Thin-Walled Structures ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Contents lists available at ScienceDirect

Thin-Walled Structures
journal homepage: www.elsevier.com/locate/tws

Residual stress influence on material properties and column behaviour

of stainless steel SHS
M. Jandera n, J. Machacek
Faculty of Civil Engineering, Czech Technical University in Prague, Thakurova 7, 166 29 Praha 6, Czech Republic

art ic l e i nf o a b s t r a c t

The paper describes the influence of forming-induced residual stresses in stainless steel SHS (Square
Keywords: Hollow Sections) on both the material itself and the behaviour of compressed members. The residual
Stainless steel stress contribution to the stress–strain diagram concerning the initial modulus of elasticity and non-
Square hollow section linearity is shown by the comparison of as-delivered and stress-relieved material. An analytical model
Residual stress covering the influence of residual stresses on material behaviour was developed and verified
Cold forming numerically. The FE study using the Abaqus software determines the influence of residual stresses in
compressed members both on local and global buckling. Finally, the study on behaviour of member
includes the influence of a varying degree of material non-linearity as an independent parameter
representing the behaviour of various steels in long and stub columns with residual stresses.
& 2014 Elsevier Ltd. All rights reserved.

1. Introduction was also proposed. Measurements for a new lean duplex stainless
steel grade were published in 2012 by Huang and Young [9].
Stainless steels are increasingly being used as structural mem- In 2011, predictive formulae were published by Jandera and
bers, particularly in the form of rectangular or circular cold-formed Machacek [10] for longitudinal (in the direction of the member axis)
hollow sections. Investigations on the performance of stainless bending (sb.pl) and membrane (sm), as well as transverse (in the
steel oval sections were published [1,2] recently. direction of section webs) residual bending (sb.pl.t) stresses. The
This paper is focused on square hollow sections (SHS) made of bending components were simplified by a plastic distribution pattern
austenitic steel Grade 1.4301 which is the most commonly used through the thickness of the material. This simplification was also
grade of stainless steel. However, the results presented are used by Cruise and Gardner [8]. A plastic-like stress distribution was
believed to be valid for other austenitic grades, and also ferritic previously presented in the analytical model of Quach et al. [11,12]
and duplex stainless steels. (except for hollow sections) and experimentally in [13].
The process of forming thin-walled structures induces residual Measurements were made for three webs of two SHS, namely
stresses which may have a significant effect on structural beha- SHS 100
100
3 and 120
120
4. Significant simplifications
viour. Residual stresses in cold-formed sections are generally were made, particularly for the transverse bending component
expected to have a substantial bending component and a com- (the transverse stresses were measured in one position only, at the
paratively low membrane component; these residual stresses are centre of the web). The resulting equations for all components are
thermally induced in welded or hot-rolled sections [3]. Large given below
residual longitudinal bending stresses in stainless steel sections
sm ¼ ð  0:253 þ 1:483ðx  x2 ÞÞs0:2 ð1Þ
subject to compression and bending have been measured during
experimental investigations [4,5]. sb:pl ¼ ð0:833  1:866ðx  x2 ÞÞs0:2 ð2Þ
The residual stresses for carbon steel SHS were measured by Key
and Hancock [6] and a model of their distribution in both directions sb:pl:t ¼  0:376s0:2 ð3Þ
was proposed. The first measurements of stresses for all stainless steel
SHS/RHS were published by Young and Lui [7] and Cruise and Gardner where x is the relative distance along the web width (with x ¼0
[8], where the distribution of the longitudinal bending component and x ¼1 for the edges of a flat web, x ¼0.5 for the centre of the
web) and s0.2 is the 0.2% proof strength of the material in the
centre of the web. The plus sign is used for tension and the minus
n
Corresponding author.
sign for compressive stress. For residual bending stresses, the sign
E-mail addresses: [email protected] (M. Jandera), is related to the outer surface of SHS (plus means tension on the
[email protected] (J. Machacek). outer surface and compression on the inner surface). Stresses in

http://dx.doi.org/10.1016/j.tws.2014.03.013
0263-8231/& 2014 Elsevier Ltd. All rights reserved.

