40 JULY 2015 | Ci | www concreteinternational com Learning through Hands On Reinforced Concrete Beam Experiments Students in senior level design class gain valuable insights into beam behavior by Ian[.]
Learning through Hands-On Reinforced Concrete Beam Experiments Students in senior-level design class gain valuable insights into beam behavior by Ian N Robertson and Gaur P Johnson R einforced concrete design can be a relatively dry topic for instructors and their students While the introduction of outstanding reinforced concrete construction projects and current laboratory research projects can pique students’ interest, the basic theory and design requirements for reinforced concrete beams, for example, can be difficult to convert to an exciting topic Possibly more importantly, students seldom gain an appreciation for the anticipated performance of reinforced concrete members when loaded to failure It is difficult for them to appreciate the variability of actual performance or the intrinsic margin of safety built into the ACI 318-111 design requirements In an attempt to augment classroom instruction, a laboratory section was added to the senior-level reinforced concrete design class at the University of Hawaii (UH) at Manoa Students work in groups of six or seven, and each group fabricates and tests a large-scale reinforced concrete beam by loading it to failure Each beam is designed with a particular objective in mind, with the intent of highlighting a specific failure mode During the Fall 2014 semester, 88 students participated in this course and laboratory section, constructing 13 beams with various types and amounts of flexural and shear reinforcement In addition to conventional steel deformed reinforcing bars used in most of the beams, two beams were fabricated with glass fiber-reinforced polymer (GFRP) bars as flexural reinforcement to demonstrate the differences in design and behavior All beams were cast using the same concrete mixture, except that polypropylene (PP) fibers were added to the concrete prior to placement of two beams All beams were tested to failure in three-point bending, either with a relatively long span to induce flexural failure or a short span to induce shear failure 40 JULY 2015 | Ci | www.concreteinternational.com Beam Details All beams had rectangular, 10 in wide and 18 in deep (254 x 457 mm) cross sections Cover to the reinforcing steel was set at in (25 mm) on all sides Longitudinal and transverse reinforcement varied per Table Each student group was responsible for cutting and bending the longitudinal reinforcement required for their beam Because the amount of shear reinforcement varied significantly between beams, each group was assigned an equal portion of the total stirrup fabrication required for the 13 beams Beams through were designed to demonstrate various flexural limit states Beams 1, 2, and were designed to demonstrate flexural response of beams with less than minimum reinforcement, tension-controlled, and transition zone flexural reinforcement, respectively Beam was designed per ACI 440.1R-062 recommendations, and the reinforcement was selected to provide the same design flexural strength as Beam with steel bars (to demonstrate that varying the modulus of elasticity of the reinforcement affects the flexural behavior of beams) Beam 3F was identical to Beam except that PP fibers were added to the concrete (to evaluate the effects of fibers on beam flexural behavior) Beam was identical to Beam 2, except that two of the three tension bars were spliced at midspan using splice lengths that did not comply with ACI 318-11 Code requirements (to demonstrate the importance of meeting splice requirements) Finally, Beam had the same tension reinforcement as Beam 5, but it also included significant compression reinforcement (to demonstrate that compression reinforcement can modify the beam response from transition zone to tension-controlled behavior) Beams through 12 were designed to demonstrate shear limit states Beams and had no shear reinforcement, but Table 1: Beam span, reinforcement, and parameters of interest Beam No Span, in Bottom bars Top bars Shear reinforcement Comments 128 (2) No (2) No No @ in o.c Minimum reinforcement, As < As,min 128 (3) No (2) No No @ in o.c Tension-controlled flexure 