You are on page 1of 6

EUROSTEEL 2014, September 10-12, 2014, Naples, Italy

STAINLESS STEEL CROSS-SECTIONS UNDER COMBINED LOADING


Ou Zhao a, Barbara Rossi b, Leroy Gardner a, Ben Young c
a
Imperial College London, Dept. of Civil and Environmental Engineering, London, UK
ou.zhao11@imperial.ac.uk, leroy.gardner@imperial.ac.uk
b
KU Leuven, Dept. of Civil Engineering, Belgium
barbara.rossi@set.kuleuven.be
c
The University of Hong Kong, Dept. of Civil Engineering, Hong Kong, China
young@hku.hk

INTRODUCTION
Stainless steel has been gaining increasing use in a variety of engineering applications due to its
unique combination of mechanical properties, durability and aesthetics. Significant progress in the
development of structural design guidance has been made in recent years, underpinned by sound
research. However, an area that has remained relatively unexplored is that of combined loading.
Testing and analysis of stainless steel cross-sections under combined axial load and bending is
therefore the subject of the present paper. A comprehensive experimental study was carried out to
investigate the cross-sectional behaviour of tubular sections under combined bending and
compression. The experimental results are analysed and then compared to the current EN 1993-1-4
[1] capacity predictions for stainless steel, which neglect the strain hardening effect and thus yield
conservative results. The Continuous Strength Method (CSM) is a deformation-based design
approach accounting for strain hardening and has been shown to provide accurate predictions of
cross-sectional resistance under compression and bending, acting in isolation. Initial proposals to
extend the CSM to the combined loading cases are made in this paper, of which the applicability
and accuracy are assessed. Finally, a modified EC3/CSM method, using the EC3 interaction
expressions but with CSM end points, is proposed and also shown to provide accurate failure load
predictions.

1 EXPERIMENTAL INVESTIGATIONS
1.1 Introduction
An experimental study covering austenitic and lean duplex stainless steels and a range of cross-
section sizes was conducted in the University of Liege and Imperial College London, with the aim
of investigating the cross-sectional behaviour of tubular structural sections under combined loading
conditions. In total, the laboratory testing programme comprised geometric imperfection
measurements, material testing, 5 stub column tests, 20 uniaxial bending plus compression tests and
4 biaxial bending plus compression tests. 5 section sizes were employed, which were SHS
1201205, SHS 1001005, RHS 1501006, and RHS 1501008 of austenitic grade (EN
1.4571, 1.4301, 1.4307 and 1.4404, respectively) and SHS 1501508 of lean duplex grade (EN
1.4162). All the cross-sections are at least Class 3 sections according to the slenderness limits in EN
1993-1-4. The testing apparatus, experimental procedures and test results are detailed in the
following sections.
1.2 Material testing
A comprehensive description of the material testing was given by Afshan et al. [2] whilst only a
brief summary is presented herein. For each section size, two flat coupons and two corner coupons
were tested. The weighted average (based on face width) measured properties from flat coupon tests
and corner coupon tests for each section size are reported in Table 1, where E is Youngs modulus,
0.2 is the 0.2% proof stress, u is the ultimate tensile stress, and n and n0.2,1.0 are the strain
hardening exponents used in the compound Ramberg-Osgood material model [3], in which n is
determined based on the 0.01% and 0.2% proof stresses.
Table 1. Weighted average coupon material properties
Flat coupon tests Corner coupon tests
Section E 0.2 u E 0.2 u
2 2 n n0.2,1.0 2 2 n n0.2,1.0
(N/mm ) (N/mm ) (N/mm ) (N/mm )
SHS 1201205 192550 343 605 6.7 2.4 192900 526 687 10.8 3.0
SHS 1001005 193400 434 683 4.7 2.9 192200 599 810 3.6 19.6
RHS 1501006 193250 341 642 6.6 2.2 189750 607 808 8.7 3.0
RHS 1501008 196109 335 608 5.9 2.6 200700 559 725 4.8 3.9
SHS 1501508 198700 519 728 5.3 2.8 206750 831 920 8.9 6.1