Please cite this article as: Jandera M, Machacek J. Residual stress influence on material properties and column behaviour of stainless
steel SHS. Thin-Walled Structures (2014), http://dx.doi.org/10.1016/j.tws.2014.03.013i
2 M. Jandera, J. Machacek / Thin-Walled Structures ∎ (∎∎∎∎) ∎∎∎–∎∎∎

on the stress–strain diagram and also to analyse the results of the

FE model.

2.1. Model description

The analytical model comprised two layers (Fig. 3) through the

web thickness. Plastic material response assuming von Mises's
yielding and the Prandtl–Reuss flow rule [17] was considered. The
assumptions simplify the true material response of stainless steel
behaviour, neglecting for example anisotropy, non-symmetry and
an increase of ultimate strength due to cold-forming. The model
was set up in a Maple mathematical software; the full code was
published in Ref. [18].
A similar approach was used by Quach et al. [11,12] for the
prediction of residual stresses due to cold-forming of stainless
steel sections, or by Rossi et al. [19] for strength enhancement
prediction.

2.2. Material properties

The accuracy of the model was studied using tensile tests of

coupons taken from two SHS bars (SHS 100
100
3 and SHS
Fig. 1. Residual longitudinal membrane stresses sm around a square hollow
section, Eq. (1).
120
120
4), which were previously investigated for residual
stresses. From both of them, one specimen was stress-relieved
using 650 1C annealing. The material properties are taken from
coupons in the web centre and are listed in Table 1. The properties
define a compound stress–strain diagram developed by Gardner
and Nethercot [20] as a modification of the Mirambell and Real
compound diagram [21], where

E0 is the initial modulus of elasticity,

s0.2 the 0.2% proof strength,
s1.0 the 1.0% proof strength,
su the ultimate tensile strength,

Fig. 2. Residual longitudinal bending stresses sb.pl around a square hollow section,
Eq. (2).

the longitudinal direction around the cross-section according to

Eqs. (1) and (2) are illustrated in Figs. 1 and 2. This distribution was
considered in the numerical studies presented hereafter.
Besides residual stresses, the material characteristics are also
affected by the cold-forming process. The material properties of
cold-formed hollow section have been investigated by many Fig. 3. Bending stress approximation considered in the analytical model.
researchers. The first result to give the corner strength increase
and define the corner region was published by Gardner [14]. In the
recent publication by Afshan et al. [15,16] where more sections Table 1
Material properties for flat parts of SHS (web centre).
were tested, a strength increase for flat parts was also given and
previous predictive formulae were modified. Specimen E0(GPa) s0.2(MPa) s1.0(MPa) sυ(MPa) n(  ) n00:2;1:0 (  )

100
100
3-F 205.8 417 457 753 7.1 2.3

2. Analytical model 100
100
3-FAa 211.5 429 456 753 13.4 1.5
120
120
4-F 192.0 429 479 783 4.3 2.7
120
120
4-FAa 205.5 405 441 762 8.1 2.1
Together with the FE model described below, an analytical
model was constituted to study the influence of residual stresses a
Stress relieved specimen.

Please cite this article as: Jandera M, Machacek J. Residual stress influence on material properties and column behaviour of stainless
steel SHS. Thin-Walled Structures (2014), http://dx.doi.org/10.1016/j.tws.2014.03.013i
 M. Jandera, J. Machacek / Thin-Walled Structures ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 3

n the Ramberg–Osgood hardening exponent,

n00:2;1:0 the modified Ramberg–Osgood hardening exponent for
the second stage of the diagram using s0.2 and s1.0.

This compound diagram was subsequently used to study the

accuracy of modelling up to 1% proof strength.
A significant difference in the material non-linearity (a different
Ramberg–Osgood hardening exponent n) and a slight change of
initial stiffness are obvious from comparison with as-delivered
(F) and stress-relieved (FA) specimens. The difference was assigned
to the presence of residual longitudinal bending stresses in as-
delivered specimens (which were reintroduced by the tensile test).

