and Jiang (2004) experimentally studied the local plastic de-
formation of a carbon steel subjected to monotonic tension. Small strain gages can
provide direct, accurate, and reliable local strain measurements. It was found that
the strain at the L€
uders front was lower than the full L€uders strain (the length of the
plateau on the stress–strain curve). The local strain was inhomogeneous even at the
work-hardening stage. By using a similar experimental method to measure the local
strain, Onodera et al. (2000) studied the plastic deformation of an aluminum alloy
during serrated flow. It was found that two-stage L€ uders deformation occurred with
a fast deformation stage followed by a slow deformation process. Shaw and Ky-
riakides (1995) used four miniature extensometers within the test section to track the
inhomogeneous deformation in NiTi wires under uniaxial loading.
Most studies of the L€ uders bands phenomenon were concentrated on monotonic
tension. Recently, Aguirre et al. (2004) investigated L€
uders bands of steel tubes under
bending. The propagation of L€ uders bands in torsion has rarely been studied. To the
authors’ knowledge, L€ uders bands phenomenon under multiaxial stress state has
never been reported. The current effort was aimed at an experimental investigation of
the inhomogeneous plastic deformation of mild steel under torsion and combined
tension and torsion by using small strain gages to characterize the local strains.

2. Experiments

The as-received SAE 1045 steel rods were in a hot rolled state. After normali-
zation at 815 C for 8 h, the rods were machined into tubular specimens. The
654 J. Zhang, Y. Jiang / International Journal of Plasticity 21 (2005) 651–670

specimens were normalized again at 815  C for 8 h to eliminate the possible residual
stresses introduced by the machining process. After the second normalization, the
chemical compositions were analyzed and the results are listed in Table 1. A typical
microstructure after the second normalization is shown in Fig. 1. The volume
fraction of pearlite was approximately 55% and the mean ferrite grain size was 10
lm. The gauge length surface of the specimens was polished with fine sand papers
with grit size 600. The uniform gage section of the specimen had a length of 25.4 mm.
The outer and inner diameters of the gage section were 23.5 and 20.2 mm,
respectively.
To measure the local strains, a custom-made 12-element strip gage was bonded on
the specimen surface. The dimensions and arrangement of the strain gages in the
strip gage are shown in Fig. 2. The total length of the strip gage was 20 mm. Each
gage element had a electrical resistance of 120 X and had a sensing area of 1.5
mm · 1.7 mm. The spacing between two neighboring strain gages was 1.77 mm. Six
strain gages aligned along the specimen axis are referred to as ‘‘AX’’ (axial) gages
and are numbered with 1, 3, 5, 7, 9, and 11. The strain output of an AX gage is
denoted by e0 . The other six strain gages were aligned at an angle of 45 with respect

Table 1
Chemical composition of 1045 steel
Element C Si Mn P S Fe
Wt% 0.487 0.28 0.74 0.009 0.027 Balanced

Fig. 1. Microstructure of 1045 steel.

 J. Zhang, Y. Jiang / International Journal of Plasticity 21 (2005) 651–670 655

Fig. 2. Test specimen and strain gages: (a) Specimen and arrangement of small strain gages, (b) AX strain
gage, FF strain gage, and determination of orientation of L€uders front.

to the specimen axis. They are referred to as ‘‘FF’’ (45) gages and are numbered
with 2, 4, 6, 8, 10, and 12. The strain output of an FF strain gage is denoted by e45 .
During an experiment, a 25.4 mm gage length extensometer was attached to the
specimen in the gage section to measure the bulk deformation.
Four different types of monotonic experiments were conducted at room temper-
ature in air. The testing conditions are listed in Table 2 and the loading paths are
schematically illustrated in Fig. 3. All the experiments were conducted with the
656 J. Zhang, Y. Jiang / International Journal of Plasticity 21 (2005) 651–670

Table 2
Experimental loading conditions
Loading paths Testing conditions
Controlled Axial strain Controlled Shear strain
displacement rate rotation rate
I: Monotonic tension 4 mm/1000 s 1.11 · 10
4 /s – –
II: Monotonic torsion – – 2/1000 s 1.15 · 10
4 /s
III: Biaxial tension–torsion: 3 mm/1000 s 6.16 · 10
5 /s 4/1000 s 3.44 · 10
5 /s
Case 1
IV: Biaxial tension–torsion: 3 mm/1000 s 4.85 · 10
5 /s 6.5/1000 s 4.55 · 10
5 /s
Case 2
V: Biaxial tension–torsion: 1.25 mm/1000 s 2.67 · 10
5 /s 8/1000 s 4.19 · 10
5 /s
Case 3
VI: Biaxial tension–torsion Axial stress: 260 2.27 · 10
5 /s 8/1000 s 4.85 · 10
5 /s
with fixed axial load MPa

