1

Recent Topics of Japanese Design Codes for Steel and Composite Bridges

Yoshiaki OKUI Saitama University, [email protected] Abstract First an overview of Japanese design codes for steel and composite bridges are given. Then, some topics discussed in Standard Specifications for Steel and Composite Structures published by JSCE are introduced. Keywords: Design codes, Bridges 1. Introduction There are three major design codes for steel and composite bridges in Japan: namely 1) Specifications for Highway Bridges [SHB] (JRA, 2002), 2) Design Standards for Railway Structures and Commentary [DSRS] (RTRI, 2008), 3) Standard Specifications for Steel and Composite Structures [SSSCS] (JSCE, 2007). The first two design codes are used in design practice, and published by the associated governing legal authorities. On the other hand, the last code SSSCS is a model code published by an academic society. It suggests a future direction of the design codes in Japan rather than current design practice. The current SHB employs the performance-based allowable stress design (ASD) method except for seismic design, which has been based on the performance-based limit state design (LSD) method. The SHB is scheduled to revise completely into the performance-based LSD method in 2011. The DSRS, which is the official design code for railway bridges, has been revised from the ASD to LSD method since 1992. In the performance-based design method, first, required performance levels for individual bridges functions must be clarified. Then, it is required to achieve the designated performance level. The methods to check the performance level is not specified in the main body of the performance-based design codes. Instead several appropriate procedures are shown in commentary. Hence, designers can select an appropriate design procedure from several design procedures that meet their purposes. This framework of the perforatedbased design method is considered to offer greater flexibility to designers, and expected to promote new technologies, such as new materials and innovative structures. The SSSCS is based on the performance-based LSD method. The SSSCS consists of the following 6 parts: Part I General Provision, Part II Structural Planning, Part III Design, Part IV Seismic Design, Part V Construction, Part VI Maintenance and Management. The first five parts have been published already, and the Part VI will be issued in 2010. The English version of Part I to III will be open within this year. In this paper, background research work for design criteria adopted in SSSCS will be introduced, and a comparison the proposed design criteria with other design code, such as AASHTO and Euorcode will be given. In the next chapter, criteria for section classification of composite girders will be discussed. The criteria are used in bending moment capacity. The shear strength of composite sections, and interaction between the bending moment capacity and the shear capacity will be discussed in Chapter 3. 2. Section classification for composite girder In SSSCS, in order to evaluate bending moment capacity of composite section, first, sections are classified into several classes depending upon their ultimate bending moments and flexural deformation capacity. In AASHTO (AASHTO, 2008) and Eurocode (CEN, 2004, 2005), the classification requirements for composite sections are originally derived from experimental and analytical studies for steel sections without concrete slabs. In fact, the same requirements as those for steel sections are used for composite sections in these specifications.

2 For composite sections under positive bending, since concrete slabs suppress buckling of the compression flange and the compressive part of the web plate, there is a possibility of relaxing the section classification requirements compared to those for steel sections only. Another issue considered is the effect of initial bending moment on section classification. In unshored construction method for composite girder bridges, steel girders have to support the initial bending moment due to self-weight of steel girders and wet concrete. After hardening concrete, steel-concrete composite sections support the superimposed dead loads and live loads. However, this initial bending moment effect on section classification is not considered explicitly in conventional design codes. In order to clarify these issues, a parametric study with the three dimensional finite element analyses of composite I girder sections is carried out. The web slenderness limit for section classification is investigated based on the parametric study (Gupta, 2006, Gupta et al., 2006). 2.1. FE Parametric Study A three-dimensional finite element model was used to analyze the composite girders under positive bending. The steel I girder and the concrete deck are modeled with 4-node thin shell elements and with 8-node solid elements, respectively. In order to consider the initial bending moment in the unshored construction method, two phases analysis is carried out. In the first phase, the initial bending moment M1 is applied to a steel section only. In the second phase after installing a concrete deck, additional bending moment M2 is loaded to a composite section (Fig.1). The stress and strain states at the last loading step in the first phase are taken over as the initial conditions in the second phase analysis. uniaxial stress-strain relationship used in the FE analysis for steel and concrete, respectively. Since a simplified stress-strain relationship without descending regime used in the analysis as shown in Fig. 3 the load step in which concrete strain attains 0.0035 is regarded as the ultimate state. In the FE model, the initial out-of-plane displacement of the web plate is considered as an imperfection. The distribution profile of the initial outof-plane displacement is assumed to be sinusoidal with a maximum displacement of bw 250 , where bw is the web depth; see for more details (Gupta et al., 2006).