2.3. Introduction of residual stresses

Fig. 5. SHS 120
120
4 specimen: as-delivered (F), stress-relieved (FA) and
stress–strain curves where longitudinal bending stresses are introduced by the
Along with the comparison of coupon tests, analytical model- analytical model.
ling was performed. The model was based on a stress-relieved
material model where the longitudinal bending component of
residual stresses was employed. analytical model was also confirmed by the FE model of coupon
As all the coupons were taken from the web centre, the stress tests, and an almost identical curve was observed.
was considered for this part of the section. According to the
residual stress measurement, a slightly different magnitude was 2.5. Results
used for SHS 100
100
3 and SHS 120
120
4. For SHS 100

100
3, the stresses were calculated in Eq. (4), and for SHS 120
The analytical model results confirm that the modulus of
120
4 in Eq. (5) from the 0.2% proof strength of as-delivered elasticity and material nonlinearity depends strongly on the
specimens residual stress magnitude (increasing with the magnitude of
residual stresses). Therefore, the material non-linearity is different
sb:pl ¼ 0:354s0:2 ¼ 0:354
416:5 ¼ 147:4 MPa ð4Þ
even for sections having a similar magnitude of cold forming but
different residual stresses, i.e. hollow or opened sections.
sb:pl ¼ 0:380s0:2 ¼ 0:380
429:0 ¼ 163:0 MPa ð5Þ
As an example, the measured Ramberg–Osgood hardening
The other components (membrane and transverse bending exponent n for flat parts and corners of the SHS could be used.
stresses) may be assumed to be zero during the tensile coupon The coefficient n for the flat parts was on an average about 18%
tests. Therefore, the longitudinal bending stress component was lower (which means larger non-linearity) than for corners in the
the only one to be considered. To obtain the stress–strain diagram, measurements of six SHS tested by Jandera et al. [13]. Even though
which may be compared with the tensile test, the stress was the corner areas are cold-formed which increases the nonlinearity
calculated iteratively via the previously described method (Section 2.1) (as proven before in Ref. [8]), the significantly larger residual
simulating a uniaxial strain increase. stresses in the flat parts lead to even greater nonlinearity.
The results presented above imply the suitability of the
2.4. Model validation previously mentioned simplification of the residual stress distri-
bution, which was also used in the following FE study.
Figs. 4 and 5 compare the tensile coupon test results and the The other purpose of the analytical model is to allow estimation
analytical model results. Considering the model simplifications, such of the residual stress free material behaviour when only coupons for
as taking a perfectly plastic residual bending stress distribution as-delivered sections are tested. As the residual stresses were known,
through thickness and a slightly simplified stress–strain behaviour the residual stress free material characteristics were calculated
characterised by the compound diagram, the residual stress influence iteratively. Due to the iterative calculation procedure, the analytical
was reasonably approximated. Nevertheless, the prediction is pretty model was found to be more time-efficient than the FE model of the
accurate for the initial part, which is essential for the buckling of coupon (which, of course, may also be used for analysis).
compressed members. The mathematical correctness of the
3. FE model

The proposed model of the residual stress distribution was

subsequently introduced into the FE model, which was success-
fully validated in Ref. [13] and contains a detailed description of
the model. Parametric studies of the influence of residual stresses
were divided into two parts: one focused on the influence of
residual stresses on the global column buckling where a pinned
column was modelled, the other on the local web buckling only,
which was represented by a stub column model. The boundary
conditions are shown in Fig. 6.

3.1. Initial geometric imperfections

Fig. 4. SHS 100
100
3 specimen: as-delivered (F), stress-relieved (FA) and
Local geometric imperfections were considered in the shape of
stress–strain curves where longitudinal bending stresses are introduced by the the lowest local buckling mode with the amplitudes measured on
analytical model. the specimens before tests [13] (for SHS 120
120
4 it was

Please cite this article as: Jandera M, Machacek J. Residual stress influence on material properties and column behaviour of stainless
steel SHS. Thin-Walled Structures (2014), http://dx.doi.org/10.1016/j.tws.2014.03.013i
4 M. Jandera, J. Machacek / Thin-Walled Structures ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Fig. 6. FE models of stub and long columns used in parametric studies. Fig. 7. Parametric study of the residual stress influence on long column's
load-carrying capacity.

