Jtam-A4.dvi JOURNAL OF THEORETICAL AND APPLIED MECHANICS 56, 4, pp. 1043-1053, Warsaw 2018 DOI: 10.15632/jtam-pl.56.4.1043 ASSESSMENT OF THE STRENGTH REDUCTION FACTOR IN PREDICTING THE FLEXURAL STRENGTH Sema Alacali, Guray Arslan Yildiz Technical University, Department of Civil Engineering, Istanbul, Turkey e-mail: semanoyal@gmail.com; aguray@yildiz.edu.tr In the design of flexural strength, the strength reduction factor φ decreases from tension- -controlled sections to compression-controlled sections to increase safety with decreasing ductility. This paper presents how to determine the reduction factor for flexural strength of reinforced concrete beams according to ACI code. In the reliability-based design, the reliable prediction of the flexural strength of reinforced concretemembers is assured by the use of reduction factors corresponding to different target reliability index β. In this study, for different β and coefficients of variation of the flexural strength parameters, the flexural strength reduction factor has been investigated by using experimental studies available in the literature. In the reliability analysis part of the study, the first-order second moment approach (FOSM)has beenused to determine the reduction factor. It has also been assumed that the random variables are statistically independent. Keywords: reinforced concrete, beam, flexure strength, reduction factor, target reliability 1. Introduction In the design of flexural strength, tension-controlled sections are desirable for their ductile be- havior for giving sufficient warning prior to failure. Hence, reinforced concrete (RC) elements are designed to behave in a ductile manner, whenever possible. This behavior can be ensured by limiting the amount of reinforcement such that tension reinforcement yields prior to concrete crushing. InACI 318 codes (1995, 1999, 2002, 2005, 2008, 2011, 2014), a lower strength reduction (φ) factor is used for compression-controlled sections compared to the one for tension-controlled sections because the compression-controlled sections are less ductile. Naaman (2004) noted that changes made from the ACI 318 (1999) to the ACI 318 (2002) codes relocated the limits for tension and compression controlled sections and added a transition region between the two. The flaw lies in this definition for these regional boundaries. In the codes, it is intended to provide the target failure probability by means of safety fac- tors that are load factors and strength reduction factors (Arslan et al., 2017). Safety factors depend on the selected target reliability index β, which is established in terms of the accepta- ble probability of failure varying with the considered loading condition, type of failure mode and material (Arslan et al., 2016). According to Du and Au (2005), the reliability indexes ba- sed on the requirements of the strength limit state for bridge girders are 3.9-4.4, 5.2-5.3 and 3.4-3.5 according toAASHTO(1998), theChineseCode (1991) and theHongKongCode (2002), respectively. Nowak et al. (2001) compared the reliability levels of prestressed concrete girders designed using Spanish Code (1998), Eurocode ENV 1991-3 (1994), and AASHTO (1998), and indicated that the reliability indexes varied considerably for the three codes. The reliability indexes for bridge girders were 7.0-8.0, 5.1-6.8 and 4.5-4.9 according to Eurocode ENV 1991-3 (1994), the SpanishCode (1998) andAASHTO (1998), respectively. In this study, the change in the strength reduction factor considered in predicting the flexural strength of tension-controlled 1044 S. Alacali, G. Arslan sections according toACI318 (2014) is investigated andcompared for different reliability indexes and coefficients of variation of the flexural strength parameters. 