ACI STRUCTURAL JOURNAL

Title no. 109-S31

TECHNICAL PAPER

Compressive Strength of Concrete Masonry Beams

by Thomas Ring, Sreekanta Das, and David Stubbs
Concrete masonry construction comprises concrete block units and mortar joints in horizontal and vertical directions, known as bed joints and head joints, respectively. The compressive force in beams acts in the direction parallel to the bed joint. It is assumed that masonry is weaker when compression load is applied in the direction parallel to the bed joint, although the existing research data provide no unique answer. Moreover, there exists a perception that the interruption by block webs in grout continuity has a severe detrimental effect on the compressive strength, although no research validates this. Therefore, this study was undertaken to determine the compressive strength of concrete masonry parallel to the bed joint and how the presence of webs in the grout influences this strength. It was found that the concrete masonry could be stronger when compression load is applied normal to the bed joint and the effect of interruption in grout continuity is insignificant.
Keywords: beam; compressive strength; concrete masonry unit; loading direction; web interruption.

Fig. 1Head and bed joints in beam. aspect ratio of prism constructions used in one study were very different from the others. A literature review shows that two studies were undertaken on masonry beam specimens (Khalaf et al. 1983 and Suter and Fenton 1986). The first study was undertaken by Khalaf et al. (1983), and the primary objective of this study was to investigate the effect of an unfilled head joint on the compressive strength of a masonry beam. The study used beam specimens of three and four courses high and suggested that the compressive strength parallel to the bed joint may, in fact, be higher than the compressive strength normal to the bed joint. The increase in strength parallel to the bed joint may be due to the presence of a strain gradient in the beam sections. The other study was undertaken by Suter and Fenton (1986) and was undertaken to develop a better shape for the equivalent rectangular compression stress block that can accurately predict the strength of masonry beams. This study did not make any attempt, however, to determine the value of compressive strength parallel to the bed joint. Nonetheless, it is generally believed that the compressive strength of masonry construction parallel to the bed joint is much lower than the compressive strength normal to the bed joint. Because there is no general agreement on how the compressive strength of masonry is dependent on the loading direction, various masonry standards treat the effect of loading direction on compressive strength differently. For example, CSA S304.1 (2004a) recommends reducing the compressive strength parallel to the bed joint by 30 to 50%, which seems to be very conservative. Eurocode 6 (2001) suggests determining the compressive strength by loading the block unit in the direction of the compression force as it is expected in the beam assemblage. Other standards, such as MSJC (2008) and AS 3700 (2001), do not address this issue. Therefore, one objective of this study is to determine the compressive strength of masonry constructions when the load is applied parallel to the bed joint and how this strength differs from
ACI Structural Journal, V. 109, No. 3, May-June 2012. MS No. S-2010-179.R2 received November 18, 2010, and reviewed under Institute publication policies. Copyright 2012, 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 March-April 2013 ACI Structural Journal if the discussion is received by November 1, 2012.

INTRODUCTION Walls, beams, lintels, columns, and pilasters are common structural elements built by concrete masonry blocks. The behavior of masonry beams and lintels is typically different from other masonry structural members. In beams and lintels, the compression load acts parallel to bed joint, whereas in most other structural elements, the primary compression load acts normal to the bed joint (Fig. 1). Traditionally, the compressive strength of masonry construction is determined by applying a monotonically increasing concentric axisymmetric axial compressive load on a small segment of a masonry assemblage, which is commonly known as a prism specimen or, simply, a prism. The load in the prism is applied normal to the bed joint. Five or more prisms are tested to determine the characteristic compressive strength of masonry construction (fm ). Previous research indicates that there may be a reduction in the compressive strength of masonry structures when compressive load acts parallel to the bed joint; this is often known as the effect of loading direction on compressive strength of masonry construction. The levels of reduction in compressive strength in those studies, however, are not consistent. For example, Lee et al. (1984), Wong and Drysdale (1985), and Khalaf (1997) considered the effect of the loading direction on the compressive strength of concrete masonry. In these studies, prism specimens of various aspect ratios and various configurations were loaded concentrically in the direction parallel and normal to the bed joint to study the effect of the loading direction on the compressive strength. These studies indicate that the reduction in strength parallel to the bed joint can widely vary from 12 to 55% of the compressive strength normal to the bed joint. Therefore, the reduction in compressive strength of concrete masonry when it is loaded parallel to the bed joint greatly varied from one study to another. This is because the geometry and ACI Structural Journal/May-June 2012

