Sei sulla pagina 1di 10

ACI STRUCTURAL JOURNAL

Title no. 103-S61

TECHNICAL PAPER

Strength of Struts in Deep Concrete Members Designed Using Strut-and-Tie Method


by Carlos G. Quintero-Febres, Gustavo Parra-Montesinos, and James K. Wight
Results from an experimental investigation aimed at evaluating the adequacy of the strength factors for concrete struts in strut-and-tie models given in Appendix A of the 2002 ACI Building Code are presented. The main design variables considered were: the angle between primary strut-and-tie axes, amount of reinforcement crossing the strut, and concrete strength. A total of 12 deep beams were tested, eight with normal strength concrete and four with high-strength concrete. The ratio between experimentally obtained failure loads and the strengths predicted using the strut strength factors given in Appendix A of the ACI Code ranged between 1.00 and 1.22, and between 0.91 and 1.02 for normal and high-strength concrete beams, respectively. Inconsistencies were found in the provisions for minimum reinforcement crossing a strut in Sections A.3.3 and A.3.3.1 when applied to the test specimens, with the former leading to substantially larger reinforcement ratios. The use of a strut strength factor s = 0.60 in high-strength concrete bottle-shaped struts without web reinforcement led to strength predictions approximately 10% higher than the experimental failure loads. The limited test results suggest that, as a minimum, an effective reinforcement ratio of 0.01, calculated according to ACI Code, Section A.3.3.1, should be provided in high-strength concrete members when a strength factor s = 0.60 is used. Additional test data, however, are required before a definite recommendation can be made in this regard.
Keywords: beam; cracking; shear strength.

RESEARCH SIGNIFICANCE The adequacy of the strength factors for concrete struts given in Appendix A of the 2002 ACI Code was evaluated. Strut strengths determined from Appendix A of the ACI Code were compared with experimental data obtained from tests of several deep reinforced concrete beams. The data generated in this study also provides valuable information to ACI Committees 445 and 318-E for evaluating current design specifications for concrete members using the strut-and-tie method. STRENGTH FACTORS FOR CONCRETE STRUTS A conservative estimation of the compressive strength of struts in strut-and-tie models is crucial to ensure adequate structural safety and the ability to design deep beams such that yielding of longitudinal reinforcement occurs prior to failure of concrete struts. If either no yielding or only limited yielding occurs prior to failure of a concrete strut, a sudden failure could take place with the drastic consequences that have been observed in past shear failures.1 During the last two decades, several researchers have conducted studies aimed at evaluating the strength of concrete struts for use in strut-and-tie models.1-6 Generally, the approach has been to lump the effect of strut stress and strain conditions, reinforcement details, concrete strength, and uncertainties in the truss model into a single factor, commonly referred to as the efficiency factor or, more recently, as the strength reduction factor v 1.0.4,6,7 Therefore, the strength of a concrete strut is expressed as fc = vfc (1)

INTRODUCTION The use of strut-and-tie models for the design of reinforced concrete members was incorporated in the 2002 edition of the ACI Building Code (herein referred to as the ACI Code) as Appendix A. Some of the provisions for the design of concrete members using the strut-and-tie method, however, are still the subject of discussion and validation, such as that related to the strength factors for concrete struts, which are assumed to be dependent on strut cracking condition and reinforcement provided, but independent of concrete strength and strut angle. Discussions on the adequacy of current strength factors for concrete struts have been triggered by the significant discrepancies that exist between the proposed values in the ACI Code and those suggested by several researchers.1-6 In some cases, the latter ones are substantially lower than those given in the ACI Code, especially for higher strength concrete. This paper presents results from an investigation aimed at evaluating the adequacy of the strength factors given in Appendix A of the ACI Code for the determination of strength of concrete struts in strut-and-tie models. In particular, the ultimate capacity of diagonal compression struts in deep concrete beams with either minimum or no transverse reinforcement was experimentally evaluated. For this purpose, several concrete beams with various shear span-to-depth ratios, concrete strengths, and reinforcement layouts were tested. ACI Structural Journal/July-August 2006

where fc is the concrete compressive strength. To make the application of strut-and-tie models practical, single values for the efficiency factor have been proposed, which are based on the cracking conditions expected in the struts, that is, struts with cracks parallel to or skewed with respect to the strut axis, but are independent of concrete strength. Proposed values for the efficiency factor typically range from 0.3 for heavily cracked struts to 0.85 for struts in which the concrete is subjected to uniaxial or biaxial compression.4 Appendix A of the ACI Code8 adopted this format in Article A.3.2, where the efficiency factor is expressed as 0.85s. Values of s range between 0.40 and 1.0, depending on the strut geometry, reinforcement provided, and stress conditions in the member. Thus, the specified values for the efficiency factor
ACI Structural Journal, V. 103, No. 4, July-August 2006. MS No. 05-104 received May 6, 2005, and reviewed under Institute publication policies. Copyright 2006, American Concrete Institute. All rights reserved, including the making of copies unless permission is obtained from the copyright proprietors. Pertinent discussion including authors closure, if any, will be published in the May-June 2007 ACI Structural Journal if the discussion is received by January 1, 2007.

