Professional Documents
Culture Documents
Efficient Seismic Design of 3D Asymmetric and Setback RC Frame Buildings For Drift and Strain Limitation
Efficient Seismic Design of 3D Asymmetric and Setback RC Frame Buildings For Drift and Strain Limitation
Efficient Seismic Design of 3D Asymmetric and Setback RC Frame Buildings For Drift and Strain Limitation
Abstract: This paper presents an analysis–redesign-type approach for the efficient seismic design of three-dimensional (3D) irregular RC
frame structures for bidirectional ground motions. The designs obtained using the approach satisfy interstory drift and ductility limits while
Downloaded from ascelibrary.org by Ryerson University on 11/01/16. Copyright ASCE. For personal use only; all rights reserved.
having the minimum total moment capacity of all seismic members. This leads to cost-efficient designs because the total amount of steel as
well as the base shear and overturning moments are relatively low. While doing that, the approach is very intuitive and is relatively simple to
implement in practice because it only requires analysis tools, not background knowledge or tools related to optimization theory. The examples
show that the proposed methodology requires only a few analyses and converges to designs that exactly satisfy the design objectives with
limited amounts of steel assigned only where required. This makes the design approach feasible for practice in terms of computational effort
and time. DOI: 10.1061/(ASCE)ST.1943-541X.0001689. © 2016 American Society of Civil Engineers.
Author keywords: Seismic effects; Irregular frame structures; Optimal seismic design; Efficient seismic design; Analysis–redesign.
i i ∀ i ¼ seismic element
drifts. Similarly, constraints are assigned on seismic beam and col-
grav
umn base ductility demands. M yb ≤ M yb ∀ b ¼ seismic beam
The behavior of irregular structures is often characterized M grav ∀ c1 ¼ seismic column bending axis 1
yc1 ≤ M yc1
by highly concentrated local deformation demands in regions of
irregularity (both heightwise and planwise). Asymmetric and gen- M grav
yc2 ≤ M yc2 ∀ c2 ¼ seismic column bending axis 2 ð1Þ
erally irregular 3D buildings, for example, often experience the
largest deformation demands in their peripheries. Thus, local Here, M yi = nominal flexural yield strength of seismic element
damage measures in these locations are to be constrained so as (beam or column) i considering its gravity axial force. When i re-
to regularize the response of the building. In this paper, the peak fers to a column, Myi equals the sum of M yc1 and M yc2 , which are
interstory drift of each story of each peripheral frame is limited to the nominal flexural yield strengths of column c about Bending
an allowable value. This value is adjusted for each level of seismic- Axes 1 and 2, respectively. The interstory drift performance index,
ity considered. Suites of realistic ground motions are adopted to PIθj ¼ maxk ðθjk =θLIMk Þ = maximum over all intensities k, where
describe the various levels of seismic hazard. For each level of θjk is the selected peak interstory drift response in bent j for inten-
seismic hazard, the mean peak interstory drift at each location is sity k, and θLIMk is the interstory drift limit for intensity k. The se-
constrained. Alternatively, the mean plus standard deviation or lected interstory drift response could be, for example, the mean
the envelope interstory drift from all ground motions for a peak drift response to all ground motions considered for that inten-
selected seismic hazard level could be adopted. If the latter is sity, the maximum peak drift response, or the mean plus standard
chosen, the active set of ground motions approach (Lavan and Levy deviation. A bent refers in this paper to one story of a plane
2006) could be used to appreciably reduce the computational frame including columns and the beams connected to the top of
effort. the columns, for example, all ground-story columns and first-
According to the capacity design philosophy, elastic response is floor beams of a peripheral frame (e.g., Fig. 2 in “Examples”).
desired in the columns above their bases. Thus, additional con- The local seismic beam ductility response performance index,
straints in the form of minimum column-to-beam strength ratios PIμb ¼ maxk ðμbk =μLIMk Þ, is the maximum over all intensities k,
are assigned to minimize column yielding above their bases. where μbk is the maximum (of the two potential plastic hinges,
As mentioned previously, the elements are presized, with the one at each beam end) selected peak plastic hinge ductility response
of seismic beam b for intensity k, and μLIMk is the maximum
option of changing their sizes in very limited and specific cases.