0.84 mm). These web deflections were introduced into the models residual bending stress component was removed within the analytical
of stub columns as well as for long columns. In addition, global model. The resulting stress–strain diagram, therefore, corresponded to
initial geometric column deflections were introduced into the long the stress-relieved material relationship. The enhanced corner proper-
column models by the lowest global buckling eigenmode with the ties were assumed for the corner regions which consisted of curved
amplitude of L/2000, where L is the column length. This value was corner sections plus a distance of twice the web thickness on both
proposed by Gardner [14] as an average value, and lately con- sides of the corners (proposed for SHS by Gardner [14]).
firmed by Cruise and Gardner [22].
To study the residual stress influence, the average magnitudes 4.1. Influence of residual stresses on global buckling
of imperfections were introduced. Using safe (larger) amplitudes
of imperfections would mean a smaller residual stress contribu- By analysing column models of varying length, the influence of
tion to the load capacity. residual bending and membrane stresses on the global buckling
capacity over the slenderness range was assessed. The results of
3.2. Residual stresses the study are presented in Fig. 7.
For non-dimensional slenderness (defined as the square root of
Residual bending stresses were introduced in six integration the ratio of yield load and elastic buckling load) of up to 1.3,
points with a quadratic integration through the web thickness. The residual stresses may be seen to have a positive influence on load-
proposed residual stresses pattern was introduced in five modes: carrying capacity. Beyond this slenderness, a negative influence is
evident. Over the investigated slenderness range, the inclusion of
 Membrane: longitudinal membrane stresses only, residual stresses causes a variation in resistance between  16% to
 Longitudinal: longitudinal membrane and bending stresses, þ10% and  20% to þ 14%, if the upper bound of the 95% predictive
 Max. longitudinal: longitudinal membrane and bending stres- interval for residual longitudinal bending stresses is considered.
ses which were taken by the upper bound of the 95% predictive The influence of residual membrane stresses was negligible for the
interval [10], whole employed slenderness range.
 All: longitudinal membrane and bending stresses as well as The variation of resistance results principally from the effect of
transverse bending stresses, residual bending stresses on the non-linearity of the stress–strain
 Max. all: longitudinal membrane and bending stresses as well as curve. A positive influence of residual stresses arises when the column
transverse bending stresses, residual longitudinal bending stresses failure strains coincide with the increased tangent modulus region.
were taken as the upper bound of the 95% predictive interval [10]. This is illustrated for residual longitudinal bending stresses in Fig. 8
where the material response with and without residual longitudinal
Residual bending stresses in corners, where no measurement bending stresses is depicted (the average magnitude of residual
was carried out, were neglected in the analysis. As previously stresses for the web is employed). The material stress–strain curve
shown by Cruise and Gardner [8], residual bending stresses in containing residual stresses may be seen to be consistently below the
corners are low. This implies, together with increased strength in residual stress free curve (i.e. the secant modulus is always lower).
the corner and the comparatively small area of the corner, that However, this is not valid for the tangent modulus, which is known to
their effect on the resistance of the whole member is insignificant. be fundamental in controlling the column buckling resistance. Below
Residual longitudinal membrane stresses in corners were always approximately 0.12% strain, the tangent modulus of the stress free
calculated from the condition of equilibrium over the whole cross- curve is higher than that of the residual stresses containing curve.
section. Their magnitudes were very low. This corresponds with the Conversely, for higher strains, the reverse is true.
conclusions presented by Cruise and Gardner [8] and the patterns It follows from Fig. 9, where the load–strain curves of columns
proposed by Key and Hancock [4] for carbon steel SHS. are shown (strain denotes the ratio of the axial deformation to the
column length; similarly in Fig. 11), that the failure strains of
columns with slenderness of up to 1.3 reach more than 0.12%. That
4. Parametric study based on SHS 120
120
4 is, the range where residual longitudinal bending stresses were
found to have a positive influence on the tangential modulus.
The study was based on SHS 120
120
4 tested at the Czech For higher column slenderness (λ Z 1:3), lower strains are
Technical University in Prague. All principal geometrical and material reached at ultimate load. Therefore, the tangent modulus for
characteristics were measured. The material characteristics for the a material where residual bending stresses are included is lower
corner were taken as measured. For flat parts, the influence of the than the tangent modulus of a material without residual stresses

