Академический Документы
Профессиональный Документы
Культура Документы
a r t i c l e i n f o a b s t r a c t
Article history: In this work we present an optimization procedure for achieving best beam-to-column reduced beam section
Received 5 January 2017 connections in steel moment frames under static loading. The whole scheme is based on selecting European I-
Received in revised form 5 September 2017 profiles for beams and columns, using the component method of Eurocode 3 for evaluating the design moment
Accepted 20 September 2017
resistance of the joint, regression analysis to correlate the geometrical and inertial characteristics of the various
components and elementary principles of Mechanics. In doing this, were are led to a constraint optimization
Keywords:
Steel moment frames
problem, which is formed according to the design restrictions of European and North American Standards for
Reduced beam section RBS connections, regarding expected seismic behavior. Adopting advanced numerical procedures, optimum re-
Connections sults are obtained, regarding either combinations of connection components or existing steel frames, a fact vali-
Optimization dating the flexibility and applicability of the proposed method. The calibration, verification and validation of the
European I-profiles excellent seismic performance of the results obtained will be demonstrated in a companion paper, via FE analyses
Monotone (static) loading under cyclic loading. Hopefully, this work will contribute in adopting RBS connections in the design of steel mo-
ment frames in European engineering practice, as well as in the relevant codes.
© 2017 Elsevier Ltd. All rights reserved.
1. Introduction where db is the depth of the beam, while the radius of cut is equal
to
Τhe reduced-beam-section (RBS) concept in steel moment resisting
frames is based on the selective removal of beam material adjacent to 2
4c2 þ b
the joint, either from the web or from the flanges of the beam. This re- R¼ ð3Þ
8c
duction in cross-sectional area will lead to the consecutive reduction
of the moment bearing capacity at a discrete location of the beam,
where yielding will be concentrated, protecting in this manner the con- Extending a recent work by the authors [8], the goal of the present
nection from early fracture. After extensive theoretical and experimen- study is to optimize symmetric extended end-plate RBS connections
tal work, as overviewed by the authors [1], the radius-cut RBS with European I-profiles (for beams and columns) under monotone
configuration was prequalified in both USA and Canadian Standards loading, based on the above limitations and in conjunction with ele-
[2–6]. Contrary, it was only incorporated in EC8-Part 3 [7] – and not in mentary principles of Mechanics.
Eurocode 3 – for retrofitting, as a mere follow-up of the North More specifically, utilizing advanced regression analysis and
American Standards mentioned above, without any indication graphics software [9], in order to approximately produce functions
concerning the type of connection or the profiles used for the beams that relate the overall characteristics of the I-profiles used, the whole
and the columns. The geometry of such a RBS is depicted in Fig. 1, problem is lead to the formulation of an objective function, to be numer-
while the corresponding size limitations are given in expressions ically minimized. This function accounts for required ductility of the
(1) and (2) for AISC/CISC and EC8 respectively. connection, desirable plastic deformation and formation of the plastic
hinge within the RBS, without exceeding the actual static design mo-
0:5bfb ≤ α ≤ 0:75bfb ; 0:65db ≤ b ≤ 0:85db ; 0:10bfb ≤ c ≤ 0:25bfb ð1Þ ment resistance of the joint Mj,Rd, which is calculated via the component
method [10], with guidelines and programming techniques based on a
α ¼ 0:60bfb ; b ¼ 0:75db ; c ¼ 0:20bfb ð2Þ relevant book [11]. Shear yielding and local buckling phenomena are
also avoided. The optimized results obtained will be validated and ver-
⁎ Corresponding author.
ified via FEM simulation under cyclic loading in a companion paper,
E-mail addresses: dimsof@civ.uth.gr (D.S. Sophianopoulos), nagoritsa@civ.uth.gr aiming in the whole to provide a tool for accepting the RBS as a reliable
(A.E. Deri). seismic connection for European Steel Design Practice.
https://doi.org/10.1016/j.jcsr.2017.09.028
0143-974X/© 2017 Elsevier Ltd. All rights reserved.
