Engineering Structures 24 (2002) 11051117 www.elsevier.

com/locate/engstruct

A new design equation for predicting the joint shear strength of monotonically loaded exterior beam-column joints
P.G. Bakir ab,, H.M. Boduroglu a
a b

Istanbul Technical University, Civil Engineering Department, Maslak 80626, Istanbul, Turkey Postal address: Yazmaci Tahir sok, Derya apt. no 11/4, Catalcesme, Suadiye, Istanbul, Turkey Received 11 June 2001; received in revised form 8 March 2002; accepted 8 March 2002

Abstract In this study a new design equation for predicting the shear strength of monotonically loaded exterior beam column joints is proposed. For this purpose, the inuence of several key variables on the behaviour of beam-column joints are inspected using results of parametric studies on an experimental database compiled from a large number of exterior joint tests. The design equation suggested has three differences from the previously proposed equations. First, the equation proposed considers the inuence of beam longitudinal reinforcement ratio, which was not taken into account in previously suggested design equations. Second, as the inuence of this parameter is taken into account, a more realistic estimate of the inuence of joint aspect ratio is obtained. Third, the inuence of stirrups is considered differently for joints with low, medium and high amount of stirrup ratios, in a way, which was not considered in previously suggested equations. The results showed that the proposed design equation predicts the joint shear strength of exterior beam column connections accurately with minimal standard deviation and is more reliable than the previously suggested equations. 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Reinforced concrete; Shear strength; Monotonically loaded exterior beam-column connections

1. Introduction It is now generally believed that beam-column joints can be critical regions in reinforced concrete frames under severe seismic effects. Beam-column joint failures have been observed in the 1980 El Asnam [1], 1985 Mexico [2], 1986 San Salvador [3], 1989 Lome Prieta [4] and 1999 Kocaeli earthquakes [5]. During the past forty years, signicant amount of research has been carried out on seismic behaviour of beam-column joints all over the world. However, compared to cyclically loaded joints, little information exists in literature for predicting the shear strength of monotonically loaded exterior joints. Remarkable differences exist in the design of joints for seismic loading or monotonic loading. Parameters such as column axial load, concrete cylinder strength, stirrup ratio, stirrup index, joint aspect ratio, beam

reinforcement detailing, the ratio of beam longitudinal reinforcement etc inuence the joint shear strength differently for interior or exterior and monotonically loaded or cyclically loaded joints. This investigation was planned with the objective of adding useful data to the understanding of the inuences of the above-mentioned parameters on the joint shear strength of monotonically loaded exterior beam-column joints. Surprisingly, none of the previously suggested design equations for monotonically loaded exterior joints consider the factors that inuence joint shear strength such as beam longitudinal reinforcement ratio, joint aspect ratio and concrete cylinder strength together.

2. Previously suggested design equations for exterior beam column joints In this section, the existing empirical design equations for predicting the shear strength of the monotonically loaded exterior beam-column joints are reviewed.

Corresponding author. Fax: +90-216-386-9742. E-mail address: gundes@itu.edu.tr (P.G. Bakir).

0141-0296/02/$ - see front matter 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 1 4 1 - 0 2 9 6 ( 0 2 ) 0 0 0 3 8 - X

1106

P.G. Bakir, H.M. Boduroglu / Engineering Structures 24 (2002) 11051117

2.1. The design equation of Sarsam and Phillips Sarsam and Phillips [6] have proposed the following equation for the design of monotonically loaded exterior beam column connections. Vcol dc 5.08fcurc0.33 1.33bcdc db 1 0.29 Nu Ag (1)

where Vcol is the column shear force at the column-joint interface (N); fcu is the concrete cube strength (MPa); rc is the column longitudinal reinforcement ratio rc Aso / bcdc (2)

where Asje is the cross-sectional area of the joint stirrups within the top ve eighths of the beam depth below the main beam reinforcement (mm2); a is a coefcient that depends on factors including column load, concrete strength, stirrup index, and joint aspect ratio, which is conservatively taken as 0.2 in Vollums equation. Vc is the joint shear strength without stirrups (N); Vj is the total joint shear strength (N); hb is the section depth of the beam (mm); hc is the section depth of the column (mm); fc is the concrete cylinder strength (MPa); beff is the average of the beam and column widths (mm); fy is the yield strength of stirrups (MPa). 2.3. Present design guidelines The ACI-ASCE Committee 352 [8], and EC8 [9] recommend the following design equations for the shear strength of monotonically loaded joints. Vjd 1.058 fcbeffhc (ACI ASCE Committee 352) Vjd 0.525f 2/3beffhc (EC8 ductility class DCL) c (10) (9)

where Aso is the area of the layer of steel furthest from the maximum compression face in a column (mm2); Ag is the gross cross-sectional area of the column at the joint (mm2); Nu is the axial column load (N);dc is the effective depth of the layer of steel furthest away from the maximum compression face in a column (mm); db is the effective depth of beam tension reinforcement (mm); bc is the width of column section at the joint (mm). The shear force resisted by the links is taken as: Vsd Vud 0.87Ajsfyv Vcd Vsd (3) (4)

where Ajs is the total area of horizontal link reinforcement crossing the diagonal plane from corner to comer of the joint between the beam compression and tension reinforcement (mm2); fyv is the tensile strength of the link reinforcement (MPa); Vsd is the design link shear force resistance (N); Vcd is the design shear force resistance of concrete in a joint (N); Vud is the design ultimate shear capacity of joint (N) All the joint stirrups are considered to be effective in increasing the joint shear capacity. 2.2. The design equation of Vollum Vollum [7] proposed the following equation on the design of exterior beam column connections: Vc Vj 0.642b 1 Vc 0.555(2 hb ) b h f hc eff c c (5) (6)

In order to investigate the reliability of the above design equations, the authors carried out several parametric studies on monotonically loaded exterior beamcolumn joints. The parametric studies are explained in the next section followed by a comparison of the above design equations with the parametric studies of authors.