577

ACI member Carlos G. Quintero-Febres is a Professor of Civil Engineering at the Universidad de los Andes, Mrida, Venezuela. He received his degree in civil engineering from the Universidad de los Andes, Venezuela; his MS from the University of Iowa, Iowa City, Iowa; and his PhD from the University of Michigan, Ann Arbor, Mich. He is a recipient of the ACI Structural Research Award in 2003. His research interests include the earthquake-resistant design of reinforced concrete structures and structural reliability. ACI member Gustavo Parra-Montesinos is an Assistant Professor of Civil Engineering at the University of Michigan. He is Secretary of ACI Committee 335, Composite and Hybrid Structures, and a member of ACI Committees 318-F, New Materials, Products and Ideas (Structural Concrete Building Code), and 544, Fiber Reinforced Concrete. He is also a member of Joint ACI-ASCE Committee 352, Joints and Connections in Monolithic Concrete Structures. His research interests include the seismic behavior and design of reinforced concrete, hybrid steel-concrete, and fiber-reinforced concrete structures. James K. Wight, FACI, is a Professor of Civil Engineering at the University of Michigan. He is Chair of ACI Committee 318, Structural Concrete Building Code, and a member of Joint ACI-ASCE Committees 352, Joints and Connections in Monolithic Concrete Structures, and 445, Shear and Torsion. He has received several ACI awards, including the Delmar L. Bloem Award in 1991, the Joe W. Kelly Award in 1999, and the ACI Structural Research Award in 2003. His research interests include the earthquake-resistant design of reinforced concrete structures.

in the ACI Code range between 0.34 and 0.85. As might be expected, these discrete values of the efficiency factor are somewhat arbitrary and primarily based on experimental evidence. Background information about provisions for concrete struts given in Appendix A of the ACI Code can be found elsewhere.7 Other more sophisticated expressions for estimating the strength of concrete struts include the effect of concrete compressive strength,3,5 in which the value of the efficiency factor decreases as the concrete strength is increased. One such expression is given in Eq. (2)5 f c - f ( MPa ) f c = 0.7 ------- 200 c (2)

where 1 and 2 are the principal tension and compression strains, respectively, and o is the concrete strain at peak compressive strength. In Eq. (3), the principal strain ratio 1/2 needs to be known in order to determine the effective concrete strength in the strut. Equation (4) represents a simpler version, in which only the principal tension strain needs to be determined. If Eq. (4) is evaluated for principal tensile strains of 0.002 and 0.006 assuming o = 0.002, the efficiency factor would be equal to 0.88 and 0.55, respectively. More recently, Vecchio11 proposed a reduction in the degree of concrete softening predicted by Eq. (3) and (4) to account for the fact that some of the softening, which was attributed to the concrete in previous tests, was actually due to crack slip and not to transverse tensile strains. Thus, for cases where crack reorientation does not occur, higher values of the efficiency factor would be obtained compared to those determined from Eq. (3) and (4). A simple expression for the concrete efficiency factor as a function of the angle of inclination of the strut with respect to the tie direction was recently proposed by Matamoros and Wong,12 with the values of the efficiency factor decreasing as the angle of inclination of the strut decreases. That model gave lower strength estimates compared with the ACI Code for the range of strut angles considered in their study (between 30 and 60 degrees). From the previous discussion, it is clear that, depending on the model used, significant differences can be obtained in the values of the efficiency factor used for concrete struts. Values as low as 0.3 have been proposed in the past, which is less than the lowest factor specified in the ACI Code. Therefore, an evaluation of the adequacy of the strength factors for concrete struts given in Appendix A of the ACI Code, in particular those corresponding to struts crossed by no or low amounts of transverse reinforcement, was performed. EXPERIMENTAL INVESTIGATION The experimental program reported herein was carried out at the Structural Engineering Laboratory of the University of Michigan. It comprised the testing to failure of 12 reinforced concrete deep beam specimens designed with various shear span-to-depth ratios, concrete strengths, and reinforcement layouts. Detailed information about this testing program can be found elsewhere.13 All beams were 2440 mm (8 ft) long, 460 mm (18 in.) deep, and 150 mm (6 in.) wide, except for the high-strength concrete specimens, in which the beam web width in the test region was reduced to 100 mm (4 in.) to force a diagonal compression strut failure prior to yielding of the longitudinal reinforcement. The clear-span-to-totaldepth ratio for all beams was less than 4.0, which classifies them as deep beams, according to the ACI Code. The beams were loaded at a single point on the top face close to one of the supports in order to develop a single concrete strut acting approximately at a predetermined angle. Figure 1 shows the load and support configuration for the beams. Specimen design parameters Three main design parameters were considered in this study, primary strut angle with respect to the member longitudinal (tie) axis, concrete compressive strength, and amount of web reinforcement crossing the primary strut. Shallow and steep strut angles were considered, with values of 25 to 29 degrees, and 35 to 44 degrees, respectively, based on a strut-and-tie model constructed following the provisions in Appendix A of the 2002 ACI Code. The ACI Structural Journal/July-August 2006

