Self-Consolidating Concrete for Cast-in-Place Bridge Components (2016)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

30 Results, Interpretation, and Application This chapter presents test results and provides interpreta- tion of the findings. Proposed changes to the current AASHTO LRFD design and construction specifications are presented in Attachment A. Proposed guidelines for proportioning, quality control testing, and acceptance criteria for SCC applications in cast-in-place bridge components are provided in Attachment B. Details of test results are presented in Appendices C, D, E, and F. 2.1 Fresh Concrete Properties 2.1.1 Rheology Figure 2-1 shows the rheological properties of the SCC and CVC mixtures obtained from a mortar rheometer after sieving out the coarse aggregate. Figure 2-1 indicates that the dynamic yield stress of SCC mixtures is significantly lower than that of CVC mixtures, which makes them flow more easily. SCC mix- tures also have a wider range of plastic viscosity compared to CVC mixtures because of the larger range of water-powder ratios and SCM/filler types used. Figure 2-2 shows rheologi- cal properties of the SCC and CVC mixtures obtained using a concrete rheometer (including the coarse aggregate). In com- parison to CVC mixtures, SCC mixtures have lower yield torque (which represents yield stress) and a wider range of slope (which indicates plastic viscosity). Also, Figure 2-2 shows that the SCC mixtures containing gravel aggregate had higher yield torque and lower viscosity than the SCC mixtures containing lime- stone aggregate. In comparison to the round shape of gravel particles, the angularity of limestone particles causes more particle-to-particle interlock, which results in higher viscosity and increased packing density improving the flow and reducing the yield torque (Erdogan and Fowler, 2005; Lu, 2008). The flow curves for the SCC and CVC mortar and concrete mixtures are provided in Appendix C. 2.1.2 Workability Properties Tests were conducted to evaluate the FA, PA, and stability of the SCC mixtures; the results are presented in Figures 2-3 through 2-8. Figures 2-3 and 2-4 show the DD and DH versus slump flow for the J-ring test on SCC mixtures with differ- ent NMSA. Figures 2-3 and 2-4 indicate that most mixtures had high PAs (DD ≤ 2 in. and DH ≤ 0.6 in.). A few mixtures, mostly those with ¾ in. NMSA, had low PAs (DD = 2 to 4 in. and DH = 0.6 to 0.8 in.). These mixtures may not be suitable for components with high congestion of reinforce- ment and/or narrow sections (e.g., box girder), but may be appropriate for components with large sections and a low level of reinforcement (e.g., footing) (Khayat and Mitchell, 2009). Figure 2-5 shows filling capacity versus slump flow obtained from the caisson test on SCC mixtures with differ- ent NMSA. Figure 2-5 indicates that all mixtures had either high (> 80%) or moderate (70 to 80%) filling capacity. Fig- ure 2-6 shows penetration depth versus slump flow for SCC mixtures with different NMSA. Figure 2-6 indicates that most SCC mixtures had high static stability (penetration ≤ 0.5 in.), only a few mixtures had moderate static stability (penetration of 0.5 to 1.0 in.), and most mixtures with low slump flow had higher static stability than those with high slump flow. Figure 2-7 shows the column segregation ver- sus slump flow for SCC mixtures with different NMSA. Fig- ure 2-7 indicates that the majority of SCC mixtures had high static stability (column segregation ≤ 10%), and only a few mixtures (mostly with ¾ in. NMSA) had moderate static stability (column segregation between 10% and 15%) or low static stability (column segregation between 15% and 20%). These mixtures might be suitable for shallow and short components with simple and uncongested sections (e.g., grade beam). All SCC mixtures had VSI and HVSI values of 0 or 1, indicating adequate static stability in both fresh and hardened conditions. Figure 2-8 shows dynamic stability measured using a modified flow trough for SCC mixtures with high slump flow only. Figure 2-8 indicates that most mixtures exhibited either high dynamic stability (segregation ≤ 20%) or moderate dynamic stability (segre- gation ≤ 30%). Most SCC mixtures with high slump flow C H A P T E R 2

31 0.0 0.5 1.0 1.5 2.0 2.5 0 50 100 150 200 Pl as tic V isc o sit y (P a-s ) Dynamic Yield Stress (Pa) SCC CVC Figure 2-1. Rheological properties of mortar mixtures. 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 Sl op e (N -m - s) Yield Torque (N-m) SCC (Limestone) SCC (Gravel) CVC (Limestone) CVC (Gravel) Figure 2-2. Rheological properties of concrete mixtures. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 22.0 23.0 24.0 25.0 26.0 27.0 28.0 29.0 30.0 31.0 J- Ri ng ∆D (i n.) Slump Flow (in.) 3/4" NMSA 1/2" NMSA 3/8" NMSA Figure 2-3. J-ring reduction in slump flow diameter (D) of SCC versus slump flow.

