https://doi.org/10.3311/PPci.

17150
566|Creative Commons Attribution b
Periodica Polytechnica Civil Engineering, 65(2), pp. 566–576, 2021

Reliability Based Design Optimization of Reinforced Concrete

Frames Using Genetic Algorithm

Mohammadreza Seify Asghshahr1*

1
Department of Civil Engineering, Faculty of Engineering, Azarbaijan Shahid Madani University, Tabriz P.O. Box 53714-161, Iran
Corresponding author, e-mail: [email protected]
*

Received: 03 September 2020, Accepted: 21 January 2021, Published online: 03 February 2021

Abstract
This paper introduces a new framework for reliability based design optimization (RBDO) of the reinforced concrete (RC) frames. This
framework is constructed based on the genetic algorithm (GA) and finite element reliability analysis (FERA) to optimize the frame weight
by selecting appropriate sections for structural elements under deterministic and probabilistic constraints. Modulus of elasticity of
the concrete and steel bar, dead load, live load, and earthquake equivalent load are considered as random variables. Deterministic
constraints include the code design requirements that must be satisfied for all the frame elements according to the nominal values
of the aforementioned random variables. On the other hand, this framework provides the minimum required reliability index as
the probabilistic constraint. The first-order reliability method (FORM) using the Newton-type recursive relationship will be used to
compute the reliability index. The maximum inter-story drift is considered as an engineering demand parameter to define the limit-
state function in FORM analysis. To implement the proposed framework, a mid-rise five-story RC frame is selected as an example.
Based on the analysis results, increasing the minimum reliability index from 6 to 7 causes an 11 % increase in the weight of the selected
RC frame as an objective function. So, we can obtain a trade-off between the optimized frame weight and the required reliability index
utilizing the developed framework. Furthermore, the high values of the reliability index for the frame demonstrate the conservative
nature of code requirements for interstory drift limitations based on the linear static analysis method.
Keywords
reinforced concrete frame, reliability based design optimization, first-order reliability method, genetic algorithm

1 Introduction
In the majority of structural optimization problems in rein- analysis should be performed during the optimization pro-
forced concrete (RC) framed structures, it is mainly focused cess so that various sources of uncertainties [4] in the design
to optimize the objective function according to determinis- process can be properly accounted for. This approach is
tic constraints. In this way, Kaveh and Sabzi [1] optimized called reliability based design optimization (RBDO) that
the design of RC frames using a big bang-big crunch algo- has been emerged as a new optimization procedure for var-
rithm. Esfandiari et al. [2] optimized the seismic design ious structural systems [5, 6].
of RC frames subjected to time-history loadings using an Reliability analysis in RBDO can be performed by var-
algorithm combining multi-criterion decision-making and ious methods. Gholizadeh et al. [7] proposed a neural net-
particle swarm optimization. The considered determinis- work-based method for reliability assessment of optimally
tic constraints in these problems usually consist of code seismic designed moment frames. Grubišić et al. [8] con-
requirements such as lateral displacement limitation. In this ducted reliability analysis of the RC frame by finite ele-
procedure, uncertainties in the structural analysis will be ment method considering implicit limit state functions.
ignored that can lead to unsafe design. These uncertainties In RBDO of RC framed structures, in addition to the
in RC framed structures come from improper design and deterministic constraints in the optimization practice, prob-
construction practices and other sources like time-depen- abilistic constraints must also be examined in which failure
dent deterioration induced by corrosion [3]. To solve this probability or corresponding structural reliability index of
problem and to design a safe structure, structural reliability the structure is determined. Consequently, it can be argued

Cite this article as: Asghshahr, M. S. "Reliability Based Design Optimization of Reinforced Concrete Frames Using Genetic Algorithm", Periodica
Polytechnica Civil Engineering, 65(2), pp. 566–576, 2021. https://doi.org/10.3311/PPci.17150
 Asghshahr
Period. Polytech. Civ. Eng., 65(2), pp. 566–576, 2021
|567

that the frame is designed to meet a specific reliability The remainder of this paper is prepared as follows:
index uniformly, assuring its performance over the struc- In Section 2 an RBDO problem of RC frames is defined.
ture service life under the involved uncertainties. In this The GA technique is then presented in Section 3 for solv-
way, Aoues and Chateauneuf [9] examined the RBDO of ing the RBDO of RC frames. An illustrated example is
RC frames by adaptive target safety. Khatibinia et al. [10] introduced in Section 4. Analyses results are presented in
introduced a discrete gravitational search algorithm and Section 5. The final section concludes the paper.
a meta-modelling framework for RBDO of RC structures
including soil-structure interaction. Shayanfar et al. [11] esti- 2 Definition of RBDO problem of RC frames
mated the corrosion occurrence in RC structures using reli- In the RBDO problem of the RC frames, the objective is to
ability based particle swarm optimization. Léger et al. [12] find the sections (S) for the structural members such that
performed an RBDO of RC structures with elastomeric iso- the frame weight (W) is optimized considering the involved
lators using adaptive sparse polynomial chaos expansion. deterministic and probabilistic constraints. Selected cross
Zou et al. [13] formulated a reliability-based performance sectional areas are the discrete design variables extracted
design optimization for seismic retrofit of RC buildings from a prepared section list (PSL) where all of the existing
with fiber-reinforced polymer composites. sections in this list have satisfied the preliminary geomet-
Genetic algorithm (GA) is a very powerful optimization rical and reinforcement detailing requirements. Therefore,
method used in structural optimization problems [14–16]. RBDO problem of RC frame structures can be formulated
The design of RC frames has been optimized using a GA as follows:
considering the deterministic constraints [17, 18]. Shahraki
find S
and Noorossana [19] demonstrated that multi-objective
min W (S )
GA can be used to solve discrete RBDO problems. There
is a limited practical application of GA in the structural  Si ∈ PSL i = 1, 2,..., n
 max + r+
RBDO problems. In this regard, Biabani Hamedani and  M j ( S , Nom X ) ≤ M j (S ) j = 1, 2,..., nb
Kalatjari [20] performed structural system reliability-based  max − r− (1)
 M j ( S , Nom X ) ≤ M j (S ) j = 1, 2,..., nb
optimization of truss structures using a GA through the s.t.  max + r+
branch and bound method. Shayanfar et al. [21] developed  N k ( S , Nom X ) ≤ N k (S ) k = 1, 2,..., nc
 N max − ( S , Nom ) ≤ N r − (S ) k = 1, 2,..., nc
a genetic algorithm based method for RBDO of struc-  k X k