2. Design of RC beams for flexure According to ACI 318 (2014), the nominal flexural strength Mn of a beam section is computed from internal forces at the ultimate strain profile when the extreme compressive fiber strain is equal to 0.003. Sections in flexure exhibit differentmodes of failure depending on the strain level in the extreme tension reinforcement. According to Section 21.2 of ACI 318 (2014), thesemodes aredefinedas tension-controlled sections, compression-controlled sections anda transition region between the tension- and compression-controlled sections. Tension-controlled sections have the net tensile strain in the extreme tension steel either equal to or greater than 0.005. Compression- -controlled sections have the net tensile strain in the extreme tension reinforcement either equal to or less than the compression-controlled strain limit when the concrete in compression reaches the strain limit of 0.003. The compression-controlled strain limit is the net tensile strain in the reinforcement at balanced strain conditions. Compression-controlled sections have strains equal to or less than the yield strain,which is equal to 0.002 forGrade 420 reinforcement. There exists a transition region between the tension- and compression-controlled sections. Thenominalflexural strengthofa rectangular sectionwith tension reinforcement is computed fromthe internal force couple for tension failurebytheyieldingof the reinforcement.Thenominal flexural strength of the beamsMn can be calculated as Mn =Asfyd−0.59 A2sf 2 y bfc (2.1) inwhichAs is the area of the flexural reinforcement, fy is the yield strength of the reinforcement, fc is the compressive strength of concrete, d and b are the effective depth and beam width, respectively. The governing equation given by ACI 318 (2014) states that the reduced (design) strength φMn must exceed the ultimate (factored) moment Mu, and the safety criteria for flexural design of the RC beams can be defined as φMn ­Mu (2.2) in which φ is the strength reduction factor for flexure. According to ACI 318 (2014), the φ for an element depends on parameters such as the ductility and the importance of the element in terms of the reliability of the entire structure. For tension-controlled sections, aφ of 0.90 is used. Compression-controlled sections are defined as having strain limit at the nominal strength at or below the yield strain of the reinforcement. For compression-controlled sections, the φ is either 0.65 or 0.75 depending on the nature of the lateral confinement reinforcement. For sections with reinforcement strains between the aforementioned two limits, the strength reduction factor φ is determined by a linear interpolation between the value of φ for tension- and compression- -controlled sections. 3. Reliability analysis In reliability analysis, the main objective of engineering planning and design is to insure the performance of an engineering system.Under conditions of uncertainty, the assurance of the per- formance is possiblewith the use of safety factors. The reliability assessment requires knowledge of the performance function to define the safety factors (Ang andTang, 1984). The performance Assessment of the strength reduction factor... 1045 function,Z = g(X1,X2, . . . ,Xn), can be determined in terms of many random variables as load components, resistance parameters, material properties. In this equation, Xi are basic random variables influencing the limit state. The failure surface can be defined as Z =0. The safety or reliability is defined by Z > 0, and the failure state is Z < 0. In the reliability based design, the problem is to determine the partial safety factors of the variables according to the target reliability index β. In this study, the first-order second moment approach (FOSM) is used and the design points γimXi corresponding to the target reliability indexβ are obtained. In the space of reduced variates, β being ameasure of reliability is defined as the shortest distance from the failure surface to the origin. The limit state function can be defined with Eq. (3.1) by multiplying the safety factor γi with each of the basic design variables g(γ1mX1,γ2mX2, . . . ,γimXi)= 0 i=1,2, . . . ,n (3.1) x∗i(= γimXi) is the most probable failure point on the failure surface, and the determination of x∗i requires an iterative solution. In the space of reduced variates, the most probable failure point is x ′ ∗ i =−α ∗ iβ. The sensitivity coefficient α ∗ i is defined by α∗i = ∂g ∂X′i / √ √ √ √ n ∑ i=1 ( ∂g ∂X′i )2 ∗ (3.2) The partial safety factors required for the given β are defined as γi(=x ∗ i/mXi). The original variates are given by x∗i =mXi(1−α ∗ iβVXi), in whichmXi and VXi are themean value and the variance coefficient of the original variableXi with normal distribution, respectively. VXi is the ratio of standard deviation σXi to themean valuemXi. The partial safety factors are calculated as (Nowak and Collins, 2000) γi =1−α ∗ iβVXi (3.3) In this study, the distributions of random variables in the performance function are given in Table 1. In lognormal and extreme type I distributions, mXi and σXi are replaced by the equivalent normal mean mNXi and standard deviation σ N Xi . In addition, it is also assumed that the random variables are statistically independent. 3.1. Establishment of performance function According to ACI 318 (2014), the strength reduction factor for flexure ranges from 0.70 to 0.90 depending on the nature of the lateral confinement reinforcement and the strain level in the extreme tension reinforcement. The reduction factors for RC beams have been investigated by considering the reliability indexes β (5.2, 4.75, 4.27, 3.72, 3.5, 3.09 and 2.33) corresponding to various failure probabilities pF (10 −7, 10−6, 10−5, 10−4, 2.33 · 10−4, 10−3 and 10−2). The performance function used in the calculations is given by g(X) = γ1Mn−γ2Mu (3.4) in which Mu is the ultimate (factored) moment at the RC beam section that can be taken as the test result andMn is the nominal flexural strength of the beam defined in ACI 318 (2014). γ1 and γ2 are the strength reduction factors for the corresponding variables. 1046 S. Alacali, G. Arslan 3.2. Coefficients of variation of design parameters The ultimate (factored) and nominal flexural strength of the beams obtained through expe- riments and equation have been modeled as random variables to perform a probability-based analysis. In modeling of those parameters as random variables, the values of coefficients of va- riations have been determined based on the studies available in the literature and codes. They are summarized in Table 1. In the literature review (Table 1), it has been observed that the coefficient of variation of the concrete compressive strength Vfc varies between 0.10 and 0.21, depending on the construction quality (Arslan et al., 2015). By taking advantage of studies in the literature and codes, it is assumed that Vfc is 0.05, 0.10 and 0.15, respectively, in this study. Table 1.Coefficients of variation of the variables Cases Coefficients of variation fy fc As b d Mu Case 1 0.05 0.04 0.03 0.03 0.04 Case 2 0.03 0.10 Case 3 0.15 Case 4 0.05 Case 5 0.05 0.10 Case 6 0.15 Case 7 0.05 Case 8 0.07 0.10 Case 9 0.15 Case 10 0.05 Case 11 0.10 0.10 Case 12 0.15 Distribution Log- Log- Normal Normal Normal Extreme type -normal -normal type I The coefficient of variation of the reinforcement yield strength Vfy has also been reported by many researchers, and Vfy ranges from 0.05 to 0.15 (Arslan et al., 2016). Vfy was taken as 0.03 byNowak et al. (2005), 0.05 by JCSS (2000), 0.06 by Soares et al. (2002), 0.07 byAkiyama et al. (2012), 0.08 by Val et al. (1997), Hosseinnezhad et al. (2000) and Low and Hao (2001), 0.08-0.11 by Ostlund (1991), 0.12 by Enright and Frangopol (1998), 0.15 by Mirza (1996). In the present study, model variations of fy are taken as 0.03, 0.05, 0.07 and 0.10, respectively. The coefficients of variation of the effective depth Vd, width Vb and tensile reinforcement area VAs of beams have also been reported by many researchers. Vd was taken as 0.02 by Lu et al. (1994), 0.03 by Wieghaus and Atadero (2011), 0.04 by Nowak and Szerszen (2003) and Szerszen et al. (2005). Vb was taken as 0.04 by Nowak and Szerszen (2003) and Szerszen et al. (2005). It is assumed that the Vd, Vb and VAs are 0.03, 0.03 and 0.04, respectively, in this study. To carry out the reliability analysis of RC beam specimens, a meaningful probability di- stribution for the nominal flexural strength parameters and ultimate flexural strength is also necessary. In the present study, randomness of the applied load is described using Extreme ty- pe I distribution. In the studies by Hognestad (1951) and Mirza (1996), it was assumed that the coefficient of variation of strength due to test procedure was 0.04, which is the value used in this study. 