In Eq. (2), the efficiency factor is represented by the terms in parenthesis. Evaluating this expression for fc = 28 and 70 MPa (4000 and 10,000 psi), values for the efficiency factor of 0.56 and 0.35 are obtained, respectively. Ramirez and Breen3 proposed an efficiency factor for concrete struts in beams of 2.5/fc MPa. In this case, is equal to 0.47 and 0.30 for 28 and 70 MPa concrete, respectively. Extensive research has also been conducted at the University of Toronto by Vecchio and Collins9,10 to evaluate the compression behavior of cracked concrete. Expressions for effective concrete strength developed by Vecchio and Collins are based on the strain conditions in the concrete, and thus their application to strut-and-tie models requires an accurate estimation of the state of strain in the strut. Two expressions they developed are 1 - f f c f c = ------------------------------1 c 0.85 0.27 --- 2 and 1 f c = -------------------------------- f c f c 1 0.80 + 0.34 --- o 578

(3)

(4)

smallest angle between a strut and a tie is limited to 25 degrees in Section A.1 of the ACI Code. Therefore, beams tested with shallow strut angles were at or close to the lower limit, and thus likely to depart from a deep beam behavior (that is, shear resistance through a single direct strut between the load point and support). Beams tested with steep strut angles did represent deep beam behavior. Normal-strength concrete (in the order of 28 MPa [4000 psi]) and high-strength concrete (in the order of 55 Mpa [8000 psi]) were used to evaluate the adequacy of current strength factors for a wide range of concrete strengths. Based on these design parameters, the test specimens were grouped in three series of four specimens each, as follows: Series A having normal strength concrete and a main strut angle of approximately 28 degrees; Series B with normal strength concrete and a primary strut angle ranging between 41 and 44 degrees; and Series H with high-strength concrete, two specimens having a strut angle of approximately 25 degrees and two with a strut angle of either 35 or 38 degrees. Table 1 lists the primary strut angle, , for each test specimen. Two amounts of reinforcement crossing the primary strut were considered in this study. For Series A and B, two specimens had no reinforcement crossing the strut, and two specimens had minimum reinforcement as per Section A.3.3.1. For each strut angle considered in Series H, one specimen had no web reinforcement and one had web reinforcement crossing the primary strut. The web reinforcement provided in the specimens of Series H, however, did not satisfy the minimum reinforcement ratio required in Section A.3.3 of the ACI Code, as will be explained in detail later. Design of test specimens The test specimens were designed using the strut-and-tie model shown in Fig. 2. In this model, a single strut is assumed between the load point and the closer support, which is located at approximately one-third or less of the clear span away from the load. For the other portion of the beam specimen (left portion of beam in Fig. 2) a more complex truss is required to avoid a strut angle less than

25 degrees. With this configuration, the strength of the primary strut to the closer support was computed based on Section A.3 of the ACI Code. Bearing strengths at nodal zones were then checked and a prediction of the failure load was obtained for the selected strut-and-tie model. Calculations for determining the beam strength for the critical shear span of one of the test beams in Series A is provided in the Appendix of this paper. In addition, a complete design example for a deep beam using a strut-and-tie model similar to that shown in Fig. 2 can be found elsewhere.14 Reinforcement details Details of the reinforcement for the test specimens are given in Fig. 3 through 5. The longitudinal reinforcement for

Fig. 1Overall dimensions and load configuration of deep beam specimens.

Fig. 2Strut-and-tie model for deep beam design.

Table 1Properties of test specimens and summary of test results


Specimen A1 A2 A3 A4 B1 B2 B3 B4 HA1 HA3 HB1 HB3
*

bw, mm 150 150 150 150 150 150 150 150 100 100 100 100

d, mm 370 370 370 370 375 375 375 375 380 380 380 380

a, mm 525 525 525 525 335 335 305 305 595 545 340 310

, degree 28.5 28.5 27.9 27.9 41.3 41.3 43.8 43.8 25.0 25.0 35.0 37.6

Effective reinforcement ratio provided 0.0042


*

Effective reinforcement ratio required 0.003 (0.0096) 0.003 (0.0096) 0.003 (0.013) 0.003 (0.013) NA (0.014) NA (0.021)

Failure mode DS DS DS DS DC DC DC DC SC SC DC DC

Vu, kN 251 237 221 196 456 426 468 459 265 292 484 460

vu /fc, MPa 0.95 0.90 0.84 0.74 1.40 1.31 1.44 1.41 0.64 0.71 1.18 1.12

VACI , kN 215 215 185 185 425 425 385 385 275 320 475 505

Vu /VACI 1.17 1.13 1.20 1.06 1.07 1.00 1.22 1.19 0.96 0.91 1.02 0.91

0.0042 0.0038 0.0038 0.0056 0.0085

Based on Eq. (5). Based on ACI 318, Section A.3.3.1. Based on ACI 318, Section A.3.3. DS = diagonal splitting; DC = diagonal compression; SC = shear compression; Vu = shear force at failure; vu = average shear stress at failure (Vu/bwd), VACI = calculated strength based on ACI 318, Appendix A. Notes: NA = not applicable; 1 mm = 0.394 in.; 1 kN = 0.225 kips; and 1 MPa = 0.145 ksi.