tures with discrete and continuous design variables using  p ( g ( S , X ) ≤ 0) ≤ p t where p tf = Φ((− β t )
 f
OpenSees and TCL. However, no research has been con-
ducted on the RBDO of the RC frames using a GA. Where Si is the selected section for ith structural element
This paper presents a new framework for RBDO of in frame; n is the number of structural elements in frame;
the RC frames employing finite element reliability anal- Mjmax+ and Mjmax– are the maximum moment at the midspan
ysis (FERA) and genetic algorithm (GA) to optimize the and each end of jth beam in frame, respectively; Mjr+ and
frame weight incorporating the deterministic and proba- Mjr– are the moment strength at the midspan and each end of
bilistic constraints. Deterministic constraints include the jth beam in frame, respectively; NomX is the nominal value
code design requirements while probabilistic one provides vector of random variables; nb is the number of beams in
the minimum required reliability index. The first-order frame; Nkmax+ and Nkmax– are the maximum compressive and
reliability method (FORM) using the Newton-type recur- tensile axial force of kth column in frame, respectively; Nkr+
sive relationship is used to compute the reliability index or and Nkr– are the compressive and tensile axial strength of kth
corresponding probability of failure. The maximum inter- column in frame, respectively; nc is the number of columns
story drift is considered as an engineering demand param- in frame; p(g(S, X) ≤ 0) represents the failure probability of
eter to define the limit-state function in FORM analysis. the frame; X is the vector of random variables; g(S, X) ≤ 0
The linear static analysis is used to calculate the struc- is the limit state function; pft is the allowable probability
tural response during the optimization process. To evalu- of failure of the frame; Φ(.) is the standard cumulative
ate the plastic performance of the frame, reliability based function of the normal distribution and βt is the minimum
limit and shakedown analyses [22–24] can be utilized. required reliability index of the frame. A detailed presenta-
However, this research is limited to the elastic perfor- tion of the applied approach to solve the RBDO problem
mance of the framed structure. reported above can be found in following sections.
568|Asghshahr
Period. Polytech. Civ. Eng., 65(2), pp. 566–576, 2021

3 Solving RBDO problem of RC frames using GA where κ is the vector of constant coefficients consists of
In present study, the RBDO problem as formulated in Eq. (1) κβ, κb and κ c for reliability index of the frame, beam moment
will be solved using the genetic algorithm. GA is a heuristic and column axial force constraints, respectively; ψ is the
method that uses the natural selection concept in the optimi- vector of penalty terms consists of ψβ, ψbi and ψck for reli-
zation process. Due to the involving discrete decision vari- ability index of the frame, jth beam moment and kth col-
ables and also implicit form of the constraints, evolution- umn axial force, respectively. Other involved parameters
ary algorithms like GA is more suitable than gradient-based were introduced in Eq. (1). It must be noted that κβ, κb and κ c
methods to solve the introduced RBDO problem. The aim are selected large enough to penalize the infeasible design
is to find the sections for structural members of RC frame solutions. To compute ψbi and ψck terms, refer to Section 3.1
for optimizing its weight considering the deterministic and for deterministic constraints and to compute ψβ term refer
probabilistic constraints. Due to variety of available RC to Section 3.2 for probabilistic constraint. Therefore, a fit-
sections according to their dimensions, bar sizes, and bar ness value of each individual will be inversely proportional
arrangements; set of selected sections (design solution) for to its total weight.
structural members during the optimization process can be - Step 4: Perform the single-point crossover operation.
semi-infinite. This selection process entails high computa- - Step 5: Perform the mutation operation.
tional efforts. To mitigate this problem, a prepared section - Step 6: Reproduce the next generation. Generated indi-
list (PSL) is constructed that consists of finite number of viduals in steps 2, 4 and 5 are mixed to construct the total
sections for structural members of the RC frame. In a PSL, population. These individuals compete with each other to
all sections satisfy the preliminary code requirements survive in the next generation. To this end, the individuals
include the minimum and maximum dimensions of the sec- are sorted corresponding to their fitness values.
tion, maximum aspect ratio, rebar sizes, the minimum and - Step 7: Repeat the above mentioned steps until niter is
maximum spacing of the bars, the cover thickness, the min- achieved.
imum and maximum steel bar area, and etc. An identifica-
tion number is assigned to each section in PSL. To consider 3.1 Deterministic constraints
the construction practices, the beam and column groups in The deterministic constraints in the RBDO problem as for-
the specific lines are selected in the frame. Therefore, the mulated in Eq. (1) are related to beam moments and column
RBDO problem is solved by selecting the specific identi- axial forces. These constraints were fixed in a way that the
fication numbers from a PSL for each beam and column maximum moment at the midspan and each end of the jth
groups in the frame to optimize its weight considering the beam (Mjmax+, Mjmax–) do not exceed the moment strength
constraints. So, the number of dimensions (nd) of design at the midspan and each end of the jth beam (Mjr+, Mjr–),
solution in the GA will be equal to sum of the number of respectively; and as well the maximum compressive and
beam groups (nBG) and column groups (nCG). The main steps tensile axial forces of the kth column (Nkmax+, Nkmax–) do
of GA in this research can be listed as follows: not exceed the compressive and tensile axial strength of
- Step 1: Select the number of dimensions (nd), number the kth column (Nkr+, Nkr–), respectively. To compute the
of population (np), number of crossovers (ncr), number of M max+, M max–, N max+ and N max–, the finite element (FE) anal-
mutations (nmu), and number of iterations (niter). ysis of the RC frame is conducted using OpenSees soft-
- Step 2: Generate the initial population (individuals). ware under the code specified load combinations. On the
The number of positions in each individual (chromosome) other hand, to calculate the M r+, M r– , N r+ and N r–, a sim-
is equal to the number of dimensions of the problem (nd). plified approach is used based on the provisions specified
- Step 3: Evaluate the fitness of generated individuals. in Iranian design code for RC structures [25]. In this sim-
To handle the constraints, the constrained optimization plified approach, reinforcing bars is used only in two sides
problem in Eq. (1) is transformed into the unconstrained (top and bottom) of the RC section with symmetric layout.
problem using penalty function as Eq. (2) to penalize infea- Furthermore, the specified yielding strength of tensile ( f y)
sible design solutions by increasing their total weight (Wtot) and compressive ( f y') bars is assumed to be same. Also,
as objective function which decreases their fitness values. effect of the compressive bars in reduction of the compres-
nb nc sion zone in RC section is neglected. In the adopted limit
Wtot (S , κ ,Ψ ) = W (S ) + κ β .ψ β2 +∑ κ b .ψ bj2 + ∑ κ c .ψ ck2 , (2) state design method, ultimate compressive strain of the
j =1 k =1
concrete in extreme compressive fiber (εcu) is considered
 Asghshahr
|569
Period. Polytech. Civ. Eng., 65(2), pp. 566–576, 2021