369

Thomas Ring is a Research Assistant at the University of Windsor, Windsor, ON, Canada, where he received his BASc and MASc. Sreekanta Das is an Associate Professor in the Department of Civil and Environmental Engineering at the University of Windsor. He received his BS from the University of Calcutta, Calcutta, India; his MS from the University of Wollongong, Wollongong, NSW, Australia; and his PhD from the University of Alberta, Edmonton, AB, Canada. His research interests include the behavior of structural concrete masonry. David Stubbs is the Director of the Canadian Masonry Design Centre, Mississauga, ON, Canada. He received his BS and MS from McMaster University, Hamilton, ON, Canada.

RESEARCH SIGNIFICANCE This research was undertaken to study the effects of loading direction and web interruption on the compressive strength of concrete masonry. The Canadian standard (CSA S304.1 [2004a]) specifies strength reduction when load is applied parallel to the bed joint, but other design standards make no suggestions. In addition, CSA S304.1 (2004a) recommends further strength reduction when the grout is interrupted by webs of the blocks, although no research data are available to validate this recommendation. The masonry industry and structural engineers are keen in knowing what actually happens when load is applied parallel to the bed joint and when grout continuity is interrupted by the webs. This study answers these questions and, hence, is significant for the safety and economy of structural masonry constructions. EXPERIMENTAL PROGRAMS Test materials and test setup A total of nine full-scale reinforced concrete masonry beams were constructed and tested in this study. The bottom course was made of lintel blocks, whereas the upper three courses were constructed from standard stretcher blocks. The block units were 16 in. (400 mm) long, 8 in. (200 mm) wide, and 8 in. (200 mm) deep (Fig. 2). These beams were four courses high and, as a result, the total height of each beam was 31.5 in. (800 mm). The total length and span for each beam were 25.4 and 24 ft (7720 and 7315 mm), respectively. Each beam was reinforced with six Canadian No. 25 bars, resulting in a total cross-sectional area of 4.65 in.2 (3000 mm2) of longitudinal flexural steel bars. Thus, a large amount of reinforcement was used to ensure that an over-reinforced brittle failure occurred due to crushing of the concrete masonry and/or grout in the compression zone. These bars were placed in three layers at depths of 24.6, 26.4, and 27.8 in. (625, 670, and 705 mm) measured from the top compression fiber of the beam. Special care was taken to ensure the presence of a minimum of 0.4 in. (10 mm) of clear spacing between each layer of bars and, at the same time, ensure that a good bond between the steel and concrete grout developed. Inspection after failure of the beam specimens confirmed that a good bond between the steel and concrete grout material developed and was maintained throughout the test. Steel stirrups were placed in every core, with the exception of the constant moment region at the center of the beam, where no shear reinforcement was provided. The placement of shear reinforcement away from the constant moment region prevented shear failure in the beam specimens. A schematic of the test setup is shown in Fig. 3. A fourpoint bending load was applied through a 100 kip (445 kN) capacity actuator onto a 47 in. (1200 mm) long spreader steel beam. The load and deformation readings were recorded from the actuator load cell and linear variable displacement transformers (LVDTs), respectively. The spreader beam was supported by a pin at one end and a roller on the other end. A 50 kip (222 kN) capacity load cell was placed under each support to record the reactions at two ends of the spreader beam. The distance between these two supports was 39.4 in. (1000 mm); hence, a constant moment region of the same dimension was created in each beam specimen. Two ends of the masonry beam also rested on a set of pin-roller supports, which were 24 ft (7315 mm) apart from each other. A total of nine beams were tested, and the test matrix used in this study is shown in Table 1. Three different amounts of web interruption were used to study the effect of web interACI Structural Journal/May-June 2012

Fig. 2Lintel and stretcher blocks with nominal dimensions.

