Numerical study of the cyclic load behavior of AISI 316L stainless steel shear links for seismic fuse device

Abstract

This paper presents the results of nonlinear finite element analyses conducted on stainless steel shear links. Stainless steels are attractive materials for seismic fuse device especially for corrosion-aware environment such as coastal regions because they are highly corrosion resistant, have good ductility and toughness properties in combination with low maintenance requirements. This paper discusses the promising use of AISI 316L stainless steel for shear links as seismic fuse devices. Hysteresis behaviors of four stainless steel shear link specimens under reversed cyclic loading were examined to assess their ultimate strength, plastic rotation and failure modes. The nonlinear finite element analysis results show that shear links made of AISI 316L stainless steel exhibit a high level of ductility. However, it is also found that because of large over-strength ratio associated with its strain hardening process, mixed shear and flexural failure modes were observed in stainless steel shear links compared with conventional steel shear links with the same length ratio. This raises the issue that proper design requirements such as length ratio, element compactness and stiffener spacing need to be determined to ensure the full development of the overall plastic rotation of the stainless steel shear links.

Keywords

hysteretic damper / eccentrically braced frame / energy dissipation / seismic / stainless steel / shear link

Introduction

Growing interests in seismic design have been given to a strategy that involves the use of fuse-like energy dissipation devices with highly ductile performance while other members of the primary structural system are designed to have little or no damage. Examples of such energy dissipation devices are buckling-restrained brace [ 1], metallic yield damper [ 2], and shear links [ 3] in eccentrically braced frames (EBFs).

EBFs combine the advantages of moment resisting frames and concentrically braced frames which provide high ductility and stiffness of the structures, and make them a very competitive seismic force resistant system to use in high seismic regions. Nonlinear behavior of the EBFs is limited to the link beams while the other structural members remain elastic. Link beams are capable of developing large plastic deformation and are divided into three categories: shear links (short links), mixed shear-flexural links (intermediate links), and flexural links (long links). Using shear links is preferred because of its highly ductile performance under cyclic loads. The remarkable hysteresis behavior of shear links in EBFs is very appealing to seismic design and they have also been used in other structures such as bridges. It is worth noting that shear links were used in the new San Francisco-Oakland Bay Bridge for improving the seismic performance [ 4]. Shear links were installed to connect the four separate steel legs of the 525-foot-tall bridge tower of the self-anchored suspension span to absorb most of the seismic energy during an earthquake.

Stainless steels have been used in construction ever since they were invented over 100 years ago. They are attractive and highly corrosion resistant, while having good strength, toughness and fatigue properties in combination with low maintenance requirements [ 5]. The high ductility of stainless steels is a useful property where resistance to seismic loading is required since greater energy dissipation is possible. Stainless steel (SS) are also available in the form structural shapes and has been used for structural applications such as bridges. A recent example is the 55 m long Cala Galdana Bridge in Spain, which has its main structure entirely made of duplex stainless steel. Another stainless steel arch bridge in Europe is the 120 m long Piove di Sacco Bridge of Padua in Italy, which has two 1300 mm diameter main arches completely made of duplex stainless steel.

Critical structural components of bridges such as shear links in bridge tower, could potentially benefit from the high ductility and corrosion resistance properties of stainless steels, especially for bridges located in coastal regions. Despite the stainless steel’s higher initial cost, the life cycle cost of critical structural component made of stainless steel could be lowered due to its low maintenance.

This study examines the cyclic load behavior of shear links made of AISI 316L stainless steel based on numerical simulation results. There are a few parameters that affect the load capacity and ductility of shear links such as section size, compactness ratio, stiffener spacing, material properties and loading protocols. Four different sections of link beams with their length ratio value equal to 1.3 are used for this comparative study of AISI 316L stainless steel shear links and A992 steel shear links. Stiffener spacing and thickness are designed in accordance with the requirements prescribed in the AISC Seismic Provisions [ 6] for A992 steel link beams. It is observed that both types of shear links have high strength values and can undergo large inelastic deformations. The failure mode of stainless steel shear links are found to be mixed shear-flexural and eventually lose capacity due to local instability of the flanges despite the link length ratio is 1.3 (lower than 1.6 according to the definition for shear link in AISC Seismic Provisions 2010 [ 6]).