Figure 2. Stress-strain relationship for steel (JIS-SM490Y, yield strength: 355MPa)

QuickTime LvO sN`Kv- B

Figure 3. Stress-strain relationship for concrete (compressive strength: 40MPa) As an example, Fig.4 shows the bending moment and curvature relationships obtained form the FE analysis. In this figure, the first less steep part in the momentcurvature curve corresponds to the phase 1 analysis for only steel section. The obtained buckling mode after a peak bending moment is shown in Fig.5. In the parametric study, more than 300 sections are analyzed by changing dimensions of sections and the magnitude of the initial bending moment. The thickness of concrete deck and depth of web plate are fixed 300mm and 3000mm, respectively. The width to

Figure 1. Phase 2: composite section

In the analysis material and geometric nonlinearity is considered. For the constitutive model of metal plasticity, Prandtl-Reuss equation is employed. On the other hand, for concrete, the Mohr-Coulomb yield criterion with the associated flow rule is used. Figs. 2 and 3 show the

3 thickness ratio of web plate ranges from120 to 231. The initial bending moment is assigned to 0, 20, 30, 40% of the yield bending moment of steel section only. 2.2. Web slenderness limit for compact section In Fig.6, the width-thickness ratios bw t of web plate are plotted against parameter obtained from the numerical analysis. The parameter stands for the position of plastic neutral axis and defined in Fig.7. The numerical results obtained from the parametric study are expressed by square symbols for compact sections and open circle symbols for noncompact sections. The plotted results are compared with AAHSTO's compact and Eurocode's class 1 and class 2 limits. boundary.

250
w w

Width-to-thickness ratio b /t

AASHTO Compact 200 Proposed Compact limit

150 EC Class 1 EC Class 2 50

Compact sections for (FE Analysis) Noncompact sections for ' (FE Analysis)

100

0.1

0.2

0.3

0.4

0.5

or '

Figure 6. Web slenderness limit between compact and slender sections (M1=0)

Figure 4. Moment-curvature relationship obtained from two phases FE analysis

Figure 7. Stress distribution of compact section at full plastic moment Mpl, f cd =concrete compressive strength.

On the basis of the numerical results, the following web slenderness limit is proposed as compact section requirement: bw 59 (1) tw where = 235 / f y , and f y =yield strength of steel [MPa]. The proposed web slenderness limit is depicted as the dash curve in Fig.6. Figure 5. Web buckling in slender section 2.3. Web slenderness limit for noncompact section Fig.8 shows the web slenderness limit for noncompactslender boundary. The vertical axis is the width-thickness ratio of web, while the horizontal axis stands for the position of the neutral axis as shown in Fig.9. For noncompact sections, some portions of a steel girder section yields and the other is still elastic at the ultimate state. Thus, the position of the neutral axis is calculated from strain distribution of a steel section obtained from the numerical analysis.

Comparing AASHTO with Eurocode, since compact sections in AASHTO corresponds to class 2 in Eurocode, AASHTO allows thinner web plates than Eurocode as compact sections. Our numerical results support AASHTO's limit, and the web slenderness limit in Eurocode's class 2 seems to be conservative. It is confirmed that the initial bending moment has no effect on the web slenderness limit for compact-noncompact

Figure10. Definition of and ; M1=initial bending moment, M2=bending moment due to superimposed dead load and live load

(a) M1=0

In Fig.8, noncompact and slender sections are denoted by open circle and solid triangle symbols, respectively. It is apparent that the web slenderness limit moves upward with increasing initial bending moment. It should be noted that in the case M1=0, AASHTO gives unsafe judgment of noncompact sections. The numerical results show that noncompact sections based on AASHTO design code may buckle elastically, especially under condition 0.5. In the following, first we will set up a new web slenderness limit for noncompact sections in the case of M1=0, and then will take the initial bending moment effect into account. For M1=0, we set a new web slenderness limit by relaxing Eurocodes web slenderness limit for class 3:
bw 52 t w 0.67 + 0.33 bw 77 (1 ) tw for > 1.0

(2)
for < 1.0

(b) M1=0.3Mys

where the definition of is given in Fig.10. Next, in order to include the initial bending moment effect, a new coefficient is introduced into Eq.(1):
52 bw for > 1.0 t w 0.67 + 0.33 (3) bw 77 (1 ) for < 1.0 tw Difficulty in identification of is that once initial bending moment is introduced then the stress gradient also changes. To decompose the initial bending moment effect and the stress gradient effect, we plot the ratio of the right hand side to the left hand side of Eq.(2) against the initial bending moment. If Eq.(2) accounts for the stress gradient effect on the web slenderness properly, this ratio should be independent of the stress gradient and be a function of merely the initial bending moment. To examine this assumption, all numerical results for noncompact and slender sections are plotted into the ratio versus the initial bending moment diagram as shown in Fig.11. In this figure, the initial bending moment M1 is normalized with respect to the yield

(d) M1=0.4Mys Figure 9. Web slenderness limit between noncompact and slender sections; Effect of initial bending moment on web slenderness limit

5 bending moment for only steel section Mys. The lower limit of slender sections is practically important. For both cases > 1.0 and 1.0 , the lower limits give the almost same boundary, which is expressed by M M 2 1 1 = 1 0.1 (4) M + 2.3 M ys ys Consequently, we propose Eq.(3) with Eq.(4) as the web slenderness limit for noncompact sections. The proposed web slenderness limit for M1=0 is 25% larger than Eurocode's class 3 limits. Furthermore, the proposed web slenderness limit increases up to 33% with increasing initial bending moment up to 40% of the yield bending moment of steel section only.