128 (6) No GFRP (2) No No @ in o.c Flexure with GFRP bars 3F 128 (6) No GFRP (2) No No @ in o.c Effect of PP fibers in flexure 128 (3) No (2) No No @ in o.c Inadequate splice at midspan* 128 (5) No (2) No No @ in o.c Transition zone flexure 128 (5) No (3) No No @ in o.c Effect of compression reinforcement 78 (4) No (2) No None Shear strength of concrete, Vc 78 (4) No (2) No Fibers Effect of PP fibers in shear 78 (4) No (2) No No @ in o.c Minimum shear reinforcement 10 78 (4) No (2) No No @ in o.c Effect of shear reinforcement 11 78 (5) No (3) No No @ in o.c Over-reinforced in shear 12 78 (5) No (3) No No @ in o.c Effect of bar anchorage† Two of the No bottom bars were spliced with 10 in laps at midspan No hook at end of No bottom bars Note: in = 25.4 mm * † the former comprised plain concrete and the latter comprised concrete with PP fibers (to evaluate the effects of fibers on shear strength) Beams 9, 10, and 11 had increasing amounts of shear reinforcement—from slightly more than Av,min (per ACI 318-11) in Beam to more than Av,max in Beam 11 (the steel contribution exceeded the maximum of 8√fc′(bwd)) While all of the other beams had hooks at both ends of all tension reinforcement, Beam 12 had straight bars with no hooks (to demonstrate the necessity for well-anchored tension reinforcement to achieve the beam shear capacity) The groups assembled their own reinforcing cages and placed them in the beam forms with appropriate cover blocks and lifting hoops (Fig 1) Each group was also responsible for placing, vibrating, screeding, and curing the concrete in their beam The concrete required for all beams was delivered in a single mixer truck After the as-delivered concrete had been placed in 11 of the beams, PP fibers and a high-range water-reducing admixture were added to the remaining concrete, and the truck was run at mixing speed for minutes Prior to testing any of the beams, each student group was required to submit a draft report including the ACI 318-11 Code nominal strength for (a) all beams except those with GFRP reinforcing bars and those with PP fibers They were also required to predict the failure load for each of these beams, taking into account the level of conservatism inherent in ACI Code predictions The group with the closest predictions was promised prizes: a $25 music gift card for each group member (courtesy of the instructor) (b) Fig 1: Beam fabrication: (a) closing forms with reinforcing cages assembled by students; and (b) concrete placement and finishing www.concreteinternational.com | Ci | JULY 2015 41 Table 2: Table 3: Concrete material properties Reinforcing steel properties Concrete mixture Unit weight,* lb/ft3 Compressive strength* ( fc′), psi Modulus of rupture† ( fr), psi As-delivered 145.4 5090 620 With PP fibers 143.1 5290 490 Average of three x 12 in (150 x 300 mm) cylinder tests † Average of two x x 24 in (150 x 150 x 600 mm) beam tests Note: lb/ft3 = 16 kg/m3; psi = 0.00689 MPa Bar size 140 in No P 70 in No * No 64 in (a) 70 in 140 in P 10 in 14 in 18 in d´ 74.0 10 in 75.0 14 in 72.0 119.0 18 in 114.0 71.5 N/A* N/A* N/A* Exceeded capacity of test frame Note: ksi = 6.89 MPa * 90 in P 10 in 14 in 64 in Tensile strength ( fu), ksi 60.0 No 128 in No 45 in d2 d Yield strength ( fy), ksi 18 in d´ d2 d 39 in 128 in (b) (a) 78 in Fig 2: Test beam dimensions and loading: (a) for long beams; and (b) for short beams (Note: in = 25.4 mm) 90 in Material Properties Concrete 45 in P 10 in d´ Concrete was ordered from a local ready mixed concrete d2 d 14 in 18 in supplier, with a specified compressive strength of 4000 psi (27.6 MPa) During 39 in the concrete placement, a number of standard x 12 in (150 x 300 mm) cylinders were made from 78 in samples (b) of the as-delivered concrete and the mixture modified with PP fibers One of the cylinders from the latter mixture was immediately washed out to determine the actual fiber content With 1.133 oz (32 g) of fiber collected from the cylinder, the fiber content in the concrete was computed to be 9.74 lb/yd3 (5.78 kg/m3) Three cylinders were tested at 28 days to determine the compressive strengths per ASTM C39/C39M, “Standard Test Method for Compressive Strength of Cylindrical Concrete Specimens”; see Table Four x x 24 in (150 x 150 x 600 mm) beams were fabricated (two from each