Finite element modeling

A total of four shear link specimens with a constant link length ratio of 1.3 for A992 steel based on AISC Seismic Provisions 2010 [ 6] have been selected to study the cyclic load behavior of AISI 316L stainless steel shear links in comparison with ASTM A992 Grade 50 steel: W10 × 33, W12 × 50, W14 × 74, and W16 × 77. The link length ratio, ρ, is defined as the ratio between link length L and the ratio of the plastic moment capacity, M_p, and the plastic shear capacity, V_p, of the link section, i.e., ρ = L/(M_p/V_p). This link length ratio of 1.3 is first used to determine the link beam length for A992 steel shear links, and then the same link beam length value is used for the corresponding AISI 316L stainless steel shear links. This is decided considering the special stress-strain curve features of AISI 316L stainless steel, that is, no pronounced yield plateau and significant strain hardening effect. AISC Seismic Provisions [ 6] defines the link beam types based on the value of this link length ratio, that is, shear link for ρ<1.6, flexural link for ρ>2.6, and intermediate link for 1.6<ρ<2.6. Table 1 lists the details of these four shear link specimens. It is worth noting that these wide-flange steel shapes are available in hot rolled shape made of A992 steel, but can be built up for stainless steel link beams even they are not currently available in hot-rolled stainless steel shapes. The thickness and the spacing of the web stiffeners are designed based on A992 steel according to the AISC Seismic Provisions [ 6], and for comparison purpose, the same web stiffener spacing and plate thickness values are adopted for the stainless steel link beams. It should be noted that currently no codes specifications for stainless steel link beams are available.

Material modeling

Accurate finite element analysis requires correct modeling of the corresponding material characteristics. As aforementioned, two types of materials are considered in this study of shear link specimens: ASTM A992 steel Grade 50 and AISI 316L stainless steel. For A992 steel, the cyclic material property adopted for this study was calibrated using experimental data from cyclic coupon tests performed by Kaufmann et al. [ 7] on ASTM A572 Grade 50. Steel yielding has been modeled using the Von Mises yield criterion, associated flow and the assumption of isotropic hardening. For AISI 316L stainless steel, the stress-strain relationship under cyclic loading is based on the experiments by Shit et al. [ 8].

To characterize the stress-strain relationship for stainless steel, two-stage models generally have been used (e.g., Ref. [ 9– 11]). The two stage model modified by Rasmussen [ 9] is adopted here to model the stress-strain relationship of 316L stainless steel, as shown in Eq. (1).

(1) ϵ=\{\begin{array}{l}{\displaystyle \frac{\sigma }{E_{\mathrm{0}}}}+\mathrm{0.002}{\left({\displaystyle \frac{\sigma }{{\sigma }_{\mathrm{0.2}}}}\right)}^{\mathrm{n}},\mathrm{f}\mathrm{o}\mathrm{r}\sigma \le {\sigma }_{\mathrm{0.2}}\\ {\displaystyle \frac{\sigma -{\sigma }_{\mathrm{0.2}}}{E_{\mathrm{0.2}}}}+{ϵ}_{\mathrm{u}}{\left({\displaystyle \frac{\sigma -{\sigma }_{\mathrm{0.2}}}{{\sigma }_{\mathrm{u}}-{\sigma }_{\mathrm{0.2}}}}\right)}^m+{ϵ}_{\mathrm{0.2}},\mathrm{f}\mathrm{o}\mathrm{r}\sigma >{\sigma }_{\mathrm{0.2}}\end{array}

where ϵ_u is the plastic strain at ultimate strength, and m is an additional strain hardening exponent defined as m = 1+ 3.5(σ_0.2/σ_u). E_0.2 is the tangent stiffness at σ_0.2, which can be calculated using the following formula, E_0.2 = E₀ /(1+ 0.002n/e). The term {ϵ}_{\mathrm{0.2}} is the strain value at {\sigma }_{\mathrm{0.2}} .The values of E₀, {\sigma }_{\mathrm{0.2}} , {\sigma }_{\mathrm{u}} , and {ϵ}_{\mathrm{u}} used for this study are given in Table 2; and e, n, m are material constants, which are set as e = 0.0017, n = 5.88, m = 2.8, respectively. Key parameter values required in this model for AISI 316L stainless steel are calibrated by Rasmussen [ 9] and listed in Table 2. The stress-strain curve for AISI 316L stainless steel under monotonic loading derived from this two-stage constitutive model is shown in Fig. 1.