Please cite this article as: Jandera M, Machacek J. Residual stress influence on material properties and column behaviour of stainless
steel SHS. Thin-Walled Structures (2014), http://dx.doi.org/10.1016/j.tws.2014.03.013i
 M. Jandera, J. Machacek / Thin-Walled Structures ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 5

Fig. 8. Stress–strain relationship (left) and tangent modulus (right) of material with and without residual longitudinal bending stresses valid for the average web value
of SHS 120
120
4.

Fig. 10. Parametric study of the residual stress influence on stub column
Fig. 9. Load–strain diagrams of long columns. load-carrying capacity.

(Fig. 8), and thus residual longitudinal bending stresses were stresses was also found to be insignificant in comparison to the
found to reduce the load-carrying capacity. influence of the bending components.
The magnitude of residual stresses clearly influences the variation However, the main difference is that for local buckling, no
in load-carrying capacity. It was found that by considering mean negative influence of residual stresses occurred. This is due to the
residual longitudinal bending stress values rather than the upper post-buckling behaviour which increases the failure strain for very
bound of the predictive interval, the sensitivity of the column slender webs. In opposite to the post-buckling behaviour of
response to the presence of residual stresses was nearly halved for columns, the local buckling failure strain was always greater than
the most sensitive slenderness range (λ ¼ 0:6 to 1:2). For high 0.12% (Fig. 11), where residual stresses had a positive influence on
slenderness (λ Z1:3), the difference was lower. The difference was the tangential modulus.
higher when the transverse bending stress was also included. The highest effect of residual bending stresses is observed for
medium slenderness (λ p ¼1.0 to 1.3) where the sensitivity of columns
4.2. Influence of residual stresses on local buckling is large and the difference in the tangential modulus at failure strain
between a material with and without bending stresses is the highest.
The influence of residual stresses on the local buckling capacity
was assessed in a similar manner to the above studied global
buckling. Stub column models of varying local web slenderness
with and without residual stresses were examined. The results are
shown in Fig. 10. The maximum influence of residual stresses in 5. Parametric study on the influence of Ramberg–Osgood
terms of load-carrying capacity was 9% and 11% if the upper bound hardening exponent
of the 95% predictive interval for residual longitudinal bending
stresses was considered. Although the influence was slightly The second parametric study was performed in order to explain
less sensitive than that in the column buckling results, similar the influence of residual stresses in materials of various nonlinea-
conclusions could be drawn. The influence of residual membrane rities. For simplicity, only the one-stage Ramberg–Osgood diagram

Please cite this article as: Jandera M, Machacek J. Residual stress influence on material properties and column behaviour of stainless
steel SHS. Thin-Walled Structures (2014), http://dx.doi.org/10.1016/j.tws.2014.03.013i
6 M. Jandera, J. Machacek / Thin-Walled Structures ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Fig. 11. Load–strain diagrams of stub columns.

Fig. 13. Influence of residual stresses on load-carrying capacity of long columns
with λ ¼1.8 according to the Ramberg–Osgood factor n.

Fig. 12. Influence of residual stresses on load-carrying capacity of long columns

with λ ¼ 1.0 according to the Ramberg–Osgood factor n.
Fig. 14. Influence of residual stresses on load-carrying capacity of stub columns
with λ p ¼ 1.0 according to the Ramberg–Osgood factor n.
of Eq. (6) was used for the whole cross-sectional area in all models
 
s s n
ε ¼ þ 0:002 ð6Þ
E0 s0:2
where ε is the strain and s the stress. The 0.2% proof strength was
considered according to Eurocode 1993-1-4, i.e. s0.2 ¼230 MPa for
Grade 1.4301. The hardening exponent n was taken as:

 n ¼4, representing cold-formed austenitic steels,

 n ¼6, representing annealed austenitic steels,
 n ¼16, representing the lowest nonlinearity of stainless steels,
 n ¼1, a bilinear stress–strain diagram representing common
carbon steels.