3. Shear capacity and interaction between bending and shear Loading tests of 3 composite girder specimens with different span lengths were carried out to investigate the shear capacity and interaction between bending moment and shear capacities. Fig. 12 shows one of the girder specimens with a span length of 5.6m. The other two specimens have the same cross section but 7.0m and 9.2m span lengths in order to change the ratio between the bending moment and shear force. These specimens were loaded at the center of the specimens

Figure 12. Composite girder experimental specimen with a span length of 5.6m

Table 1 Summary of experimental results and comparison with design capacities

Experiment results at max. load Design capacity Full plastic moment Mpl 3061 Shear capacity by Basler Qb 1149

(a) > 1.0

Bending moment Mu (kNm) Spec. 1 (5.6m) Spec. 2 (7.0m) Spec. 3 (9.0m) 3259 709 3264 933 3298 1178 Shear force Qu (kN)

3033

1149

3127

1171

(b) 1.0 Figure 11. Effect of initial bending moment

Table 1 shows summary of experimental results, in which both the ultimate bending moment and shear force are calculated from the maximum load based on the beam theory. As the bending moment capacity, the full plastic moment is calculated using material test results for concrete and steel plate. Baslers equation is employed to calculate the shear capacity neglecting shear capacity contribution from the concrete slab. Spec. 1 is the shortest specimen with 5.6m span length. In this

6 specimen, shear failure of steel web plate governs the ultimate failure mode, and accordingly the Baslars shear capacity is very close to the maximum shear force in the experiment. On the other hand, for the longest specimen (Spec. 3), the bending moment governs the failure mode and the maximum bending moment closes to the full plastic moment. Figure 13 shows the experimental results in (M u / M pl ) vs. (Qu / Qb ) plot along with 4th order interaction equation: M 4 Q 4 u u (5) M + 1.0 pl Qb which is employed in SSSCS; M pl and Qb are the full plastic bending moment and the shear capacity by Baslar, respectively. Judging from Fig.13, there is no interaction effect of the shear force on the bending moment capacity. The interaction effect between the bending moment and shear capacities seems to be negligible in design practice. sections, which are independent of initial bending moment, seem to be inappropriate for composite girders. The initial bending moment, which is firstly applied to steel section only in unshored construction, has considerable effect on the noncompact web slenderness limits. The web slenderness limits for compact and noncompact sections are proposed on the basis of the parametric study. The interaction effect between the bending moment and shear capacities is investigated experimentally. The loading test results for 3 composite girders shows that the interaction effect between the bending moment and shear capacities is negligible in design practice.

Figure 13. Interaction between bending moment and shear capacities

4. Summary The positive bending moment capacity of composite steel girders is examined through parametric study employing elasto-plastic finite displacement analyses. The effects of initial bending moment in unshored construction on the bending moment capacity and on the web slenderness limit for section classification are investigated. Observations made during the numerical study indicate that the noncompact web slenderness limits in conventional design standards, which are based on tests of steel I-sections, are conservative for composite sections. Many sections, which are classified as slender by current specifications, demonstrate sufficient flexural capacity as noncompact. The conventional web slenderness limits for noncompact

References American Association of State Highway and Transportation Officials (AASHTO), LRFD bridge design specifications, Washington, D.C. 2008 CEN, EN 1993-1-1:2005 Eurocode 3: Design of steel structures - Part 1.1: General rules and rules for buildings, 2005. CEN, EN 1994-1-1:2004 Eurocode 4: Design of composite steel and concrete structures - Part 1.1: General rules and rules for buildings, 2004. Gupta, V. K., Development of section classification criterion and ultimate flexural equation for composite I-girders, Doctoral Dissertation, Saitama University, Japan, 2006. Gupta, V. K., Okui, Y., Nagai, M., Development of web slenderness limits for composite I-girders accounting for initial bending moment, Doboku Gakkai Ronbunshuu A, Vol. 62, No. 4, pp.854-864, 2006. Japan Road Association (JRA) [Part II, Steel bridges] (2002) Specifications for Highway Bridges (in Japanese). Japan Society of Civil Engineers (JSCE) (2007) Standard Specifications for Steel and Composite Structures [I General Provision, II Structural Planning, III Design] (in Japanese). Railway Technical Research Institute (RTRI) (2008) Design Standards For Railway Structures and Commentary [Steel-Concrete Hybrid Structures] (English version).