Yielding is assumed to occur when the distortion energy in a unit material volume reaches a critical value. Based on the Von Mises yield criterion, this critical value of the distortion energy would be equal to that of the same volume of the material uniaxially loaded to yield point. In the principal stress space, the Von Mises yield criterion is represented by a yielding surface. Any stress state within the yielding surface implies that the material is still within the elastic range, while any stress state outside of the prior yield surface indicates that the material is in the plastic range. Once a stress state is outside the prior yield surface, a new yielding surface is formed and stresses are redistributed. The yielding surface is thus subject to change in its size or position [ 12].

Change of the yielding surface with plastic deformation is governed by hardening rule. Previous researchers have hypothesized two basic hardening rules—isotropic hardening and kinematic hardening, in characterizing the stain hardening behaviors of metallic materials in the plastic range [ 12]. The isotropic hardening rule prescribes that the yielding surface only changes in its size with plastic deformation while the kinematic hardening rule prescribes that the yielding surface only changes in its position with plastic deformation. Plastic behaviors of most metallic materials do not exactly follow either of these two basic strain hardening rules. Instead, their plastic behaviors fall in between the behaviors predicted by the two basic strain hardening rules. Hence the combined strain hardening rule is preferred for metallic materials, as illustrated in Fig. 2.

The combined strain hardening rule of nonlinear kinematic hardening (Chaboche model [ 13, 14]) and nonlinear isotropic hardening (model based on Voce hardening law [ 23]) has been verified for its accuracy in predicting the plastic behavior of stainless steel under large cyclic loading [ 8, 15, 16]. Therefore, the combined hardening rule is adopted for this numerical study to characterize the cyclic load behaviors of AISI 316L stainless steel. The governing equations for the combined hardening model of AISI 316L stainless steel are given below as Eq. (2). The combined hardening model for AISI 316L stainless is calibrated with experimental data and the model parameter values are given in Table 3.

(2)

where \chi is the back stress computed from two superimposed nonlinear kinematic hardening models (Chaboche model), i.e., \chi ={\chi }_{\mathrm{1}}+{\chi }_{\mathrm{2}} , {\chi }_{\mathrm{1}}={\displaystyle \frac{C_{\mathrm{1}}}{{\gamma }_{\mathrm{1}}}}\left(\mathrm{1}-e^{-{\gamma }_{\mathrm{1}}{ϵ}_{\mathrm{p}}}\right) , {\chi }_{\mathrm{2}}={\displaystyle \frac{C_{\mathrm{2}}}{{\gamma }_{\mathrm{2}}}}\left(\mathrm{1}-e^{-{\gamma }_{\mathrm{2}}{ϵ}_{\mathrm{p}}}\right) , {ϵ}_{\mathrm{p}} denotes plastic strain. C₁, γ₁, C₂, and γ₂ are material constants calibrated from the cyclic loading tests conducted by Kaufmann et al. [ 7] for A992 steel and from the cyclic loading tests conducted by Shit et al. [ 8] for AISI 316L stainless steel. R represents the yielding surface expansion due to isotropic hardening effect based on the Voce hardening law [ 23], R=R_{\mathrm{0}}\cdot {{ϵ}^{\prime }}_{\mathrm{p}}+R_{\infty }\cdot \left(\mathrm{1}-e^{-b\cdot {{ϵ}^{\prime }}_{\mathrm{p}}}\right) , {{ϵ}^{\prime }}_{\mathrm{p}} denotes the accumulated plastic strain; R₀, R₁, and b are material constants that can be calibrated using cyclic test results. k is the material’s initial yield strength, which is taken as 389.3 MPa for A992 steel from Kaufmann et al. [ 7] and taken as 190 MPa for AISI 316L stainless steel from Rasmussen [ 9]. sgn(x) is the signum function. f=\left|\sigma -\chi \right|-R-k ; f is used to distinguish the elastic domain (f<0) and the inelastic domain (f≥0) [ 12].

FE model details for link beams

The finite element program ANSYS (ANSYS, Inc. 14.5 [ 17]) was employed here to perform FE analysis of the shear links. The FE models are intended to predict the seismic behavior of the links such as strength, ductility, and strength degradation due to local buckling in flange and web. The strength degradation resulting from material fracture is not considered in this study.