102 D.S. Sophianopoulos, A.E. Deri / Journal of Constructional Steel Research 139 (2017) 101–109
in which The second term in the right hand side of the above is directly corre-
lated to MRBS by writing
M RBS 2MRBS
V seismic
RBS ¼ ⇒V seismic
RBS ¼ ð8aÞ Mfc ¼ 1:1MRBS þ μM RBS ð12Þ
L−2s L−2s
2
Parameter μ is evaluated in order that the following two require-
and ments are met: (a) for Eq. (12) to be valid for a wide range of values
of L and w, and (b) no effect of shear force on the bending resistance ap-
pears at the position of the 1st plastic hinge. The 2nd requirement dic-
wL
Vw
RBS ¼ ð8bÞ tates that
2
V Ed
Combining Eqs. (5), (7) and (8a), (8b) we get ≤ 0:50 ð13Þ
V pl;Rd
2MRBS wL Since the acting shear is equal to VRBS, using Eq. (12) we can write
Mfc ¼ M RBS þ þ s ð9Þ
L−2s 2 that
In Eqs. (6a), (6b) fRBS is the maximum expected stress that will μΜRBS W pl;b −ck f RBS
V Ed ¼ ⇒V Ed ¼ μ ð14Þ
develop in the RBS. Under cyclic loading conditions and large in- s s
elastic deformations this stress – at the weakest RBS section – pffiffiffi
may reach the ultimate strength f u . In the sequel, we Taking into account that V pl;Rd ¼ Apv ffiffif3u and after setting γ ¼ μ 3
choose maxfrbs = fu, satisfying in this manner the overstrength re- inequality (13) is expressed as
quirements of clauses 6.5.2, 6.5.5 and 6.6.4 of Eurocode 8 – Part 1
[14]. γ W pl;b −ck W pl;b 0:5sAv
For an optimum connection as in Fig. 2, it is required that ≤ 0:50 ⇔ c ≥ − ð15Þ
sAv k γk
ΔΜ ¼ M j;Rd −Mfc ≥ 0 and ΔΜ minimum ð10Þ Since for every beam cross-section the quantities s, Wpl, b, Av and k
are known, the smaller accepted value of c depends only on parameter,
Obviously, one should avoid a situation that corresponds toΔΜ= 0, namely γ, to be determined. As it can be perceived from the limitations
i.e. simultaneous failure of joint and RBS. For this reason, a safety factor given in expressions (1)–(3), c must always be b 0.25bfb. If this restric-
equal to 1.1 is introduced in Eq. (9) yielding tion is combined with inequality (13), then one may establish the max-
imum value of γ, which will satisfy the 2nd requirement regarding the
shear force. The evaluation of this value is performed graphically, and
2M RBS wL its outcome is shown in Fig. 5 for IPE beam profiles and in Fig. 6 for
Mfc ¼ 1:1MRBS þ þ s ð11Þ
L−2s 2 HEA beam profiles.
104 D.S. Sophianopoulos, A.E. Deri / Journal of Constructional Steel Research 139 (2017) 101–109
Fig. 6. Graphical evaluation of the maximum value of γ for HEA beam profiles.
The maximum values found are γ = 0.64 and γ = 0.58 for IPE and
HEA beams respectively, which are equivalent to μ = 0.37 and μ =
0.335. Thus, based on Eq. (12), it is evident that
Fig. 4. Generalized forces in the beam of a typical RBS steel moment frame.
In the present work, and without loss of generality of the proposed
methodology, parameter σ is varied between σ = 2 (neglecting the
gravity forces on the beam) and σ = 4, a situation where the seismic
shear is equal to the shear of the gravity loads.
In the 2nd task, we calculate the values of L, as a function of s, for
which Eqs. (16a) or (16b) are valid, where appropriate. The results ob-
tained are given in Table 1.
Furthermore, as mentioned in the Introduction, RBS connections are
prequalified in the US Standards, and more specifically for Special Mo-
ment Frames (SMF) and Intermediate Moment Frames (IMF). For W-
beams, these Standards recommend that the ratio between the length
of the beam L and the height (depth) of the beam hb should be kept
not smaller than 7 for SMF and not smaller than 5 for IMF. These recom-
mendations may be also extended for the use of RBS with European pro-
files, with slight modifications. Keeping the lower limit of the
aforementioned ratio as benchmark, one may plot the curves (hb, Lmin)
Table 1
Results concerning the calculation of L as a function of s.
IPE beams
μ Lmin, σ = 2 Lmin, σ = 4
0.37 7.40541s 12.8108s
HEA beams
μ Lmin, σ = 2 Lmin, σ = 4
0.335 7.97015s 13.9403s
Fig. 5. Graphical evaluation of the maximum value of γ for IPE beam profiles.