3. Parametric studies on joint shear strength The authors carried out a parametric investigation of exterior beam-column joint behaviour based on 58 tests conducted in Europe. The loading in all the tests was monotonic. The authors realised that there might be important interactions between variables that at rst were supposed to act independently and this necessitated viewing the entire population of results as a single parameter. For this purpose, monotonically loaded exterior beam column joints are investigated through test results assembled and reconsidered as a unied whole. Examining a large number of individual series of tests as a single database has the advantage of observing which variables have a signicant inuence on joint shear strength in all tests and which variables interact with each other. A new design equation is proposed based on the parametric studies carried out on the experimental database. Table 1 shows the experimental database used in this study. The database comprises of the tests of Ortiz [10], Kordina [11], Scott [12], Scott & Hamill [13], Taylor [14], and Parker & Bullman [15]. The specimen forms included in the database are shown in Fig. 1(a). Fig. 1(b) shows the notations used in this study.

(Asjefy abeffhc fc)

a is taken as 0.2 for low, medium and high amount of stirrups. b is 0.9 for U detail beam reinforcement and 1 for L bent down. The joint shear strength should be limited to: Vj 0.97beffhc fc 1 1.33beffhc fc 0.555(2 hb ) hc (7) (8)

Table 1 Experimental database H (mm) L (mm) hb/hc bc/bb Beam rein.ratio Column rein. ratio fc(MPa) Column axial load SI (MPa)0.5 Vjpredicted/Vjactual Failure modes

Investigator

Specimen

Detail

Ortiz

Kordina

Taylor

P.G. Bakir, H.M. Boduroglu / Engineering Structures 24 (2002) 11051117

Scott

BCJ 1 BCJ 2 BCJ 3 BCJ 4 BCJ 5 BCJ 6 BCJ 7 RE 2 RE 3 RE 4 RE 6 RE 7 RE 8 RE 9 RE 10 P1/41/24 P2/41/24 P2/41/24A A3/41/24A D3/41/24 B3/41/24 C3/41/24BY C3/41/13Y C1AL C4 C4A C4AL C7 C3L C6 C6L C9

L bar L bar L bar L bar L bar L bar L bar L bar L bar L bar L bar L bar U bar U bar U bar L bar L bar L bar L bar L bar L bar U bar U bar L bar L bar L bar L bar L bar U bar U bar U bar U bar

2000 2000 2000 2000 2000 2000 2000 3000 3000 3000 3000 3000 3000 3000 3000 1290 1290 1290 1290 1290 1290 1290 1290 1700 1700 1700 1700 1700 1700 1700 1700 1700

1050 1100 1100 1100 1100 1100 1100 1000 1000 1000 1000 975 975 975 975 470 470 470 470 470 470 470 470 750 750 750 750 750 750 750 750 750

1.33 1.33 1.33 1.33 1.33 1.33 1.33 2.00 1.50 1.50 1.50 1.40 1.40 1.40 1.56 1.43 1.43 1.43 1.43 1.43 1.43 1.43 1.43 1.40 1.40 1.40 1.40 2.00 1.40 1.40 1.40 2.00

1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.40 1.40 1.40 1.40 1.40 1.40 1.40 1.40 1.36 1.36 1.36 1.36 1.36 1.36 1.36 1.36 1.36

1.1 1.1 1.1 1.1 1.1 1.1 1.1 0.9 1.8 1.2 1.2 1.3 1.3 1.3 1.2 2.4 2.4 2.4 2.4 2.4 2.4 2.4 1.4 1.1 2.1 2.1 2.1 1.4 2.1 2.1 2.1 1.4

2.19 2.19 2.92 3.65 3.65 3.65 3.65 2.41 2.41 2.41 2.41 1.61 1.61 1.61 1.61 4.10 4.10 4.10 4.10 4.10 4.10 4.10 4.10 4.29 4.29 4.29 4.29 4.29 4.29 4.29 4.29 4.29

34 38 33 34 38 35 35 25 40 32 32 26 28 28 24 33 29 47 27 53 22 32 28 33 41 44 36 35 35 40 46 36

0 0 0 0 300 300 300 240 400 51 213 650 525 770 551 240 240 240 240 60 240 240 240 50 275 275 50 275 50 275 50 275

0 0.16 0 0.33 0 0 0.76 0 0.26 0.19 0.38 0.43 0.42 0.41 0.45 0.3 0.3 0.26 0.3 0.24 0.75 0.31 0.33 0.188 0.203 0.196 0.218 0.22 0.22 0.21 0.19 0.22