Shell elements with six degrees of freedom per node (Shell element 181 in ANSYS) were used in the models. This type of element has four nodes with six degrees of freedom (DOFs) at each node, 3 translational DOF and 3 rotational DOFs. The geometry dimensions of the finite element model are determined based on the centerlines of the shear link cross section. The geometry of shell elements corresponded to the centerline dimensions of the links. Stiffener welding was not modeled explicitly.

Mesh refinement studies were conducted to determine the level of refinement necessary to achieve reasonable accuracy for seismic behavior study of shear links. Figure3 illustrates the meshed finite element model of the four link beam specimens with different section size. The mesh size can be estimated from the figures since the section size is known. Large displacement option was taken into account in FE analysis to capture local buckling. All nodes at the right end section of the shear link model were restrained against translational movement and rotations.

To simulate the loading condition of shear links, a short length of rigid segment (100 mm long) was added to the model on the left side to model loading fixture attached to the shear link specimen. This rigid section is supposed to restrain the left end section against rotations, while the translational (lateral and longitudinal) movement of the shear link is unconstrained. To make the rigid segment stiff enough to minimize its own deformation, the thickness of the flange and web plates were 20 and 40 times those of the shear link’s flange and web plates respectively. The left side end section nodes of the rigid segment’s web were restrained against rotations. In addition, the left end nodes of the links were restrained against out-of-plane movement. In this way the fixture is simulated as a rigid test fixture which is only allowed to move in the transverse loading direction (along y-axis in the finite element model in Fig. 3).

According to Richards and Uang [ 18], monotonic loading underestimate buckling load amplitude and strength degradation. To determine ductility ratio, performing cyclic load analysis is essential. In other word, cyclic loading was necessary to consider local buckling and associated strength degradation accurately. A cyclic load testing provides a more realistic loading to structures under earthquake loading. Iwasaki et al. [ 19] concluded that loading rate has insignificant effect on hysteresis loops and for large displacement the energy dissipation capability is smaller when loading rate is lower.

Load on the shear link specimen was applied by controlling transverse displacements at the left end section nodes of the shear link model. Link rotation γ is defined as the ratio of the transverse displacement between the two ends of the link beam and its initial length according to AISC Seismic Provisions [ 6]. To conduct quasi-static load testing, cyclic loading protocol adopted by AISC Seismic Provisions [ 6] was used in finite element analysis. The loading protocol in the finite element analysis includes 6 cycles of link rotation γ at an amplitude of 0.00375 radians, 6 cycles of link rotation γ at an amplitude of 0.005 radians, 6 cycles of link rotation γ at an amplitude of 0.0075 radians, 6 cycles of link rotation γ at an amplitude of 0.01 radians, 4 cycles of link rotation γ at an amplitude of 0.015 radians, 4 cycles of link rotation γ at an amplitude of 0.02 radians, 2 cycles of link rotation γ at an amplitude of 0.03 radians, 1 cycle of link rotation γ at an amplitude of 0.04 radians, 1 cycle of link rotation γ at an amplitude of 0.05 radians, and the following each cycle of link rotation γ increased in the amplitude by 0.02 radians. The load history is shown in Fig. 4.

Nonlinear FE analysis of link beams under cyclic loading

Cyclic loading analysis was conducted on eight link beam specimens, four specimens each for two link beam materials — AISI 316L stainless steel and A992 steel, in order to evaluate the hysteresis behavior of 316 stainless steel link beams. Failure of link beams is generally due to the occurrence of buckling or fracture. Okazaki and Engelhardt [ 20] conducted cyclic loading tests on W10 × 33 link beam with the length ratio of 1.04, in which web fracture occurred at a link rotation angle of 0.12 radians. Corte et al. [ 21] conducted finite element simulation on three shear link specimens with section size of W10 × 33 (length ratio 1.04), W18 × 40 (length ratio 1.02), and W10 × 68 (length ratio 1.25) respectively and observed that buckling occurred in the link beams when link rotation angle reached 0.09 radians.