allowable beam plastic hinge ductility for intensity k. The local
Nonetheless, limits on their minimum and maximum values are
seismic column base ductility response performance index, PIμc ¼
also assigned. In addition, modern seismic codes assign limits
maxk ðμck =μcLIMk Þ, is the maximum over all intensities k, where μck
on the minimum and maximum reinforcement ratios allowed in
is the maximum selected peak plastic hinge ductility response of
the beams and columns. These limits are set so as to ensure a ductile
seismic column base c for intensity k, and μcLIMk is the maximum
behavior. Such limits are adopted in this work too. Because the
allowable column base plastic hinge ductility for intensity k. The
frames are also part of the gravity load–resisting system, a limit
selected peak ductility could be the mean to all ground motions
on the minimum stiffness and strength that is required to carry considered for that intensity, the maximum, or the mean plus stan-
the gravity loads is assigned as well. dard deviation. Additionally, ρi is the steel reinforcing ratio of
seismic element i; bi and di are the cross section width and depth,
Optimization Problem respectively, of seismic element i; the superscripts min i and max
i
denote minimum and maximum allowable values for the particular
The optimization problem to follow is based on the previous sec- parameter (ρi , bi , and di ) for seismic element i; M grav
yb is the nominal
tions. The nominal flexural strengths in the various elements serve flexural yield strength of seismic beam b required by the gravity
as the primary design variables while some changes in elements load case(s); and M grav grav
yc1 and M yc2 are the nominal flexural yield
dimensions are allowed under some circumstances. The total flexu- strengths about Bending Axes 1 and 2 of seismic column c required
ral strength is to be minimized. Constraints are added on the inter- by the gravity load case(s).
story drift of each story and each peripheral frame, as well as beam As the problem formulation implies, shear capacity of the col-
strains, under each of the seismic levels considered (e.g., DBE, umns and beams is assumed to be sufficient. Thus, once the flexural
MCE). The columns are intended to yield at their bases only, by strengths are obtained, the shear capacity of the columns and beams
virtue of predefined column-to-beam strength ratios. Side con- is to be designed to prevent shear failure. This also implies that the
straints are assigned to the sizes of the elements and the reinforce- problem formulation and the proposed solution scheme are not
ment ratios. In addition, constraints are added so as to ensure applicable to geometries in which shear failure is expected to con-
acceptable behavior under gravity loads. trol the behavior of elements (e.g., short beams or columns).
Here, the minimum gauge for beam flexural strength is deter- height.
mined either by reinforcement ratio and size limitations or by For the initial seismic design of the examples presented in this
gravity (or other) load cases. The performance measure is the maxi- paper, minimum section flexural strengths were set by applying
mum of PIθj of the peripheral frame in the story below the floor typical factors to moment demands obtained from a linear elastic
and the maximum PIμb in the beams of that bent. For the first floor, static analysis of a structural model. Initial section sizes were set by
the maximum PIμc in the direction of the bent, of all column bases relating story shears to drifts using Wilbur’s formula (Stafford
in that bent, is also considered in the performance measure. This Smith and Coull 1991, section 7.4.1), attributing half the allowed
optimality criterion could be achieved by a generalization of the AR drift to beam deformations and half to column deformations. The
procedure as presented in the next section. story shears were derived from a preliminary equivalent lateral load
In the preceding discussion, interstory drifts were taken as the assessment similar to that found in most current design codes. For
damaging deformations and were assumed to be strongly related this purpose, cracked cross sections were assumed to have second
to the story stiffnesses and strengths. In general, this is more ap- moments of area equal to half the gross section second moments
parent in low-rise to midrise buildings. These deform mainly of area.
due to the total story shear. In high-rise buildings, on the other
hand, the axial contraction of columns leads to nonnegligible Step 2: Analysis
deformations due to the external overturning moment. Thus, the With a given design at hand, the structure is subjected to seismic
interstory drifts (the relative horizontal displacement of two ad- analysis for the various ground motion intensities considered. Be-
jacent floors) are comprised of two contributors: a rigid body dis- cause the methodology is tailored for the design of nonlinear 3D
placement and a racking (shear) deformation (CTBUH 2008). generally irregular RC MRF buildings, an analysis method captur-
Here, racking is the main contributor to damage. It is also the one ing the unique behavior of such structures, as elaborated in the “In-
strongly related to the story stiffness and strength. Thus, it seems troduction,” should be adopted. Specifically, the analysis method
reasonable to use the racking deformation rather than the inter- should reasonably predict the interstory drifts at the peripheries.
story drift in tall buildings. This, however, was not verified as part The ductility demands in the beams, and column bases of the seis-
of the study. mic frames, are also required, although strains very rarely govern
the design (Priestley et al. 2007).
In the examples presented in this work the authors adopted a
Analysis–Redesign Scheme
3D nonlinear response history analysis (NRHA) to predict the
As per the optimality criterion stated previously, strength and stiff- responses of interest. This analysis approach is becoming more
ness are concentrated in the peripheral frames. Hence, inner frames widespread for the purpose of design verification when it comes
are considered as gravity frames and are designed to sustain gravity to nonlinear 3D generally irregular buildings. Because this analysis
loads only. The gravity columns typically have smaller sections type is required by some codes for design verification of the type of
and, in practice, would be detailed for ductilities congruent with structure considered, such a model has to be built in any case. The
the expected peak deformations. Appropriate details are to be de- analysis–redesign procedure does not usually require more that
signed for the columns of these frames so as to ensure their ability 10–15 iterations for practical convergence. Thus, it is the authors’
to sustain the gravity loads while experiencing displacements due opinion that adopting NRHA for the design as well would not
to the seismic action. Because the yield displacements of these col- require excessive additional effort. With the proposed design
umns depend mainly on their geometry and material yield strains approach, the engineer can use any analysis tool the engineer be-
(Paulay 2002), their ductility demands do not, in general, depend lieves would lead to reasonable response prediction. However, it
on their strength. Hence, the addition of longitudinal reinforcement should be kept in mind that the designs generated using the pro-
would not directly reduce their ductility demands. Steel reinforcing posed methodology satisfy the acceptance criteria when verified
details that provide sufficient ductility capacity should be adopted. using the same analysis tool that was used for design. Hence, the
The last statement, in the authors’ opinion, holds regardless of the quality of the designs is as good as the quality of the structural
method of design adopted. model, seismic hazard modeling, and analysis tools used.
An overview of the proposed AR method is presented in Fig. 1.
The various steps will now be elaborated. Step 3: Stopping Criteria
If the stopping criteria are met, the process ends. If not, the process
Step 1: Initial Design continues until convergence. Three criteria were employed simul-
In this step, the seismic intensities as well as corresponding accep- taneously. The first checked the seismic performance against the
tance criteria should be defined. The seismic hazard is modeled in target performance. The second two checked convergence. The first
the examples to follow using a suite of ground motion histories to criterion was that the performance indices (PIs) (responses or
describe each seismicity level. Nonetheless, the methodology could limits of all seismic frame interstory drifts and all seismic beam
and column base plastic hinge ductilities) must all be less than 1 yield curvature [as per Paulay (2002), φyi ¼ 2.1εsy =di for bending
(plus tolerance). The second criterion was that the objective about axis parallel to bi dimension and φyi ¼ 2.1εsy =bi for bending
function value (the total volume of steel reinforcing) must have about axis parallel to di dimension; εsy = longitudinal reinforcing
stabilized (converged) to a tolerable change per iteration for two steel design yield strain].
consecutive iterations. The third criterion was similar to the second, Strength distribution within a bent (all columns and beams in
being that all the performance indices must have stabilized one story of a frame) using this recurrence relation may depend
(converged). This was checked for each location by comparing on the initial design. Various relative beam strength distributions
the change of each PI during each of the last two iterations against depending on relative beam lengths were investigated (results
the selected change tolerance. These criteria allowed drifts to be not shown). However, the total volume of reinforcing steel was
less than allowable but not greater. This was necessary to allow found to be insensitive to variations of such distributions. Priestley
gravity considerations to govern elements where appropriate. et al. (2007) suggested that constant beam cross section within
bents may be a rational solution considering the dependence of
Step 4: Redesign Beams ductility demand and capacity ratios on relative beam length and
The beams in each seismic frame in each story are redesigned using ease of construction. However, other rational distributions may
the following recurrence relation: be used.
ðpþ1Þ ðpÞ
ðpÞ
ðpÞ Q Ductility capacities rarely govern the design (Priestley et al.