σ
δ σ
τ
VI
I III 260 MPa
IV
τ
V
II θ θ
σ

(a) (b)

Fig. 3. Loading paths and stress state: (a) stress state; (b) loading paths.

stroke-control mode. In Fig. 3, d represents the axial displacement and h is the ro-
tation angle. The loading rates listed in Table 2 were the rate of the actuator
movement during an experiment. The strain rates listed were the average strain rates
measured by the attached extensometer during the propagation of L€ uders bands. To
distinguish the deformation measured by the extensometer and the strain gages, the
strains obtained from the extensometer are referred to as extensometer strains while
the strain gage measurements are denoted as local strains. During an experiment, the
axial load, torque, extensometer strains, and the outputs of all the 12 strain gages
were recorded simultaneously. The extensometer strains represent the average de-
formation within the gage section of the specimen.
The average axial stress, r, and the average shear stress, s, were calculated from
the measured axial load, P, and torque, T, using the following formulas:
4P
r¼ ; ð1Þ
pðD2
d 2Þ
 J. Zhang, Y. Jiang / International Journal of Plasticity 21 (2005) 651–670 657

12T
s¼ ; ð2Þ
pðD3
d 3 Þ
where D and d are the outer and inner diameters in the gage section, respectively. The
mid-section shear strain was used as the average shear strain. A linear shear strain
distribution over the wall thickness of the specimen was assumed. The extensometer
measured the surface strains. It should be noted that Eqs. (1) and (2) were obtained
by assuming uniform axial and shear stress distributions over the wall thickness of
the specimen. With elastic–plastic deformation under combined axial-torsional
loading, neither the axial stress nor the shear stress is uniformly distributed over the
wall thickness of the thin-walled tubular specimen. However, those average values
obtained by using the equations can serve as a good approximation.
For monotonic tension, the local axial strain is equal to e0 . The local surface
shear strain is equal to 2e45 under torsion. Under biaxial tension–torsion, the local
axial strain is equal to e0 . To determine the local surface shear strain, the local axial
strain of the area covered by an FF strain gage, e0;f , is fitted from the strain re-
sponses of the two neighboring AX strain gages. The specific algorithm will be
discussed later. The local surface shear strain, cloc , is calculated using the following
equation:
r
cloc ¼ 2e45
0:5e0 ;f
ð0:5
lÞ ; ð3Þ
E
where E is the Young’s modulus, and l is the Poisson ratio of the material. For 1045
steel, E ¼ 208 GPa and l ¼ 0:3. Eq. (3) considers the different contraction ratio due
to elastic–plastic deformation. The ratio of transverse strain over the axial strain
under uniaxial loading is the Poisson’s ratio for elastic deformation and the value is
0.5 for plastic deformation.
The detailed dimensions and geometry of the strain gages together with the signal
outputs from the individual strain gages are used to determine the nucleation sites,
the orientation, and velocity of the L€ uders fronts. The L€
uders front is assumed to be
sharp and planar. Fig. 2(b) schematically illustrates the method to determine the
orientation of a L€ uders front propagating downward. The rectangles ABCD and
EFGH represent the sensing areas of an AX strain gage and its neighboring FF
strain gage. Assume that the L€ uders front is propagating at an average velocity,
v,
with an angle of h with respect to the specimen axis. The propagation velocity,
v, is
determined by knowing the distance between the two neighboring AX strain gages
and the time taken to travel such a distance. At time t1 , the L€ uders front reaches
Point A and the AX strain gage starts to sense plastic deformation. At time t2 , the
L€uders front reaches Point E and the area covered by the FF strain gage starts
plastic deformation. The spacing between Points C and F is very small and can be
ignored. When h is greater than 45, the following equation can be obtained:
L1 sin a þ L2
tan h ¼ ; ð4Þ

vðt2
t1 Þ
L1 ð1
cos aÞ
where L1 and L2 are the dimensions of the sensing area of a strain gage (refer to Fig.
2(b)). For the strip gage, L1 ¼ 1:5 mm, L2 ¼ 1:7 mm, and a ¼ 45. When h < 45 or
658 J. Zhang, Y. Jiang / International Journal of Plasticity 21 (2005) 651–670

the L€
uders front propagates upward, similar formulas can be derived to determine the
angle h. Generally, a trial and error method is necessary to determine the value of h.