3.3. Properties of beams In the determination of the flexural strength reduction factors, 84 beamswith flexural failure collected from 3 different researches (Johnson and Cox, 1939; Ashour, 2000; Pam et al., 2001) Assessment of the strength reduction factor... 1047 have been evaluated. The number of beams produced fromnormal-strength concrete (NSC) and high-strength concrete (HSC) with fc ­ 55MPa are 52 and 32, respectively. The beams have a broad range of design parameters: 22.0¬ fc ¬ 48.6MPa, 0.17¬ ρ¬ 2.37%, 200¬ b¬ 305mm and 215 ¬ d ¬ 305mm for NSC beams and 57.1 ¬ fc ¬ 107.1MPa, 1.03 ¬ ρ ¬ 4.04%, 120¬ b¬ 200mm and 208¬ d¬ 260mm for HSC beams. 4. Investigating the strength reduction TheACI 318 code imposes aφ factor of 0.65when the strain in the tension reinforcement equals 0.002 for Grade 420 reinforcement. The φ increases linearly to the maximum value of 0.90 as the tension strain increases from 0.002 to 0.005. A tension-controlled section is defined as a cross section in which the tensile strain in the extreme tension reinforcement at the nominal strength is greater than or equal to 0.005. Tension-controlled sections are desirable for their ductile behavior, which allows redistribution of the stresses and sufficient warning against an imminent failure. It is always a good practice to design RC elements to behave in a ductile manner, whenever possible. For tension-controlled sections, a φ factor of 0.9 has been used. In the design of RC beams, to apply a higher resistance factor φ of 0.9, the member should exhibit desirable behavior. In this study, φ factors of the ACI 318 code are investigated for tension-controlled beam sections. For different Vfc and Vfy, the value of φ corresponding to β (2.33, 3.09, 3.50, 3.72, 4.27, 4.75 and 5.20) and different Vfc and Vfy are summarized for NSC, HSC and all beams (NSC and HSC) in Table 2. For a given β and different Vfc and Vfy, the value of φ for HSC beams is found to be smaller than the one for NSC beams, so it can be inferred that φ for NSC beams is more safe than that for HSC beams. Saatcioglu (2014) indicated that theACI 318 (2005) adopted strength reduction factors that were compatible with ASCE7-02 (2002) load combinations, except for the tension controlled section for which the φwas increased from 0.80 to 0.90. In this study, it is founded that 0.80 value of φ corresponds to the target values of β=3.5, Vfc =0.05 and Vfy =0.10 in all analyzed beams. In ACI 318 (2014), φ considered in predicting flexural strength of beams is updated as 0.90, which corresponds to the target values of β=3.5, Vfy =0.05 and Vfc =0.05, in all analyzed beams. It is observed that this value is conservative for β in the range from 2.33 to 5.20 for Vfy = 0.05 and Vfc ¬ 0.15 in NSC beams, and it can also be noted that it is conservative for β in the range from 2.33 to 5.20 for Vfy = 0.03 and Vfc ¬ 0.15 in HSC beams. The values of φ obtained from the analyses which have been performed by considering diffe- rent Vfc (0.05, 0.10 and 0.15), Vfy (0.03, 0.05, 0.07 and 0.10) and β (5.2, 4.75, 4.27, 3.72, 3.50, 3.09 and 2.33) values of the beam sections are shown in Fig. 1. Fig. 1. Effect of variation in the β on φ; (a) NSC, (b) HSC 1048 S. Alacali, G. Arslan Table 2.Average φ for different values of COV and β values Beams Coefficients of β variation 2.33 3.09 3.50 3.72 4.27 4.75 5.20 NSC Vfy =0.03 Vfc =0.05 0.934 0.929 0.927 0.926 0.925 0.925 0.924 Vfc =0.10 0.932 0.927 0.926 0.925 0.924 0.923 0.923 Vfc =0.15 0.930 0.925 0.923 0.922 0.921 0.921 0.920 Vfy =0.05 Vfc =0.05 0.911 0.903 0.901 0.900 0.899 0.898 0.897 Vfc =0.10 0.910 0.902 0.900 0.899 0.898 0.897 0.896 Vfc =0.15 0.908 0.900 0.898 0.897 0.895 0.894 0.894 Vfy =0.07 Vfc =0.05 0.880 0.869 0.866 0.865 0.862 0.861 0.860 Vfc =0.10 0.880 