D.S. Sophianopoulos, A.E. Deri / Journal of Constructional Steel Research 139 (2017) 101–109 105
Table 2
Connection data involved in the optimization procedure.
Connection Data
combination
Column Beam End plate Bolt diameter Bolt
thickness (mm) (mm) quality
Table 3
Results of the optimization procedure.
Table 4
Design dimensions and properties of the optimized connections.
x y c O = A or B R = O/Mpl,b⁎⁎ V RBS
(mm) (mm) (mm) (kNm)⁎ V pl;RBS
(%)
1 IPE220/HEB400 100 75 25 3.98 0.05934 42
2 IPE270/HEB450 95 90 33 2.3 0.02022 42
3 IPE220/HEA550 100 70 27 4.2 0.06262 39
4 IPE270/HEA500 85 95 33 2.398 0.02108 42
5 HEA220/HEB400 110 110 53 2.314 0.01732 33
6 HEA220/HEB500 100 90 47 10.182 0.07621 37
7 HEA220/HEA450 100 100 52 10.134 0.07585 34
8 HEA220/HEA360 100 90 48 4.68 0.03503 36
⁎ A = Mj,Rd − 1.470MRBS (IPE beams), B = Mj,Rd − 1.435MRBS (HEA beams).
Fig. 8. Limits of beam lengths with HEA profiles. ⁎⁎ M pl;b ¼ Wγpl;b f y .
Μ0
106 D.S. Sophianopoulos, A.E. Deri / Journal of Constructional Steel Research 139 (2017) 101–109
Local instability phenomena within the RBS are also unlikely to occur, For more details, the reader may refer to the link http://www.
since IPE and HEA beam sections under pure bending (excluding rare stalbyggnadsinstitutet.se/uploads/source/files/Artiklar/Tvarsnittsklasser
combinations of profiles and steel grade) are of Class I and II. %20fr%20valsade%20IPE-%20och%20HE-tvarsnitt.pdf.
Table 5
Results of the optimization scheme for the exemplary moment frame.
Optimum values
6.70/0.05891 11.58 28.17 M24/10.9 127.10 128.43 21.3
Design values
8.95/0.07869 12 30 M24/10.9 130 130 21
a
R defined in Table 4.
Fig. 11. Diagrams of generalized forces of the frame dealt with for (a) serviceability and (b) ultimate limit states.
108 D.S. Sophianopoulos, A.E. Deri / Journal of Constructional Steel Research 139 (2017) 101–109
Consequently, Eq. (11) can be written as The moment resistances of the final connection design are equal to:
2s wL M j;Rd ¼ 166:76 kNm; Mfc ¼ 157:81 kNm and M RBS ¼ 134:15 kNm;
Mfc ¼ ðA1 þ 1:1ÞM RBS þ A2 ; A1 ¼ ; A2 ¼ s ð21Þ
L−2s 2
while – as expected – VRBS/Vpl,b,Rd = 0.122 b 0.50.
In Eq. (21) the only free parameter is c, which for an IPE 270 beam Moreover, the values of parameters σ, μ and γ (defined in earlier sec-
should be less or equal to 33.75 mm. The objective function, according tions) for this example are found equal to 37.42, 1.21 and 2.096
to all the aforementioned analysis, takes the following form: respectively.
The proposed optimum design scheme can be extended to include
f ðp; cÞ ¼ M j;Rd ðpÞ−Mfc ðcÞ ð22Þ axial vs. moment interaction, by replacing the component method of
Eurocode 3 for the evaluation of Mj,Rd with the one proposed by Del
The rest of the free parameters involved in the calculation of Mj,Rd Savio et al. [15]. In their work, a generalized component-based method
are: was demonstrated, to account for M-N interaction in semi-rigid moment
• The thickness of the continuity plates ttp frame connections. As far as the RBS is concerned – and since fRBS = fu (as
• The thickness of the end-plate tep stated earlier) – this interaction will have no effect on the level of the MRBS
• The diameter of the bolts d, which will affect the diameters of the value, for monotone loading. For a cyclic (seismic one) Mpl,RBS will be re-
holes d0 placed by MN,RBS according to EC3 Part1.1 §6.2.9.1(4) [16], considering the
• The bolt quality (fyb, leading to fub or vice versa) weakest section as an individual I-beam section.