Ductility

During finite element analysis, nodal stresses and strains over web area of the shear link, as well as the nodal reaction forces and displacements at both ends of the link beam specimens were recorded. Based on the finite element analysis results, the hysteresis loops of the four shear link specimens are plotted in Figs. 5–8. The backbone curves of the corresponding hysteresis loops of the shear link specimens are shown in Fig. 9 to evaluate the ductility and overstrength ratio. Ultimate link rotation γ_u is taken as the point when local buckling occurrence is imminent in the shear link model. This imminence point of local buckling can be identified by the excessive local plastic strains in the link beam model’s strain contour plot. The shear force value at this point is defined as the ultimate strength of the shear link. The over-strength factor for the selected link beam in this research is calculated as the ratio of its ultimate strength value to its yield strength value. Shown in Figs. 5 to 9, shear links made of AISI 316L stainless steel have lower yield strength than the corresponding A992 steel shear links because the yield stress of AISI 316L is lower. The ultimate strength values of the stainless steel shear links are much larger, which reflects that the isotropic strain hardening effect is more significant in AISI 316L stainless steel than A992 steel. The yield strength, ultimate strength, and the over-strength factor of the four link beams are given in Table 4.

Ductility is the ability of the structure to sustain large permanent deformation without significant reduction in strength. Earthquake energy absorbed through inelastic deformation is one of the important seismic resistant structures characteristics. Ductility of properly designed EBF is directly related to the ductility of the shear links [ 3]. Shear links (short links) dissipate energy primarily through shear distortion which provides more ductility than longer links (flexural links). Well detailed shear links exhibited stable and ductile cyclic behavior without brittle failure before reaching a plastic rotation of 0.1 rad. Ductility ratio provides a metric so that quantifies the maximum inelastic deformation capacity of seismic resistant structures or components. Ductility factor is defined by Park [ 22] as the ratio of post-peak displacement at 20% reduction in load carrying capacity ( {\mathrm{\Delta }}_{\mathrm{u}} ) to displacement at yield ( {\mathrm{\Delta }}_{\mathrm{y}} ). When local buckling occurs, load carrying capacity of link beams may not drop substantially. The ductility ratio of the four AISI 316L stainless steel link beams is calculated as 13.4 as they all failed in flange local buckling at link rotation value of 0.13. Once the local buckling occurred in flange at link rotation value of 0.13, the link beam is considered to fail and numerical simulation is terminated. Web local buckling was not observed at this link rotation value.

Based on the finite element analysis results shown in Figs. 5–9, it is seen that shear links made of 316L stainless steel have lower yield strength than the A992 steel shear links with same section size and length ratio. The ultimate strength values of the stainless steel shear links are generally greater than that of the corresponding A992 steel shear links. The shear links made of stainless steel are capable of undergoing large inelastic link rotation up to 0.13 radians before local instability occurs in flange. The differences between the ultimate deformations of the two types of shear link specimens are due to their different failure phenomena. Based on the finite element simulation results, for 316L stainless steel shear links with a length ratio of 1.3, local buckling occurs at the flange or the web area when link rotation γ reaches approximately 0.13 radians, while the plastic shear of link web is still far below the plastic deformation capacity of 316L stainless steel. This reveals that for stainless steel link beam with a length equal to 1.3(M_p/V_p), mixed shear-flexural behavior is observed instead of the shear yielding behavior in A992 steel shear link. This can be explained by the significant strain hardening effect and associated large overstrength ratio of 316L stainless steel, as shown in Fig. 1. Therefore, for stainless steel link beams, different limit values to classify short or long links need to be defined.

Stiffness

Stiffness is a key parameter in characterizing the behaviors of the shear links. Elastic stiffness of the link beam can be calculated analytical using Eq. (3) below:

(3) K_{\mathrm{e}}={\displaystyle \frac{K_{\mathrm{b}}{K}_{\mathrm{s}}}{K_{\mathrm{b}}+K_{\mathrm{s}}}},

where K_e is the equivalent elastic stiffness, K_b = 12EI/L³ is the bending stiffness; E is the elastic modulus; I is the second moment of area. K_s = GA_lw/L is the shear stiffness; G is the shear modulus; A_lw is the web area. The calculated K_e and elastic stiffness obtained from FE analysis was compared and verified to be of very similar value. The stiffness of link beam specimens before yielding and after yielding can be identified from the hysteresis loops and corresponding backbone curves plotted in Figs. 5–9. It is seen that the shear link specimens with both material types have similar initial stiffness values in the elastic range while in the plastic range, the hysteresis loops of the stainless steel shear links have greater post-yield stiffness values due to its significant stress hardening effect.