ACI Structural Journal/July-August 2006

579

all beam specimens was provided by four equal size bars distributed in two layers whose centroid coincided with the theoretical centroid of the main tie. This reinforcement was chosen such that the stress in the steel corresponding to the maximum predicted tie force for a particular strut strength would be at or below the yield strength of the steel fy. The reinforcement selected to carry the maximum tie force was kept constant along the specimen length. At both beam ends, the longitudinal reinforcement was extended beyond the supports and terminated with either standard 90- or 180degree hooks to ensure proper development. Transverse reinforcement over the portion of the beam away from the primary strut consisted of U-shaped, No. 10M (area = 71 mm2) deformed bars. More stirrups than required for shear strength were used to force a failure in the strut being tested. The amount of reinforcement used in each case was checked using the strut-and-tie model of Fig. 2. The extensions of the beam beyond the supports were reinforced with U-shaped stirrups, as per ACI Code Section 12.5, to ensure proper development of reinforcement. Transverse reinforcement crossing the primary diagonal strut was provided through No. 6M (area = 32 mm2) horizontal bars,

and either U-shaped stirrups (Series A and B) or two single-leg ties (Series H) (Fig. 3 through 5). For Series A and B, which had a specified concrete strength less than 41 MPa (6000 psi), ACI Code Eq. (A-4) was used to select the reinforcement - sin i 0.003 -----bs i A si (5)

where Asi is the total area of reinforcement at a spacing si in a layer of reinforcement i with bars at an angle i to the axis of the strut. For Series H (concrete strength above 41 MPa [6000 psi]), a relaxation of the requirements of Section A.3 was made. In this case, instead of assuming that the compressive force in the strut would spread at a 2:1 slope (ACI Code, Section A.3.3 and Fig. RA.1.8(b)), a shallower slope of 6:1 was used to avoid excessive amounts of web reinforcement. For all beams, a single U-shaped stirrup (Series A and B) or a set of two single-leg ties (Series H) was placed at the location of the applied load and the reaction force. Table 1 gives the web reinforcement ratios provided in the critical shear span, calculated according to Eq. (A-4) of the ACI Code. Materials The concrete for the three series was obtained in different batches from a local ready mixed concrete supplier. The maximum coarse aggregate size was 10 mm (3/8 in.). For each batch, 12 100 x 200 mm (4 x 8 in.) cylinders were taken for later evaluation of concrete compressive strength. All longitudinal reinforcement was made of Grade 60 (nominal fy = 410 MPa (60 ksi) deformed steel. Main strut transverse reinforcement consisted of smooth round bars (not compliant with ASTM A 615M standards), while all other beam transverse reinforcement consisted of Grade 60 steel deformed bars. Reinforcing steel for each series was obtained separately from a local steel supplier. Measured concrete compressive strengths, obtained from at least three cylinder tests, and steel yield strengths are given in Table 2. Fabrication of test units Both the longitudinal and transverse reinforcement were bent by a local steel supplier. After strain gauges were attached to the bars at the desired locations, steel reinforcement cages were assembled in the laboratory and placed into the forms. The beams were cast following common procedures and wet-cured for seven to 10 days before being removed from their wood formwork. Beams from Series A and B, which had a uniform cross-section, were cast upright, while those from Series H, which had a reduction of the cross section in the shear span, were cast upside down to ease the casting process and to ensure proper consolidation of the concrete in the forms. Once the concrete had gained sufficient strength, the beams were placed in the testing rig. Table 2Material properties
Series A B fc, MPa (ksi) 22.0 (3.2) 32.4 (4.7) 50.3 (7.3) Reinforcement yield strength fy , MPa (ksi) No. 6M No. 10M No. 13M No. 19M No. 22M 407 (59.0) 545 (79.1) 586 (85.0) 441 (64.0) 476 (69.0) 476 (69.0) 455 (66.0) 427 (61.9) 462 (67.1) 434 (63.0)

Fig. 3Reinforcement details for Series A.

Fig. 4Reinforcement details for Series B. 580

ACI Structural Journal/July-August 2006

Test setup and instrumentation The beams were tested in two different setups. A test bed with a 490 kN (110 kip) actuator was used for Series A, while for Series B and H, a steel loading frame with a 1780 kN (400 kip) capacity was used. The same load configuration was used in both test setups, as shown in the schematic drawing of Fig. 1. All beams were simply supported through the use of restrained and free rollers, and were loaded on the top face at a single location. The applied load and the reaction forces were distributed on the beam surface through 25 mm (1 in.) thick by 150 mm (6 in.) square steel plates. High strength grout was used for proper seating of the steel plates on the beams surfaces. The load was applied monotonically up to failure in increments that corresponded to approximately one-tenth of the predicted failure load for each beam. Load cells, displacement transducers, and electrical resistance strain gauges were used to monitor the applied loads, deflections and deformations, and strains in the steel reinforcement, respectively. Load cells provided readings for the loads imposed by the actuator and hydraulic jack used in the beam tests. A displacement transducer was placed under the beam at the load point to record the imposed displacements. An arrangement of three pairs of displacement transducers, oriented vertically, horizontally, and at 45 degrees, was used in each beam to monitor average strains in the critical shear span. Cracks were also marked at selected load points