Myb ¼ M yb max PIθj ; PIμb ð2Þ 2007, p. 226, section 5.3.1). Hence, interstory drifts are the leading
criteria for design. If the ductility limit in a single beam governs,
ðpþ1Þ ðpÞ
where M yb and M yb = nominal flexural yield strengths of there are two possible redesign approaches. The first decreases the
ðpÞ depth of that beam. This would increase the yield curvature and
seismic beam b for iterations p þ 1 and p, respectively; PIθj and
ðpÞ decrease the ductility demand for a given maximum curvature
PIμb = parameters PIθj and PIμb defined in “Optimization Problem” (Priestley et al. 2007). Alternatively, increasing the total strength
computed for iteration p, respectively. The new flexural stiffness is [and stiffness, which is proportional to strength (Priestley et al.
ðpþ1Þ
computed from the new flexural strength M yb and corresponding 2007)] of all beams in that bent by adding reinforcement would
area equal to half the gross section second moments of area. The
better the initial seismic design, the fewer the iterations required to
achieve convergence.
modified using the program SeismoMatch to match the design re-
sponse spectrum. This was used as the seismicity for Limit State 2. Discussion of AR Design Performance
The period range of interest for scaling purposes was 0.6–3 s. As The analysis–redesign procedure leads to a fully stressed design
will be shown subsequently, the fundamental periods of the final solution within the constraints on flexural strengths. Minimum
designs obtained using both the AR designs and force-based strengths are based on the gravity load design. Maximum flexural
designs in this example and in the next example fall in the range strengths were based on reasonable beam–column dimension
of 0.63–1.68 s. The thin gray lines are the response spectra of the limits and allowable steel reinforcing ratios. The analysis–redesign
individual scaled ground motions. The final design is expected to algorithm was then executed. Fig. 4 shows the convergence of the
be well into the constant velocity region, thus the matched spectra objective function (total steel volume) and the maximum constraint
agree well with the design spectra for the relevant period range. violation (MCV) in percent for a Q value of 0.7 [Eq. (2) in “Step 4:
For simplicity, the same ground motion records were simply Redesign Beams”]. It can be seen that after approximately 10
factored by 0.5 to model the seismicity for the serviceability limit iterations, the MCV reaches approximately 10%, which is probably
state. acceptable for most design purposes considering the uncertainty
usually inherent in the input data (e.g., characteristics of materials
Modeling and Analysis and seismicity).
The modeling and analysis was done in Ruaumoko 3D (Carr 2006). Fig. 2 shows the final component dimensions and relative
Finite-element models were built using one (Giberson beam) line flexural strengths (darker is stronger). The AR design verification
element for each beam–column member. These elements have one results (interstory drifts and beam and column ductility demands
potential plastic hinge of a specified constant length at each end. normalized by allowable values) are shown in Fig. 5.
Uniformly distributed line loads were applied to the beams In general, most of the peripheral frames achieved the interstory
assuming reasonable gravity loads. Moment–axial force interaction drift limit in most stories for one or both of the limit states
was modeled using the Reinforced Concrete Type 1 (IBEAM = 2) considered. The black drift bars show the serviceability mean peak
option. The modified Takeda hysteresis model (Otani 1974) was demands and the gray bars the ultimate limit state (ULS) demands.
employed (α ¼ 0.25, β ¼ 0.3, NF ¼ 1, and KKK ¼ 1), as well The vertical lines in the interstory drift figures show the drift limits
as strength degradation (ILOS ¼ 6) considering peak ductility (tolerance of 2% shown each side).
demand and number of inelastic cycles. Column plastic hinges In Figs. 5(e and f), one horizontal bar is shown for each seismic
were assigned post-yield and elastic moment-curvature stiffness column story element and beam element. Only results for seismic
of 5%, whereas 2% was assigned to the beam plastic hinges. columns and beams are shown, hence the number of columns
Rotational fixity of column bases was modeled by moment- shown (18) is less than the total number of columns (19), and the
resisting connections to elastic foundation beams of relatively same for the beams. The column ductility performance indices
low flexural rigidity connecting the column bases. If desired, the (demand and limit) show that most of the eight base plastic hinges
foundation beams could also be designed in the AR procedure. in the ground-story seismic columns achieved more than half their
However, for these examples, they were assumed to be designed as ductility capacities. The curvature ductility limits of the column
floor slab edge beams and ties between columns, so their constant
properties were preselected. Column uplift was modeled using
compression-only springs having axial stiffnesses resulting in
approximately 5-mm settlement under gravity loads at each
ground-story column base. As recommended by Carr (2006),
5% damping was assigned to all modes using the Wilson and
Penzien (1972) damping model (ICTYPE ¼ 2). This damping
model uses the initial elastic modal properties and so is constant
throughout the NRHA.