32 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 22 23 24 25 26 27 28 29 30 31 J- R in g ∆H (i n.) Slump Flow (in.) 3/4" NMSA 1/2" NMSA 3/8" NMSA Figure 2-4. J-ring difference in average height (H) versus slump flow. 50% 55% 60% 65% 70% 75% 80% 85% 90% 95% 100% 22 23 24 25 26 27 28 29 30 31 Fi lli ng C ap ac ity (% ) Slump Flow (in.) 3/4" NMSA 1/2" NMSA 3/8" NMSA Figure 2-5. Caisson filling capacity of SCC versus slump flow. 0.00 0.25 0.50 0.75 1.00 1.25 22 23 24 25 26 27 28 29 30 31 Pe ne tra tio n (in .) Slump Flow (in.) 3/4" NMSA 1/2" NMSA 3/8" NMSA Figure 2-6. Penetration versus slump flow.

33 and ¾ in. NMSA showed poor dynamic stability, making them inappropriate for long or deep components. The T50 values for all mixtures were very close (approximately 2 sec), which helps speed placement and produce a formed surface with a good quality. 2.1.3 Workability Retention This investigation showed that the rate of slump flow loss is directly proportional to the initial slump flow when no workability retaining admixtures (WRAs) or additional dosage of HRWRA are used. Figure 2-9 shows that the rate of workability loss for SCC mixtures with initial slump flow of 30 in. and 24 in. averaged 7 and 3.5 in. per hr, respectively. These rates could vary depending on the mixture composi- tion, temperature, and type of chemical admixtures used. Using WRAs during batching is a recommended practice for cast-in-place applications requiring workability retention for an extended period (e.g., 90 minutes). Adding dosages of HRWRA at the job site is not desired and should only be used to address unexpected interruptions to SCC placement operations. 0% 5% 10% 15% 20% 25% 30% 22 23 24 25 26 27 28 29 30 31 Co lu m n Se gr eg at io n (% ) Slump Flow (in.) 3/4" NMSA 1/2" NMSA 3/8" NMSA Figure 2-7. Column segregation versus slump flow. 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 22 23 24 25 26 27 28 29 30 31 D yn am ic S eg re ga tio n (% ) Slump Flow (in.) 3/4" NMSA 1/2" NMSA 3/8" NMSA Figure 2-8. Flow trough dynamic segregation versus slump flow.

34 2.2 Early-Age Concrete Properties 2.2.1 Formwork Pressure Figure 2-10 shows the ratio of maximum exerted lateral pressure to hydrostatic pressure (Pmaximum/Phydrostatic) for SCC and CVC mixtures. Generally, SCC mixtures generated higher lateral pressure (93 to 100% of the hydrostatic pressure) than CVC mixtures (88 to 95% of the hydrostatic pressure). Larger differences between the lateral pressure of SCC and CVC mix- tures are expected when low placement rates (< 15 ft/hr) are used. Figure 2-10 also indicates linear relationships between the ratio of Pmaximum/Phydrostatic and thixotropy and yield torque for all mixtures as reported in the literature (Assaad, Khayat, and Mesbah, 2003; Khayat and Assaad, 2012). Mixtures with high thixotropy and yield torque exerted lower lateral pres- sure than those with low thixotropy and yield torque. These relationships support the use of rheological properties of SCC mixtures to predict formwork pressure. 2.2.2 Heat of Hydration Figure 2-11 shows the maximum increase in temperature obtained from semi-adiabatic calorimetry versus time for SCC and CVC mixtures. Figure 2-11 indicates that the temperature rise for SCC and CVC mixtures was similar (20 to 40°F) but SCC mixtures generally took a longer time to reach peak tem- perature. The difference in time needed to reach peak tem- perature depends on the type of SCM/filler (relationships of temperature change versus time for different types of SCMs/ fillers are provided in Appendix D). Also, it was observed that using Class C fly ash delays the start of the acceleration phase (Figure 2-12) as reported in earlier studies (Schindler and Folliard, 2005). Figure 2-13 shows the peak rate of energy gen- eration during hydration obtained from isothermal calorime- try for mortar sieved from SCC and CVC mixtures. There was no significant difference in the peak rate of energy generation for CVC and SCC mortar mixtures, but there was a significant delay in reaching the peak value for SCC mixtures. The tem- perature rise and rate of energy generation of all mixtures over a 24-hr period are provided in Appendix D. 2.2.3 Time of Setting Figure 2-14 shows the time of initial setting for SCC mix- tures versus slump flow at two ambient temperatures (60 and 80°F). Figure 2-14 indicates that SCC mixtures with high slump flow have longer time of setting than SCC mixtures with low slump flow, possibly due to the retarding effects of HRWRA. The ambient temperature has also a significant effect on the time of setting as higher temperatures result in shorter times of setting. The wide range in time of setting for SCC mixtures (4.5 to 11 hr) may be attributed to the effect of SCM/filler type (a similar range was reported by Khayat and Mitchell, 2009). Mixtures with Class C fly ash had the longest time of setting and those with Class F fly ash had the shortest. For purposes of comparison, CVC mixtures had an average time of initial setting of 6 hr at 80°F and 7 hr at 60°F. 2.3 Hardened Concrete Properties 2.3.1 Mechanical Properties Compressive Strength Figure 2-15 shows the relationships between the average 28-day compressive strength and the average compressive strength at 7, 14, and 56 days for all SCC mixtures. The best fit lines indicate that the average ratios of 7-day, 14-day, and 56-day compressive strength to 28-day compressive strength were 0.77, 0.88, and 1.12 respectively; these values are close to the values of 0.70, 0.88, and 1.09 predicted by the ACI 209 model (ACI y = 0.60x − 10.71 R² = 0.86 0 1 2 3 4 5 6 7 8 9 10 22 23 24 25 26 27 28 29 30 31 32 33 R at e o f S lu m p Fl ow L os s (in ./h r) Initial Slump Flow (in.) Figure 2-9. Initial slump flow versus rate of slump flow loss.