The initial geometric imperfections and residual stresses were

employed as above.

5.1. Residual stresses and material nonlinearity in global buckling

Fig. 15. Influence of residual stresses on load-carrying capacity of stub columns
with λ p ¼ 1,8 according to the Ramberg–Osgood factor n.
The parametric study is carried out for a column slenderness
range from 0.8 to 1.8, but only the results of the two most
representative slenderness values λ ¼ 1.0 and 1.8 are shown here SHS by Key and Hancock [6]. This is due to a significant drop in the
(Figs. 12 and 13). stiffness of a material which offers no or low strain hardening
The results indicate that residual stresses for a nonlinear material beyond the yield strength. The influence of the membrane
(n¼4 to 6) may increase the load-carrying capacity for medium component of residual stresses is always low.
slenderness (λ ¼1.0) but for high slenderness (λ ¼ 1.8) the opposite The magnitude of the influence of residual stresses, however, is
is true. rather overestimated in this study due to the simplification of
For materials with a less non-linear stress–strain diagram, material characteristics. The increased strength of a material in the
especially for the bilinear stress–strain diagram (e.g. carbon steel), corner area as well as a slightly different stress–strain diagram
the inclusion of residual stresses always causes a decrease in the (better represented by the compound Ramberg–Osgood diagram)
load capacity as generally known and numerically confirmed for would suppress the residual stress influence significantly.

Please cite this article as: Jandera M, Machacek J. Residual stress influence on material properties and column behaviour of stainless
steel SHS. Thin-Walled Structures (2014), http://dx.doi.org/10.1016/j.tws.2014.03.013i
 M. Jandera, J. Machacek / Thin-Walled Structures ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 7