up to approximately 75% of the predicted peak load and numbered according to the loading stage. ANALYSIS OF TEST RESULTS The behavior of the test specimens was evaluated in terms of load versus deflection and load versus shear span distortion responses, cracking pattern, failure mode, and strain history at selected reinforcement locations. Overall response All beam specimens exhibited a nearly linear response up to failure, as shown by the load versus deflection curves of Fig. 6 through 8. Even though there was a larger region of low stiffness for the specimens in Series A (due to support seating), it is clear that the specimens in Series A (shallow strut angle) experienced larger load-point deflections than those in Series B (steep strut angle) at peak load. Also, from Fig. 6 it is seen that the specimens of Series A with web reinforcement had a larger strength than those without web reinforcement, while no major differences in strength were observed for the beams in Series B and H (Fig. 7 and 8, respectively). Average shear distortions in the critical shear span (main strut) were computed based on the readings obtained from an arrangement of linear potentiometers, as explained earlier. The load versus shear distortion response was characterized

Fig. 6Load versus displacement response (Series A).

Fig. 5Reinforcement details for Series H. ACI Structural Journal/July-August 2006

Fig. 7Load versus displacement response (Series B). 581

by a linear uncracked response up to first diagonal cracking (loads in the order of 25% of the peak load), followed by a nearly linear cracked-elastic response up to peak load, which corresponded roughly to an average shear distortion of 0.004 to 0.005 rad for all test specimens. Once the peak load was attained, a rapid decrease in the applied load with increasing shear distortions was observed. Providing minimum web reinforcement crossing the strut did not seem to noticeably affect the deformation capacity of the struts. Cracking patterns and failure modes Shear force at failure Vu and failure modes for the test specimens are listed in Table 1. Figure 9 through 11 show typical crack patterns at failure for the specimens in Series A, B, and H, respectively. In general, a few flexural cracks formed first, which remained at a narrow width throughout the tests. Diagonal cracks then formed at approximately 25% of the peak load, as mentioned earlier, defining the direction of the main concrete strut. Failure for all test specimens was brittle and the failure modes were identified as follows: 1) Diagonal-splitting failure, in which diagonal cracks that formed initially at mid-depth of the beam in the direction of the main strut propagated to the outside edge of the loading plate and the inside edge of the bearing plate at the support

(Fig. 9). This type of failure was typical for Series A, regardless of the amount of web reinforcement provided; 2) Strut crushing failure at beam mid-depth following the formation of several diagonal cracks (Fig. 10). This failure mode was typical for Series B and Specimens HB-1 and HB-3 from Series H; and 3) Shear-compression failure near the loading point (Fig. 11) after formation of one or two main diagonal cracks, which had developed at mid-depth of the beam and propagated toward the outside edge of the loading plate. Specimens HA1 and HA3 (shallow strut angle with and without web reinforcement) had this type of shear-compression failure. Strains in longitudinal tension (tie) reinforcement The strain profile in the main longitudinal reinforcement was very similar for all specimens. The formation of a tie action was confirmed by the nearly uniform strain distribution in the longitudinal reinforcement over the shear span. The strains measured in the lower reinforcement layer, however, were slightly larger than those measured in the upper reinforcement layer. A significant reduction in longitudinal reinforcement strains between the load point and the support away from the failure region was observed, following a trend similar to that predicted by the strut-and-

Fig. 10Cracking pattern at failure for Specimen B3. Fig. 8Load versus displacement response (Series H).

Fig. 9Cracking pattern at failure for Specimen A1. 582

Fig. 11Cracking pattern at failure for Specimens HA1 and HA3. ACI Structural Journal/July-August 2006

tie model shown in Fig. 2 for the left side of the beam. The longitudinal reinforcement in all specimens of Series A and H remained elastic throughout the tests, while one or two strain gauges had just reached the yield point when a strut failure occurred in the specimens of Series B. No signs of anchorage problems were observed in any of the tests. Strains in longitudinal compression reinforcement Typically, two strain gauges were attached to the compression reinforcement, one just under the load point and the other at the outer face of the theoretical upper compression node (right side inclined face in Fig. 2). Strains recorded at these two locations were well below yielding for all specimens, with the strains under the load point being larger. In some cases, one or both of the strain gauges recorded tensile strains at later stages of the test when the top reinforcement was crossed by diagonal cracks (Fig. 9). Strains in web or strut reinforcement As shown in Fig. 3 through 5, minimum web reinforcement, calculated according to Appendix A of the ACI Code, was provided in Specimens A1 and A2 (shallow strut angle), and B1 and B2 (steep strut angle), but less than the minimum specified web reinforcement was placed in Specimens HA1 (shallow strut angle) and HB1 (steep strut angle). Strains were recorded at various locations along the main strut on the vertical stirrup legs and on the horizontal reinforcing bars. Strain measurements and visual observations indicated that the web reinforcement was effective in controlling crack opening. Figure 12 shows the measured strain response for vertical and horizontal web reinforcement for Specimen A1. The strain response for the web reinforcement of all test specimens was similar and can be summarized as follows: 1) strains in the horizontal reinforcement were lower than those in the vertical reinforcement; 2) for the specimens with a shallow angle strut, the level of strains in the vertical stirrups was higher than that for specimens with a steep angle strut; and 3) for specimens with web reinforcement in Series A and B, the vertical reinforcement was close to or had just exceeded the yield point at the moment of failure, while for Specimens HA1 and HB1, the strains measured at failure corresponded to a stress of approximately 0.7fy and 0.5fy, respectively. It should be mentioned that smooth bars, as opposed to deformed bars, were used as web reinforcement in the critical shear span. Strain measurements, however, indicated good bond between the bars and the concrete and thus, no major differences in behavior were expected if deformed reinforcement had been provided. PREDICTED VERSUS EXPERIMENTAL STRUT STRENGTHS The main objective of this research project was to experimentally evaluate the adequacy of the expressions in Section A.3 of the ACI Code for the strength of concrete struts in strut-and-tie models. The nominal compressive strength of a strut without reinforcement in the direction of the strut axis is taken as (ACI Code Eq. (A-2)) Fns = fcu Ac (6)