Plastic strain distribution

Plastic shear strain distributions of link beams at link rotation γ of 0.015 rad. (slightly after yield point) are shown in Figs. 10–13, for two types of materials. The plastic shear stain distributions of link beams at link rotation γ of 0.13 rad. (local flange buckling is imminent) are also shown in Fig. 14. However, for A992 steel shear link, the plastic shear stain distributions at only one deformation level —link rotation γ equal to 0.015 rad., are shown in Figs. 10–13. This is because it is difficult to determine the fracture point of A992 steel shear link from finite element analysis, which typically controlled the ultimate strength of shear link with a length ratio of 1.3. As the loading direction is in the plane of the link web, and buckling only occurs toward the end of the loading process, the shear strain {ϵ}_{yz} in the web panel (y-z plane in the finite element model) has the highest value among the 6 strain components during the entire loading process. Hence the shear strain {ϵ}_{yz} in the web plane is selected as the indication for the strain distribution pattern of the shear link over the web area. It is observed from Figs. 10–13 that stainless steel link beams have plastic shear strain over a larger web area than corresponding A992 steel shear links. This indicates that more energy can be dissipated by stainless steel link beams in the form of hysteretic energy or plastic deformation at link rotation angle γ of 0.015. It is also seen that the values of the plastic shear strain are higher in the mid-length portion of the link web than the areas near the end sections, probably due to the constraints provided by the rigid end loading fixtures.

The high plastic strain area will expand with increasing load. At link rotation γ of 0.13 rad. (when the flange local buckling is about to occur), very high plastic shear strains are observed in the web areas close to the flanges and end sections of all link beam specimens, as shown in Fig. 14. These areas are found to be the locations where buckling occurs shortly and also the locations where high compressive stress due to flexural action. When buckling occurs, the buckled area deforms significantly, resulting in the excessive plastic shear strains. Although local flange buckling did not cause sudden drop in strength in the hysteresis loops shown in Figs. 5–8, the finite element simulation is terminated at the buckling point. Therefore, the substantial strain hardening of 316L stainless steel is attributed to cause this mixed flexural-shear failure mode even though its length ratio is 1.3.

Conclusions

In this study, cyclic load behaviors of 316L stainless steel shear links are investigated by nonlinear finite element simulation. For comparison purpose, A992 steel link beam with the same section sizes and length ratio are also examined. Combined strain hardening rule is used for simulating the AISI 316L stainless steel’s cyclic load material property. Cyclic loading analysis was conducted on eight link beam specimens, four specimens each for two link beam materials — AISI 316L stainless steel and A992 steel, in order to evaluate the hysteresis behavior of AISI 316 stainless steel link beams. Link beams made of AISI 316L stainless steel is found to be capable of developing high ductile performance up to link rotation of 0.13, when flange local buckling was observed and numerical analysis was terminated.

Based on the finite element analysis results, shear links made of 316L stainless steel are found to have lower yield strength than the A992 steel shear links with same section size and length ratio. The ultimate strength values of the stainless steel shear links are generally greater than that of the corresponding A992 steel shear links. The link beams made of stainless steel are capable of undergoing large inelastic link rotation up to 0.13 radians before local instability occurs in flange. Based on the finite element simulation results, for 316L stainless steel shear links with a length ratio of 1.3, local buckling occurs at the flange when link rotation γ reaches approximately 0.13 radians, while the plastic shear of link web is still far below the plastic deformation capacity of 316L stainless steel. This reveals that for stainless steel link beam with a length equal to 1.3(M_p/V_p), mixed shear-flexural behavior is observed instead of the shear yielding behavior in A992 steel shear link. This can be explained by the significant strain hardening effect and associated large overstrength ratio of 316L stainless steel. Therefore, different classification criteria need be developed for stainless steel link beams to differentiate between flexural link, intermediate link, and shear link. It is also observed that the stainless steel link beams have plastic shear strain over a larger web area than corresponding A992 steel shear links at link rotation value of 0.015. This is due to the lower yield stress value of AISI 316L stainless steel compared to A992 steel.

Clearly, AISI 316L stainless steel shear link provides a very promising seismic fuse devices, especially for applications in corrosion sensitive environments such as coastal structures or bridges on coastlines. However, due to of lack of experimental test results of stainless steel link beams, other effects such as residual stress, fracture due to manufacturing defect (e.g., welding) and complicated stress condition at the interconnection of stiffeners and web plates, are not considered in the conclusions drawn from this study.