mixture type) to evaluate modulus of rupture per ASTM C78/C78M, “Standard Test Method for Flexural Strength of Concrete (Using Simple Beam with Third-Point Loading).” It is noteworthy that the presence of fibers increased the compressive strength of the concrete slightly, but also appeared to have a detrimental effect on bending strength (Table 2) A larger test program would be required to determine if this is a valid result Reinforcing steel Coupons of the various sizes of reinforcing steel were tested in tension to determine yield and tensile strengths 42 JULY 2015 | Ci | www.concreteinternational.com (Table 3) Because of limitations of the test equipment, the No bar tests did not reach the tensile strength, and the No bars could not be tested The yield strength of the No bars was assumed to be the same as the No bars for purposes of calculation of beam capacity per ACI 318-11 The GFRP bars were not tested, so the elastic modulus and tensile strength and strain reported by the manufacturer were used in determining the GFRP beam capacities Test Setup The beams were tested in three-point bending over a 128 in (3250 mm) span for the long beams with anticipated flexural failure and over a 78 in (1980 mm) span for the short beams with anticipated shear failure (Fig 2) The load was applied through a 14 in (355 mm) wide bearing plate by a 300 kip (1335 kN) MTS actuator Linear potentiometer displacement transducers were placed over the supports and adjacent to the loading plate to measure midspan deflection Load was applied under displacement control, and the load was periodically held to allow for crack identification (Fig 3) Flexural Limit State Test Results Figure shows midspan load-deflection curves for the seven long beams with various flexural failure mechanisms, while Fig shows each beam after testing As expected, flexural strength increased with flexural reinforcement (from Beam to Beam to Beam 5) However, the relatively low deflection exhibited by Beam (Fig 4) and the compression d´ d2 d Beam Beam Beam Fig 3: Students highlight cracks forming during beam testing Beam 3F 160 140 Midspan Load, kip 120 Beam ρ = 2.46% 100 Beam Beam 3F ρ = 0.96% GFRP + PP fiber Beam ρ = 0.96% GFRP Beam ρ = 0.83% 80 60 40 0.5 1.5 2.5 Midspan Deflection, in Beam Beam Beam ρ = 0.83% with 10 in splice 20 Beam ρ = 2.46%; ρ' = 1.48% Beam ρ = 0.14% 3.5 Fig 5: Long beams after testing to failure Fig 4: Load-deflection response for long beams (Note: kip = 4.45 kN; in = 25.4 mm) failure in the concrete (Fig 5) demonstrates the low ductility associated with transition zone flexural members The response of Beam (Fig 4) and the lack of a crushing failure in the concrete (Fig 5) shows how the addition of compression reinforcement can convert a non-ductile transition zone beam to a tension-controlled beam The poor performance of Beam 4, with identical reinforcement to Beam 2, demonstrated the consequence of an inadequate splice at the location of maximum moment demand Bond failure along the short splice resulted in a large tension crack developing at one end of the splice along with significant splitting of the cover within the length of the splice as the bars pulled through the concrete (Fig 5) While Beam 3, which had GFRP tension reinforcement, exhibited strength comparable to the equivalent steel-reinforced Beam 2, it also had considerably lower post-cracking stiffness and overall ductility (Fig 4) Beam 3F also had GFRP reinforcement, but included 2.25 in (57 mm) long PP fibers at a dosage of 9.74 lb/yd3 (5.78 kg/m3) While it exhibited a slightly higher cracking strength than Beam 3, it had essentially the same post-cracking stiffness, strength, and ductility In both Beam and Beam 3F, the GFRP bars remained intact and within the elastic range up to crushing of the compression zone concrete The beams therefore exhibited almost complete recovery of deflection after the test in contrast to the permanent deformation due to yielding of the reinforcement, as seen in Beam (Fig 5) Shear Limit State Test Results Figure shows midspan load-deflection curves for the six short beams, while Fig shows each beam after testing Beam 7, with no shear reinforcement, provides a measure of the concrete shear capacity The effect of adding PP fibers at 9.74 lb/yd3 (5.78 kg/m3) is evident