0.68 is 0.77 is 0.64 is 0.78 is 0.72 Is 0.68 is 0.68 b 0.66 is 0.96 is 0.83 is 0.91 is 0.87 is 0.90 is 0.86 is 0.94 is 0.97 is 0.94 is 0.92 is 0.88 is 0.89 is 0.92 is 1.04 is 0.95 is 0.87 is 0.89 is 0.86 is 0.86 is 0.90 is 1.03 is 1.05 is 0.94 is 0.93 is (continued on next page)

1107

1108

Table 1 (continued) H (mm) L (mm) hb/hc bc/bb Beam rein.ratio 2.1 4.29 42 50 0 0.88 p Column rein. ratio fc(MPa) Column axial load SI (MPa)0.5 Vjpredicted/Vjactual Failure modes

Investigator

Specimen

Detail

Scott Hamil

& C4ALN0

L bar

1700

750

1.40

1.36

P.G. Bakir, H.M. Boduroglu / Engineering Structures 24 (2002) 11051117

Parker

C4ALN1 C4ALN3 C4ALN5 C4ALH0 C6LN0 C6LN1 C4ALH1 C4ALH3 C4ALH5 C6LN3 C6LN5 C6LH0 C6LH1 C6LH3 4a 4b 4c 4d 4e 4f 5a 5b 5d 5e 5f

L bar L bar L bar L bar U bar U bar L bar L bar L bar U bar U bar U bar U bar U bar L bar L bar L bar L bar L bar L bar L bar L bar L bar L bar L bar

1700 1700 1700 1700 1700 1700 1700 1700 1700 1700 1700 1700 1700 1700 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000

750 750 750 750 750 750 750 750 750 750 750 750 750 750 850 850 850 850 850 850 850 850 850 850 850

1.40 1.40 1.40 1.40 1.40 1.40 1.40 1.40 1.40 1.40 1.40 1.40 1.40 1.40 1.67 1.67 1.67 1.67 1.67 1.67 1.67 1.67 1.67 1.67 1.67

1.36 1.36 1.36 1.36 1.36 1.36 1.36 1.36 1.36 1.36 1.36 1.3 1.36 1.36 1.20 1.20 1.20 1.20 1.20 1.20 1.20 1.20 1.20 1.20 1.20

2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.1 0.9 0.9 0.9 0.9 0.9 0.9 0.9 0.9 1.4 1.4 1.4

4.29 4.29 4.29 4.29 4.29 4.29 4.29 4.29 4.29 4.29 4.29 4.29 4.29 4.29 1.09 1.09 1.09 4.38 4.38 4.38 2.67 2.67 2.67 2.67 2.67

46 42 50 104 51 51 95.2 105.6 98.4 49 37 101 102 97 39 39 37 39 40 38 42 43 43 45 43

50 50 50 100 50 100 100 100 100 50 50 100 100 100 0 300 600 0 300 600 0 300 0 300 600

0.229 0.478 0.718 0 0 0.19 0.159 0.302 0.469 0.44 0.765 0 0.153 0.472 0 0 0 0 0 0 0.404 0.4 0.6 0.589 0.6 Average Standard deviation

0.85 0.78 0.85 0.86 0.92 0.96 0.93 0.97 1.00 0.92 0.74 0.72 0.98 0.93 1.05 0.83 0.97 0.92 0.78 1.08 0.86 0.88 0.10

js js js p js js b b b js js js js js c js js js js js c js c c js

Note: C is column failure; b is beam failure; js is jointshear failure; p is connection zone reinforcement pullout failure.

 P.G. Bakir, H.M. Boduroglu / Engineering Structures 24 (2002) 11051117

1109

Fig. 2. The inuence of concrete cylinder strength on joint shear strength.

cylinder strength is related to the joint shear strength by the following equation. nj 0.9155 fc (11)

where fc is the uniaxial concrete cylinder strength without any factors of safety and vj is in MPa. C4ALHO and C4ALNO of Scott have been investigated in order to nd the possible inuences of the interaction of other parameters with the concrete cylinder strength. These specimens were deliberately chosen so as to eliminate other factors such as stirrups, column axial load and joint aspect ratio. Because they had no stirrups, they had similar joint aspect ratios and they belonged to the same group of tests. The analysis conrmed the above relationship between the concrete cylinder strength and the joint shear strength. Hamill has also suggested that the joint shear strength is proportional to the square root of the concrete cylinder strength.
Fig. 1. (a) Typical specimen shape in the experimental database. (b) Typlical elevation and notations used for exterior beam colum joints. (c) The strut and truss mechanisms.

5. Determining the joint shear strength The joint shear is calculated using the following procedure:

It is commonly accepted that beam column joints resist shear by the strut and truss mechanisms as suggested rst by Paulay [16] as shown in Fig. 1(c). The strut mechanism accounts for the contribution of the concrete, whereas the truss mechanism represents the contribution of stirrups to joint shear strength. In this study, the resistance of the concrete to the joint shear will be determined rst and then the inuence of the stirrups will be added.