equal to 0.0035. The selected RC sections for beams are Furthermore, if e is less than eb, the compressive axial
designed to yield the tensile bars but not to compressive strength of the column can be calculated as follows based
bars. Based on these assumptions, the absolute value of on the Whitney simplified assumptions (compression-con-
moment strength at the midspan (M r+) and each end (M r–) trolled region):
of rectangular RC beam will be same and can be calcu-
φc f c bh φs As f y
lated as follows: N r+ = + . (7)
2.87eh e
a + 1. 18 + 0 .5
M r ± = 0.85φc f c ab(d − ) + φs As' f s' (d − d ' ) , (3) d2 d − d'
2
where ϕc and ϕs are the strength reduction factors for the The obtained compressive axial strength of the col-
concrete and steel, respectively; fc is the specified com- umns based on the Eqs. (6) and (7) must be less than the
pressive strength of the concrete; d and d' are the distance maximum axial strength (Nrmax) according to [25]:
from the extreme compressive fiber to the tensile and com-
N r max = 0.85 α1φc f c ( Ag − Ast ) + φs f y Ast  , (8)
pressive bars, respectively; b is the section width; A's is the
total area of compressive reinforcement; the depth of the in which α1 = 0.85 − 0.0015 f c , Ag is the total area of the
compressive stress block (a) can be derived by solving the section and Ast is the total reinforcement area in a section.
following equation: In the case of tensile axial force and moment interac-
tion, the tensile axial strength of the column (N r–) can be
φs (700 As' − As f y ) φ (700 β1 As' d ' )
a2 + a− s =0, (4) calculated as follows:
φc (0.85 f c b) φc (0.85 f c b)
1 e
where As is the total area of tensile reinforcement; β1 is N r− = 1 / ( + ), (9)
No M o
equal to 0.97 – 0.0025 fc; and stress of the compressive bars
( f's) is evaluated as 700(a – β1d')/a. So, the penalty term in which No is equal to ϕsAstf y, and Mo is the pure bending
ψbj for the jth beam in Eq. (2) can be computed as follows: moment in tension-controlled region. So, the penalty term
ψck for the kth column in Eq. (2) can be computed as Eqs.
0 if max(M max j
+
, M max
j
−
) ≤ M rj ± (10) and (11) in the case of compressive and tensile axial

ψ bj =  max ( M max
j
+
, M max
j
−
) − M rj ± , (5) forces, respectively:
 otherwise
 M rj ±  N kmax + − N kr +
 if N kmax + > N kr +
max+ max– ψ ck = N kr + (10)
where max(M , |M |) denotes the maximum absolute
j j 0 if N max +
≤N r+
value of maximum moment at the midspan and each end  k k

of the jth beam.

 N kmax − − N kr −
In the case of compressive axial force and moment  if N kmax − > N kr −
interaction, if the axial load eccentricity (e) of the column ψ ck = N kr − (11)
is greater than the eccentricity in balanced condition (eb),  max − r−
0 if N k ≤ N k
the compressive axial strength of the column (N r+) can be
calculated as follows (tension-controlled region): It must be noted that the mentioned procedure for calcu-
 lating the axial strength of the columns has been validated
e' e' d' 
N r + = 0.85φc f c bd (1 − ) + (1 − ) 2 + 2 ρ m(1 − )  (6) by the P-M interaction diagram approach [26].
 d d d 

where ρ is the ratio of tension reinforcement equals to 3.2 Probabilistic constraint

As/bd; e' is equal to e + d – h/2; e is equal to M/P; M and The probabilistic constraints in the RBDO problem as for-
P are the applied moment and axial load on the col- mulated in Eq. (1) are related to reliability index of the
umn under the selected load combinations; eb is equal to frame. This constraint was considered to provide the min-
(0.2 + 0.77 ρ g m)h ; ρg is the ratio of total reinforcement in imum allowable reliability index for the frame. To com-
section; m is equal to φs f y / (0.85φc f c ) and h is the section pute the reliability index of the RC frame or correspond-
height. It should be noted that compressive force in con- ing probability of the failure, the finite element reliability
crete is approximately assumed as 0.85ϕc fc ab. analysis (FERA) of the frame is done. In this research, the
570|Asghshahr
Period. Polytech. Civ. Eng., 65(2), pp. 566–576, 2021

FORM method 2 [27] will be utilized that Newton-type transformation matrix into uncorrelated reduced space
recursive formula is used in this method to find the design including the eigenvectors of the correlation matrix [C']
point without need to solve the limit state function. as Eqs. (14) and (15):
The main steps of the FORM method 2 can be explained
 1 ρ X1 , X 2 ⋯ ρ X1 , X n 
as follows:  
 ρ X 2 , X1 1 … ρ X2 ,Xn 
- Step 1: Define the proper limit state function g( ). [ C ′] =  , (14)
- Step 2: Assume initial value of the design point X*. ⋮ ⋮ ⋮ 
 
- Step 3: Calculate the mean (μXN) and standard devia-  ρ X n , X1 ρ X n , X 2 … 1 
tion (σXN) at the design point of the equivalent normal dis-
tribution for non-normal random variables. The coordi- λ1(1) λ1( 2 ) … λ1( n ) 
 (1) 
nates of the design point in the equivalent standard normal λ λ2( 2 ) … λ2( n ) 
space (X' *) are: [T ] =  2 , (15)
⋮ ⋮ ⋮ 
 
X * − μ NX λn(1) λn( 2 ) … λn( n ) 
X ′∗ = . (12)
σ NX
where pXi, Xj is the correlation coefficient of the Xi and Xj
The FORM procedure can be used when all of the ran- random variables; {λ(i)} is the eigenvector of the ith mode;
dom variables are uncorrelated. If the random variables λ1(i), λ2(i), …, λn(i) are the components of the ith eigenvector.
are correlated, these variables should be transformed - Step 4: Calculate the partial derivative ∂g/∂X at the
into uncorrelated variables {Y}. Because the joint PDF design point X *. Finite difference method (FDM) is used
of basic random variables are not available in this study, to calculate the response sensitivities.
Rosenblatt transformation [28] cannot be used. On the - Step 5: Calculate the partial derivative ∂g/∂X' in the
other hand, as similar to requirements of current study, equivalent standard normal space as follows:
when the joint PDF of basic random variables is unknown
∂g ∂g ∂X ∂g
while the marginal PDFs and correlation matrix are = = σXN . (16)
∂X ′ ∂X ∂X ′ ∂X
known, the Nataf transformation [29] can be used to con-
vert correlated variables to uncorrelated random variables. - Step 6: Calculate the new values for the design point
It has been concluded that the major obstacle for Nataf in the equivalent standard normal space using the recur-
transformation is to evaluate the equivalent correlation sive formula as [32]:
matrix in standard normal space that should be computed 1
by solving nonlinear integral equations which requires X k′*+1 = 2
∇g ( X k′* )t X k′* − g ( X k′* )  ∇g ( X k′* ) , (17)
∇g ( X k′* )
tedious calculations [27, 30]. Therefore, to avoid these
complex calculations in the FORM analysis, a method where ∇g ( X k′* ) is the gradient vector of the limit state
proposed by Haldar and Mahedavan [27] has been used function at Xk' * corresponding to kth iteration. Note that t
in this study to modify the original limit state function represents the transpose of the vector.
expressed in terms of correlated variables into a function - Step 7: Calculate the reliability index of the frame (β)
of uncorrelated random variables. In this method, Eq. (13) as distance to the obtained design point from the origin:
is evaluated based on the obtained uncorrelated variables. n
This transformation can be performed according to fol- β= ∑(X ′ ) i
* 2
. (18)
i =1
lowing equation:
- Step 8: Calculate the new values for the design point
{ X } =  σ XN  [T ]{Y } + { μ XN } , (13)
in the original space as:
where {X} is vector of the correlated random variables;
X * = μ NX + σ NX X ′* . (19)
[σXN] is the diagonal matrix containing the equivalent nor-
mal standard deviations; {μXN} is vector of the equiva- Then, value of the limit sate function is computed for
lent normal means; σXN and μXN are calculated using two- this new design point. The aforementioned steps will be
parameter equivalent normal transformation suggested repeated until convergence criteria in steps (7) and (8)
by Rackwitz and Fiessler [31]; [T] is an orthogonal are satisfied. This procedure is developed in MATLAB
 Asghshahr
|571
Period. Polytech. Civ. Eng., 65(2), pp. 566–576, 2021

software calling OpenSees software to finite element anal-

ysis of the frame. So, the penalty term ψβ for the frame in
Eq. (2) can be computed as follows:
β − βt
 if β < β t
ψβ =  βt . (20)
0 if β ≥ β t

The mentioned method to evaluate the probabilis-
tic constraint in the RBDO problem is named reliability
index approach (RIA). Performance measure approach
(PMA) is identified as another method to examine the
probabilistic constraint. The main difference between
the RIA and the PMA is the type of optimization prob-
lem which is solved in each case. It was concluded that
PMA is inherently robust for RBDO and is more efficient
in evaluating the inactive probabilistic constraint. In con-
trast, RIA may diverged and yield singularity when uni-
Fig. 1 The selected RC frame for RBDO problem
form or Gumble random variables were employed, though
it is more efficient in evaluating the violated probabilis- Table 1 Prepared section list (PSL)
tic constraint [33–35]. One of the most important goals of Bars*
Section number Width (mm) Depth (mm)
this research is to monitor the deterministic and probabi- (top and bottom)

listic constraints during the optimization process, and to Beams

#1 300 300 3 Φ 16
determine which of these constraints is more critical in
#2 300 300 3 Φ 18
the RBDO problem. However, it is not possible to directly
monitor the reliability index during the optimization pro-
#16 400 400 5 Φ 22
cess in PMA approach, because during iterations the reli-
#17 400 400 4 Φ 25
ability index is kept fixed at β t [33]. Therefore, the RIA
Columns
method is used in this study. It will be discussed more
#1 300 300 3 Φ 16
about this issue later in Section 5.
#2 300 300 3 Φ 18

4 An illustrated example
#36 450 450 4 Φ 30
A five-story intermediate RC frame as illustrated in Fig. 1
#37 450 450 5 Φ 28
is utilized to implement the proposed framework. The story * Φ represents the bar diameter in mm
height and bay width of the frame are 3 and 4 m, respec-
tively. As presented in Fig. 1, four column group (C1, C2, random variables will be employed to determine the struc-
C3 and C4) and three beam group (B1, B2 and B3) are used ture response using the developed model in OpenSees
to consider the construction practices. (Please see Section 4.1).
Therefore based on the mentioned definitions, n, nb, The specified compressive strength of the concrete
nc, nBG, nCG and nd will be equal to 35, 15, 20, 3, 4 and 7, and yield strength of the steel bars are 25 and 400 MPa,
respectively. During the optimization process, the sections respectively. The concrete cover thickness is equal to
for structural elements are extracted from prepared sec- 50 mm. The strength reduction factors for the concrete (ϕc)
tion list (PSL) consists of finite number of prequalified and steel (ϕs) are considered 0.65 and 0.85, respectively.
sections. This list for beams (17 sections) and columns (37 Earthquake load (base shear) is calculated by using equiv-
sections) are listed in Table 1. alent static method and distributed as joint load in frame
Modulus of elasticity of the concrete (Ec), modulus of height. According to [36], the earthquake coefficient is
elasticity of the steel (Es), dead load (D), live load (L) and equal to 0.1554. In the probabilistic constraint as defined
earthquake load (E) are considered as random variables (X). in Eq. (1), the maximum inter-story drift (θmax) is employed
In deterministic constraints, the nominal value of these as engineering demand parameter. So, the limit state
572|Asghshahr
Period. Polytech. Civ. Eng., 65(2), pp. 566–576, 2021

function g ( ) = θ a − θ max is utilized in reliability analy- a Newton-Raphson solution algorithm is used to run the
sis of the frame in which θa is the code-defined allowable linear analysis. A plain constraints are utilized to handle
drift. According to [36], θa is equal to 0.0045 for interme- the boundary conditions. To provide the mapping between
diate RC frames. Table 2 represents the characteristics of the degrees of freedom at the nodes and the equation num-
the aforementioned random variables. bers, plain numberer is employed. In order to store and
The nominal value of the resistance parameters is kR solve the system of equations in the analysis, BandGeneral
standard deviation blow the mean value, and nominal system is used. Finally, to ensure the numerical conver-
value of single load parameters is kS standard deviation gence of the solution over the maximum 10 iterations, the
above the mean value. It must be noted that for all the ran- norm of displacement increment (NormDispIncr test) is
dom variables we have kR = kS = 5. The correlation coeffi- checked with a defined tolerance of 10 –8.
cient between dead load and earthquake load is considered
equal to 0.8. 5 Results
The parameters of the genetic algorithm must be selected
4.1 RC frame modelling according to the sensitivity analysis results. Also with a
A two-dimensional finite element model of the RC frame genetic algorithm, different results may be obtained based
is constructed using the OpenSees platform. All the base on the random seeds used. Therefore, 20 different runs
nodes of the foundation are modeled as an ideal fixed bear- are conducted for each selected set to examine the vari-
ing. The uniaxial Material Elastic is used to model the con- ability of the results with different random seeds. Results
crete and steel bar materials. To consider the reduced stiff- indicate the ability of the GA with np ≥ 30 to estimate the
ness of RC members due to cracking, the moment of inertia optimum value with negligible variability. Based on the
of the cross section for beams and columns is reduced by results, np = 30, ncr = 18 and nmu = 5 will be utilized in the
0.35 and 0.7, respectively. In order to consider the geomet- following sections. Also based on the convergence consid-
ric nonlinearities in the RC frame analysis, geomTransf erations, niter is set to 150.
P-Delta transformation has been utilized. Force-based ele- For comparison purposes, two levels of β t equal to 6 and
ments are the most common type of the element formula- 7 will be used in this example. Table 3 shows the identifi-
tions that are extensively adopted in the analysis of framed cation numbers of optimal design solutions for both cases
structures. So, a nonlinearBeamColumn element with five of β t equal to 6 and 7. Corresponding characteristics of the
integration points is used to model the RC columns and obtained optimal sections have been reported in Table 4.
beams whose cross sections are discretized into a number The convergence history of the best and mean weight
of fibers including cover concrete, steel bar, and core con- as an objective function in both cases has been shown
crete. Based on the sensitivity analysis to achieve suitable in Fig. 2. The difference between best and mean objec-
accuracy, number of fibers in the cover and core concrete tive values at initial generations is related to penalize the
patches in the two sides of the rectangular RC sections is infeasible individuals by the penalty function.
considered as equal to 10. The convergence is achieved after 53 and 68 iterations
A linear static analysis procedure was adopted in this for β t equals to 6 and 7, respectively. It is obvious that
study to calculate the structural response during the opti- optimized frame weight will increase as the minimum
mization process. Dead loads, live loads, and distributed required reliability index for frame increases. The opti-
lateral earthquake load in the frame height are applied to mized frame weights are 32394.66 and 36003.78 kg for β t
the structure with linear plain pattern. A load control with equals to 6 and 7, respectively.
In deterministic constraints, it is expected that the
Table 2 Characteristics of the random variables
maximum moment of beams (M max±) and maximum axial
Random Nominal Coefficient force of columns (N max±) will not be significantly affected
Distribution Unit
variable value of variation
Ec Lognormal 25 × 109 0.1 N/m 2 Table 3 Optimal design solutions

Es Lognormal 200 × 10 9
0.05 N/m 2 Column groups Beam groups
βt
D Lognormal 18000 0.15 N/m C1 C2 C3 C4 B1 B2 B3

L Lognormal 6000 0.4 N/m 6 #21 #18 #26 #19 #13 #10 #2

E Lognormal 220181 0.5 N 7 #31 #18 #32 #25 #14 #12 #2

 Asghshahr
|573
Period. Polytech. Civ. Eng., 65(2), pp. 566–576, 2021

Table 4 Optimal design sections

Bars*
Section number Width (mm) Depth (mm)
(top and bottom)
Beams
#2 300 300 3 Φ 18
#10 300 350 4 Φ 20
#12 350 350 4 Φ 22
#13 350 400 3 Φ 25
#14 350 400 4 Φ 22
Columns
#18 400 400 3 Φ 20
#19 400 400 4 Φ 18 (a)

#21 400 400 4 Φ 20

#25 400 400 4 Φ 25
#26 450 450 3 Φ 22
#31 450 450 5 Φ 22
#32 450 450 4 Φ 25
* Φ represents the bar diameter in mm

(b)
Fig. 3 The history of maximum moment of beams in different stories
(a) β = 6 and (b) β = 7
t t

by Mmax in Fig. 3. For the introduced RC frame, no col-

umn under the defined load combinations undergoes the
(a) tensile axial force. Therefore, in this case study, interac-
tion between the compressive axial force and moment has
been occurred for all of the columns. Fig. 4 depicts the
history of maximum compressive axial force of columns
in different stories for both values of β t. It should be noted
that these results belong to the members specified with red
color in Fig. 1. As shown in Figs. 3 and 4, the obtained
results confirm the predicted trends.
On the other hand, it is expected that section changes
during the optimization have a considerable effect on the
moment strength of beams (M r±) and axial strength of
(b)
columns (N r±). Fig. 5 illustrates the history of the max-
Fig. 2 The convergence history of (a) best and (b) mean objective
imum moment and moment strength of the beam in the
by section changes in the optimization process. However, third story for both values of β t. Also, the history of max-
these section changes have more effects on the maximum imum axial force and axial strength of the column in the
moment of beams than the maximum axial force of col- third story for both cases of β t has been shown in Fig. 6.
umns. Fig. 3 depicts the history of the maximum abso- As shown in Figs. 5 and 6, the obtained results confirm
lute value of maximum moment at the midspan and each the expected trends. Surely, the frame design will be con-
end of the beams max(Mjmax+, |Mjmax–|) in different stories servative for a higher value of the minimum required reli-
for both values of β t. For brevity, this parameter is shown ability index in the optimization process. As shown in
574|Asghshahr
Period. Polytech. Civ. Eng., 65(2), pp. 566–576, 2021

Fig. 5(b), the optimized maximum moment is significantly

less than the moment strength for β t = 7, but these two val-
ues are slightly different for β t = 6. A similar trend is seen
in Fig. 6(b) for the third story column as an instance.
In probabilistic constraints, the aim is to provide the
minimum required reliability index (β t) for the frame.
The history of reliability index (β) during the optimiza-
tion process is illustrated for both values of β t in Fig. 7.
It is worth to note that simultaneous satisfaction of deter-
ministic and probabilistic constraints in the case of β t = 6
is not possible for available discrete design solutions.
(a)
As shown in Fig. 7, results indicate that the optimized

(b)
Fig. 4 The history of maximum axial force of columns in different (a)
stories (a) βt = 6 and (b) βt = 7

(b)
(a) Fig. 6 The history of maximum axial force and axial strength of third
story column (a) βt = 6 and (b) βt = 7

(b)
Fig. 5 The history of maximum moment and moment strength of third
Fig. 7 The reliability index history for both values of βt
story beam (a) βt = 6 and (b) βt = 7
 Asghshahr
Period. Polytech. Civ. Eng., 65(2), pp. 566–576, 2021
|575

minimum reliability index for satisfying all of the con- RC frame according to Iranian design codes, the proposed
straints is equal to 6.09. On the other hand, for the case framework can also be applied to other types of RC frames
of β t = 7, the probabilistic constraint is more critical than utilizing other code provisions. The obtained high values
deterministic ones. Therefore, providing the β t = 7 will of reliability index for the frame demonstrate the conserva-
lead to satisfying the deterministic constraints as well. tive nature of code requirements for interstory drift limita-
In the introduced case study in this paper, all of the ran- tions based on the linear static analysis method. According
dom variables have a lognormal distribution and no sin- to the proposed method, the relationship between the objec-
gularity and divergence are happened during the optimi- tive function and the minimum required reliability index
zation process. Hence, RIA is a robust method for solving can be studied. For example, the results of present research
the introduced RBDO problem. In this case study, due to demonstrate that increasing the minimum reliability index
the obtained high values of reliability index, PMA is more from 6 to 7 resulted in an 11 % increase in the weight of the
efficient in evaluating the inactive probabilistic constraint. structure as an objective function. So, we can obtain a trade-
However, as a limitation of the current study, using differ- off between the frame weight and minimum required reli-
ent approaches for solving the RBDO problem of the RC ability index. It is worth to note that comparison of differ-
frame is not investigated in this paper. ent approaches to evaluate the probabilistic constraint is
not a main goal of the present research. This study can be
6 Conclusions extended in the future by following aspects:
This paper presents a new framework for reliability based • Using nonlinear analysis procedures to estimate the
design optimization (RBDO) of reinforced concrete (RC) realistic lateral capacity of RC frames considering
frames using a genetic algorithm (GA) and finite element the material and geometric nonlinearities.
reliability analysis (FERA). Results indicate the ability of • Investigating the effectiveness of the RIA and PMA
the proposed GA-FERA method to solve the RBDO prob- approaches to evaluate the probabilistic constraint in
lem of the studied RC frame. Although this method is the RBDO problem of RC frame in the cases of lin-
applied only to the optimization of the mid-rise intermediate ear and nonlinear analyses.

References
[1] Kaveh, A., Sabzi, O. "Optimal design of reinforced concrete frames [6] Lógó, J., Rad, M. M., Knabel, J., Tauzowski, P. "Reliability based
using big bang-big crunch algorithm", International Journal of Civil design of frames with limited residual strain energy capacity",
Engineering, 10(3), pp. 189–200, 2012. [online] Available at: http:// Periodica Polytechnica Civil Engineering, 55(1), pp. 13–20, 2011.
ijce.iust.ac.ir/browse.php?a_id=572 https://doi.org/10.3311/pp.ci.2011-1.02
[2] Esfandiari, M. J., Urgessa, G. S., Sheikholarefin, S., Deghan [7] Gholizadeh, S., Aligholizadeh, V., Mohammadi, M. "Neural net-
Manshadi, S. H. "Optimization of reinforced concrete frames sub- work-based reliability assessment of optimally seismic designed
jected to historical time-history loadings using DMPSO algorithm", moment frames", International Journal of Optimization in Civil
Structural and Multidisciplinary Optimization, 58, pp. 2119–2134, Engineering, 4(1), pp. 1–26, 2014. [online] Available at: http://ijoce.
2018. iust.ac.ir/article-1-159-en.pdf
https://doi.org/10.1007/s00158-018-2027-y [8] Grubišić, M., Ivošević, J., Grubišić, A. "Reliability Analysis of
[3] Shi, X., Xie, N., Fortune, K., Gong, J. "Durability of steel reinforced Reinforced Concrete Frame by Finite Element Method with Implicit
concrete in chloride environments: An overview", Construction and Limit State Functions", Buildings, 9(5), Article No. 119, 2019.
Building Materials, 30, pp.125–138, 2012. https://doi.org/10.3390/buildings9050119
https://doi.org/10.1016/j.conbuildmat.2011.12.038 [9] Aoues, Y., Chateauneuf, A. "Reliability-based optimization of struc-
[4] Zaman, K., Mahadevan, S. "Reliability-based design optimization of tural systems by adaptive target safety- Application to RC frames",
multidisciplinary system under aleatory and epistemic uncertainty", Structural Safety, 30(2), pp. 144–161, 2008.
Structural and Multidisciplinary Optimization, 55, pp. 681–699, https://doi.org/10.1016/j.strusafe.2006.10.002
2017. [10] Khatibinia, M., Salajegheh, E., Salajegheh, J., Fadaee, M. J.
https://doi.org/10.1007/s00158-016-1532-0 "Reliability-based design optimization of reinforced concrete struc-
[5] Nikolaidis, E., Burdisso, R. "Reliability based optimization: A safety tures including soil–structure interaction using a discrete gravita-
index approach", Computers and Structures, 28(6), pp. 781–788, tional search algorithm and a proposed metamodel", Engineering
1988. Optimization, 45(10), pp. 1147–1165, 2013.
https://doi.org/10.1016/0045-7949(88)90418-X https://doi.org/10.1080/0305215X.2012.725051
576|Asghshahr
Period. Polytech. Civ. Eng., 65(2), pp. 566–576, 2021