Nomenclature

References

[1] Uang C M, Nakashima M, Tsai K C. Research and application of buckling-restrained braced frames. International Journal of Steel Structures, 2004, 4: 301–313

[2] Tsai C, Tsai K. TPEA device as seismic damper for high-rise buildings. Journal of Engineering Mechanics, 1995, 121(10): 1075–1081

[3] Kasai K, Popov E. General behavior of WF steel shear link beams. Journal of Structural Engineering, 1986, 112(2): 362–382

[4] McDaniel C C, Uang C M, Seible F. Cyclic testing of built-up steel shear links for the New Bay Bridge. Journal of Structural Engineering, 2003, 129(6): 801–809

[5] Baddoo N. AISC Design Guide 27: Structural Stainless Steel. American Institute of Steel Construction (AISC), Chicago, Illinois, USA, 2013

[6] AISC. Seismic Provisions for Structural Steel Buildings. ANSI/AISC 341–05. Chicago, IL, 2010

[7] Kaufmann E J, Metrovich B R, Pense A W. Characterization of cyclic inelastic strain behavior on properties of A572 Gr. 50 and A913 Gr. 50 rolled sections. ATLSS report, No. 01–13. Bethlehem, PA: Lehigh University, 2001

[8] Shit J, Dhar S, Acharyya S. Characterization of cyclic plastic behavior of SS 316 stainless steel. International Journal of Engineering Science and Innovative Technology (IJESIT), 2013, 2

[9] Rasmussen K J R. Full-range stress–strain curves for stainless steel alloys. Journal of Constructional Steel Research, 2003, 59(1): 47–61

[10] Gardner L, Nethercot D A. Numerical modeling of stainless steel structural components—a consistent approach. Journal of Structural Engineering, 2004, 130(10): 1586–1601

[11] Ashraf M, Gardner L, Nethercot D A. Finite element modeling of structural stainless steel cross-sections. Thin-walled Structures, 2006, 44(10): 1048–1062

[12] Chaboche J L. A review of some plasticity and viscoplasticity constitutive theories. International Journal of Plasticity, 2008, 24(10): 1642–1693

[13] Chaboche J L, Rousselier G. On the plastic and viscoplastic constitutive equations—Part I: rules developed with internal variable concept. J. Pressure Vessel Technol, 1983, 105(2): 153–158

[14] Chaboche J L, Rousselier G. On the plastic and viscoplastic constitutive equations—Part II: application of internal variable concepts to the 316 stainless steel. J Pressure Vessel Technol, 1983b, 105(2): 159–164

[15] Okazaki T, Engelhardt M D. Cyclic loading behavior of EBF links constructed of ASTM A992 steel. Journal of Construction Steel Research. 2007, 63: 751–765

[16] Olsson A. Stainless steel plasticity-material modeling and structural applications. Dissertation for the Doctoral Degree. Lulea: Lulea University of Technology, 2001

[17] Hyde C J, Sun W, Leen S B. Cyclic thermo-mechanical material modeling and testing of 316 stainless steel. International Journal of Pressure Vessels and Piping, 2010, 87(6): 365–372

[18] ANSYS. ANSYS Mechanical APDL, version 14.5, ANSYS Academic Teaching Introductory. ANSYS, Inc, Canonsburg, Pennsylvania, USA, 2014

[19] Richards P W, Uang C M. Development of testing protocol for short links in eccentrically braced frames. Report No. SSRP–2003/08, Department of Structural Engineering, University of California at San Diego, La Jolla, Calif, 2003

[20] Iwasaki T, Kawashima K, Hasegawa K, Koyama T, Yoshida T. (1987). “Effect of Number of Loading Cycles and Loading Velocity on Reinforced Concrete Bridge Piers”. 19th Joint Meeting US-Japan Panel on Wind and Seismic Effects, UJNR, Tsukuba

[21] ANSYS. ANSYS Mechanical APDL Material Reference. ANSYS, Inc, Canonsburg, Pennsylvania, USA, 2013

[22] Della Corte G, D’Aniello M, Landolfo R. Analytical and numerical study of plastic overstrength of shear links. Journal of Constructional Steel Research, 2013, 82: 19–32

[23] Park R. Evaluation of ductility of structures and structural assemblages from laboratory testing. Bulletin of the New Zealand National Society for Earthquake Engineering, 1989, 22(3): 155–166

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg