Académique Documents
Professionnel Documents
Culture Documents
World J of Engineering and Pure and Applied Sci. 2012;2 2(3):98 SS! 22"9#0$82 ):98 al.' 2012. Safe&( of )ap )eng&*+ Predic&ion in ,einforced -oncre&e S&ruc&ure+ Sar%i e& al
Affilia&ion+:
1 <epar&5en&
of -i0il Engineering' Bacul&( of Engineering' A*5adu ;ello :ni0er+i&(' CA, A of -i0il Engineering' Bacul&( of Engineering' A*5adu ;ello :ni0er+i&(' CA, A of -i0il Engineering' Bacul&( of Engineering' A*5adu ;ello :ni0er+i&(' CA, A
2 <epar&5en&
3 <epar&5en&
Addre++ for -orre+pondence3 Adre++e pour la -orre+pondance: fa&ino(i200DE(a*oo.co.u% Accep&ed3Accep&F: 9arc*' 2011 -i&a&ion: Sar%i 7A' 9urana AA' A2e4ide S6. A++e++ing +afe&( in predic&ion+ of lap leng&*+ in reinforced concre&e +&ruc&ure+. World Journal of Engineering and Pure and Applied Science+ 2012;2(3):98#10G.
!=,6<:-= 6!
In the construction of reinforced concrete, due to the limitations in available length of bars and due to constraints in construction, there are numerous occasions when bars have to be joined, some of which are detailed and hence the essence of overlapping two bars over at least a minimum specified length called lap length [1, 2, 3]. This lap length varies depending on the bars si es as there are various bar si es and where the bars are lapped and!or which structural member or element the lapping occurs. "s a result of load transfer, the steel bars ma#be either in a$ial tension or a$ial compression [%, &]. 'le$ure, shear and torsion ma# occur as effects [(], but due to limitations of results and other factors, this paper would focus solel# on tension and compression and how the#
var# with different lap lengths and bar si es. )ence, the distribution of tensile stresses in the concrete normal to the a$is of the bars is relevant. The overlap on the other hand transfers or generates additional forces in the concrete which tend to push the bars apart, so concrete cover must be strong enough to overcome this *bursting force+ [2, ,]. This paper attempts to deal basicall# with the variables of the lap splice which include lap length, the head-si e and shape, and the bar spacing.
OPEN ACCESS
SS! 22"9#0$82 World J of Engineering and Pure and Applied Sci. 2012;2 2(3):99 ):99 Sar%i e& al.' al 2012. Safe&( of )ap )eng&*+ Predic&ion in ,einforced -oncre&e S&ruc&ure+
and reference documents and subse.uentl# adopted is given [2] in e.uation 1 below.
OPEN ACCESS
/ 2here / diameter of reinforcement. / specified #ield stress of the reinforcement. / specified compressive strength of masonr#. / re.uired splice length of reinforcement.
011
distributions are 7nown ver# precisel# as a result of a ver# long series of measurements. Det R be a variable representing the variations in strength between nominall# identical structures, whereas S represents the ma$imum load effects in successive T-#r periods, then the probabilit# that the structure will collapse during an# reference period of duration, T #ears, is given b# [A, 13]5 031 The reliabilit# of the structure is the probabilit# that it will survive when the load is applied, given b# [A, 13]5 0%1 2here, FR is the probabilit# distribution function of R and fs is the probabilit# densit# function of E. Fote that R and S are statisticall# independent and must necessaril# have the same dimensions.
/ lesser of the masonr# cover, clear spacing between adjacent reinforcement or & times / adherence factor 03.&, 1.3, 1.& 4 2.31. The minimum length of lap splices for reinforcing bars in tension or compression, , is calculated using e.uation 1 above but shall not be less than 333 mm 012 in1 The metric form of e.uation 1 above is given as5 / 021
6.uations 011 and 021 have been evaluated in a probabilistic setting to determine the safet# of their provisions when used in reinforced concrete. The safet# chec7ing is carried out as suggested b# [8]. This procedure is described in the ne$t section.
,elia2ili&( anal(+i+
9eliabilit# is defined [%] as the probabilit# of a performance function g0X1 greater than ero i.e. P:g0X1 ; 3<. In other words, reliabilit# is the probabilit# that the random variables =i / 0X1, >., Xn1 are in the safe region that is defined b# g0X1 ; 3. The probabilit# of failure is defined as the probabilit# P:g0X1 ? 3< .@r it is the probabilit# that the random variables Xi / 0X1, >>., Xn1 are in the failure region that is defined b# g0X1 ? 3. In a mathematical sense, structural reliabilit# can be defined as the probabilit# that a structure will attain each specified limit state 0ultimate or serviceabilit#1 during a specified reference period and set of conditions [A]. The idea of a Breference periodC comes into pla# because the majorit# of structural loads var# with times in an uncertain manner. )ence the probabilit# that an# selected load intensit# or criterion will be e$ceeded in a fi$ed interval of time is a function of the length of that interval. Thus, in general, structural reliabilit# is dependent on the time of e$posure to the loading environment.
OPEN ACCESS
OPEN ACCESS
World J of Engineering and Pure and Applied Sci. 2012;2 2(3):100 SS! 22"9#0$82 ):100 Sar%i e& al.' al 2012. Safe&( of )ap )eng&*+ Predic&ion in ,einforced -oncre&e S&ruc&ure+
081 The probabilit# of failure is given b# I0g0$1 ? 31 and therefore the reliabilit# inde$ Q can be related to probabilit# of failure b# the following e.uation / 1 K R 0) 01%1
Pf / I [M09, E1 L 3]
where M0$1 is the *limit state function+ and the probabilit# of failure is an identical with probabilit# of the limit state violation. 'or an# random variable X, the cumulative distribution function Fx0x1 is given b# 6.uation 0A1 provided that x N y.
It follows for the common, but special case where R and S are independent, the e$pression for the probabilit# of failure is Pf / P(R S 0) / 0131 6$pression 0131 is 7nown as the *convolution integral+ and FR0x1 is the probabilit# that R L x, or the probabilit# that the actual 9esistance R of the member is less than some value x. fs0x1 represents probabilit# that the load effect S acting in the member has a value x and x O Px in the limit as Px 3. Honsidering all possible value of $, total failure probabilit# is obtained as follows5 ()fs()dx
i.e. sum of all the cases of resistances for which the loads e$ceed the resistances.
Therefore, 0131 where is the joint probabilit# distribution function of x. The region of integration of the function g0$1 is stated as follows5 g0$1 ; 3 5 represents safet#J g0$1 / 3 5 represents attainment of the limit stateJ g0$1 ? 3 5 represents failure.
R s
2 R
+ s2 )
1/ 2
01(1
"ssume that the safet# margin S is linear in the basic variable =l,>, =n
m = C0 + C1 1 + ........+ Cn n
and
OPEN ACCESS
SS! 22"9#0$82 World J of Engineering and Pure and Applied Sci. 2012;2 2(3):101 ):101 Sar%i e& al.' al 2012. Safe&( of )ap )eng&*+ Predic&ion in ,einforced -oncre&e S&ruc&ure+
m = C + ..........+ C n n + Ci C j i j
2 2 2 1 1 2 2 i = j j =1 j i n n
OPEN ACCESS
Cov X i , X j
xi xj
[ ]
2here the failure surface is non-linear, calculation for is done using iterative procedures. " relationship can be drawn between the probabilit# of failure, If, and the reliabilit# inde$, . "ll methods are appro$imates and the problems become more difficult as the number of random variables and the comple$it# of the limit state functions increasesJ and when statistical dependence between random variables is present. 9eliabilit# anal#sis consists of the calculation that resistance will effectivel# e$ceed the applied load of a structure during its design life time. 2ith this reason, safet# becomes a measure of the probabilit# of survival. " limit state function g0$1, thus defines the resistance and load effect and various regions can be defined as failure of safet# which, is dependent on whether the g0$1 ? 3 or not.
01,1 can thus be computed. In the above, the safet# margin, S, is assumed linear in the basic variables. If on the other hand, it is non-linear in = / :=i,>, =n<, then appro$imate values for m and m can be obtained b# using a lineari ed safet# margin, S. Det
M = g ( x ) = g ( X 1 ,......., X n )
0181
Tsing Ta#lor series to e$pand about 0=i, >>., =n1 / 0i,>.., n1 and retaining onl# the linear terms, we obtain
M = g ( x ) = g ( 1 ,......., n ) +
01A1 2here,
g
xi
i =1
(X
i )
m g( 1 ,......, n )
0231
2 m
i =1 j =1
xi
xj Cov xi , x j
comple$it# of the limit state function increases and when statistical dependence between random variable is present. @f the two broad classes of methods of structural reliabilit# anal#sis 0level 2 and level 3 method of safet# chec7ing1, level 2 method shall be emplo#ed. Devel 2 is 7nown as second moment, 'irst @rder 9eliabilit# Sethod 0'@9S1. The random variables are defined as terms of means and variance and are considered to be normall# distributed. The measure of reliabilit# is based on the reliabilit# inde$. It involves use of certain iteration correlation procedure to obtain an appro$imation to the probabilit# of failure of a structure or structural s#stem. This generall# re.uires an ideali ation of failure domain and it is often associated with a simplified representation of the joint probabilit# distribution of a variable. The necessit# to have a method of reliabilit# anal#sis which is computationall# fast and efficient and which produces result with degree of accurac# prompted the use of this level 2 method.
0211 'rom the aboveJ the point of lineari ation is the so-called mean point 0I, >.., n1. )owever, a more reasonable point will be one on the failure surface. 6$perience shows that an e$pansion based on the mean point should not be used [8, A, 13]. Gecause of this objection, the reliabilit# inde$,, proposed [11], is particularl# useful. 9eliabilit# inde$, , is defined as the shortest distance from the origin to the failure surface in the normali ed Ucoordinate s#stem. The point where this occurs on the failure plane is called the design point [8, A, 13, 11]. 'irst of all, the set of basic variables is normali ed.
X i xi , Zi = i = 1,2,... n, xi
0221 where $i and $i are the mean and standard deviation of the random variable, =i, and the normali ed set, U / 0Ui, >, Un1 i / 3 and i / 1, i / 1, 2,>.,n The reliabilit# inde$ is thus formulated as below5 0231 Eince the basic variables are normall# distributed,
,ES:)=S
The stochastic models generated are anal#sed using the 'irst @rder 9eliabilit# Sethod as proposed [8], to give values for safet# inde$, Q, for the various diameter of reinforced concrete bars. "lgorithms developed into
= min Zi 2 z F i = 1
1/ 2
02%1
'@9T9"F modules were designed for failure modes in relation to some building codes, namel#, the Gritish [12, 13] and the current 6uropean [1%] codes. The diameter and compressive strength of masonr# were varied for all
OPEN ACCESS
OPEN ACCESS
World J of Engineering and Pure and Applied Sci. 2012;2 2(3):102 SS! 22"9#0$82 ):102 Sar%i e& al.' al 2012. Safe&( of )ap )eng&*+ Predic&ion in ,einforced -oncre&e S&ruc&ure+
the current 6urocode [1%] for design and construction of reinforced concrete members. @ne of the Gritish codes [13] has attempted a single value provision. )owever, the safet# intrinsic in the single value provisions in the 6urocode and Gritish code is not commensurate with the econom# in bar curtailment achieved. Thus, it is necessar# to prescribe a single value provision that will be safe while at the same time achieving the re.uired econom#, which will be applicable to all codes for the design and construction of reinforced concrete structures. This is a bid to reduce failure of structures which have occurred in recent times in man# parts of the worldJ hence the review of the current international codes of practices.
#ield stresses in the algorithms to get the various safet# levels. These safet# criteria are plotted against the various respective diameters of reinforced concrete bars. Eome of the results of the safet# values against their respective diameters are shown in figures 1 to 18.
-6!-):S 6!
There are separate formulations for bar lap lengths in compression and tension of reinforced concrete members in the Gritish codes, while onl# one provision is made in
OPEN ACCESS
SS! 22"9#0$82 World J of Engineering and Pure and Applied Sci. 2012;2 2(3):103 ):103 Sar%i e& al.' al 2012. Safe&( of )ap )eng&*+ Predic&ion in ,einforced -oncre&e S&ruc&ure+ Bigure 3:5 This figure
shows effect of Gar Ei e on Eafet#
OPEN ACCESS
OPEN ACCESS
OPEN ACCESS
World J of Engineering and Pure and Applied Sci. 2012;2 2(3):10" SS! 22"9#0$82 ):10" Sar%i e& al.' al 2012. Safe&( of )ap )eng&*+ Predic&ion in ,einforced -oncre&e S&ruc&ure+
shows effect of Gar Ei e on Eafet#
OPEN ACCESS
SS! 22"9#0$82 World J of Engineering and Pure and Applied Sci. 2012;2 2(3):10$ ):10$ Sar%i e& al.' al 2012. Safe&( of )ap )eng&*+ Predic&ion in ,einforced -oncre&e S&ruc&ure+ Bigure 18:5 This figure Bigure 1$:5 This figure
shows effect of Gar Ei e on Eafet# of Dap Dengths at f# / %(3 and V / 3.& 06H21 of Dap Dengths at f# / &33 and V / 2 06H21
OPEN ACCESS
,E-699E!<A= 6!
It is therefore suggested that for ever# diameter of
reinforced concrete bar, the lap length could be &3 times the diameter of the smallest reinforcement bar irrespective of the prevailing compressive or tensile strength and #ield stressJ that isJ Lap Length where in 50 / diameter of reinforcement bar. concrete designs and construction
"lso, this suggestion is applicable to lap lengths of bars reinforced notwithstanding the design code used.
,EBE,E!-ES
[1] "lling 6E. Eome Homment @n 'le$ural and "nchorage Gond Etresses, T.E. Wepartment of "griculture Eoil EerviceJ 6ngineering Wivision, Wesign Granch, Wesign FoteJ1A(8JFo.&,pp.1K%. [2] Thompson SX, Dedesma "D, Yersa Y@, Green Y6, Xlinger 96, "nchorage Gehavior of )eaded 9einforcement, Iart "5 Dap Eplices andIart G5 Wesign Irovisions and Eummar#J 2332Jpp.1-%&. [3] 6rico and Hagle# "ssociates. Sechanical vs Dap Eplices [%] [&] in 9einforced Honcrete. 9oc7ville 233&J151K(. Hanba#, 6, 'rosch, Y9., Wesign of Dap Epliced Gars5 Is Eimplification IossibleZ 233(Jpp. 1-,. Sosle# 2), Gunge# YH. 9einforced Honcrete Wesign. &th 6dition. Sacmillan Iress Dtd, Dondon. 1AA3. [(] Gilal E), "hmad "9, Xhaled "E. Gond Etrength of Tension Dap Eplices in )igh Etrength Honcrete Geams Etrengthened with Mlass 'iber 9einforcement. Y. Hompos. 'or Honstr. 233%J 851%-21.
OPEN ACCESS
OPEN ACCESS
[,]
World J of Engineering and Pure and Applied Sci. 2012;2 2(3):10G SS! 22"9#0$82 ):10G Sar%i e& al.' al 2012. Safe&( of )ap )eng&*+ Predic&ion in ,einforced -oncre&e S&ruc&ure+
and Honstruction. Gritish Etandards InstitutionJ Dondon. 1A,2. [13] Gritish Etandards Institution. The Etructural Tse of Honcrete, Iart 15 Hode of InstitutionJ Dondon. 1AA,. [1%] 6uropean Hommittee, Wesign of Honcrete Etructures, Iart 1.1. 6uropean Hommittee for Etandardi ationJ Grussels. 2338. [1&] "merican Honcrete Institute. Guilding Hode 9e.uirement 9einforced for Etructural Honcrete and Honcrete. "merican Honcrete Iractice for Wesign and Honstruction. Gritish Etandards for Etructural Honcrete and
"merican Honcrete Institute, Guilding Hode of 9e.uirements Hommentar#. "HI, Sichigan. 2332. p.32 K 3(.
[8]
Mollwitt er E, "bdo T, 9ac7wit 9. 'irst @rder 9eliabilit# Sethod,0'@9S1 Sanual, 9HI Mmbh, F#mphenburger Etr. 13%, S[FH)6FJ Merman#. 1A88.
[A]
Witlevsen @, Sadsen )@. Etructural 9eliabilit# Sethod. 'irst edition. Yohn 2ile# 4 Eons DtdJ Hhichester. 1AA(. 9etrieved from5 http5!!www.me7.dtu.d7!staff!od!boo7s.htm.
[13]
"##ub G, ScHuen 9). Irobabilit#, Etatistics, and 9eliabilit# for 6ngineers H9H Iress DDHJ Few \or7. 1AA,.
[11]
)asofer, "S, Dind, FH. 6$act and Invariant Eecond K moment Hode 'ormat. Yournal of 6ngineering Sechanics 1A,%J15111K21.
[12]
Gritish Etandards Institution. The Etructural Tse of Honcrete, Iart 15 Hode of Iractice for Wesign
-6!B) -= 6B !=E,ES=
Fo conflicts of interests were declared b# authors.
OPEN ACCESS