[11] Shayanfar, M. A., Barkhordari, M. A., Ghanooni-Bagha, M. [23] Movahedi Rad, M., Lógó, J. "Plastic behavior and stability con-
"Estimation of Corrosion Occurrence in RC Structure Using straints in the reliability based shakedown analysis and optimal
Reliability Based PSO Optimization", Periodica Polytechnica Civil design of skeletal structures", Asian Journal of Civil Engineering
Engineering, 59(4), pp. 531–542, 2015. (Building and Housing), 12(4), pp. 395–413, 2011. [online] Available
https://doi.org/10.3311/PPci.7588 at: https://www.sid.ir/En/Journal/ViewPaper.aspx?ID=247977
[12] Léger, N., Rizzian, L., Marchi, M. "Reliability-based design optimi- [24] Rad, M. M. "A Review of Elasto-Plastic Shakedown Analysis
zation of reinforced concrete structures with elastomeric isolators", with Limited Plastic Deformations and Displacements", Periodica
Procedia Engineering, 199, pp. 1193–1198, 2017. Polytechnica Civil Engineering, 62(3), pp. 812–817, 2018.
https://doi.org/10.1016/j.proeng.2017.09.216 https://doi.org/10.3311/PPci.11696
[13] Zou, X., Wang, Q., Wu, J. "Reliability-based performance design [25] MHUD "Iranian national building code for reinforced concrete
optimization for seismic retrofit of reinforced concrete buildings structures, Part 9", Ministry of Housing and Urban Development,
with fiber-reinforced polymer composites", Advances in Structural Tehran, Iran, 2013. [in Persian]
Engineering, 21(6), pp. 838–851, 2018. [26] Kwak, H.-G., Kim, J. "Optimum design of reinforced concrete plane
https://doi.org/10.1177/1369433217733760 frames based on predetermined section database", Computer-Aided
[14] Jenkins, W. M. "Structural optimization with the genetic Design, 40(3), pp. 396–408, 2008.
algorithm", The Structural Engineer, 69(24), pp. 418–422, https://doi.org/10.1016/j.cad.2007.11.009
1991. [online] Available at: https://www.istructe.org/jour- [27] Haldar, A., Mahadevan, S. "Probability, Reliability, and Statistical
n a l / vol u m e s / vol u m e - 69 - ( p u bl i s h e d -i n -19 91) / i s s u e -2 4 / Methods in Engineering Design", Wiley, New York, NY, USA,
structural-optimisation-with-the-genetic-algorithm/ 2000.
[15] Rajeev, S., Krishnamoorthy, C. S. "Discrete Optimization of [28] Rosenblatt, M. "Remarks on a Multivariate Transformation", Annals
Structures Using Genetic Algorithms", Journal of Structural of Mathematical Statistics, 23(3), pp. 470–472, 1952.
Engineering, 118(5), pp. 1233–1250, 1992. https://doi.org/10.1214/aoms/1177729394
https://doi.org/10.1061/(ASCE)0733-9445(1992)118:5(1233) [29] Nataf, A. "Determination des distribution donl les marges sont
[16] Baradaran, M. R., Madhkhan, M. "Application of an Improved donees", Comptes Rendus de l'Academie des Sciences, 225, pp.
Genetic Algorithm for Optimal Design of Planar Steel Frames", 42–43, 1962. (in French)
Periodica Polytechnica Civil Engineering, 63(1), pp.141–151, 2019. [30] Xiao, Q. "Evaluating correlation coefficient for Nataf transforma-
https://doi.org/10.3311/PPci.13039 tion", Probabilistic Engineering Mechanics, 37, pp. 1–6, 2014.
[17] Lee, C., Ahn, J. "Flexural Design of Reinforced Concrete Frames by https://doi.org/10.1016/j.probengmech.2014.03.010
Genetic Algorithm", Journal of Structural Engineering, 129(6), pp. [31] Rackwitz, R., Fiessler, B. "Note on discrete safety checking when
762–774, 2003. using non-normal stochastic models for basic variables", Load
https://doi.org/10.1061/(ASCE)0733-9445(2003)129:6(762) Project Working Session, MIT, Cambridge, MA, USA, 1976.
[18] Govindaraj, V., Ramasamy, J. V. "Optimum detailed design of [32] Rackwitz, R., Fiessler, B. "Structural reliability under combined
reinforced concrete frames using genetic algorithms", Engineering load sequence", Computers & Structures, 9(5), pp. 489–494, 1978.
Optimization, 39(4), pp. 471–494, 2007. https://doi.org/10.1016/0045-7949(78)90046-9
https://doi.org/10.1080/03052150601180767 [33] Tu, J., Choi, K. K., Park, Y. H. "A New Study on Reliability-Based
[19] Shahraki, A. F., Noorossana, R. "Reliability-based robust design Design Optimization", Journal of Mechanical Design, 121(4), pp.
optimization: A general methodology using genetic algorithm", 557–564, 1999.
Computers & Industrial Engineering, 74, pp. 199–207, 2014. https://doi.org/10.1115/1.2829499
https://doi.org/10.1016/j.cie.2014.05.013 [34] Youn, B. D., Choi, K. K., Park, Y. H. "Hybrid Analysis Method for
[20] Biabani Hamedani, K., Kalatjari, V. R. "Structural system reliabil- Reliability-Based Design Optimization", Journal of Mechanical
ity-based optimization of truss structures using genetic algorithm", Design, 125(2), pp. 221–232, 2003.
International Journal of Optimization in Civil Engineering, 8(4), https://doi.org/10.1115/1.1561042
pp. 565–586, 2018. Available at: http://ijoce.iust.ac.ir/article-1- [35] Lopez, R. H., Beck, A. T. "Reliability-based design optimization
363-en.html strategies based on FORM: a review", Journal of the Brazilian
[21] Shayanfar, M., Abbasnia, R., Khodam, A. "Development of a Society of Mechanical Sciences and Engineering, 34(4), pp. 506–
GA-based method for reliability-based optimization of structures 514, 2012.
with discrete and continuous design variables using OpenSees and https://doi.org/10.1590/S1678-58782012000400012
Tcl", Finite Element in Analysis and Design, 90, pp. 61–73, 2014. [36] BHRC "Iranian code of practice for seismic resistant design of
https://doi.org/10.1016/j.finel.2014.06.010 buildings, Standard no. 2800", Building and Housing Research
[22] Lógó, J., Vásárhelyi, A., Rad, M. M., Stocki, R. "Reliability based Center, Tehran, Iran, 2013. [in Persian]
limit analysis of steel frames with limited residual strain energy
capacity", Proceedings in Applied Mathematics and Mechanics,
Special Issue: 79th Annual Meeting of the International
Association of Applied Mathematics and Mechanics, 8(1), pp.
10041–10044, 2008.
https://doi.org/10.1002/pamm.200810041