1.3 Stub column tests


For each cross-section, one stub column test was performed. In each test, displacement-control was
used to drive the hydraulic actuator at a constant speed of 0.15mm/min. The test setup is shown in
Fig. 1. The measured geometric dimensions and the key test results are summarised in Table 2,
where L is the member length, B and H are the outer section width and depth respectively, t is the
thickness, ri is the internal corner radius, Fu is the failure load and u is the corresponding end
shortening at failure. The modified true loadend shortening curves for all the specimens, following
procedures proposed by [4], are shown in Fig. 2, whilst the typical local buckling failure mode is
depicted in Fig. 3.

Table 2. Stub column test data


L H B t ri Fu u
Specimens
(mm) (mm) (mm) (mm) (mm) (kN) (mm)
SHS 1201205 399.9 120.3 120.1 4.65 5.79 928.4 3.6
SHS 1001005 349.9 100.0 99.9 4.65 2.08 1057.0 5.1
RHS 1501006 450.1 150.6 100.0 5.87 7.05 1323.7 7.1
RHS 1501008 450.0 150.1 100.2 7.75 9.65 1825.1 12.8
SHS 1501508 449.8 150.4 150.0 8.04 11.17 3257.9 10.3 Fig. 1. Stub column test setup

3500
3000
SHS 1501508
2500
2000 RHS 1501008
Fu(kN)

1500
RHS 1501006
1000
SHS 1001005
500
SHS 1201205
0
0 4 8 12 16 20
End shortening (mm)
Fig. 3. Typical stub column
Fig. 2. Stub column load-end shortening curves failure mode

1.4 Eccentric compression tests


A total of 24 eccentric compression tests were carried out 20 uniaxial bending plus compression
tests and 4 biaxial bending plus compression tests to investigate the cross-sectional behaviour
under combined bending and compression. The combined loading tests were conducted using an
AMSLER 5000kN hydraulic testing machine with hemispherical bearings at both ends providing
pinned end conditions in any direction. Fig. 4 depicts the test setup, consisting of two string
potentiometers to measure the lateral deflection along each principal axis as adopted in [5], two
inclinometers to determine the end rotations and four stain gauges to capture the longitudinal strains.
Table 3 and Tables 45 summarise the specimen ID, the initial eccentricity e0, the geometric
measurement, and the key test results including the failure load Fu, the generated lateral deflection
at failure load, referred to as the second order eccentricity e, the failure moment Mu=Fu(e0+e) and
the corresponding end rotation at failure u for the uniaxial bending plus compression tests and the
biaxial bending plus compression tests, respectively. The experimental loadend rotation curves for
each cross-section under uniaxial bending plus compression with varying initial eccentricities are
shown in Figs. 59, whist those under biaxial bending plus compression are depicted in Figs. 1013.
Typical local buckling failure modes for SHS and RHS under combined loading are displayed in
Fig. 14.