3. Experimental results and discussion

3.1. Monotonic tension

Fig. 4 show the experimental results under monotonic tension. The typical stress–
extensometer strain curve is shown in Fig. 4(a). The upper yield stress is 380 MPa
and the lower yield stress is 370 MPa. The plateau corresponds to the formation and
propagation of the L€ uders bands. The length of the plateau (full L€ uders strain) is
0.81%. The variations of the local axial strains with time are shown in Fig. 4(b). The

Fig. 4. Monotonic tension behavior of 1045 steel: (a) tensile stress–extensometer strain curve; (b) variation
of local axial strains.
 J. Zhang, Y. Jiang / International Journal of Plasticity 21 (2005) 651–670 659

number shown on each curve denotes a specific AX strain gage. Corresponding to

the stress–strain curve, three stages can be identified in Fig. 4(b). They are elastic
deformation, propagation of L€ uders bands, and work hardening. Approximately
70 s were taken for the L€ uders bands to propagate over the gage section of the
specimen. The inserted bars in Fig. 4(a) schematically illustrate the initiation and
propagation of the L€ uders bands. The total length of a bar represents the length of
the gage section of the specimen and the shaded area symbolizes L€ uders bands. At t1
(t ¼ 146 s), a L€ uders band, LB1 as shown in Fig. 4(a), was initiated and the area
covered by Gages 9, 10, and 11 started plastic deformation almost simultaneously.
The L€ uders front propagated with an angle of near 45 to the specimen axis. The
orientation of the L€ uders front is close to the direction of the maximum shear stress.
At t2 (t ¼ 165 s), a new L€uders band, LB2 as shown in Fig. 4(a), was initiated and the
area covered by Gages 5, 6, and 7 deformed plastically simultaneously. At t3 (t ¼ 180
s), a third L€
uders band, LB3, was formed in the area covered by Gages 2 and 3. The
initiation and propagation of L€ uders bands were very irregular.
During the propagation of L€ uders bands, the deformation of the area covered by
Gages 11, 9, and 5, experienced three segments. In Segment 1, from Point A to Point
B for Gage 9 (Fig. 4(b)), the local axial strain increased linearly and rapidly with
time, indicating the propagation of the L€ uders front over the area. A large amount of
ferrite grains within this area experienced plastic deformation. In Segment 2, from
Point B to Point C for Gage 9 (Fig. 4(b)), the strain increased slowly and asymp-
totically reached a temporary saturation value. The deformation is possibly caused
by creep in the plastically deformed L€ uders bands. For a mild steel, Kyriakides and
Miller (2000) pointed out that the strain rate at the L€ uders front was rather high and
some dislocations unable to move at the high strain rate can continue to move at a
slow velocity after the front has passed them. Some deformed ferrite grains may
deform further and some undeformed ferrite grains may experience a delayed plastic
deformation. Such actions result in creep deformation. Similar phenomenon was
observed in a monotonic tension of a mild steel (Zhang and Jiang, 2004) and an
aluminum alloy (Onodera et al., 2000). In Segment 3, from Point C to Point D for
Gage 9 (Fig. 4(b)), the strain kept constant before the start of work hardening.
Certain area such as that covered by Gages 1 and 3 experienced an obvious plastic
deformation before the L€ uders fronts reached or a new L€ uders band was initiated in
the area (Fig. 4(b)). The initial plastic strain can be attributed to microstrain and
microcreep before the L€ uders band is formed (Friedel, 1964).
A vertical line in Fig. 4(b) denotes the end of the plateau in the stress–extens-
ometer strain curve. A general concept is that the end of the plateau is the starting
point of homogeneous deformation in the gage section. However, the results shown
in Fig. 4(b) clearly indicate that at the end of the plateau, the local deformation was
still inhomogeneous. The inhomogeneous deformation persisted in the work-hard-
ening stage. This observation is consistent with that observed on an SAE 1045 steel
with a different heat treatment (Zhang and Jiang, 2004).
Also shown in Fig. 4(b) are the average local strains and the extensometer strains.
The average local strain was the average of the strain values taken from the six AX
strain gages. The small difference between the average local strain value and the
660 J. Zhang, Y. Jiang / International Journal of Plasticity 21 (2005) 651–670

extensometer measurement was mainly due to the difference between the extens-
ometer gage length (25.4 mm) and the length of the strain strip (20 mm), knowing
that the deformation within the gage section was non-uniform.