Mb

P(L

d1)

(12)

4. Inuence of concrete cylinder strength The relation between maximum joint shear stress at the instant of joint failure and the uni-axial compressive stress of concrete fc is shown in Fig. 2. The authors carried out a regression analysis and found that concrete

where P is the failure load (N); L is the distance from the point of application of the load to the face of the column (mm); d1 is the cover (mm). A value is assumed for the strain in the beam tensile reinforcement. The force in the beam tensile reinforcement and the moment produced by it are calculated and if this is equal to the moment calculated in Eq. (2), the procedure is stopped. If not, the strain assumed is increased in small increments up until, the moment is equal to the moment given in Eq. (2). The reinforcement was assumed to have an elastic modulus of 200 GPa. The joint shear strength is calculated as below: Vj Tb Vcol (13)

where Vj is the joint shear force (N); Tb is the tensile

1110

P.G. Bakir, H.M. Boduroglu / Engineering Structures 24 (2002) 11051117

force in the beam longitudinal reinforcement (N); Vcol is the shear force in the upper column (N). The normalised joint shear strength is determined as: nj Vj beffhc fc (14)

and consequently have lower joint shear strengths with respect to joints that fail by joint shear failure.

7. Inuence of beam longitudinal reinforcement ratio The authors having investigated the specimens of Parker and Bullman decided that the possible low strength of Parker and Bullman specimens apart from detailing and high joint aspect ratio as well as low radius of bend could be the very low beam longitudinal reinforcement ratio. The relation between the beam longitudinal reinforcement ratio and the normalised joint shear strength is demonstrated in Fig. 4. Fig. 4 shows that the ratio of the beam longitudinal reinforcement is related to the normalised joint shear strength by the following equation: nj Asb b bd
0.4289

where beff is the average of the beam and the column width; fc is the concrete cylinder strength; hc is the height of the column. The unit of v j is (MPa)0.5. 6. Inuence of column reinforcement ratio The authors made a parametric study on the specimens in the experimental database which failed by joint shear, which had low amount of stirrups and which were detailed by L bars bent down detail. Parker and Bullmann specimens were deliberately excluded from this parametric study because they have high joint aspect ratios and low beam reinforcement ratio. Furthermore due to low column longitudinal reinforcement ratio and low column axial stress, some of them have failed due to column failure. The Kordina specimens were also excluded as they are provided by inclined reinforcement in their joints which signicantly increased their joint shear strength. The parametric study in Fig. 3 clearly shows that joint shear strength is independent of the column longitudinal reinforcement ratio in joints that fail by joint shear failure. The authors also analysed the joints of Parker and Bullman. The analysis of specimens 4a and 4d of Parker and Bullman showed that these two specimens were identical except for their column longitudinal reinforcement ratios. The provision of 75% less column reinforcement than specimen 4d, did not only decrease the joint shear strength of specimen 4a, but also caused the specimen 4a to fail by column failure instead of joint shear failure. It can be concluded that, the joints that have low column longitudinal reinforcement ratios and column axial stresses are more likely to fail by column hinging

(15)

where Asb is the total area of beam reinforcement; bb is the breadth of the beam; d is the depth of the beam. In order to understand the possible inuence of the ratio of beam longitudinal reinforcement on the joint shear strength, the authors investigated the specimens C4AL and C1AL of Scott, which were nearly identical except the beam reinforcement ratios they had. The joint shear strength of C1AL that had a beam longitudinal reinforcement ratio of 1.1 was nearly 30% lower than C4AL that had a beam reinforcement ratio of 2.1. The analysis of these two specimens showed that the beam longitudinal reinforcement is related to the normalised joint shear strength by the above equation.

8. Inuence of beam reinforcement detailing The beam reinforcement detailing affects both the normalised joint shear strength and the failure modes of joints. The authors analysis of two identical specimens of Kordina with different beam reinforcement detailing

Fig. 3. The inuence of the column longitudinal reinforcement ratio on the normalised joint shear strength.

Fig. 4. The inuence of the beam longitudinal reinforcement ratio on the normalised joint shear strength.

 P.G. Bakir, H.M. Boduroglu / Engineering Structures 24 (2002) 11051117

1111

showed that the joint shear strengths of monotonically loaded exterior beam column connections decrease by 15% if detailed by U bars rather than L bars. This is shown in Fig. 5. Inspection of Scott and Hamill specimens C4ALH3 (L bars bent down detail) and C6ALH3 (U bars detail) which were identical except for their beam reinforcement detailing shows that providing L bars bent down detail reinforcement can change the failure modes of monotonically loaded exterior beam column connections from joint shear to beam failure. Further inspection of Kordina specimens with that of Ortiz or Scott shows that the normalised joint shear strengths of Kordina specimens are signicantly higher than that of other researchers specimens. The authors are of the opinion that there are two important reasons for this. First, the specimens of Kordina except for RE4 are detailed by inclined reinforcement, which signicantly increases the normalised joint shear strength. Second, the concrete cylinder strengths of Kordinas specimens have been underestimated because they are based on the minimum rather than the average cube strengths. The consequence of this is that the normalised joint shear strengths may have been overestimated.

10. Inuence of joint aspect ratio The authors have carried out a parametric study to investigate the inuence of the joint aspect ratio on the normalized joint shear strength. The specimens which fail by failure modes other than joint shear failure, the U bar specimens, the specimens with inclined bars, all of Parker and Bullmann specimens, all specimens with medium and high amount of stirrups are excluded from this study. In order to minimise the possible interaction of parameters such as the concrete cylinder strength and the beam longitudinal reinforcement ratio, the joint shear strength is normalised by the square root of the concrete cylinder strength and the 0.4289 power of the beam longitudinal reinforcement ratio. Fig. 6 shows the relation between the joint aspect ratio and the normalised joint shear strength. It is evident that the joint aspect ratio is related to the normalised shear strength as shown in Eq. (16). Vj beffhc fc hb hc
0.61

(16)

9. Inuence of the vertical anchorage length and the radius of bend The analysis of BCJ3 of Ortiz showed that the joint shear strength of BCJ3 was considerably increased compared to other specimens of Ortiz with low amount of stirrups. The authors are of the opinion that this increase is due to higher vertical anchorage length and higher radius of bend of beam reinforcement of Ortiz specimen BCJ3. Thus, it can be concluded that there is evidence from Ortiz tests that the shear strength of joints can be increased if the vertical anchorage length is higher than 26db (db is the diameter of beam reinforcement) and the radius of bend is higher than 8db.

where hb is the cross-sectional height of the beam; hc is the cross-sectional height of the column. In order to investigate the possible interactions of different parameters, the authors also inspected C7 and C4AL of Scott, which were the only test data available that investigates the joint aspect ratio. The notable difference between C7 and C4AL was that they had got different beam reinforcement ratios as well as different joint aspect ratios. Analysis of the test data conrmed the reliability of Eq. (16). 11. Inuence of Joint Stirrups Without the inuence of transverse reinforcement, the authors design equation takes the following form: 0.71bg Vc 100Asb b bd hb hc
0.61 0.4289

bc 2

bb

hc fc

(17)

Fig. 5. The normalised joint shear strength of identical specimens with different beam reinforcement detailing.

Fig. 6. The inuence of the joint aspect ratio on the joint shear strength normalised by the concrete sylinder strength and the beam reinforcement ratio.

1112

P.G. Bakir, H.M. Boduroglu / Engineering Structures 24 (2002) 11051117

where b 0.85 for joints detailed by U bars and b 1 for joints detailed by L bars. and g 1.37 for inclined bars in the joint and g 1 for others. The constant 0.71 is a capacity reduction factor determined empirically. The authors have plotted the joint shear strengths of the specimens in the experimental database and their stirrup ratios in Fig. 7. It is very apparent from the gure that the relation between the stirrup ratio and the normalised joint shear strength is tri-linear. Up to stirrup ratios of 0.003, the joints are named as joints with low amount of stirrups. Joints with joint stirrup ratios between 0.003 and 0.0055 are named as joints with medium amount of stirrups. All the stirrups yield within this region. When the stirrup ratio is higher than 0.0055, not all the stirrups yield as evident from the analysis of specimen BCJ7 of Ortiz and C4ALN3 of Scott and Hamill and the contribution of stirrups to joint shear strength substantially reduces under the yield capacity of stirrups. The stirrups that are considered effective in this study are those that are situated above beam compressive chord and below the top beam reinforcement. This is because, the reported stirrup strains of Ortiz specimen BCJ7 and BCJ4 both showed that the stirrup strains substantially reduce if positioned between the beam tensile reinforcement and the beam compressive chord. Fig. 1 shows this. It is commonly assumed in the literature [16] that joint shear strength is given by the addition of concrete resistance to shear, symbolising the resistance of the concrete strut mechanism and stirrup yield capacity, symbolising the resistance of the truss mechanism: Vj Vc Asjefy (18)

Vj beffhc fc

vc

aAsjefy behc fc

(19)

Therefore the nal equation can be formulated as: 0.71bg 100 Vj Asb 0.4289 bc bb hc fc bbdb 2 hb 0.61 hc

(20)

aAsjefy where a 0.664 for joints with low amount of stirrups; a 0.6 for joints with medium amount of stirrups; a 0.37 for joints with high amount of stirrups The stirrups not only affect the normalised joint shear strength but also the failure modes of joints. The authors inspection of joints C4ALNO (no stirrup in joint) and C4ALN1 (single stirrup in joint) of Scott and Hamill has shown that provision of a single stirrup in joints which have hc/db ratios less than 10 changes the failure mode from connection zone reinforcement pull out to joint shear failure. Furthermore, it is apparent from the database that anchorage failures are not anticipated in joints which have medium or high amount of stirrups.

12. Inuence of column axial stress The authors have depicted the relation between the normalised joint shear strength and the column axial stress in Fig. 8. There is considerable scatter in the experimental data. The results show that the column axial load certainly does not inuence the joint shear strength of monotonically loaded exterior beam column connections. Vollum reaches a similar conclusion. Nevertheless, the authors investigation of the specimens of Parker and Bullman showed that column axial load inuences the behaviour of joints by changing their failure modes. The authors comparison of Parker and Bullmann specimen 4a with specimens 4b, 4c showed that the former failed by column failure while the latter two specimens all failed by joint shear failure. The

where Vc is the joint shear strength of the concrete (without stirrups); Asje is the area of the stirrups; fy is the stirrup yield strength. As mentioned above, the analysis of tests showed that when the joints have a stirrup ratio higher than 0.0055, not all the stirrups yield within the joint. The authors are of the opinion that Eq. (18) should be corrected as:

Fig. 7.

The inuence of stirrup ratio on the joint shear strength.

Fig. 8. The inuence of the column axial stress on the normalised joint shear strength.

 P.G. Bakir, H.M. Boduroglu / Engineering Structures 24 (2002) 11051117

1113

experimental data indicates that high column axial stresses and high column longitudinal reinforcement ratios are necessary in order to avoid column failures.

13. Critique of previously suggested design equations for exterior beam column joints In this section, the previously suggested empirical equations are discussed and compared with the design equation of authors. Because the inuence of beam longitudinal reinforcement ratio is not taken into account, previous researchers estimates of the inuence of the joint aspect ratio on the joint shear strength are awed. None of the previously suggested design equations take account of the inuence of the stirrups realistically. Using the authors new proposal, a more realistic estimate of joint shear strength was obtained. 13.1. Remark on Sarsam and Phillips equation The authors as well as Vollums analysis showed that the column axial load does not inuence the joint shear strength. Contrary to the experimental evidence, Sarsam and Phillips equation suggests that column axial load inuences the joint shear strength. Furthermore, all stirrups are considered as effective in Sarsam and Phillips equation while the tests of Ortiz showed that the effective stirrups are those that are situated above the beam compressive chord and below the beam reinforcement. Sarsam and Phillips equation predicts that all the transverse reinforcement yield. However, inspection of BCJ7 of Ortiz demonstrates clearly that not all the stirrups yield in some joints. Sarsam and Phillips equation predicts that column longitudinal reinforcement ratio affects the joint shear strength. However, the parametric study of authors in Fig. 6 showed that there is no apparent relationship between the column longitudinal reinforcement ratio and the normalised shear strength of joints that fail by joint shear failure. The equation predicts that there is a linear relationship between the concrete cube strength and the normalised joint shear strength. The authors parametric study in Fig. 3 showed that the normalised joint shear strength is proportional to the square root of the concrete cylinder strength. The equation of Sarsam and Phillips overestimates the inuence of the concrete cylinder strength on the normalised joint shear strength. 13.2. Remark on the design equation of Vollum As mentioned in the above paragraphs, beam reinforcement ratio signicantly inuences the joint shear strength of exterior joints. Vollums equation neglects the inuence of beam longitudinal reinforcement

Fig. 9. The inuence of the beam reinforcement ratio on the predicted normalised joint shear strength of the Vollum equation.

ratio on the joint shear strength. Because the effect of beam longitudinal reinforcement ratio is disregarded, the equation incorrectly predicts a linear relationship between the joint aspect ratio and the joint shear strength. If the above parameters are taken into account, joint shear strength is predicted as proportional with the 0.61 power of the joint aspect ratio as given by the suggested equation in this paper. Vollum has not given the basis of the upper limits chosen. It is not clear what the basis of the constant 0.97 is. Moreover, the author suggests that the other limit 1.33fcbeffhc is chosen because it is the highest shear strength of the database used by the author. So the accuracy of the above constraint is limited by the experimental database used. The authors are of the opinion that more experiments are needed to propose upper limits for joint shear strength. The authors plotted Vjpredicted/Vjactual values for both Vollum and the authors equation against beam reinforcement ratio, joint aspect ratio, stirrup ratio and the stirrup index in Fig. 916. The gures clearly show that the authors model is an improvement on the design equation of Vollum. The suggested design equation is more conservative than Vollum equation under varying beam reinforcement ratio, joint aspect ratio, stirrup ratio and the stirrup index.

Fig. 10. The inuence of the beam reinforcement ratio on the predicted joint shear strength of authors equation.

1114

P.G. Bakir, H.M. Boduroglu / Engineering Structures 24 (2002) 11051117

Fig. 11. The inuence of the joint aspect ratio on the normalised joint shear strength predicted by Vollum.

Fig. 15. The inuence of the stirrup index on the predicted joint shear strength of Vollums equation.

Fig. 12. The inuence of the joint aspect ratio on the predicted shear strength of the authors equation.

Fig. 16. The inuence of the stirrup index on the predicted joint shear strength of the authors equation/

14. Present design guidelines Both the ACI-ASCE Committee 352 [8], and EC8 [9] methods calculate the joint shear strength on the assumption that the tensile reinforcement yield and for both equations factors of safety are included. The above methods are considered to be inadequate by the authors because they neglect the inuence of the stirrups, beam longitudinal reinforcement ratio, and joint aspect ratio on the joint shear strength.
Fig. 13. The inuence of the stirrup ratio on the predicted joint shear strength of the Vollum equation.

14.1. Comparison of the parametric studies with the established principles of joint mechanics In order to investigate the reliability of the parametric studies in Fig. 29, the authors investigated the established equations on the basic mechanics of reinforced concrete beam-column joints. This has been also previously discussed by Paulay [17] and by Bonacci & Pantazopoulou for interior joints [18] who have also taken into account the joint deformations. Both of the authors use the average stresses for equilibrium as shown in Fig. 18. The typical loading system considered in analysis of exterior beam-column joints is shown in Fig. 17. Fig. 18 depicts the equilibrium of vertical and horizontal forces. Fig. 18 shows that equilibrium of forces in the horizontal direction require the average transverse compressive stress in the joint sx dened as:

Fig. 14. The inuence of the stirrup ratio on the predicted joint shear strength of the authors equation.