Table 3. Summary of uniaxial bending plus compression tests


e0 L H B t ri Fu e' Mu u
Section ID
(mm) (mm) (mm) (mm) (mm) (mm) (kN) (mm) (kN m) (deg)
1A 10 399.9 120.2 120.1 4.65 5.79 793.5 2.6 10.0 0.98
1B 38 400.0 120.0 120.1 4.61 5.70 550.0 2.8 22.4 1.24
SHS 1201205
1C 68 400.0 120.0 120.0 4.61 5.81 424.0 3.0 30.1 1.55
1Da 120 399.8 120.0 120.1 4.59 5.75 296.1 3.2 36.5
2A 18 350.0 100.0 99.9 4.65 2.08 743.5 4.4 16.6 1.61
2B 26 350.0 100.0 100.0 4.70 2.20 622.2 4.8 19.0 1.72
SHS 1001005
2C 53 350.0 100.0 100.0 4.66 2.14 472.7 6.1 27.9 2.07
b
2D 120 350.0 100.0 100.1 4.66 2.15
3A 45 350.1 150.1 100.1 5.85 7.00 825.2 6.2 42.1 2.04
RHS 1501006- 3B 64 449.8 150.4 100.0 5.85 7.05 685.3 7.2 48.9 2.60
MA 3C 92 450.1 150.1 99.9 5.82 7.01 575.7 7.4 57.5 2.60
3D 128 450.0 150.2 100.0 5.90 7.05 473.4 8.8 64.9 3.16
4A 20 450.0 150.1 100.0 7.73 9.61 1173.8 9.5 34.5 3.06
RHS 1501008- 4B 52 450.2 150.1 100.1 7.70 9.64 800.1 12.4 51.2 3.93
MI 4C 75 450.0 150.1 100.0 7.71 9.70 626.9 13.2 55.3 4.66
b
4D 140 450.1 150.0 100.0 7.68 9.60
5A 30 449.8 150.2 150.0 8.00 11.10 2186.7 7.2 80.3 2.40
5B 52 450.0 150.1 150.0 7.99 11.15 1814.9 7.2 106.7 2.76
SHS 1501508
5C 84 450.0 150.0 150.0 8.02 11.15 1403.6 7.3 128.4 3.19
b
5D 116 450.1 150.0 150.0 8.04 11.17 1186.9 >3.4 142.2
a. No rotation data was obtained.
b. Fracture of welded end plate occurred prior to specimen failure.
MA indicates bending about the major axis and MI means bending about the minor axis.

Table 4. Geometric dimensions of biaxial bending plus compression


specimens
L H B t ri
Section ID
(mm) (mm) (mm) (mm) (mm)
2E 350.0 120.1 120.0 4.60 5.75
SHS 2F 349.9 120.1 120.1 4.70 5.78
1001005 2G 350.0 120.0 120.1 4.65 5.80
2H 350.0 120.0 120.1 4.65 5.75
Fig. 4. Combined loading test setup
Fu (kN) Fu (kN) Fu (kN)
1000 800 1000
e0=10mm e0=18mm e0=45mm
800 600 750 e0=64mm
600 e0=38mm e0=26mm
400 500 e0=92mm
400 e0=53mm
e0=128mm
200 e0=68mm 200 250
0 0 0
0 0.5 1 1.5 2 2.5 3 3.5 0 1 2 3 4 5 0 1 2 3 4 5 6
End rotation (deg) End rotation R (deg) End rotation R (deg)

Fig. 5. Test curves for SHS Fig. 6. Test curves for SHS Fig. 7. Test curves for SHS
1201205 (specimens: 1A to 1C) 1001005 (specimens: 2A to 2C) 1501006 (specimens: 3A to 3D)
Fu (kN) Fu (kN) Fu (kN)
1400 2500 800
1200 e0=20mm 2000 e0=30mm
e0=52mm 600
1000 e0=52mm e0=84mm
800 1500
400 e0y=e0z=20mm
600 1000 e0=116mm
400 e0=75mm 200
200 500
0 0 0
0 2 4 6 8 0 1 2 3 4 5 6 7 0 1 2 3 4
End rotation (deg) End rotation (deg) End rotation (deg)

Fig. 8. Test curves for RHS Fig. 9. Test curves for SHS Fig. 10. Test curves for SHS
1501008 (specimens: 4A to 4C) 1501508 (specimens: 5A to 5D) 1001005 (specimen: 2E)
Fu (kN) Fu (kN) Fu (kN)
600 500 400
500 400
300
400 300 e0z=60mm e0y=23mm e0z=78mm
300 200
e0y=23mm e0z=43mm 200
200 e0y=20mm
100 100
100
0 0 0
0 1 2 3 4 5 0 1 2 3 4 5 6 0 1 2 3 4 5
End rotation (deg) End rotation (deg) End rotation (deg)