fcu = 0.85s fc

(7)

where s = 0.75 for a bottle-shaped strut with reinforcement satisfying Section A.3.3 (A.3.2.2(a)); and s = 0.60 for a bottle-shaped strut without reinforcement satisfying Section A.3.3 (A.3.2.2(b)), = 1 for normal weight concrete. The design of the test specimens was carried out using the strut-and-tie model shown in Fig. 2 and following the provisions in Appendix A of the ACI Code. The procedure followed to calculate the strength in the critical span of Specimen A1 is demonstrated in the Appendix of this paper. It should be mentioned that strength factors used for concrete struts (s in the ACI Code) are intended to account not only for the effect of strut stress and strain conditions and reinforcement details, but also for uncertainties in the truss model,7 as discussed earlier. Therefore, any strut failure mode was considered for evaluation of the adequacy of the ACI strut strength factors. Table 1 and Fig. 13 show a comparison between the calculated and experimental failure loads. It is worth mentioning that in order to properly evaluate the adequacy of

Fig. 12Strains in web reinforcement crossing primary diagonal strut (Specimen A1).

where Ac is the smaller cross-sectional area at either end of the strut, and fcu is the effective compressive strength of the concrete in a strut, given by Eq. (7) (Code Eq. (A-3)) ACI Structural Journal/July-August 2006

Fig. 13Experimental versus calculated strengths. 583

the provisions in Appendix A of the ACI Code, the predicted strengths reported in Table 1 correspond to the maximum values obtained after considering various admissible strut angles. Thus, the reported values represent the admissible solution that led to either the least conservative or more unconservative prediction. As can be seen in Fig. 13 and Table 1, the calculated strengths were lower than or equal to the experimental results for the specimens in Series A and B with normal strength concrete. For the high-strength concrete beams, however, a strut strength factor s = 0.60 was found to be inadequate for use in struts not crossed by web reinforcement, as indicated by the failure loads in Specimens HA3 and HB3 (appear as a single point in Fig. 13), which were approximately 10% lower than the predicted strengths. For the case of Specimens HA1 and HB1, which had effective web reinforcement ratios of 0.0056 and 0.0084, respectively, strengths calculated according to ACI Code Section A.3.3.1 and using a strength factor s = 0.60 led to predicted values nearly equal to the failure loads. Thus, these limited test results suggest that, as a minimum, an effective reinforcement ratio of 0.01, calculated according to ACI Code Section A.3.3.1, should be provided in high-strength concrete members when a strength factor s = 0.60 is used. Additional test data, however, are required before a definite recommendation can be made in this regard. Peak average shear stress demands for each specimen are also given in Table 1. Regardless of the amount of web reinforcement provided, test specimens with shallow and steep struts sustained peak average shear stress demands exceeding 0.6 and 1.1fc (MPa) (7.2 and 13.2fc [psi]), respectively. For the specimens subjected to average shear stresses greater than the limit given for deep beams in Chapter 11 of the ACI Code (0.83fc [MPa]), however, the calculated strut and node dimensions were believed to be excessive. EVALUATION OF MINIMUM STRUT REINFORCEMENT REQUIREMENTS IN APPENDIX A OF ACI CODE Minimum strut reinforcement requirements are defined in Section A.3.3 of Appendix A in the ACI Code. In that section, sufficient steel reinforcement is required to resist a tension force perpendicular to the strut axis, which is calculated by assuming that the strut compression force spreads out at a slope of 2:1 (longitudinal:transverse direction, refer to ACI Code Fig. RA.1.8(b)). On the other hand, Section A.3.3.1 specifies that this requirement can be assumed to be satisfied for members with a concrete compressive strength fc not greater than 41 MPa (6000 psi), if sufficient web reinforcement is provided such that Eq. (5) (Eq. (A-4) in ACI Code) is satisfied. The requirements in Sections A.3.3 and A.3.3.1, when applied to the test specimens, were found to be neither consistent nor equivalent. Specimens in Series A and B, with normal-strength concrete, were designed to satisfy the requirements of Section A.3.3.1. If, instead, Section A.3.3 were to be satisfied, a much larger amount of transverse reinforcement would have been required. Table 1 shows the amount of web reinforcement required by Appendix A of the ACI Code and the amount provided in the test specimens. As can be seen, the provisions in Section A.3.3 required more than three times the minimum reinforcement specified in Section A.3.3.1 for the specimens in Series A and B. For the specimens of Series H, ACI Code, Section A.3.3, required reinforcement ratios of 1.4 and 2.1%, respectively, 584