3.2. Monotonic torsion

The experimental results under monotonic torsion are summarized in Fig. 5. The
shear stress–extensometer shear strain is shown in Fig. 5(a). The curve is similar to
the stress–strain curve under monotonic tension. The upper yield shear stress is 218
MPa and the lower yield shear stress is 208 MPa. The full L€ uders shear strain (the
length of the plateau) is 1.47%.
In Fig. 5(b), both the local strains and the extensometer strain are shown with
respect to time. Three stages can be identified for the shear deformation: elastic
deformation, propagation of L€ uders bands, and work hardening. With the

Fig. 5. Monotonic torsion behavior of 1045 steel: (a) shear stress–extensometer strain curve; (b) variation
of local shear strains.
 J. Zhang, Y. Jiang / International Journal of Plasticity 21 (2005) 651–670 661

controlled stroke displacement, it took 140 s for the L€ uders bands to propagate
through the gage section of the specimen. Therefore, the average velocity of the
propagation of the L€ uders bands was 0.183 mm/s. In certain area, such as that
covered by Gage 12, creep strain after the propagation of L€ uders band was obvi-
ous. After the plateau in the stress–strain curve, the local shear strain was still in-
homogeneous.
The initiation and propagation of the L€ uder bands are schematically illustrated by
the inserted bars in Fig. 5(a). The inserts correspond to the time moments indicated
in the stress–strain curve. The L€
uders fronts were perpendicular to the specimen axis.
This suggests that the L€ uders fronts were parallel to the material plane of the
maximum shear stress. After the shear stress reached the upper yield shear stress, a
L€uders band, LB1, was initiated in the area outside of Gage 12. After the L€ uders
front reached the area covered by Gage 8, a new L€ uders band, LB2, was formed
within the area covered by Gage 2, and propagated in the two opposite directions.
Since the displacement of the two ends of the specimen was controlled at a constant
rate, the formation of additional L€ uders bands reduced the propagation velocities of
the L€uders fronts. Consequently, the propagation velocities of the L€
uders front from
Gage 8 to Gage 6 and that from Gage 2 to Gage 4 decreased sharply after the ini-
tiation of multiple L€ uders fronts.

3.3. Proportional tension–torsion

Three proportional tension–torsion loading experiments were conducted as listed

in Table 2. The stress–extensometer strain responses were dependent on the loading
path. For the selected three cases, the stress–extensometer strain curves were similar
for Path III and Path IV (refer to Table 2 and Fig. 3), while the curve for Path V was
different from the other two loading paths.
The stress–extensometer strains are shown in Fig. 6(a) and (b) and the variations
of the local strains with time are shown in Fig. 6(c) and (d) for Path IV. Similar to
the tension and torsion, the deformation under combined tension–torsion can be
divided into three stages: elastic deformation, propagation of L€ uders bands, and
work hardening. During the propagation of L€ uders bands, the axial stress increased
with increasing axial strain. However, the shear stress decreased sharply, from 160 to
74 MPa, as the strain increased. In the third stage, the axial stress increased with the
increase of axial strain predominantly, while the shear stress firstly increased, from
74 to 95 MPa, and then kept almost constant.
The initiation and propagation of the L€ uders fronts are shown by the inserted
bars in Fig. 6(a). At t1 (t ¼ 143 s), a L€ uders band, LB1, was initiated in the area
covered by Gages 6 and 7 and propagated towards the two opposite ends. At t2
(t ¼ 232 s), the profiles and velocities of the two L€uders fronts were different. For the
L€uders front moving upward, the L€ uders front was perpendicular to the specimen
axis and the front propagated at a velocity of 0.0837 mm/s. For the L€ uders front
moving downward, the L€ uders front was at an angle of 65 with respect to the
specimen axis and it traveled at a velocity of 0.0620 mm/s. At this particular mo-
ment, the axial stress was 320 MPa and the shear stress was 101 MPa. One direction
662 J. Zhang, Y. Jiang / International Journal of Plasticity 21 (2005) 651–670

Fig. 6. Combined tension–torsion (Loading Path IV): (a) axial stress–extensometer axial strain curve; (b)
shear stress–extensometer shear strain curve; (c) variation of local axial strains; (d) variation of local shear
strains.