Fig. 11. Test curves for SHS Fig. 12. Test curves for SHS Fig. 13. Test curves for SHS
1001005 (specimen: 2F) 1001005 (specimen: 2G) 1001005 (specimen: 2H)

Table 5. Summary of biaxial bending plus compression tests


e0y e0z Fu e'y e'z Muy Muz uy uz
Section ID
(mm) (mm) (kN) (mm) (mm) (kN m) (kN m) (deg) (deg)
2E 20 20 633.1 3.1 3.1 14.6 14.6 1.48 1.48
SHS 2F 23 43 484.3 2.8 3.8 12.5 22.7 0.96 1.95
1001005 2G 20 60 428.1 2.3 4.1 9.6 27.5 0.92 2.38
2H 23 78 363.2 2.3 4.2 9.2 29.9 0.82 2.53

Fig. 14. Typical failure modes for combined loading tests


2 COMPARISON BETWEEN TEST RESULTS AND DESIGN PREDICTIONS
Figs. 1519 show comparisons between the test results from the uniaxial bending plus compression
tests and three design methods EN 1993-1-4 (EC3), CSM, and a Modified EC3/CSM method,
which will be briefly discussed later. Quantitative comparisons between these three methods are
made in Table 6 in terms of the ratio of predicted resistance to test resistance.

1.2 1.2
Tests Tests
1.0 1.0
0.8 CSM 0.8 CSM
Mu/Mpl

Mu/Mpl
0.6 0.6
EC3 EC3
0.4 0.4
0.2 Modified 0.2 Modified
0.0 EC3/CSM EC3/CSM
0.0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2
Nu/Npl Nu/Npl
Fig. 15. Comparison of test results on SHS 1201205 Fig. 16. Comparison of test results on SHS 1001005
specimens to design curves specimens to design curves
1.4 1.2
Tests Tests
1.2 1.0
1.0 CSM CSM
0.8
Mu/Mpl

Mu/Mpl
0.8
0.6
0.6 EC3 EC3
0.4 0.4
0.2 Modified 0.2 Modified
EC3/CSM 0.0 EC3/CSM
0.0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 0.0 0.2 0.4 0.6 0.8 1.0 1.2
Nu/Npl Nu/Npl
Fig. 17. Comparison of test results on RHS 1501006 Fig. 18. Comparison of test results on RHS 1501008
specimens to design curves specimens to design curves
1.2
Tests
1.0
0.8 CSM
Mu/Mpl

0.6
EC3
0.4
0.2 Modified
EC3/CSM
0.0
0.0 0.2 0.4 0.6 0.8 1.0 1.2
Nu/Npl
Fig. 19. Comparison of test results on SHS 1501508 specimens to design curves

2.1 EN 1993-1-4 (EC3)


The current design guidance for stainless steel in EN 1993-1-4 does not provide specific provisions
for the cross-sectional behaviour under combined loading, but adopts the relevant carbon steel
design provisions in EN 1993-1-1 [6]. With respect to Class 1 or 2 cross-sections, stress
redistribution is allowed for by assuming full plasticity but without considering strain hardening
throughout the cross-section, with bilinear interaction formulae given for RHS under uniaxial
bending plus compression and non-linear interaction formulae given for RHS under biaxial bending
plus compression. As indicated in Table 6, the test results were under-predicted by an average of
about 19% and 20% for uniaxial bending plus compression cases and for biaxial bending plus
compression cases, respectively, indicating a high level of conservatism of EN 1993-1-4.
2.2 Continuous Strength Method (CSM)
The Continuous Strength Method (CSM) is a deformation-based design approach that relates the
strength of a cross-section to its deformation capacity and employs a bi-linear material model to
allow a rational exploitation of strain hardening. The method has been recently extended to cover
combined loading cases by Liew and Gardner [7], in which new interaction formulae have been
developed. The test results have been compared with the new proposals, as shown in Table 6, where
R refers to the cross-sectional capacity under combined loading. On average, an increase in the
resistance prediction of 6% and 5% were obtained for uniaxial bending plus compression and
biaxial bending plus compression, respectively, compared to the EN 1993-1-4.

Table 6. Comparisons of test strengths to predicted strengths using different methods


Uniaxial bending plus compression tests Biaxial bending plus compression tests
EC3 CSM Modified EC3/CSM EC3 CSM Modified EC3/CSM
Ru,EC3/Ru,test Ru,csm/Ru,test Ru,modified EC3/csm/Ru, test Ru,EC3/Ru,test Ru,csm/Ru,test Ru,modified EC3/csm/Ru, test
Mean 81% 87% 89% 80% 85% 88%
COV 0.03 0.03 0.04 0.02 0.02 0.02

2.3 Modified EC3 method with CSM resistances as end points


In this section, a new approach of adopting the CSM resistances Ncsm and Mcsm as the end points of
the cross-sectional interaction curves but keeping all the interaction formulae and coefficients in EN
1993-1-4 is proposed. Therefore, the general bi-linear format of the codified interaction curves was
maintained, but with Ncsm and Mcsm replacing Afy and Mpl or Mel. The accuracy of the modified
EC3/CSM method was assessed by comparing to the test results. The comparisons, shown in Table
6, reveal accurate predictions with the mean ratio of predicted resistance to test resistance equal to
89% and 88% for uniaxial eccentric compression tests and biaxial eccentric compression tests,
respectively.

3 CONCLUSIONS
A comprehensive experimental study of the cross-sectional behaviour of stainless steel under
combined loading has been presented. Comparisons of the test results with the current EN 1993-1-4
provisions indicated undue conservatism, attributed mainly to the lack of consideration for strain
hardening. The Continuous Strength Method for combined loading performed well for stainless
steel, yielding more accurate predictions than EN 1993-1-4. Finally, a modified EC3/CSM method
was proposed by utilizing the CSM resistances as the end points of the interaction curves but
retaining the interaction formulae and coefficients in EN 1993-1-4, which also led to a high level of
accuracy in the prediction of the cross-sectional resistance. As a consequence, more efficient design
could be obtained by adopting the latter two CSM-related methods.

REFERENCES
[1] EN 1993-1-4, Eurocode 3: Design of steel structures Part 1.4: General rules Supplementary rules
for stainless steel, European Committee for Standardization, Brussels, Belgium, 2006.
[2] Afshan, S, Rossi, B, Gardner, L, Strength enhancement in cold-formed structural sections Part I:
material testing, Journal of Constructional Steel Research, Vol. 83, No. 17, pp. 177-188, 2013.
[3] Gardner, L, Nethercot, D, A, Numerical modeling of stainless steel structural components- a consistent
approach, Journal of Structural Engineering, ASCE, Vol. 130, No. 10, pp. 1586-1601, 2004.
[4] Centre for Advanced Structural Engineering, Compression tests of stainless steel tubular columns,
Investigation Rep. S770, University of Sydney, Sydney, Australia, 1990.
[5] Fujimoto, T, Mukai, A, Nishiyama, I, Sakino, K, Behaviour of eccentrically loaded concrete-filled
steel tubular columns, Journal of Structural Engineering, ASCE, Vol.130, No. 2, pp. 203-212, 2004.
[6] EN 1993-1-1, Eurocode 3: Design of steel structures Part 1.1: General rules General rules and rules
for buildings, European Committee for Standardization, Brussels, Belgium, 2005.
[7] Liew, A, Gardner, L, Ultimate capacity of structural steel cross-sections under combined loading,
Journal of Constructional Steel Research, Submitted.

You might also like