Because all seismic and gravity beams and columns were mod-
eled, no P-delta leaning columns were required to account for
P-delta effects in the analyses. Reasonable vertical and rotational
Fig. 4. Convergence: objective function and maximum constraint
inertias were assigned to column end nodes. An analysis time step
violation
of 0.0025 s was used. The Newmark (Carr 2006) constant average
Fig. 5. Analysis–redesign design verification results (black drift bars show the serviceability mean peak demands and the gray bars the ULS demands)
bases were set equal to the beam ductility limits [5 for serviceability formulation and therefore considered geometrical nonlinearity
limit state (SLS) and 7.3 for ULS]. The above-base column (P-delta effects) directly. Sections were determined using the AR
ductility limit was 1. It can be seen that some yielding occurred approach to increase element dimensions when drifts were larger
above the column bases. However, these ductility demands were than the limits and vice versa. Here, as per code requirements
limited to less than 1.25 in all columns but one, and to less than (e.g., Eurocode 8) the cracked flexural stiffnesses were taken as
2 in all columns. Overall, the capacity design approach succeeded 0.5EIg (where EIg is the gross cross section bending stiffness).
in distributing plasticity at the column bases and beam ends in Therefore, these designs represent the best designs achievable using
accordance with the intended ductile collapse mechanism targeted typical force-based design assumptions about section stiffnesses
and in limiting column ductility demand above bases to acceptable and linear analysis with the equal displacements assumption. Final
values. plastic hinge region nominal flexural yield strengths were taken as
The absolute beam ductility limit is also shown as a vertical line the maximum responses recorded during the last AR linear re-
in the bottom right beam ductility bar chart in Fig. 5(f). As can be sponse history analysis divided by the reduced response modifica-
seen in this bar chart, the AR design was relatively efficient, uti- tion coefficient. Five values for response modification coefficient
lizing on average more than half the ductility capacity of the beams. were selected to result in five different force-based designs. These
The design of the beams in the peripheral frames that did not are R ¼ 1, 2, 3, 4, and 5. Because interstory drifts usually govern
achieve the drift limit were governed by gravity loads. In Fig. 2 the designs, the use of different R values allows the comparison of
it can be seen that some of these beams (e.g., x-direction Frame the AR method with the best possible design that could be achieved
Edge 2 Story 1) had final depths greater than the minimum allowed using force-based design. The response modification coefficients
values. This could be due to the analysis–redesign algorithm were reduced according to Eurocode 8 by 20% due to irregularity
converging to the expected gravity resistance capacities at slightly (setbacks) in elevation. The interstory drift sensitivity index, which
larger beam depths. Although the depths of these beams did not
aims to consider geometric nonlinearity, was computed for each
reach the expected minimum gravity design values, the flexural
story, and the strengths of all seismic elements were increased ac-
strengths at these locations reached the expected gravity design
cording to code requirements where necessary. Only the 6-story
values.
force-based designs (next example) for R ¼ 4 and 5 were affected
Force-Based Designs by this index.
The force-based designs used the same minimum gravity load Capacity design principles as were applied in the AR procedure
capacities and the same initial seismic design. Element dimensions were applied to the force-based designs to ensure monotonic
were set so as to achieve the allowable interstory drifts in linear column strength and size reduction with height and a minimum
response history analysis. These analyses used large displacement column–beam strength ratio of 2. Element dimensions were also
volume Vs, the total concrete volume Vc, the maximum PIθ of
all stories of all peripheral frames, the maximum PIμc of all seismic
columns, the maximum PIμb of all seismic beams, mean peak
Downloaded from ascelibrary.org by Ryerson University on 11/01/16. Copyright ASCE. For personal use only; all rights reserved.