 P.G. Bakir, H.M. Boduroglu / Engineering Structures 24 (2002) 11051117

1115

I2 I3

sxsy

sysz
2 z av

sxsz t2 av

(24b) (24c)

sxsysz s t

The tensile stress in the concrete is negligible and therefore s1 0, which consequently gives: sy t2 av sx (25)

From the Mohrs circle,

Fig. 18. Stress equilibrium.

tan2q

2tav sx sy

(26)

If Eq. (25) is substituted into Eq. (26), the following quadratic equation ensues: t2 av 1 tanq sxtav s2 x tanq 0 (27)

which gives: tav sx tanq (28)

Using Eq. (25), we have: sy

Fig. 17. Joint geometry.

tav tanq

(29)

Collins and Mitchell [19] suggest the following equation for the maximum stress in concrete panels: (21) f2max 0.8 fc f 170e1 c (30)

sx

Asb Asje fs f beffhb beffhb w

where fs is the average stress in the beam reinforcement; fw is the average stress in the transverse reinforcement. Consequently, the average normal concrete stress in the y direction sy can be expressed as : sy Ascol N f beffhc scol beffhc (22)

The principal compressive stress is given by: s2 2 e2 0.002 e2 0.002

2

f2max

(31)

s2 is also given from Mohrs circle as: s2 sx sy tav tanq 1 tanq (32)

where fscol is the average stress in the column reinforcement; N is the column axial load Dening the average joint shear stress in the joint as tav, the maximum principal stress associated with the stress tensor is given as; sx tav 0 s tav sy 0 0 0 sz (23)

Thus the average joint shear stress can be expresses as: tav tanq s2 1 tanq (33)

where sz is the conning stress provided by stirrups in the z direction. s3 I1s2 I2s I3 0 (24a)

Eq. (30)(33) show very clearly that as the principal tensile strain increases, the average joint shear stress decreases. Thus it is necessary to express the principal tensile strain in terms of the strains in the x and y directions in order to investigate the factors that inuence the joint shear strength. From Mohrs circle, it is known that: tan2q g e x ey (34)

In order to determine the principal stresses, Eq. (24a) has to be solved; where I1 sx sy sz

1116

P.G. Bakir, H.M. Boduroglu / Engineering Structures 24 (2002) 11051117

From Mohrs circle, the principal tensile strain will be: e1 (ex 2 ey) (ex ey) 2
2

g 2

(35)

e1 Es 1

1 hb hc
2

If Eq. (34) is substituted into Eq. (35) and appropriate trigonometric transformations are carried out, Eq. (36) given by Bonacci and Pantazopoulou is obtained. e1 ex eytan2q 1 tan2q (36) 1 Ascol beffhc

hb 1 tav Asje hc Asbm beffhb beffhb

(43)

hb 2 hc Ascol

The next step will be to express the strains in the x and y directions in terms of the stresses. sx tavtanq Asje f beffhb w where m fs / fw (38) Asbfs Asjefw beffhb beffhb Asb m beffhb (37)

The strain in the x direction can therefore be expressed as: ex fw Es tavtanq Asje Asbm Es beffhb beffhb (39)

The strain in the y direction can similarly be expressed as: ey fscol Es N beffhc tav tanq beffhc AscolEs (40)

The above equation shows that the principal tensile strain is increased by the joint aspect ratio and column longitudinal reinforcement ratio and the axial load on the column whereas it is decreased by increasing beam longitudinal reinforcement ratio and the stirrup ratio. The shear stress in the joint is dependent on the principal tensile strain as evident from Equations 30 and 33. It is therefore evident from Eq. (28)(29) and (43) that the joint shear strength increases as the beam longitudinal reinforcement ratio and the transverse reinforcement ratio increases. Eq. (29) shows that the joint shear strength increases as the column load and the column longitudinal reinforcement increases but Eq. (43) shows that as the longitudinal column reinforcement and the column load increases, the principal tensile stresses increase which consequently decreases the normalised joint shear strength. Therefore the increase in the joint shear strength due to Eq. (29) is offset by the increase in the principal tensile strain. The above conclusions are totally in accordance with the predictions of the authors equation.

If Eq. (39) and (40) are substituted into Eq. (36):

15. Conclusions The purpose of this investigation was to study the effect of the parameters inuencing the behaviour of beam to column connections and to determine if the present design guidelines are unconservative. From the analysis of the tests and results of the parametric studies, the following design recommendations can be made. 1. All the experimental evidence points to the fact that the U bar details should not be used in monotonically loaded exterior beam column joints. As mentioned in the upper paragraphs, providing L bars bent down detail beam reinforcement can change the failure modes of monotonically loaded exterior beam-column joints from joint shear to beam failure. Furthermore, the joint shear strength of joints is increased by 15% if detailed by L bars bent down detail beam reinforcement. 2. Experimental evidence shows that anchorage failures are not anticipated in joints with stirrups in monoton-

e1

1 t tanq Es(1 tan2q) av

1 Asb m beffhb Asje beffhb

(41)

1 Ascol beffhc

Ntan2q Ascol beffhc beffhc

It is evident from the inspection of experiments that cracks extend throughout the diagonal of the joint. So the angle of principal stresses can be expressed as: tanq hb hc (42)

If Eq. (42) is substituted into Eq. (41),

 P.G. Bakir, H.M. Boduroglu / Engineering Structures 24 (2002) 11051117

1117

3.

4.

5.

6.

7.