0.868 0.865 0.864 0.862 0.860 0.859 Vfc =0.15 0.878 0.867 0.863 0.862 0.860 0.859 0.858 Vfy =0.10 Vfc =0.05 0.828 0.808 0.803 0.801 0.797 0.795 0.794 Vfc =0.10 0.827 0.807 0.802 0.800 0.797 0.795 0.793 Vfc =0.15 0.826 0.806 0.801 0.799 0.796 0.794 0.792 HSC Vfy =0.03 Vfc =0.05 0.921 0.913 0.911 0.910 0.909 0.908 0.907 Vfc =0.10 0.920 0.912 0.910 0.909 0.907 0.906 0.906 Vfc =0.15 0.917 0.909 0.907 0.906 0.904 0.903 0.902 Vfy =0.05 Vfc =0.05 0.896 0.885 0.881 0.880 0.878 0.877 0.876 Vfc =0.10 0.895 0.883 0.880 0.879 0.877 0.875 0.874 Vfc =0.15 0.893 0.881 0.878 0.877 0.874 0.873 0.872 Vfy =0.07 Vfc =0.05 0.864 0.847 0.842 0.840 0.837 0.835 0.834 Vfc =0.10 0.863 0.846 0.841 0.839 0.836 0.834 0.833 Vfc =0.15 0.862 0.844 0.839 0.837 0.834 0.832 0.831 Vfy =0.10 Vfc =0.05 0.812 0.782 0.773 0.770 0.764 0.762 0.760 Vfc =0.10 0.811 0.781 0.772 0.769 0.764 0.761 0.759 Vfc =0.15 0.810 0.780 0.771 0.768 0.763 0.760 0.758 NSC + HSC Vfy =0.03 Vfc =0.05 0.929 0.923 0.921 0.920 0.919 0.918 0.918 Vfc =0.10 0.928 0.921 0.920 0.919 0.918 0.917 0.916 Vfc =0.15 0.925 0.919 0.917 0.916 0.915 0.914 0.913 Vfy =0.05 Vfc =0.05 0.905 0.896 0.894 0.893 0.891 0.890 0.889 Vfc =0.10 0.904 0.895 0.892 0.891 0.890 0.889 0.888 Vfc =0.15 0.902 0.893 0.890 0.889 0.887 0.886 0.885 Vfy =0.07 Vfc =0.05 0.874 0.860 0.857 0.855 0.853 0.851 0.850 Vfc =0.10 0.873 0.860 0.856 0.854 0.852 0.850 0.849 Vfc =0.15 0.872 0.858 0.854 0.853 0.850 0.849 0.848 Vfy =0.10 Vfc =0.05 0.822 0.798 0.791 0.789 0.785 0.782 0.781 Vfc =0.10 0.821 0.797 0.791 0.788 0.784 0.782 0.780 Vfc =0.15 0.820 0.796 0.790 0.787 0.783 0.781 0.779 It is seen that φ decreases with an increase in the value of Vfy. The rate of increasing in the value of φ for low values of β is higher than that for high values of β. When β becomes higher, the variation of φ versus β almost becomes a smooth curve for NSC andHSC beams. For given Vfy and β, φ for HSC beams are found to be smaller than the one for NSC beams, so it can be inferred that φ for NSC beams is more safe than that for HSC beams. For the same Vfy Vb, Vd, VAs andβ values, it can also be said thatφ values forNSC,HSCand all beams (NSC andHSC) are very close to each other for different Vfc. For some experimental beams, the effects of variations of the tensile strain in the tension reinforcement εs, the compressive strength of concrete fc, the ratio of tensile strain to yield Assessment of the strength reduction factor... 1049 strain in the tension reinforcementεs/εy, the ratio of percentage of tension reinforcement to the percentage of balanced reinforcement ρ/ρb, the ratio of neutral axis depth to the effective depthx/d, andeffective depthof thebeamdon theφareplotted inFig. 2 forβ=3.5,Vfy =0.05, Vfc =0.05, Vb =0.03, Vd =0.03 and VAs =0.04. Fig. 2. Effect of variation in εs, fc, εs/εy, ρ/ρb, x/d and d on φ The relationship of φ and εs at the nominal strength for the analyzed beams is shown in Fig. 2a.According toACI 318, if the reinforcement strain at the nominal strength is greater than 0.005, φ equals to 0.90 for the desirable behavior of beam sections. 20% of the tests (4 for NSC and12 forHSCof 84 tests) delivered relatively low εs values εs ¬ 0.005,where the corresponding strength reduction factors are mostly less than 0.90 for β = 3.5, Vfy = 0.05 and Vfc = 0.05. It is observed that the φ factor increases with εs for NSC and HSC beams. Based on the results of analyses, ACI 318 provisions are non-conservative for εs ¬ 0.02. The φ factor for the existing test data yields a large scatter in the results, especially for beams with εs ¬ 0.02. Figure 2b shows the φ-fc for the analyzed beams. Based on the studies of the stress-strain behavior of NSC and HSC, it is shown that concrete becomes increasingly more brittle as its compressive strength is increased. Despite HSC being a more brittle material compared with NSC, the x/d values of HSC sections are smaller than tkose of the NSC sections for a given ρ. 1050 S. Alacali, G. Arslan Hence,HSCflexuralmembers exhibit greater ductility owing to lowerneutral axis depths (Arslan andCihanlı, 2010). Based on the results of analyses, ACI 318 provisions are non-conservative for HSC flexural beams. The φ factor for the existing test data yields a large scatter in the results, especially for HSC beams with fc > 75MPa. The φ-εs/εy for the analyzed beams are shown in Fig. 2c. According to ACI 318, if εs is at least 2.5 times the yield strain (εy ∼=0.002= fy/Es), then the maximum value of φ=0.90 can be used. 32% of the tests (10 for NSC and 17 for HSC of 84 tests) delivered relatively low εs/εy values (εs/εy ¬ 5), where the corresponding strength reduction factors aremostly less than 0.90 for β=3.5, Vfy =0.05 and Vfc =0.05. It is observed that the φ factor increases with εs/εy for the beams. Theφ factor for the existing test data yields a large scatter in the results, especially for HSC beams with εs/εy ¬ 5. The effect of ρ/ρb on φ is illustrated in Fig. 2d. The ACI 318 (1999) and previous codes limit the tension reinforcement ratio ρ to no more than 75% of the ratio (0.75ρb) that would produce balanced strain conditions. The ACI 318 (2002) limits the net tensile strain εt of the extreme tension steel at the nominal strength to benot less than 0.004.Meanwhile, when the net tensile strain in the extreme tension steel is sufficiently large (equal to or greater than 0.005), the section is defined as tension-controlled where ample warning of failure with excessive deflection and crackingmay be expected. The effect of this limitation is to restrict ρ inRCbeams to about the same ratio as in editions of the code prior to 2002. 69% of the tests (26 for NSC and 32 for HSC of 84 tests) delivered relatively high ρ/ρb values (ρ/ρb ­ 0.25), where the corresponding strength reduction factors aremostly less than 0.90 for β=3.5, Vfy =0.05 and Vfc =0.05. The results of the φ factor of beams with ρ/ρb < 0.25 are limited for all the beams (6 for NSC of 84 tests). Figure 2e shows the φ-x/d for the analyzed beams. The design codes BS8110, EC 2 and GBJ 11 limit the neutral axis depthx to nomore than a certain fraction of the effective depth d. It can be noted that in the design of beams, using the simplified stress block BS 8110 (1997) limitsx to 0.5d for all concretewith fcu ¬ 100MPa to ensure that the section is under-reinforced and the strain in the longitudinal reinforcement is not less than 0.0035. EC 2-1 (1992) limits the x to no more than 0.45d when fcu < 50MPa or 0.35d when fcu ­ 50MPa. GBJ 11 (1989) requires x to be smaller than 0.35d for all concrete grades. The values of φ decrease significantly as x/d increases from 0.2 to 0.5. The corresponding φ of HSC beams are smaller than 0.90 for β = 3.5, Vfy = 0.05, Vfc = 0.05, Vb = 0.03, Vd = 0.03 and VAs = 0.04. Based on the results of analyses, the φ factor for x/d > 0.30 is non-conservative for 22 flexural beams (5 for NSC and 17 for HSC of 84 tests). The φ factor for the existing test data yields a large scatter in the results, especially for x/d­ 0.20. The φ-d for the analyzed beams are shown in Fig. 2f. 6% of the NSC beam tests (3 of 52 tests) have been conducted for d< 250mm and only 9% of the HSC beam tests (3 of 32 tests) have been conducted for d ­ 250mm. The φ factor for the existing test data yields a large scatter in the results, especially for HSC beams with d< 250mm. 5. Conclusion The change in the strength reduction factor for flexure according to the ACI 318 is investigated for different coefficients of variation and β values. The following conclusions can be drawn from the results of this study. • It is found that φ of 0.90, which is a value recommended by the ACI 318 (2002) and ACI 318 (2011), corresponds to the target values of β=3.5, Vfy =0.03 and Vfc =0.05 in all analyzed beams. It is observed that this value is conservative for β in the range from 2.33 to 5.20 for Vfy =0.05 and Vfc ¬ 0.15 in NSC beams, and it can also be noted that it Assessment of the strength reduction factor... 