leading to impractical and heavily congested designs. In Series H, for which Section A.3.3.1 is not applicable because of the use of high strength concrete, a 6:1 slope was assumed for the spreading of the strut force instead of the 2:1 slope specified in Section A.3.3. As discussed in the previous section, unconservative strength predictions were obtained for the high-strength concrete beams without web reinforcement, while relatively good agreement between predicted and experimental results was obtained in Specimens HA1 and HB3, which had effective reinforcement ratios of approximately 40% of those required by Section A.3.3. Clearly, additional experimental information needs to be generated to draw definite conclusions with regard to the minimum web reinforcement required in high-strength concrete members designed using strut-and-tie models. Based on the very limited results presented herein and until further data become available, a minimum effective web reinforcement ratio of 0.01, as defined by Eq. (5) (ACI Code Section A.3.3.1), is recommended for use in highstrength concrete members designed with a strut strength factor s = 0.60. The limited scope and budget for this experimental program did not permit a thorough investigation of what reasonable percentage of web reinforcement would be required to justify the use of a strut strength factor, s = 0.75, for high-strength concrete struts. As mentioned previously, the compliance with ACI Code Section A.3.3 may lead to excessive amounts of reinforcement that could result in impractical designs. SUMMARY AND CONCLUSIONS The adequacy of the strength factors for concrete struts given in Appendix A of the 2002 ACI Code was evaluated based on the results from the tests of 12 reinforced concrete deep beams. The main design variables considered were main strut angle with respect to the member longitudinal (tie) axis, amount of web reinforcement crossing the strut, and concrete strength. The following conclusions were drawn from the results of this investigation: 1. The strut strength factors given in Appendix A of the 2002 ACI Code were found to be adequate for use in normalstrength concrete bottle-shaped struts crossed by either no reinforcement (s = 0.60) or crossed by the minimum transverse reinforcement specified in ACI Code Section A.3.3.1 (s = 0.75); 2. The provisions for required minimum transverse reinforcement in normal-strength concrete members given in Appendix A of the 2002 ACI Building Code should be reevaluated. Substantial differences in the required amount of reinforcement were obtained when applying the provisions in Sections A.3.3 and A.3.3.1 to the test beams, with the former leading to substantially larger reinforcement ratios; and 3. Results from the tests of four high-strength concrete beams suggest that, as a minimum, an effective web reinforcement ratio of 0.01, calculated according to ACI Code Section A.3.3.1, should be provided in high-strength concrete members when a strength factor s = 0.60 is used. Additional test data, however, are required before a definite recommendation can be made in this regard. ACKNOWLEDGMENTS
The work presented in this paper was carried out at the Structural Engineering Laboratory, University of Michigan, Ann Arbor, Mich. Financial support was provided by the ACI Concrete Research Council and the University of Michigan. The first author would like to thank the Universidad de los Andes,

ACI Structural Journal/July-August 2006

Mrida, Venezuela, for its support during his sabbatical leave. Thanks are also extended to University of Michigan student A. L. Libbrecht, and to L. B. Fargier-Gabaldn and B. Afsin Canbolat for their valuable help during the experimental phase of this project. The conclusions and recommendations contained in this paper are those of the authors and do not necessarily represent the opinion of the sponsors.