AR design in terms of total steel volume, column ductility, base was designed using the proposed analysis–redesign method and us-
shears, and overturning moments. However, beam ductility demands ing a typical force-based design approach for a range of response
exceeded capacities by 20% and the drift limits were exceeded by modification coefficients. The structure has a total of 142 beam
40%. This essentially represents the best possible design achievable segments and 96 column segments, or 238 design variables. The
using assumptions typically accepted by design codes regarding performance of the designs is discussed and compared in this sec-
section stiffness and the equal displacements approximation. tion. The same material properties and floor unit masses were used
While the AR designs had the second to highest total concrete as in the 3-story example. The same procedure for generating the
volume, this parameter was not targeted for minimization in the AR initial seismic designed was also utilized.
algorithm because steel reinforcing is generally significantly more The same two limit states that were considered in the design of
expensive than concrete. The concrete volume varied less than 20% the preceding 3-story example were adopted in this example as
from the AR design at the most. well: a SLS and a ULS. Also, the same seismic action, modeling
As can be seen in Fig. 6, the top two stories experienced very and AR algorithm as for the 3-story example was used. Fig. 8
high peak interstory drift response, reaching values of 3.74 times shows the convergence of the objective function (total steel
the allowable in the top story for R ¼ 5, verging on collapse. volume) and the MCV in percent for a recursion exponent value
The final response modification coefficient values used in the of Q ¼ 0.7 [Eq. (1) in “Step 4: Redesign Beams”]. It can be
force-based designs are shown in parentheses in Table 1. Mean seen that, as for the 3-story example, after approximately 10 iter-
peak interstory drift demands divided by target response limits, ations, the MCV decreases to below 10%. All the selected conver-
and maximum (of the four corners of the building plan) absolute gence criteria were satisfied within 20 iterations for this 6-story
values of mean peak floor total accelerations in each horizontal case.
concrete volumes were fairly constant, varying less than 11% from
the AR design value.
The base shear and overturning moment KPIs of the AR design
Downloaded from ascelibrary.org by Ryerson University on 11/01/16. Copyright ASCE. For personal use only; all rights reserved.
“Earthquake induced torsion in buildings: Critical review and state of Levy, R., and Lavan, O. (2006). “Fully stressed design of passive control-
the art.” Earthquake Struct., 8(2), 305–377. lers in framed structures for seismic loadings.” J. Struct. Multidiscipl.
ASCE. (2010). “Minimum design loads for buildings and other structures.” Optim., 32(6), 485–498.
ASCE/SEI 7-10, Reston, VA. Liu, Q., Zhang, J., and Yan, L. (2010). “An optimal method for seismic
ASCE. (2014). “Seismic rehabilitation of existing buildings.” ASCE/SEI drift design of concrete buildings using gradient and Hessian matrix
41-13, Reston, VA. calculations.” Archiv. Appl. Mech., 80(11), 1225–1242.
Balling, R. J., Pister, K. S., and Ciampi, V. (1983). “Optimal seismic- Moehle, J. P., and Alarcon, L. F. (1986). “Modeling and analysis
resistant design of a planar steel frame.” Earthquake Eng. Struct. methods for vertically irregular buildings.” Proc., 8th European Conf.
Dyn., 11(4), 541–556. on Earthquake Engineering.
BCJ (Building Center of Japan). (1981). “Structural provisions for building NISEE. (2016). “Suites of earthquake ground motions for analysis of steel
structures.” Tokyo (in Japanese). moment frame structures.” 〈https://nisee.berkeley.edu/elibrary/files
Bertero, V. V. (1997). “Performance-based seismic engineering: A critical /documents/data/strong_motion/sacsteel/ground_motions.html〉 (Sep.
review of proposed guidelines.” P. Fajfar and H. Krawinkler, eds., 26, 2016).
Seismic design methodologies for the next generation of codes, A.A. Otani, S. (1974). “Inelastic analysis of R/C frame structures.” J. Struct.
Balkema, Rotterdam, Netherlands, 1–31. Div., 100(7), 1433–1449.
Carr, A. J. (2006). “RUAUMOKO—Program for inelastic dynamic analy- Paulay, T. (2002). “An estimation of displacement limits for ductile
sis.” Dept. of Civil Engineering, Univ. of Canterbury, Christchurch,
systems.” Earthquake Eng. Struct. Dyn., 31(3), 583–599.
New Zealand.
Priestley, M. J. N. (2000). “Performance based seismic design.” Proc., 12th
CEN (European Committee for Standardization). (2004). “Design of struc-
World Conf. on Earthquake Engineering.
tures for earthquake resistance. Part 1: General rules, seismic actions
and rules for buildings.” EN 1998-1, Eurocode 8, Brussels, Belgium, Priestley, M. J. N., Calvi, M. C., and Kowalsky, M. J. (2007).