of the maximum shear stress was at an angle of 61 degrees with respect to specimen
axis. Therefore, the L€uders front propagating downward had a profile consistent
with the maximum shear stress.
During the propagation of L€ uders bands, the plastic deformation of an FF
strain gage was not synchronous with its neighboring AX strain gages. In order to
determine the local shear strain, the local axial strain of the area covered by the FF
strain gage must be extrapolated from the strain responses of two neighboring AX
strain gages. For example, in order to determine the local axial strain of the area
covered by Gage 2, denoted by f2 ðtÞ, the strains of Gages 1 and 3 were used.
Assume that the strains of Gages 1 and 3 are f1 ðtÞ and f3 ðtÞ, respectively, and the
moments for Gages 1, 2, and 3 to start plastic deformation are T1 , T2 , and T3 ,
respectively, ðT1 > T2 > T3 Þ. An algorithm of shifting-and-weighting can be derived
to determine f2 ðtÞ
 J. Zhang, Y. Jiang / International Journal of Plasticity 21 (2005) 651–670 663

0.014 Biaxial Tension-Torsion

0.012 Propagation of Lüders Bands Work Hardening

Axial Strain 0.010

0.008
0.006
Gage 7 9
5
0.004 3 11 1
Elastic Strain Gage
0.002 Extensometer
Avg Strain Gage

0.000
100 150 200 250 300 350
(c) Time (s)

0.016 Biaxial Tension-Torsion

Work Hardening
Surface Shear Strain

Propagation of Lüders Bands

0.012

0.008 Gage 6 8 4

10 2 12
0.004 Elastic
Strain Gage
Extensometer
Avg Strain Gage
0.000
100 150 200 250 300 350
(d) Time (s)
Fig. 6 (continued)

f1 ðtÞ ðt 6 T2 Þ;
f2 ðtÞ ¼ T2
T3 ð5Þ
f ½t
T1
T3 1
þ ðT1
T2 Þ þ TT11
T

T3
2
f3 ½ t þ ðT 3
T 2 Þ  ðt > T2 Þ:

The variations of the deformation with time are shown in Fig. 6(c) for the axial
strains and Fig. 6(d) for the shear strains on the specimen surface. When the L€ uders
front reached a given area, both the local axial strain and local shear strain increased
abruptly. During the propagation of the L€ uders bands, a continuous decrease in
shear stress led to a continuous decrease in elastic shear strain in the plastically
undeformed zone. This phenomenon corresponded to a negative slope in the local
shear strain versus time curve before the arrival of the L€ uders front (Fig. 6(d)). At
the work-hardening stage, the local shear strain in the area where the L€ uders front
had reached first (Gage 6) was significantly larger than the shear strain of the area
where the plastic deformation occurred the last (Gage 12). As compared to the axial
local strains, the shear strains displayed a more obvious diversity in value at the
work-hardening stage.
664 J. Zhang, Y. Jiang / International Journal of Plasticity 21 (2005) 651–670

Table 3
Variations of axial stress and shear stress during the propagation of L€
uders bands
Loading Initiation of L€
uders bands End of propagation of L€
uders bands
path
Axial stress r (MPa) Shear stress s (MPa) Axial stress r (MPa) Shear stress s (MPa)
Path III 320 110 366 45
Path IV 258 160 340 74
Path V 60 205 195 180

The variations of the axial stress and the shear stress during the propagation of
L€uders bands for Paths III, IV, and V are listed in Table 3. When the axial strain rate
was relatively high compared to the shear strain rate, such as in Paths III and IV, the
axial stress was high at the initiation of L€uders bands. During the propagation of
L€uders bands, the axial stress increased slowly and s decreased sharply. When the
axial strain rate was low as compared to the shear strain rate, such as in Path V, the
axial stress was very small when L€ uders bands were nucleated. During the propa-
gation of L€ uders bands, the axial stress increased sharply, from 60 to about 200
MPa, while the shear stress decreased slightly. The local axial strain increased pro-
portionally with the extensometer axial strain, and the strain burst of the local axial
strain was not observed.