Fig. 3Beam test setup. the compressive strength when the load is applied normal to the bed joint. A reinforced concrete masonry beam or lintel uses lintel blocks at the bottom course. Then, stretcher blocks are used for the remainder of the courses (Fig. 2). The use of lintel blocks at the bottom course creates a continuous void necessary for the placement of longitudinal tensile reinforcement bars. Knockout blocks, which offer removal of the webs, can be used in the subsequent courses to facilitate a continuous void if required in subsequent courses. After all blocks and reinforcement are laid, the entire assemblage is fully grouted. Because stretcher blocks are used in the top courses (compression zone) of the beams and lintels, the continuity of grout in that zone of the beam is interrupted by the presence of webs of the stretcher block units. There exists a general belief that the interruption in the continuity of grout in the compression zone by the webs of the masonry blocks is detrimental, although no research has been undertaken to validate this. As a result, CSA S304.1 (2004a) recommends a higher reduction factor for the compressive strength of masonry constructions parallel to the bed joint when the grout in the compression zone is interrupted. Other standards and codes do not discuss the effect of grout interruptions on the compressive strength of masonry. Hence, the main objective of this study is to determine the effect of various amounts of grout interruptions on the compressive strength of concrete masonry. 370

ruption on the compressive strength parallel to the bed joint. The web height of the standard stretcher blocks was modified by removing part of the web to simulate three levels of web interruptions in the beam specimens. Figure 2(b) shows an unaltered stretcher block unit and Figure 4(a) and (b) shows the altered stretcher block units used to simulate web interruptions of 47% and 32%, respectively. The partial removal of webs reduced the interruption level in the continuity of grout in the compression zone of the beam (Fig. 5). Hence, the levels of web interruptions in this study were chosen as 100, 47, and 32%, as shown in Table 1. Three beams with each web interruption were tested to study the repeatability in the test data. The cross-sectional views of three different beams are shown in Fig. 5. Figure 5(a) shows the cross section for beams that were built with 100% web interruption. The standard stretcher blocks (Fig. 2(b)) were used in the top three courses, and lintel blocks (Fig. 2(a)) were used in the bottom course. Beams with 47% and 32% web interruptions were built with lintel blocks at the bottom course and modified stretcher blocks (Fig. 4(a) and (b)) in the top three courses, as shown in Fig. 5(b) and (c), respectively. The compression load in beams acts in the direction parallel to the bed joint, and the strength in that direction is called the compressive strength parallel to the bed joint. The notation fmp is used in this paper to indicate this strength. The strength normal to the bed joint could not be obtained from the beam test data. Therefore, five identical prism specimens, as shown in Fig. 6(a), were constructed and tested to determine the compressive strength normal to the bed joint, which is indicated by the notation fmn in this paper (CSA 2004a). The schematic of the test setup used for the prism tests is shown in Fig. 6(b). These prism specimens were made from the same standard stretcher blocks (Fig. 2(b)), grout, and mortar that were used in the construction of the beam specimens. Because unaltered stretcher blocks were used in constructing the prism specimens, the grout in the compression zone was fully (100%) interrupted by the webs. Monotonically increasing axisymmetric axial compression load was applied through the actuator until the prism specimens failed due to crushing. The beams and prisms were constructed by an experienced mason from the Canada Masonry Design Centre located in Mississauga, ON, Canada. Type S mortar, as recommended in CSA S304.1 (2004a), was used with a volume ratio of 1:0.5:4 for cement:lime:sand. Mortar strength was obtained by testing mortar cubes of 2 x 2 x 2 in. (50 x 50 x 50 mm), which were built in nonabsorbent molds. Three mortar cubes were tested at 28 days, and three more cubes were tested on the day of the corresponding beam test. The beams were grouted after 1 week of curing. A fine grout mixture design was used in all beam specimens. The required amount of water was added to achieve a high slump of 10.6 in. (270 mm) such that the grout flowed well into the cells of the blocks. No admixtures were used in this study. The strength of the grout was obtained from cylinder tests. For each batch of grout mixture, three cylinders were cast in nonabsorbent plastic molds. Three block cells were also filled with the same grout mixture. They were then used to obtain the cylindrical grout specimens by coring them out from the block cells. These cylindrical specimens were used to determine the in-place strength of the grout. Cylindrical grout specimens were chosen because they are easy to core from in-place grouted blocks. ACI Structural Journal/May-June 2012

Table 1Test matrix for beam specimens

Specimen Beam 1 Beam 2 Beam 3 Beam 4 Beam 5 Beam 6 Beam 7 Beam 8 Beam 9 3 32 2 47 4.65 (3000) 1 100 Group Web interruption, % Steel area, in.2 (mm2)

Fig. 4Modified stretcher blocks.