REFERENCES
1. MacGregor, J. G., Reinforced Concrete, Mechanics and Design, 2nd Edition, Prentice Hall, 1988, 848 pp. 2. Schlaich, J.; Schfer, K.; and Jennewein, M., Toward a Consistent Design of Structural Concrete, PCI Journal, V. 32, No. 3, 1987, pp. 74-150. 3. Ramirez, J. A., and Breen, J. E., Evaluation of a Modified TrussModel Approach for Beams in Shear, ACI Structural Journal, V. 88, No. 5, Sept.-Oct. 1991, pp. 562-571. 4. Yun, Y. M., and Ramirez, J. A., Strength of Struts and Nodes in StrutTie Model, Journal of Structural Engineering, ASCE, V. 122, No. 1, 1996, pp. 20-29. 5. Nielsen, M. P., Limit Analysis and Concrete Plasticity, CRC Press LLC, 1999, 936 pp. 6. Rogowsky, D. M., and MacGregor, J. G., Design of Reinforced Concrete Beams, Concrete International, V. 8, No. 8, Aug. 1986, pp. 49-58. 7. MacGregor, J. G., Derivation of Strut-and-Tie Models for the 2002 ACI Code, Examples for the Design of Structural Concrete with Strutand-Tie Models, SP-208, K.-H. Reineck, ed., American Concrete Institute, Farmington Hills, Mich., 2002, pp. 7-40. 8. ACI Committee 318, Building Code Requirements for Structural Concrete (ACI 318-02) and Commentary (318R-02), American Concrete Institute, Farmington Hills, Mich., 2002, 443 pp. 9. Vecchio, F., and Collins, M. P., The Modified Compression Field Theory for Reinforced Concrete Elements Subjected to Shear, ACI JOURNAL, Proceedings V. 83, No. 2, Feb. 1986, pp. 219-231. 10. Vecchio, F. J., and Collins, M.P., Compression Response of Cracked Reinforced Concrete, Journal of Structural Engineering, ASCE, V. 119, No. 12, 1993, pp. 3590-3610. 11. Vecchio, F. J., Disturbed Stress Field Model for Reinforced Concrete: Formulation, Journal of Structural Engineering, ASCE, V. 126, No. 9, 2000, pp. 1070-1077. 12. Matamoros, A., and Wong, K. H., Design of Simply Supported Deep Beams Using Strut-and-Tie Models, ACI Structural Journal, V. 100, No. 6, Nov.-Dec. 2003, pp. 704-712. 13. Quintero-Febres, C.; Parra-Montesinos, G.; and Wight, J. K., Evaluation of Strength Factors for Concrete Struts in Deep Concrete Members, Report No. UMCEE 05-04, Department of Civil and Environmental Engineering, University of Michigan, Ann Arbor, Mich., 2005, 78 pp. 14. Wight, J. K., and Parra-Montesinos, G., Use of Strut and Tie Model for Deep Beam Design as Per ACI 318 Code, Concrete International, V. 25, No. 5, May 2003, pp. 63-70.

Fig. A-1Strut-and-tie model in critical span of Specimen A1. Ftie = Fhorizontal at nodes = Fstruct cos = 1.20P (A-2)

Check of top node and adjacent inclined strut Strut width and strength at intersection with top node, (ws)t and (Fns)t , respectively (ws)t = (0.65 150 mm)sin + hncos = 125 mm (A-3) (Fns)t = 0.85s fc (ws)t bw (A-4)

(Fns)t = 0.85 0.75 22 MPa (215 mm)(150 mm) = 450 kN Equation (A-4) leads to an applied load P = (Fns)t /1.36 = 330 kN. Strength of vertical and horizontal faces of upper right node, (Fnn)t-v and (Fnn)t-h, respectively ( F nn ) t - v = 0.85 n f c ( 0.65 150 mm) b w (A-5)

APPENDIX Figure A-1 shows the strut-and-tie model for the critical span of Specimen A1. The selection of the strut angle and thus, the depth of the top node hn, was made such as to maximize the predicted shear strength. This was done with the purpose of determining the most critical admissible solution for the strut-and-tie model considered, according to Appendix A of the ACI Code. Other admissible solutions will be below this maximum prediction and, thus, they would represent more conservative predictions. From Fig. A-1, it can be seen that the applied load at the top of the beam, P, was divided into two forces of magnitude equal to the beam shears at the left and right beam portions, 0.35P and 0.65P, respectively, acting over a proportional plate width (refer to Reference 14 for detailed information about this procedure). The critical solution was found for a strut angle = 28.5 degrees. For this angle, hn = 195 mm. The forces acting at the nodes and struts can be expressed in terms of the applied load P as follows 0.65 P - = 1.36P Fstrut = ------------sin ACI Structural Journal/July-August 2006 (A-1)

(Fnn)t-v = 0.85 1.0 22 MPa(0.65 150 mm)(150 mm) = 275 kN

( F nn ) t - h = 0.85 n f c h n b w

(A-6)

(Fnn)t-h = 0.85 1.0 22 MPa(195 mm)(150 mm) = 545 kN Equation (A-5) leads to P = (Fnn)t-v /0.65 = 420 kN, while Eq. (A-6) leads to P = (Fnn)t-h/1.20 = 455 kN. Check of bottom node and adjacent inclined strut Strut width and strength at intersection with bottom node, (ws)b and (Fns)b, respectively (ws)b = (150 mm)sin + (180 mm)cos = 230 mm (A-7) ( F ns ) b = 0.85 s f c ( w s ) b b w (A-8)

(Fns)b = 0.85 0.75 22 MPa(230 mm)(150 mm) = 485 kN Equation (A-8) leads to P = 355 kN. 585

Strength of vertical and horizontal faces of bottom node, (Fnn)b-v and (Fnn)b-h, respectively (Fnn)b-v = 0.85 n f c ( 150 mm) b w (A-9)

( F nn ) b - h = 0.85 0.80 22 MPa (180 mm) (150 mm) = 405 kN < 715 kN Equation (A-9) and (A-10) lead to P = 515 kN and P = 340 kN, respectively. The strength of the beam is then governed by the upper strut end, which leads to a maximum applied load of 330 kN and a corresponding shear in the critical span of 215 kN.

(Fnn)b-v = 0.85 0.80 22 MPa(150 mm)(150 mm) = 335 kN (Fnn)b-h = 0.85 n f c h tie b w < A s f y (A-10)

586

ACI Structural Journal/July-August 2006

Potrebbero piacerti anche