3.4. Monotonic torsion with a constant axial load

The experimental results for the combined tension–torsion with a constant axial
load (Path VI in Fig. 3) are summarized in Fig. 7. The tubular specimen was first
loaded to a stress of 260 MPa. Torsional load was applied while the axial stress was
kept at a constant value of 260 MPa. Before the application of the torsional load, the
deformation of the material was elastic. Yielding occurred when the shear stress
reached 138 MPa. During the propagation of L€ uders bands, the shear stress in-
creased slightly from 138 MPa at the start of yielding to 158 MPa at the end of the
plateau (Fig. 7(a)). The initiation and propagation of the L€ uders fronts are sche-
matically shown by the inserted bars in Fig. 7(a). At t1 (t ¼ 85 s), two L€
uders bnads,
LB1 and LB2, were formed simultaneously. Both of the two fronts propagated in the
two opposite directions. At t2 (t ¼ 150 s), the two L€uders bands had propagated to
cover certain neighboring areas near the initiation sites. All the four L€uders fronts
were at approximate 70 to specimen axis. Since the axial stress was kept a constant
and the shear stress was approximately constant during the propagation of the
L€uders bands, the orientation of the maximum shear stress was 70 from the spec-
imen axis. Therefore, the L€ uders fronts were consistent with the material plane of
maximum shear stress.
At t3 (t ¼ 265 s), the two L€
uders bands had coalesced into one band, and a new
L€uders band, LB3, was formed outside Gage 12 and was propagating upward.
During the propagation of L€ uders bands, the L€ uders fronts kept parallel to the di-
rection of maximum shear stress. The variations of the local axial strains and local
 J. Zhang, Y. Jiang / International Journal of Plasticity 21 (2005) 651–670 665

Fig. 7. Tension–torsion with constant axial load (Loading Path VI): (a) shear stress–extensometer shear
strain curve; (b) variation of local axial strains; (c) variation of local shear strains.
666 J. Zhang, Y. Jiang / International Journal of Plasticity 21 (2005) 651–670

shear strains with time are shown in Fig. 7(b) and (c). When the L€ uders front
reached, both local axial strain and local shear strain increased sharply. For the
loading path, the local axial strains were more inhomogeneous than the local shear
strains in the work-hardening zone.

4. Further discussion

In a previous study (Zhang and Jiang, 2004), solid shaft specimens of 1045 steel
were subjected to monotonic tension. The material was heat treated slightly different
from that used in the current study. When the solid shaft specimen was subjected to
monotonic tension, only a single L€ uders band was formed and the propagation
velocity followed a regular pattern. However, a tubular specimen under monotonic
tension resulted in the formation of complex L€ uders bands and the local strain re-
sponses were more irregular.
Generally, the L€ uders front is close to the plane of maximum shear stress. The
orientations of the L€ uders fronts in tension, torsion, and biaxial tension–torsion with
a constant axial load followed this principle, even though multiple L€ uders bands
were formed. There are two planes of maximum shear stress that are perpendicular
to each other. If the normal of a plane is used to represent the direction of the plane,
the L€
uders front often occurs on the plane of maximum shear which makes a smaller
angle to the specimen axis.
Under proportional combined tension–torsion loading, the propagation of
L€uders bands is more complicated. As the L€ uders fronts advance, the axial stress and
the shear stress change continuously with time. As a result, the orientation of the
maximum shear stress varies with time and the L€ uders front changes its orientation
continuously during the propagation of the L€ uders bands. The two L€ uders fronts
with different orientations in Path IV probably imply the rotation of the L€ uders
front. Direct observation of propagation of L€ uders bands on specimen surface is
necessary to get accurate information of orientation of the L€ uders front.
In order to compare the results obtained from the different loading paths, an
equivalent stress, req , and an equivalent plastic strain, eeq;pl , are defined as follows:
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
req ¼ r2 þ 3s2 ; ð6Þ
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
eeq;pl ¼ e2pl þ c2pl =3; ð7Þ

where
r
epl ¼ e
ð8Þ
E
and
s
cpl ¼ c
; ð9Þ
G
 J. Zhang, Y. Jiang / International Journal of Plasticity 21 (2005) 651–670 667

500 1045 Steel

Combined Tension-Torsion

Equivalent Stress (MPa)

400

300

200 Path I
Path II
Path III
100 Path IV
Path V
Path VI
0
0.000 0.005 0.010 0.015 0.020
Equivalent Plastic Strain

Fig. 8. Equivalent stress–equivalent plastic strain curves for different loading paths.