Fig. 5Beam cross sections depicting web interruptions.

Fig. 6Schematic of prism and test setup. 371

Fig. 7LVDTs placed at several locations. Instrumentation Various instruments were used for acquiring the values of loads and reactions, deflections, and strains from the beam specimens. LVDTs were used to measure the displacements under the beam at the midspan of the beam, the longitudinal displacement at the roller end of the beam, and the out-ofplane displacements at the midspan and at 2.8 in. (70 mm) below the top surface of the beam (Fig. 7). They were installed at the same locations for all the beams. Spring-loaded linear potentiometers (LPs) attached by wire were used to obtain the strain gradient that developed during loading of the beam specimens. Six LPs were installed at a spacing of 3 in. (75 mm) on the face of the beam (Fig. 8(a)). Two more LPs were also mounted on the top surface of the beam to obtain compressive strains at the topmost concrete fibers (Fig. 8(b)). These LPs were mounted on small aluminum base plates with two holes used for easier mounting. Tapcon screws were attached and used to mount the base plates to the beam. A nylon-coated steel wire was attached between one end of the LP and the screw. This created a gauge length of 20.5 in. (520 mm). Strain gauges (SGs) of 1/8 in. (3 mm) gauge length were used to acquire the strain data from the flexural steel bars. One strain gauge was installed at the midspan of each bar (Fig. 9). In addition, two more strain gauges were installed on the bar at the middle row at a distance of 47 and 94 in. (1200 and 2400 mm). As a result, a total of eight SGs for each beam specimen were used. These SG data were used primarily to monitor if a brittle failure in the beam specimens actually occurred. Four load cells were used to determine the reactions under the steel spreader beam and the concrete masonry beam specimen. The loading actuator had its own load cell and LVDT. The load was applied through a loading actuator of 100 kip (445 kN) capacity. A computerized data acquisition system was used to acquire all the test data. The test setup for the prism specimens is shown in Fig. 6(b). Compression load was applied through a universal loading actuator. The load data were acquired through the same data acquisition system that was used in the beam specimens. TEST PROCEDURE A four-point bending load was applied to the beam specimens (Fig. 3). The load was applied by a universal loading actuator through a spreader steel beam, which was supported on pin-roller supports. The load was increased monotonically in displacement control. The loading was held a few times to monitor the crack initiations and growth in the beam specimen. The loading was continued until the beam specACI Structural Journal/May-June 2012

Fig. 8LPs mounted on side and top faces.

Fig. 9SGs installed on steel reinforcing bar. A total of 10 regular stretcher blocks were chosen at random from the pallets of blocks for compression testing according to CSA A165.1 (2004b). All the blocks were obtained from the same production run. Lintel blocks were not tested for their strength because they were used in the bottom course only where only tensile stress existed. The reinforcement was also tested in accordance with ASTM A615/A615M-09b (2009) to determine the yield and ultimate strength and yield strain. 372

imen failed. Failure occurred due to crushing of the concrete block masonry at the top compression fiber. A typical loadversus-midspan deflection of the beam specimens is shown in Fig. 10. In this figure, load-deformation behavior for one beam that is more representative of the group (web interruption) was chosen. TEST RESULTS Material properties The summary of properties obtained from the mortar, grout, and stretcher block unit is shown in Table 2. The yield strength, tensile strength, and yield strain for the steel reinforcement bar were obtained at 71.4 and 91.7 ksi (492 and 632 MPa) and 0.0026, respectively (ASTM 2009). The compressive strength of masonry assemblage normal to the bed joint (fmn) obtained from the prism specimens is

2.15 ksi (14.8 MPa) with a coefficient of variation (COV) of 5.7%. Load-deformation behavior of beams Each beam was loaded until failure. A total of nine fullscale beams in three different groups were tested (Table 1). Table 3 shows the summary of the test results. The maximum load value recorded from the three tests of a beam group varied slightly, even though the beams were made of the same dimensions and same materials and were cured the same way. This is expected because it is typical to have some variability in masonry constructions. The maximum deflection under the beam at its midspan also varied within each group. The primary reason for this difference depends on when the test was discontinued. For some beams, the test was continued until larger damage occurred, whereas in other beam specimens, the test was discontinued at relatively smaller damage levels. The typical load-versus-midspan deflection behaviors of various beams are shown in Fig. 10. The load-deformation curves exhibit very similar characteristics for all the beams of all three groups. At approximately 13.4 kips (60 kN), there is a change in slope, which is due to the initiation of flexural cracking of the masonry in tension. Before and after this load, the relationship is largely linear. A few dips that are present in the load-deformation curves correspond to when the load was held for monitoring the crack initiations and crack growths. The load-deformation plots indicate that the amount of web interruption has no influence on the slope of the early part of the load-deflection curves. The loaddisplacement plot obtained from beams with 100% web interruption (Group 1) lies in between those of beams with 47% web interruption (Group 2) and 32% web interruption (Group 3) at a higher load level. Failure modes of beams All of the beams failed due to crushing of the concrete masonry in the compression zone at the top of the beam. The failure mode was brittle, which was expected because the beams were over-reinforced. Crack initiation and growth were similar in all the beams. First, flexural cracks initiated at the midspan and along the lintel block head joints in the bottom course (Fig. 11(a)). Flexural cracks were visible at a moment value of approximately 90 to 100 kN-m (66 to 73.5 kip-ft). The flexural cracks then propagated through the upper course and caused the splitting of the units in this course. These cracks finally stopped progressing when they reached closer to the

Fig. 10Typical load-deflection curves. Table 2Some material properties

Beam test-day values Materials Block Mortar Grout core Grout cylinder Strength, psi (MPa) 3680 (25.4) 3090 (21.3) 3790 (26.1) 3440 (23.7) COV, % 8.1 9.3 5.5 2.6

Table 3Test results obtained from beam specimens

Load applied, kips (kN) Beam number 1 2 3 4 5 6 7 8 9 32 47 100 Web interruption, % Maximum 75.1 (334) 82.1 (365) 84.3 (375) 76.7 (341) 72.8 (324) 74.4 (331) 93.1 (414) 88.6 (394) 87.5 (389) 89.7 (399) 3.3 74.6 (332) 2.6 80.4 (358) 6.0 Average COV, % Midspan maximum deflection, in. (mm) 1.3 (34) 1.5 (38) 1.6 (40) 1.5 (39) 1.3 (34) 1.3 (34) 1.7 (44) 1.5 (38) 1.5 (39)

ACI Structural Journal/May-June 2012

373

Fig. 13Typical strain distributions (Beam 2).

Fig. 11Crack initiations and progressions. Fig. 14Equivalent stress block used in calculations. the flexural reinforcement bars at the midspan of the beam (Fig. 9). The strain distributions obtained from each beam of a specific group were very similar. The strain distribution was then used to locate the neutral axis and its depth. The average depths of the neutral axis measured from the top compression fiber for the three sets of beams were 15.0, 14.5, and 14.4 in. (380, 367, and 365 mm), respectively. Figure 13 shows typical distributions of strains along the depth of the beam. The depth and compressive strain values are indicated by a negative sign. The maximum masonry strain recorded at the top masonry fiber ranged from 0.0028 to 0.0032. However, no pattern for the maximum strain value versus percentage interruption was found. The maximum strains recorded from the flexural steel bars were much lower than the yield strain of 0.0026. The maximum strain recorded from two bars reached 0.0024; this is lower than the yield strain value of 0.0026. This, in addition to the failure mode, ensured that these beam specimens failed in a brittle manner. Compressive strength Effect of web interruptionThe beam test data were used to determine the masonry compressive strength parallel to the bed joint (fmp). The equivalent concrete compression stress block and strength calculations, as recommended in CSA S304.1 (2004a) and shown in Fig. 14, were used. However, the material resistance factors (fm and fs) and strength reduction factor (c) were excluded from the calculations. It should be noted that the compressive strength of ACI Structural Journal/May-June 2012

Fig. 12Crushing failure of beams. neutral axis. Shear cracks also initiated in the head joints in the bottom course away from the constant moment region. The shear cracks formed at approximately 35 in. (890 mm) away from the beam supports (Fig. 11(b)). These cracks progressed at approximately 45 degrees into the second and third courses. Shear cracks initiated at approximately a 230 to 250 kN-m (169 to 184 kip-ft) moment. Nonetheless, the flexural and shear cracks did not cause the ultimate failure of the beams. Rather, the beam specimens failed due to crushing of the masonry in the compression zone in the constant moment region. The face shells of the masonry blocks in this area separated from the grout and, often, the face shell fell off (Fig. 12). The test was discontinued when the face shell crushed, even though in most cases the grout was completely undamaged. Strain distributions The strain distribution along the height of the beam section at the midspan was determined from the displacement readings of the LPs installed on the beam (Fig. 8) and strain readings obtained from the SGs installed on 374

the masonry parallel to the bed joint calculated using this method is not called the characteristic compressive strength (fm rather, it is called compressive strength (fmp). This is ); because compressive strength determined from this method does not follow the traditional definition of the characteristic compressive strength (fm (CSA S304.1 2004a). ) The maximum moment M applied to the beam specimen was equated to the internal moment resistance, as shown in Eq. (1).

a M = Cd 2

(1)

where C is the compressive force; d is the depth of the reinforcement; a = b1c; b1 = 0.8; and c is the neutral axis depth measured from the strain distributions. Equation (1) can be reorganized as Eq. (2).

Fig. 15Effect of web interruption on compressive strength. in compressive strength is approximately 29%. Hence, this study shows that the current Canadian standard is conservative on this issue. Other standards and codes do not address this issue. Effect of loading directionCompression loading in beams is parallel to the bed joint; hence, the compressive strength of the concrete masonry normal to the bed joint (fmn) could not be obtained from the beam test data. Hence, prism tests were conducted to determine the compressive strength of similar concrete masonry construction but normal to the bed joint. As discussed previously, five prism specimens with 100% web interruptions were built and tested. The average compressive strength normal to the bed joint (fmn) obtained from the prism tests was 2.15 ksi (14.8 MPa) with a COV of 5.7%. The average compressive strength parallel to the bed joint (fmp) obtained from the beams, however, was 3280 psi (22.6 MPa) with a COV of 6%. Hence, it is found that the compressive strength parallel to the bed joint (fmp) is in fact much higher (by approximately 53%) than the compressive strength normal to the bed joint (fmn). CSA S304.01 (2004a), however, recommends strength reduction when compression load is applied parallel to the bed joint and the minimum reduction is 30%. Hence, this study shows that the current Canadian standard is very conservative on this issue as well. It should be noted, however, that the comparison between the two compressive strengths presented in this paper is applicable for concrete masonry constructions when grout in the compression zone is fully (100%) interrupted. CONCLUSIONS The following conclusions are made based on the test results obtained from this study. Therefore, the conclusions are limited to the type of beam and prism specimens used in this study. 1. The compressive strength of grouted concrete masonry parallel to the bed joint is higher than the compressive strength of grouted concrete masonry loaded normal to the bed joint when the grout is fully (100%) interrupted. 2. The beam test data show that the relationship between compressive strength parallel to the bed joint and the percent of web interruption is almost linear. 3. The compressive strength parallel to the bed joint decreases as the amount of web interruption increases. However, the reduction is not as severe as the current 375

C=

M 0.8c (d ) 2

(2)

However, the compressive force C can also be calculated from the equivalent rectangular stress block (Fig. 14).

C = 0.85 fmp ab

(3)

where b is the beam width. Equating C from Eq. (2) and (3), the compressive strength of the masonry parallel to the bed joint (fmp) was obtained, as shown in Eq. (4).

M 0.8c d 2 p fm = 0.85ab

(4)

The masonry compressive strength parallel to the bed joint (fmp) is then plotted against the percent of web interruption, as shown in Fig. 15. A general trend on how the compressive strength changes as the amount of web interruption changes can be found in this figure. This is shown by the linear fit of all the compressive strength data. The linear fit of the test data represents the actual test data reasonably well. It is found that the value of the compressive strength of the masonry parallel to the bed joint (fmp) is 3510 psi (24.2 MPa) when there is no interruption in the grout continuity in the compression zone if the trend is assumed to be the same when the web interruption is lower than 32%. The value of compressive strength is slightly less (3280 psi [22.6 MPa]) when the grout in the compression zone is fully (100%) interrupted. As a result, the drop in compressive strength is only approximately 7%. This value is substantially lower than what is recommended in CSA S304.01 (2004a). As per this standard, the reduction ACI Structural Journal/May-June 2012

Canadian standard specifies. The reduction in compressive strength found in this study was only 7%.
This study was completed with financial assistance from the Natural Science and Engineering Research Council of Canada (NSERC) and the Canada Masonry Design Centre (CMDC). CMDC also provided technical help and guidance. Additional support was provided by Con-tact Masonry Ltd., located in Oldcastle, ON, Canada.

ACKNOWLEDGMENTS

b C c d fm fmn fmp M T c fm fs

= = = = = = = = = = = =

width of beam compressive force depth of neutral axis measured from top compression fiber depth of flexural reinforcement measured from top compression fiber characteristic compressive strength of masonry construction compressive strength of concrete masonry normal to bed joint compressive strength of concrete masonry parallel to bed joint maximum moment applied tension force in steel strength reduction factor when load is applied parallel to bed joint material resistance factor for masonry material resistance factor for reinforcement steel

NOTATION

AS 3700-2001, 2001, Masonry Structures, Standards Australia, Sydney, NSW, Australia, 191 pp.

REFERENCES

ASTM A615/A615M-09b, 2009, Standard Specification for Deformed and Plain Carbon-Steel Bars for Concrete Reinforcement, ASTM International, West Conshohocken, PA, 6 pp. CSA S304.1-04, 2004a, Design of Masonry Structures, Canadian Standards Association (CSA), Mississauga, ON, Canada, 148 pp. CSA A165.1-04, 2004b, Concrete Block Masonry Units, Canadian Standards Association (CSA), Mississauga, ON, Canada. DD ENV 1996-1-3:2001, 2001, Eurocode 6: Design of Masonry Structures, European Committee for Standardization (ECS), Brussels, Belgium, 34 pp. Joint ACI-ASCE-TMS Committee 530, 2008, Building Code Requirements and Specifications for Masonry Structures (TMS 402/ ACI 530/ASCE 5) and Specification for Masonry Structures (TMS 602/ ACI 530.1/ASCE 6) and Related Commentaries), American Concrete Institute, Farmington Hills, MI, 236 pp. Khalaf, F.; Glanville, M.; and El Shahawi, M., 1983, Study of Flexure in Reinforced Masonry Beams, Concrete International, V. 5, No. 6, June, pp. 46-53. Khalaf, F. M., 1997, Blockwork Masonry Compressed in Two Orthogonal Directions, Journal of Structural Engineering, ASCE, V. 123, No. 5, pp. 591-596. Lee, R.; Longworth, J.; and Warwaruk, J., 1984, Concrete Masonry Prism Response due to Loads Parallel and Perpendicular to Bed Joints, Structural Engineering Report, Department of Civil and Environmental Engineering, University of Alberta, Edmonton, AB, Canada, pp. 1-101. Suter, G., and Fenton, G., 1986, Flexural Capacity of Reinforced Masonry Members, ACI JOURNAL, Proceedings V. 83, No. 1, Jan.-Feb., pp. 127-136. Wong, H., and Drysdale, R., 1985, Compression Characteristics of Concrete Block Masonry Prisms, ASTM STP 871, pp. 167-177.

376

ACI Structural Journal/May-June 2012