8.

ically loaded exterior beam column joints. In joints without stirrups on the other hand, hc/db values should be larger than 10 in order to avoid anchorage failures. Column axial load has no inuence on ultimate shear capacity of the joint but high column axial load and high column longitudinal reinforcement ratios are necessary in the joint to avoid column failures. Transverse reinforcement in the joint improves the joint shear capacity but not in the same ratio as indicated by the addition rule Vc+Vs. The authors plotted the stirrup ratio against the joint shear strength. The results showed that the diagram is trilinear. Based on this parametric study, the authors classied the stirrups into three in increasing the joint shear strength. Up to stirrup ratios of 0.003, the joints are named as joints with low amount of stirrups. Joints with joint stirrup ratios between 0.003 and 0.0055 are named as joints with medium amount of stirrups. All the stirrups yield within this region. When the stirrup ratio is higher than 0.0055, not all the stirrups yield as evident from the analysis of specimen BCJ7 of Ortiz and C4ALN3 of Scott and Hamill and the contribution of stirrups to joint shear strength substantially reduces under the yield capacity of stirrups. Parametric studies further carried out on the experimental database in Table 1 showed that the addition rule Vc+Vs, should be corrected as (Vc aVs)beffhc fc where is 0.664 for low amount of stirrups, 0.6 for medium amount of stirrups, 0.37 for high amount of stirrups. Increasing the beam longitudinal reinforcement ratio increases the joint shear strength. Because the inuence of beam longitudinal reinforcement ratio is taken into account, the proposed equation predicts that the joint shear strength is proportional to (hb/hc) 0.61. The present guidelines, design equations and code recommendations for predicting the shear strength of monotonically loaded exterior beam column joints are unconservative. The suggested design equation gives more conservative and reliable results for predicting the joint shear strength under varying beam longitudinal reinforcement ratio, joint aspect ratio, stirrup ratio as well as stirrup index. The analysis of BCJ3 of Ortiz shows that the joint shear strength will further increase if the vertical anchorage length is higher than 26db and the radius of bend of beam bars are higher than 8db. The authors are of the opinion that these gures can be used as lower limits for the radius of bend and vertical anchorage length. The authors applied their equation on the experimental database in Table 1. The results showed that the average Vjpredicted Vjtest values for the authors equ-

ation applied on all the experiments in the experimental database is 0.88 and the standard deviation is 0.1. The results show that the equation suggested gives realistic and conservative estimates of the joint shear strength. 9. The extremely good results of the proposed design equation on the experimental database conrmed and supported the value of the parametric studies. References
[1] Bertero VV, Shadh H. Algeria earthquake, October 10, 1980. Oakland, CA: Earthquake Engineering Research Institute, 1983. [2] Mitchell D. Structural damage due to the 1985 Mexico earthquake. In: Proceedings of the 5th Canadian Conference on Earthquake Engineering. Rotterdam: A.A. Balkema; 1987. p. 87111. [3] Seismology Committee, Structural Engineers Association of California. Recommended lateral force requirement and tentative commentary, 1998. [4] EERI Reconnaissance Team. Lome Prieta earthquake, October 17, 1989. Preliminary reconnaissance report. Oakland, CA: EERI. [5] Bakir PG, Boduroglu MH. Earthquake risk and hazard mitigation in Turkey. Earthquake Spectra, accepted for publication. [6] Sarsam KF, Phillips ME. The shear design of insitu reinforced beam-column joints subjected to monotonic loading. Magazine of Concrete Research 1985;37(130):1628. [7] Vollum RL. Design and analysis of exterior beam column connections. PhD thesis, Imperial College of Science Technology and Medicine-University of London, 1998. [8] Recommendations for design of beam-column joints for design of beam column joints in monolithic reinforced concrete structures, 1991 edition. Reported by ACI-ASCE Committee 352, ACI. [9] Eurocode 8: design provisions for earthquake resistance of structures. London: BSI, 1995. [10] Ortiz R. Strut and tie modelling of reinforced concrete short beams and beam column joints. PhD thesis, University of Westminster, 1993. [11] Kordina K. Bewehrungsfuhrung in Ecken und Rahmenendknoten, Deutscher Ausschuss fur Stahlbeton, Heft 354, 1984. [12] Scott RH. The effects of detailing on RC beam column connection behaviour. The Structural Engineer 1992;70(18):31824. [13] Scott RH, Hamill SJ. Connection zone strain in reinforced concrete beam column connections, In: Proceedings of the 11th International Conference on Experimental Mechanics, Oxford, UK, 1998. pp. 65-69. [14] Taylor HPJ. The behaviour of in situ concrete beam column joints. Technical report 42.492, Cement and Concrete Association, May 1974. [15] Parker DE, Bullman PJM. Shear strength within reinforced concrete beam-column joints. The Structural Engineer 1997;75(4):537. [16] Park R, Paulay T. Reinforced concrete structures. New York: John Wiley and Sons, 1975. [17] Paulay T. Critique of the special provisions for seismic design of the building code requirements for reinforced concrete (ACI 318-83). ACI Journal 1986; 83(2): 274, 283. [18] Pantazopoulou S, Bonacci J. Consideration of questions about beam-column joints. ACI Structural Journal 1992;89(1):2737. [19] Collins MP, Mitchell D. Prestressed concrete structures. Englewood Cliffs, NJ: Prentice-Hall, 1991.