are the axial plastic strain and the shear plastic strain, respectively. In these equa-
tions, r and s denote the average axial stress and the average shear stress, respec-
tively, as defined by Eqs. (1) and (2). The symbols e and c represent the average axial
strain and the average shear strain, respectively. E is the Young’s modulus and G is
the shear modulus of the material.
The relationships between the equivalent stress and the equivalent strain for all
the loading cases studied are summarized in Fig. 8. These curves are practically
identical. Most of the curves exhibited an upper yield point and a lower yield point
which corresponded to the initiation of plastic deformation in a form of L€ uders
bands, and all the curves exhibited a distinct plateau which corresponded to the
propagation of L€ uders bands. It should be mentioned that, during the propagation
of L€uders bands, the local deformation is inhomogeneous. The strains used in the
determination of the equivalent plastic strain represent the average deformation
within the gage section.
As shown in Fig. 8, the relationship between the equivalent stress and the equiv-
alent plastic strain were practically identical for all the loading cases studied. When
the equivalent stress reaches a critical value (upper yield stress), the L€uders band is
initiated. During the propagation of L€ uders bands, the equivalent stress keeps a
constant value (lower yield stress). This suggests that the equivalent stress plays an
important role in the initiation and propagation of the L€ uders bands under multiaxial
stress state. According to the dislocation theory, the dislocations are unpinned from
the Cottrell atmospheres at the upper yield stress. For a given material, the stress
magnitude necessary to unpinned dislocations from the Cottrell atmospheres is a
constant. Therefore, the equivalent upper yield stress is independent of the loading
path. The propagation of L€ uders bands can be regarded as a continuous formation of
new L€ uders band at the boundary between the plastically deformed area and the
plastically undeformed area. During the propagation of L€ uders bands, the stress
concentration at the boundary between the deformed and undeformed areas facili-
tates the dislocations to unpin from the Cottrell atmospheres at the L€ uders front
668 J. Zhang, Y. Jiang / International Journal of Plasticity 21 (2005) 651–670

(Cottrell, 1964). As a result, the equivalent stress for the propagation of L€

uders band
is reduced as compared to the required applied stress to initiate the first L€uders band
on the specimen. Such an interpretation is in agreement with that for the initiation
and propagation of L€ uders bands under monotonic tension (Hall, 1970). Once un-
pinned from the Cottrell atmosphere, the dislocations move on the plane of the
maximum shear stress. That is why the L€ uders front generally coincides with the
material plane of maximum shear stress. In summary, the formation of the L€ uders
bands is determined by the applied equivalent stress. Only when the equivalent stress
reaches a level to unpin the dislocation from the Cottrell atmospheres can L€ uders
bands be formed. The orientation of the L€ uders front is determined by the maximum
shear stress.

5. Conclusions

1. An experimental technique was developed to characterize the local deformation

during the propagation of L€ uders bands by using distributive small strain gages.
This technique has been successfully applied to investigate the inhomogeneous de-
formation under combined axial-torsion loading of tubular specimens. The signals
from the strain gages were used to determine the initiation sites and the propaga-
tion of the L€uders bands.
2. During the propagation of L€ uders bands, the local material experiences two-
staged L€uders deformation: an abrupt deformation stage followed by a slower de-
formation process. The first stage corresponds to the fast propagation of the
L€uders front through the local area and the second stage can be attributed to
creep deformation within the L€ uders bands. Small initial plastic deformation
can occur in the localized area in the gage section of the specimen before the
L€uders front reaches.
3. At the end of the propagation of L€ uders bands, the local deformation is inhomo-
geneous. Usually, the local area experienced L€ uders strain earlier has a higher lo-
cal strain than that of the area experienced L€ uders strain later. The
inhomogeneous local strain persists in the work-hardening stage.
4. Under proportional axial-torsion, the stress–extensometer strain curves are de-
pendent on the axial/torsional loading ratio. During the propagation of L€ uders
bands, the axial stress increases with increasing axial strain, while the shear stress
decreases as the shear strain increases. The process results in a near constant
equivalent stress during the propagation of the L€ uders bands.
5. During the propagation of L€ uders bands, multiple L€ uders bands can be formed.
In monotonic tension, torsion, and biaxial tension–torsion with a constant axial
load, the L€
uders fronts are approximately parallel to the plane of maximum shear
stress. Under proportional axial-torsion, the L€uders fronts can be oriented on the
material plane different from the maximum shear plane.
6. For all loading paths studied, the equivalent stress versus equivalent plastic strain
curves are practically identical. The curve exhibits an upper yield point, a lower
yield point, and a distinct plateau which corresponds to the initiation and prop-
 J. Zhang, Y. Jiang / International Journal of Plasticity 21 (2005) 651–670 669

agation of L€uders bands. Only when the equivalent stress reaches a certain level
can L€uders bands be formed. The orientation of the L€
uders front is mainly deter-
mined by the maximum shear stress.

Acknowledgements

Financial support provided by the National Science Foundation (CMS-9984857)

is greatly acknowledged.