222. Displacement-based seismic design of structures, IUSS Press, Pavia,
Cilley, F. H. (1900). “The exact design of statically indeterminate frame- Italy.
works: An exposition of its possibility but futility.” Transactions, 43(1), Ruaumoko 3D [Computer software]. Univ. of Canterbury, Christchurch,
353–407. New Zealand.
CTBUH (Council on Tall Buildings and Urban Habitat). (2008). “Recom- SeismoMatch [Computer software]. Seismosoft, Pavia, Italy.
mendations for the seismic design of high-rise buildings.” Chicago. Somerville, P., Smith, N., Punyamurthula, S., and Sun, J. (1997). “Develop-
Daniel, Y., and Lavan, O. (2014). “Gradient based optimal seismic retro- ment of ground motion time histories for Phase 2 of the FEMA/SAC
fitting of 3D irregular buildings using multiple tuned mass dampers.” steel project.” Rep. No. SAC/BD-97/04, Sacramento, CA.
Comput. Struct., 139(1), 84–97. Stafford Smith, B., and Coull, A. (1991). Tall building structures: Analysis
Daniel, Y., and Lavan, O. (2015). “Optimality criteria based seismic design and design, Wiley, New York.
of multiple tuned-mass-dampers for the control of 3D irregular Stathopoulos, K. G., and Anagnostopoulos, S. A. (2005). “Inelastic torsion
buildings.” Earthquakes Struct., 8(1), 77–100. of multistorey buildings under earthquake excitations.” Earthquake
Dooley, K. L., and Bracci, J. M. (2001). “Seismic evaluation of column-to- Eng. Struct. Dyn., 34(12), 1449–1465.
beam strength ratios in reinforced concrete frames.” ACI Struct. J., Tzan, S. R., and Pantelides, C. P. (1996). “Convex model for seismic
98(6), 843–851. design of structures—II: Design of conventional and active structures.”
Fragiadakis, M., and Papadrakakis, M. (2008). “Performance-based opti- Earthquake Eng. Struct. Dyn., 25(9), 945–963.
mum seismic design of reinforced concrete structures.” Earthquake Valmundsson, V. E., and Nau, J. M. (1997). “Seismic response of building
Eng. Struct. Dyn., 37(6), 825–844. frames with vertical structural irregularities.” J. Struct. Eng., 10.1061
Ganzerli, S., Pantelides, C. P., and Reaveley, L. D. (2000). “Performance-
/(ASCE)0733-9445(1997)123:1(30), 30–41.
based design using structural optimization.” Earthquake Eng. Struct.
Wilson, E. L., and Penzien, J. (1972) “Evaluation of orthogonal damping
Dyn., 29(11), 1677–1690.
matrices.” Int. J. Numer. Methods Eng., 4(1), 5–10.
Hajirasouliha, I., Asadi, P., and Pilakoutas, K. (2012). “An efficient perfor-
Zaghi, A. E., Soroushian, S., Itani, A., Maragakis, E. M., Pekcan, G., and
mance-based seismic design method for reinforced concrete frames.”
Earthquake Eng. Struct. Dyn., 41(4), 663–679. Mehrraoufi, M. (2015). “Impact of column-to-beam strength ratio on
Kyrkos, M. T., and Anagnostopoulos, S. A. (2011). “An assessment of code the seismic response of steel MRFs.” Bull. Earthquake Eng., 13(2),
designed, torsionally stiff, asymmetric steel buildings under strong 635–652.
earthquake excitations.” Earthquake Struct., 2(2), 109–126. Zou, X. K., and Chan, C. M. (2005a). “An optimal resizing technique
Lavan, O. (2015). “A methodology for the integrated seismic design of for seismic drift design of concrete buildings subjected to response
nonlinear buildings with supplemental damping.” Struct. Control spectrum and time history loadings.” Comput. Struct., 83(19),
Health Monitor., 22(3), 484–499. 1689–1704.
Lavan, O., and Daniel, Y. (2013). “Full resources utilization seismic Zou, X. K., and Chan, C. M. (2005b). “Optimal seismic performance-
design of irregular structures using multiple tuned mass dampers.” based design of reinforced concrete buildings using nonlinear pushover
Struct. Multidiscipl. Optim., 48(3), 517–532. analysis.” Eng. Struct., 27(8), 1289–1302.