• The geometry of the bolt arrangement (x, y)
5. Conclusions
The most important conclusions drawn from this study are the
The constraints involved are those given in expression (4), with the
following:
addition of the ones given below:
t tp ≥ t fb ⇔ t tp ≥ 10:20 mm; t ep ≥ t fc ⇔ t ep ≥ 19 mm ð23Þ • Beam-to-column RBS connections using European I-profiles may be
optimized, using standard procedures of Mechanics, advance regres-
From the nature of the problem no regression (curve fitting) is re- sion analyses, numerical procedures, and existing North-American
quired, a fact reducing to a very large extent the nonlinear nature of and European recommendations.
the objective function, while the increase of the number of constraints • The proposed scheme is theoretically justified, combines all the re-
will work favorably in seeking an optimum solution. Its existence how- quirements of postulated seismic response, satisfies all related con-
ever premises that the flange panel of the beam possesses enough shear straints and, moreover, can be rather easily extended to apply on
resistance to withstand the loads imposed. every day practice design.
In the present example, where all combinations of M20, M24 bolts • The calibration, verification and validation of the optimum results ob-
and 8.8, 10.9 qualities are used, an optimum solution did result. This cor- tained herein will be demonstrated in a companion paper, aiming to
responds to the data given in Table 5, and schematically in the connec- include the RBS concept in European Practice and Codes for the design
tion shown in Fig. 12. of steel moment resisting frames.
Fig. 12. Optimum RBS connection geometry for the exemplary steel frame.
D.S. Sophianopoulos, A.E. Deri / Journal of Constructional Steel Research 139 (2017) 101–109 109
References [9] Origin Pro, Data Analysis and Graphing Software, OriginLab Corporation, Northampton,
MA, USA, 2016.
[1] D.S. Sophianopoulos, A.E. Deri, Parameters affecting response and design of steel [10] EN 1993-1-8, Eurocode 3: Design of Steel Structures – Part 1–8: Design of Joints,
moment frame reduced beam section connections: an overview, Int. J. Steel Struct. 2005 (Brussels, Belgium).
10 (2) (2011) 133–144, https://doi.org/10.1007/s13296-011-2003-5. [11] C. Faella, V. Piluso, G. Rizzano, Structural Steel Semirigid Connections: Theory,
[2] FEMA 350, Recommended Seismic Design Criteria for New Steel Moment Frames, Design, and Software, CRC Press, Boca Raton, Florida, USA, 2000.
2000 (Washington D.C., USA). [12] Mathematica V. 11. , Wolfram Research, Oxfordshire, UK, 2016.
[3] FEMA 355D, State of the Art Report on Connection Performance, 2000 (Washington [13] A.E. Deri, Parametric Analysis and Optimization of Steel Moment Frame Beam-to-
D.C., USA). Column RBS Connections (in Greek)(Ph.D. Dissertation) University of Thessaly,
[4] ANSI/AISC 341-10, Seismic Provisions for Structural Steel Buildings, 2010 (Chicago, Volos, Greece, 2013, http://www.didaktorika.gr/eadd/handle/10442/28151.
Illinois, USA). [14] EN 1998.01, Eurocode 8: Design of Structures for Earthquake Resistance – Part 1:
[5] ANSI/AISC 358-10, s1-11, s2-14, Prequalified Connections for Special and Intermedi- General Rules, Seismic Actions and Rules for Buildings, 2005 (Brussels, Belgium).
ate Steel Moment Frames for Seismic Applications, 2010 (Chicago, Illinois, USA). [15] A.A. Del Savio, D.A. Nethercot, P.C.G.S. Velassco, S.A.L. de Andrade, L.F. Martha, Gen-
[6] CISC/ICCA, Moment Connections for Seismic Applications, 2008 (Markham, Ontario, eralized component-based model for beam-to-column connections including axial
Canada). versus moment interaction, J. Constr. Steel Res. 65 (8–9) (2009) 1876–1895,
[7] EN 1993-8, Eurocode 8: Design of Structures for Earthquake Resistance – Part 3: https://doi.org/10.1016/j.jcsr.2009.02.011.
Assessment and Retrofitting of Buildings, 2005 (Brussels, Belgium). [16] EN 1993-1-1, Eurocode 3: Design of Steel Structures – Part 1–8: General Rules and
[8] A.E. Deri, D.S. Sophianopoulos, Parametric analysis and optimization of reduced Rules for Buildings, 2005 (Brussels, Belgium).
beam section steel frame connectionsProceedings of the 7th European Conference
on Steel and Composite Structures, Naples, Italy, September, 2014.