References

Ananathan, V.S., Hall, E.O., 1989. Shear and kink angles at the L€ uders band front. Scripta Metall. 23,
1075–1078.
Ananathan, V.S., Hall, E.O., 1991. Macroscopic aspects of L€ uders band deformation in mild steel. Acta
Metall. Mater. 39, 1353–3160.
Aguirre, F., Kyriakides, S., Yun, H.D., 2004. Bending of steel tubes with L€ uders bands. Int. J. Plast. 20,
1199–1225.
Butler, J.F., 1962. L€uders front propagation in low carbon steels. J. Mech. Phys. Solids 10, 313–334.
Boxall, T.D., Hundy, B.B., 1955. Photographing stretcher-strain markings with the Vickers projection
microscope. Metallurgia 51, 52–54.
Conrad, H., Stone, G., 1964. The effect of stress on the L€uders band velocity in columbium. J. Mech. Phys.
Solids 12, 139–148.
Cottrell, A.H., 1964. The Mechanical Properties of Matter. Wiley, New York.
Cottrell, A.H., Bilby, B.A., 1949. Dislocation theory of yielding and strain aging of iron. Proc. Phys. Soc.
62/I-A, 49–62.
Fisher, J.C., Rogers, H.C., 1956. Propagation of L€ uders bands in steel wires. Acta Metall. 4, 180–185.
Friedel, J., 1964. Dislocations. Pergamon Press, New York.
Hall, E.O., 1950. An optical method for studying the deformation of mild steel. Proc. Phys. Soc. B 63,
724–726.
Hall, E.O., 1951. The deformation and aging of mild steel: II. Characteristics of the L€ uders deformation.
III. Discussion of results. Proc. Phys. Soc. B 64, 742–753.
Hall, E.O., 1970. Yield Point Phenomena in Metals and Alloys. Plenum Press, New York.
Iricibar, R., Mazza, J., 1975. Meaning of the strain profile of a propagating L€ uders band front. Scripta
Metall. 9, 1045–1050.
Iricibar, R., Mazza, J., Cabo, A., 1975. The microscopic strain profile of a propagating L€ uders band front
in mild steel. Scripta Metall. 9, 1051–1058.
Iricibar, R., Mazza, J., Cabo, A., 1977. On the L€ uders band in mild steel-I. Acta Metall. 25, 1163–1168.
Kyriakides, S., Miller, J.E., 2000. On the propagation of L€ uders bands in steel strips. J. Appl. Mech. 67,
645–654.
Liss, R.B., 1957. L€ uders bands. Acta Metall. 5, 341–342.
Lloyd, D.J., Morris, L.R., 1977. L€ uders band deformation in a fine grained aluminium alloy. Acta Metall.
25, 857–861.
Lomer, W.M., 1952. The yield phenomenon in polycrystalline mild steel. J. Mech. Phys. Solids 1, 64–73.
Louche, H., Chrysochoos, A., 2001. Thermal and dissipative effects accompanying L€ uders band
propagation. Mater. Sci. Eng. A 307, 15–22.
Miyazaki, S., Fujita, H., 1979. Dynamic observation of the process of L€ uders band formation in
polycrystalline iron. Trans. Japan Inst. Metals 20, 603–608.
Moon, D.W., 1971. Strain distribution through a propagating L€ uders band front. Scripta Metall. 5, 213–
216.
670 J. Zhang, Y. Jiang / International Journal of Plasticity 21 (2005) 651–670

Morrison, W.B., Glenn, R.C., 1968. Examination of the L€ uders front in a low-carbon steel by
transmission electron microscopy. J. Iron Steel Inst. 206, 611–612.
Onodera, R., Nonomura, M., Aramaki, M., 2000. Stress drop, L€ uders strain and strain rate during
serrated flow. J. Japan Inst. Metals 64, 1162–1171.
Prewo, K., Li, J.C.M., Gensamer, M., 1972. L€ uders band motion in iron. Metall. Trans. 3, 2261–2269.
Shaw, J.A., Kyriakides, S., 1995. Thermomechanical aspects of NiTi. J. Mech. Phys. Solids 43, 1243–1281.
Sylwestrowicz, W., Hall, E.O., 1951. The deformation and aging of mild steel. Proc. Phys. Soc. B 64, 495–
502.
Verel, D.J., Sleeswyk, A.W., 1973. L€
uders band propagation at low velocities. Acta Metall. 21, 1087–1098.
Zhang, J., Jiang, Y., 2004. A study of inhomogeneous plastic deformation of 1045 seel. ASME J. Eng.
Mater. Tech. 126, 164–171.