5.2. Residual stresses and material nonlinearity in local buckling modulus of elasticity. It is therefore proposed to estimate and use a
different (smaller) hardening exponent n for cross-sections in
The second part of the study focusing on local buckling showed which large residual bending stresses are present (notably SHS
similar results (for λ p ¼1.0 and 1.8, see Figs. 14 and 15). However, and RHS).
the magnitude is different. As in the previous study based on real
material characteristics, the effect of residual stresses is always
positive for very nonlinear stress–strain relationships, whereas for Acknowledgements
the bilinear material diagram it is always negative. The influence
of residual membrane stresses was negligible in all cases. The support of the Czech Science Foundation Grant P105/12/
P307 (“Influence of Cold-Forming on Stainless Steel Mechanical
Properties”) is gratefully acknowledged.
6. Conclusions
References
The paper describes a numerical investigation of the influence
of residual stresses on the stress–strain relationships and column
[1] Theofanous M, Chan TM, Gardner L. Flexural behaviour of stainless steel oval
behaviour of structural stainless steel hollow sections. Extensive hollow sections. Thin Walled Struct 2009;47(6-7):776–87.
numerical parametric studies using geometrically and materially [2] Theofanous M, Chan TM, Gardner L. Structural response of stainless steel oval
non-linear FE analysis were performed dealing with long and stub hollow section compression members. Eng Struct 2009;31(4):922–34.
[3] Rasmussen KJR, Hancock GJ. Deformations and residual stresses induced in
columns to determine the influence of residual stresses on global
channel section columns by pre-setting and welding. J Constr Steel Res
and local buckling. The results were verified using previous 1988;11(3):175–204.
experimental results. [4] Rasmussen KJR, Hancock GJ. Design of cold-formed stainless steel tubular
members, I: Columns. J Struct Eng 1993;119(8):2349–67.
Paradoxically, it was found that the inclusion of residual
[5] Rasmussen KJR, Hancock GJ. Design of cold-formed stainless steel tubular
stresses may lead to an increase of load-carrying capacity in some members, II: Beams. J Struct Eng 1993;119(8):2368–86.
cases. This was principally attributed to the influence of residual [6] Key PW, Hancock GJ. A theoretical investigation of the column behaviour of
bending stresses on the material stress–strain curve. It was found cold-formed square hollow sections. Thin-Walled Struct 1993;16(1–4):31–64.
[7] Young B, Lui WM. Behaviour of cold-formed high strength stainless steel
that despite the secant modulus being consistently reduced in the sections. J Struct Eng 2005;131(11):1738–45.
presence of residual stresses, the tangent modulus was increased [8] Cruise RB, Gardner L. Residual stress analysis of structural stainless steel
in some regions of the stress–strain curve. The influence of sections. J Constr Steel Res. 2008;64(3):352–66.
[9] Huang Y, Young B. Material properties of cold-formed lean duplex stainless
residual stresses was between þ 10% to  16% for global buckling steel sections. Thin-Walled Struct 2012;54:72–81.
and up to þ 9% for local buckling of SHS members using the FE [10] Jandera, M, Machacek, J . Measurement of residual stress pattern in stainless
model based on measured material characteristics. The influence steel cold-rolled SHS. In: Proceedings of the 6th international conference on
thin-walled structures. Timisoara; 2011. p. 619–624.
of residual membrane stresses in the analyses was usually very [11] Quach WM, Teng JG, Chung KF. Residual stresses in press-braked stainless steel
low and may be neglected in most cases. sections, I: Coiling and uncoiling of sheets. J Constr Steel Res 2009;68(8–9):
A parametric study with a varying Ramberg–Osgood strain 1803–15.
[12] Quach WM, Teng JG, Chung KF. Residual stresses in press-braked stainless
hardening factor n showed that the inclusion of the residual stress
steel sections, II: Press-braking operations. J Constr Steel Res 2009;68(8–9):
pattern may lead to an increase in the load-carrying capacity of 1816–26.
compressed members having a non-linear stress–strain diagram, [13] Jandera M, Gardner L, Machacek J. Residual stresses in cold rolled stainless
but always to a decrease in capacity for materials with a bilinear steel hollow sections. J Constr Steel Res 2008;64(11):1255–63.
[14] Gardner L . A new approach to structural stainless steel design [PhD thesis].
stress–strain diagram (e.g. for common carbon steel). The results Department of Civil and Environmental Engineering, Imperial College London;
proved that both residual stresses (mainly the bending compo- 2002.
nent) and the degree of non-linearity may have a significant [15] Afshan S, Rossi B, Gardner L. Strength enhancements in cold-formed structural
sections-Part I: Material testing. J Constr Steel Res 2013;83(4):177–88.
influence on the load capacity. [16] Afshan S, Rossi B, Gardner L. Strength enhancements in cold-formed structural
Although the behaviour of stainless steel columns with and sections – Part II: Predictive models. J Constr Steel Res 2013;83(4):189–96.
without residual bending stresses has been investigated in this [17] Hill R. The mathematical theory of plasticity. Norfolk, UK: Oxford University
Press; 1998.
study, it should be noted that these stresses are inherently present [18] Jandera, M . Residual stresses in stainless steel box sections [PhD thesis]. Czech
in the stress–strain behaviour of material extracted from structural Technical University in Prague; 2009 (in Czech).
sections and, therefore, need not generally be re-introduced into [19] Rossi B, Degée H, Pascon F. Enhanced mechanical properties after cold process
of fabrication of non-linear metallic profiles. Thin-Walled Struct 2009;47
numerical models. However, in the design practice as opposed to
(12):1575–89.
research, coupon tests are usually not performed and generalised [20] Gardner L, Nethercot DA. Experiments on stainless steel hollow sections – Part 1:
stress–strain relationships are used. In such cases, residual stresses Material and cross-sectional behaviour. J Constr Steel Res 2004;60(9):1291–318.
are not considered. As shown by the analytical model, the non- [21] Mirambell E, Real E. On the calculation of deflections in structural stainless
steel beams. J Constr Steel Res 2000;54(1):109–33.
linearity is increased significantly by the inclusion of residual [22] Cruise RB, Gardner L. Measurement and prediction of geometric imperfections
bending stresses, accompanied by a slight decrease in the initial in structural stainless steel members. Struct Eng Mech 2006;24(1):63–89.

Please cite this article as: Jandera M, Machacek J. Residual stress influence on material properties and column behaviour of stainless
steel SHS. Thin-Walled Structures (2014), http://dx.doi.org/10.1016/j.tws.2014.03.013i