in the increased shear capacity of Beam As expected, the addition of increasing amounts of shear reinforcement in Beams 9, 10, and 11 is associated with increases in shear capacity Beams 11 and 12 had more shear reinforcement than the maximum specified by ACI 318-11, resulting in a compression zone failure (Fig 7) Although intended to demonstrate the effect of inadequate anchorage at the ends of the tension reinforcement, Beam 12 performed almost identically to Beam 11 This is possibly because the embedment required to develop a No bar that’s partially confined by a bearing plate is less than the development length calculated per ACI 318 www.concreteinternational.com | Ci | JULY 2015 43 250 Beam 12: Vs = 9.4 fc´ with straight bottom bars 200 Beam 11: Vs = 9.4 fc´ with hooked bottom bars Midspan Load, kip Beam 150 Beam 10: Vs = 3.5 fc´ 100 Beam 9: Vs = 1.2 fc´ Beam 8: Vs = with PP fibers 50 Beam Beam 7: Vs = 0 0.2 0.4 0.6 Midspan Deflection, in 0.8 Fig 6: Load-deflection response for short beams (Note: kip = 4.45 kN; in = 25.4 mm) Beam Comparisons with Predicted Strengths Table provides a comparison between the beam test results and the predicted strengths Predicted strengths were based on the ACI 318-11 Code and ACI 440.1R-06 for beams with steel and GFRP flexural reinforcement, respectively The nominal capacities are based on the specified concrete compressive strength of 4000 psi (27.6 MPa), steel reinforcement yield strength of 60 ksi (414 MPa), and the manufacturer’s specified properties for the GFRP bars As neither ACI 318-11 nor ACI 440.1R-06 include adjustments for the addition of PP fibers, calculations for Beam 3F and Beam include concrete, steel, and GFRP properties only The ratio between the test results and the nominal capacities predicted per the ACI documents shows that for all beams with flexural limit states (Beams to 6), the test result consistently exceeds the predicted nominal capacity by 38 to 115% This provides a considerable margin of safety, which is obviously increased by inclusion of the appropriate strength reduction factor This is particularly evident for the transitionzone Beam 5, which requires a smaller reduction factor than the tension-controlled beams, and the GFRP-reinforced Beams and 3F, which require an even smaller reduction factor than the beams with steel reinforcement Including the strength reduction factor, the test results exceed the ACI prediction by 55 to 158% This relatively large margin of safety in the performance of beams subjected to flexural limit states was considerably greater than students had anticipated in their pre-test estimates For short beams (Beams to 12), the test results were again significantly higher than the nominal capacities predicted per ACI documents While the nominal strength of both Beam 11 and Beam 12 were relatively close to the test results, it’s important to note that the nominal shear strength provided by shear reinforcement in these beams exceeded the code specified maximum of 8√fc′(bwd) Beams and had no stirrups and thus had low shear capacities However, the test results exceeded the nominal capacities by 49% and 108%, respectively Beam 8, which contained PP fibers, attained a 44 JULY 2015 | Ci | www.concreteinternational.com Beam 10 Beam 11 Beam 12 Fig 7: Short beams after testing to failure peak load that was 40% greater than the peak load reached by Beam 7, which comprised plain concrete As with flexural test results, the high margins of safety evident in the ACI 318-11 shear design requirements was greater than the students had anticipated Summary and Conclusions The hands-on experience of building and testing largescale reinforced concrete beams gave students in the seniorlevel reinforced concrete design course at UH Manoa valuable exposure to design and construction aspects of reinforced concrete Perhaps more importantly, the tests physically demonstrated behaviors that would normally be covered only in the abstract Primary conclusions drawn from the beam test results include the following: Predictions of nominal flexural and shear capacity made per ACI documents provide significant margins of safety when compared with test results For flexural members, the margin of safety ranged from 38 to 115%, while for shear members meeting ACI 318-11 Code