1051 is conservative for β in the range from 2.33 to 5.20 for Vfy =0.03 and Vfc ¬ 0.15 in HSC beams. • For the given β = 3.5, Vfy = 0.05, Vfc = 0.05, Vb = 0.03, Vd = 0.03 and VAs = 0.04, φ for the HSC beams are found to be smaller than those for the NSC beams, so it can be inferred thatφ for theHSCbeams ismore non-conservative than that for theNSCbeams. • According to ACI 318, if εs is at least 2.5 times the yield strain (εy ∼= 0.002 = fy/Es), then themaximum value of φ=0.90 can be used. 32% of the tests (10 for NSC and 17 for HSCof 84 tests) delivered relatively low εs/εy values (εs/εy ¬ 5), where the corresponding strength reduction factors aremostly less than 0.90 forβ=3.5, Vfy =0.05 andVfc =0.05. It is observed that theφ factor increaseswith εs/εy for beams.Theφ factor for the existing test data yields a large scatter in the results, especially for HSC beams with εs/εy ¬ 5. • Thevalues ofφdecrease significantly asx/d increases from0.2 to 0.5. The correspondingφ of HSC beams are smaller than 0.90 for β = 3.5, Vfy = 0.05, Vfc = 0.05, Vb = 0.03, Vd =0.03 and VAs =0.04. Based on the results of analyses, the φ factor for x/d> 0.30 is non-conservative for 22 flexural beams (5 forNSCand17 forHSCof 84 tests).Theφ factor for the existing test data yields a large scatter in the results, especially for x/d­ 0.20. In order tomake amore reliable evaluation, the determination of the reduction factor for fle- xural strength ofRCbeams for a greater number of beamswith differentmaterial and geometric properties should be realized. References 1. AASHTOLRFD: BridgeDesign Specifications, 1998, AmericanAssociation of State Highway and TransportationOfficials,Washington, DC 2. ACI, 1995, ACI 318M-95:Building Code Requirements for Structural Concrete and Commentary, ACI, Farmington Hills, MI, USA 3. ACI, 1999, ACI 318M-99:Building Code Requirements for Structural Concrete and Commentary, ACI, Farmington Hills, MI, USA 4. ACI, 2002, ACI 318R-02: Building Code Requirements for Structural Concrete and Commentary, ACI, Farmington Hills, MI, USA 5. ACI, 2005, ACI 318R-05: Building Code Requirements for Structural Concrete and Commentary, ACI, Farmington Hills, MI, USA 6. ACI, 2008, ACI 318R-08: Building Code Requirements for Structural Concrete and Commentary, ACI, Farmington Hills, MI, USA 7. ACI, 2011, ACI 318R-11: Building Code Requirements for Structural Concrete and Commentary, ACI, Farmington Hills, MI, USA 8. ACI, 2014, ACI 318-14: Building Code Requirements for Structural Concrete and Commentary, ACI, Farmington Hills, MI, USA 9. Akiyama M., Matsuzaki H., Dang H.T., Suzuki M., 2012, Reliability-based capacity design for reinforced concrete bridge structures, Structure and Infrastructure Engineering, Maintenance, Management, Life-Cycle Design and Performance, 8, 12, 1096-1107 10. AngA.H.S.,TangW.H., 1984,Probability Concepts inEngineeringPlanning andDesign.Vol. II, Decision, Risk, and Reliability, Wiley, NewYork, NY, USA 11. Arslan G., Alacali S., Sagiroglu A., 2016a, Assessing reduction in concrete shear strength contribution, Proceedings of the Institution of Civil Engineers, Structures and Building, 169, 4, 237-244 1052 S. Alacali, G. Arslan 12. Arslan G., Alacalı S.N., Sagiroglu A., 2016, The investigation of the strength reduction factor in predicting the shear strength, Journal of Theoretical and Applied Mechanics, 53, 2, 371-381 13. ArslanG., Alacali S.N., SagirogluA., 2017,Determining the reduction factor in predicting the contributionof concrete to shear strengthbyusingaprobabilisticmethod, International Journal of Civil Engineering (IJCE) Transaction A: Civil Engineering, in reviewer 14. ArslanG., Cihanlı E., 2010,Curvature ductility prediction of reinforced high-strength concrete beam sections, Journal of Civil Engineering and Management (JCEM), 16, 4, 462-470 15. Ashour S.A., 2000, Effect of compressive strength and tensile reinforcement ratio on flexural behavior of high-strength concrete beams,Engineering Structures, 22, 5, 413-423 16. British Standards Institution – Part 1, 1997, Structural Use of Concrete: Code of Practice for Design and Construction, BSI, London, BS 8110 17. Chinese Design Code for Highway Bridges-Beijing, 1991, People’s Communication Press 18. Du J.S., Au F.T.K., 2005, Deterministic and reliability analysis of prestressed concrete bridge girders: comparison of theChinese, HongKong andAASHTOLRFDCodes;Structural Safety, 27, 230-245 19. Enright, M.P., Frangopol, D.M., 1998, Probabilistic analysis of resistance degradation of reinforced concrete bridge beams under corrosion,Engineering Structures, 20, 960-971 20. EurocodeENV1991-3.Eurocode1, 1994,Basis ofDesign andActions onStructures.Part 3,Traffic Loads on Bridges, Final draft, August 21. European Committee for Standardisation, Design of Concrete Structures, Part 1, 1992, General Rules and Rules for Buildings, European Committee for Standardisation, Brussels, EC 2 22. HognestadE., 1951,A studyof combinedbending andaxial load in reinforced concretemembers, Engineering Experiment Station Bulletin, 399, University of Illinois, Urbana, IL, USA 23. Hosseinnezhad A., Pourzeynali S., Razzaghi J., 2000, Aplication of first-order second mo- ment level 2 reliability analysis of presstressedconcrete bridges,7th International Congress onCivil Engineering 24. JCSS 2000,Probabilistic model code – Part III, Joint Committee on Structural Safety 25. Johnson B., Cox K.C., 1939, High yield-point steel as tension reinforcement in beams, AC1 Journal Proceedings, 36, 1, 65-80 26. LowH.Y.,HaoH., 2001,Reliability analysis of reinforced concrete slabs under explosive loading, Structural Safety, 23, 2, 157-178 27. Lu R.H., Luo Y.H., Conte J.P., 1994, Reliability evaluation of reinforced concrete beams, Structural Safety, 14, 4, 277-298 28. Minimum Design Loads for Buildings and Other Structures (SEI/ASCE 7-02), 2002, American Society of Civil Engineers, http://dx.doi.org/10.1061/9780784406243. 29. Mirza S.A., 1996, Reliability-based design of reinforced concrete columns, Structural Safety, 18. 2/3, 179-194 30. Naaman A.E., 2004, Limits of reinforcement in 2002 ACI code, transition, flaws, and solution, ACI Structural Journal, 101, 2, 209-218 31. National Standard of the People’s Republic of China, 1989,Code for Seismic Design of Buildings, GBJ11-89 (in Chinese) 32. Nowak A.S., Collins K.R., 2000,Reliability of Structures, McGrawHill, Boston,MA, USA 33. Nowak A.S., Park C.H., Casas J.R., 2001, Reliability analysis of prestressed concrete bridge girders: comparison of Eurocode, SpanishNorma IAP andAASHTOLRFD, Structural Safety, 23, 331-344 Assessment of the strength reduction factor... 1053 34. Nowak A.S., Szerszen M.M., 2003, Calibration of design code for buildings (ACI 318), Part 1 – Statistical models for resistance,ACI Structural Journal, 100, 3, 377-382 35. Nowak A.S., Szerszen M.M., Szwed S.A., Podhorecki P.J., 2005, Reliability-Based Cali- bration for Structural Concrete, Report No. UNCLE 05-03, University of Nebraska 36. OstlundL., 1991,AnestimationofT-values, [In:]Reliability ofConcrete Structures.CEBBulletin d’Information, 202, Lausanne, Switzerland 37. PamJ.H.,KwanA.K.H., IslamM.S., 2001,Flexural strengthandductility of reinforcednormal- and high-strength concrete beams, Structure and Buildings, 4, 381-389 38. Saatcioglu M., 2014,Chapter 1 – Design for Flexure, Published by Albert Path on Sep. 14 39. Soares R.C., Mohammed A., Venturini W.S., Lemaire M., 2002, Reliability analysis of nonlinear reinforced concrete frames using the response surface method, Reliability Engineering and System Safety, 75, 1-16 40. Spanish Norma IAP-98, 1998, Actions in highway bridges, Road Directorate, Spanish Ministry of PublicWorks,Madrid 41. Structures Design Manual for Highways and Railways, 1997, Highways Department, Government of the Hong Kong Special Administrative Region, 2nd ed., with Amendment No. 1/2002, Hong Kong 42. Szerszen M.M., Szwed A., Nowak A.S., 2005, Reliability analysis for eccentrically loaded columns,ACI Structural Journal, 102, 5, 676-688 43. Val D., Bljuger F., Yankelevsky D., 1997, Reliability evaluation in nonlinear analysis of reinforced concrete structures, Structural Safety, 19, 2, 203-217 44. WieghausK.T., AtaderoR.A., 2011,Effect of existing structure andFRPuncertainties on the reliability of FRP-based repair, Journal of Composites for Construction, 15, 4, 635-643 Manuscript received June 29, 2017; accepted for print March 2, 2018