requirements, the margin of safety ranged from 49 to 75%; •• Table 4: Comparison between test results and predicted strengths per ACI 318-11 and ACI 440.1R-06 Beam No. Span length, in 128 128 128 3F 128 * † Ptest, kip ACI Pn, kip Ptest/Pn fPn, kip Ptest/fPn 14.4 6.7 64.0 37.3 2.15 6.0 2.40 1.72 33.6 1.90 74.8 52.5 1.42 34.1 2.19 88.1 52.5 1.68 34.1 2.58 128 32.5 20.0 1.63 18.0 1.81 128 120.0 86.8 1.38 69.4 1.73 128 136.0 97.2 1.40 87.5 1.55 78 56.6 38.0 1.49 28.5 1.99 † 78 79.0 38.0 2.08 28.5 2.77 78 104.8 60.0 1.75 45.0 2.33 10 78 163.3 104.0 1.57 78.0 2.09 11 78 220.9 208.6 1.06 156.5 1.41 12 78 232.6 208.6 1.12 156.5 1.49 Bending strength of Beams and 3F (with GFRP bars) is based on recommendations in ACI 440.1R-06 † The predicted strengths include no considerations for the PP fibers Note: kip = 4.45 kN * PP fibers at 9.74 lb/yd (5.78 kg/m ) to the concrete •• Adding mixture increased the concrete shear capacity by 40%, but •• had only a marginal impact on the flexural response; and Replacement of steel tension reinforcement with GFRP bars required a 40% increase in reinforcement crosssectional area to provide equivalent flexural design strength The lower stiffness of the GFRP bars also resulted in larger deflections at the service load condition Acknowledgments The authors appreciate the significant assistance of laboratory technicians Mitchell Pinkerton and Kim Hertzog as well as their student assistants during the various stages of beam fabrication and testing The authors also acknowledge the assistance of Michael Bell, who served as teaching assistant for this course, and Benjie Batangan, who helped in the design of the beams Finally, the contributions of Kimo Scott of OK Hardware, Inc., who generously donated the Aslan 100 GFRP bars and Forta-Ferro PP fibers for the test program, are gratefully acknowledged References ACI Committee 318, “Building Code Requirements for Structural Concrete (ACI 318-11) and Commentary,” American Concrete Institute, Farmington Hills, MI, 2011, 503 pp ACI Committee 440, “Guide for the Design and Construction of Structural Concrete Reinforced with FRP Bars (ACI 440.1R-06),” American Concrete Institute, Farmington Hills, MI, 2006, 44 pp Note: Additional information on the ASTM standards discussed in this article can be found at www.astm.org ACI member Ian N Robertson is a Professor of structural engineering at the University of Hawaii at Manoa, HI, where he teaches senior- and graduate-level courses on structural design A licensed structural engineer in Hawaii, he received his MS and PhD in civil engineering from Rice University, Houston, TX He has 30 years of research and design experience in the performance of structures during extreme events including earthquakes, hurricanes, and tsunamis, and has participated in a number of post-disaster reconnaissance surveys His research interests include response of structures to extreme loading, durability of concrete materials, and the long-term performance of reinforced concrete structures ACI member Gaur P Johnson is an Assistant Professor at the University of Hawaii, Honolulu, HI He received his BS, MS, and PhD in civil engineering from the University of Hawaii, and is a licensed structural engineer in Hawaii He is a member of ACI Committee 437, Strength Evaluation of Existing Structures, and ACI Subcommittees 562-B, Loads, and 562-E, Education His research interests include performance evaluation of deteriorating reinforced concrete structures and field instrumentation for seismic and long-term monitoring Selected for reader interest by the editors www.concreteinternational.com | Ci | JULY 2015 45 JULY 2015 V 37 No & s r o o l F ndations Fou 37 2015 CFA Awards ... compression reinforcement can convert a non-ductile transition zone beam to a tension-controlled beam The poor performance of Beam 4, with identical reinforcement to Beam 2, demonstrated the consequence... 20 Beam ρ = 2.46%; ρ'' = 1.48% Beam ρ = 0.14% 3.5 Fig 5: Long beams after testing to failure Fig 4: Load-deflection response for long beams (Note: kip = 4.45 kN; in = 25.4 mm) failure in the concrete. .. testing Beam 3F 160 140 Midspan Load, kip 120 Beam ρ = 2.46% 100 Beam Beam 3F ρ = 0.96% GFRP + PP fiber Beam ρ = 0.96% GFRP Beam ρ = 0.83% 80 60 40 0.5 1.5 2.5 Midspan Deflection, in Beam Beam Beam