35 (a) Ratio of Pmaximum/Phydrostatic versus thixotropy. (b) Ratio of Pmaximum/Phydrostatic versus yield torque. y = −10.04x + 99.51 R² = 0.89 86 88 90 92 94 96 98 100 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 P m ax im um /P hy dr os ta tic Thixotropy (J-rev/sec) CVC SCC y = −1.93x + 101.27 R² = 0.82 86 88 90 92 94 96 98 100 0 1 2 3 4 5 6 7 P m ax im um /P hy dr os ta tic Yield torque (N-m) CVC SCC Figure 2-10. Ratio of Pmaximum / Phydrostatic versus thixotropy and yield torque. 0 5 10 15 20 25 30 35 40 45 50 0 2 4 6 8 10 12 14 16 18 20 M ax . I nc re as e of T em pe ra tu re (˚ F) Elapsed Time (hr) CVC SCC Figure 2-11. Maximum increase of temperature in semi-adiabatic condition.

36 0 5 10 15 20 25 30 35 40 0 2 4 6 8 10 12 14 16 18 20 22 24 M ax im u m In cr ea se o f T em pe ra tu re (˚ F) Elapsed Time (hr.) CVC SCC with Class C fly ash SCC with Class F fly ash SCC with GGBFS SCC with Class F fly ash plus LSP Figure 2-12. Semi-adiabatic calorimetry test results for mixtures containing ½ in. nominal maximum size limestone aggregates. 1.0 2.0 3.0 4.0 5.0 6.0 0 2 4 6 8 10 12 14 16 18 20 Pe ak R at e of E ne rg y G en er at io n (ca l/g /h) Elapsed Time (hr) CVC SCC Figure 2-13. Peak rate of energy generation in isothermal condition. T = 80oF T = 80oF y = 0.3x − 0.5 R² = 0.4 T = 60oF T = 60oF y = 0.9x − 14.2 R² = 0.7 0 2 4 6 8 10 12 22 23 24 25 26 27 28 29 30 31 32 33 Ti m e o f I ni tia l S et tin g (hr ) Slump Flow (in.) Figure 2-14. Time of initial setting versus slump flow for SCC mixtures.

37 209, 1997) for CVC with cement type I/II and moist curing conditions. The average compressive strength versus age values for all SCC mixtures are provided in Appendix E. These data indicated that SCC mixtures containing limestone aggregate had higher compressive strength than SCC mixtures contain- ing gravel aggregate, possibly because the interfacial transition zone (ITZ) is weaker in gravel particles than it is in limestone particles (Ozturan and Cecen, 1997). These data also indicated that SCC mixtures containing limestone powder had lower compressive strength than those without limestone powder (possibly due to the coarseness of the limestone powder used in this study). The particle size of limestone powder has a signifi- cant effect on compressive strength because coarser limestone particles reduce the reactivity of the system and, consequently, the compressive strength (Bentz et al., 2015). Modulus of Elasticity (MOE) The AASHTO LRFD Equation 5.4.2.4-1, for predicting MOE of CVC (Ec = 33,000 K1 wc1.5 √fc [ksi]), includes a correc- tion factor for source of aggregate (K1) to be taken as 1.0 unless determined by physical test. Since two different types of coarse aggregate were used in this study, the K1 factor was first deter- mined by comparing the MOE values of SCC mixtures contain- ing limestone aggregate to those of SCC mixtures containing gravel aggregate to determine their relative stiffness. Figure 2-16 shows the average measured MOE for SCC mixtures contain- ing the two aggregate types versus the square root of compres- sive strength times the unit weight of concrete (0.143 and 0.140 kcf for limestone and gravel mixtures, respectively) raised to the power of 1.5. Figure 2-16 indicates that the MOE of SCC y = 1.12x R² = 0.96 y = 0.88x R² = 0.90 y = 0.77x R² = 0.89 2 3 4 5 6 7 8 9 10 2 3 4 5 6 7 8 9 f ca t 7 , 1 4, a nd 5 6 da ys (ks i) 28-day fc (ksi) 56-day 14-day 7-day Figure 2-15. Relationships between average compressive strength at 7, 14, and 56 days and average compressive strength at 28 days for SCC mixtures. Figure 2-16. Comparing MOE of SCC mixtures containing gravel and limestone aggregates. Limestone y = 32,222x R = 0.71 Gravel y = 30,799x R = 0.72 0 1,000 2,000 3,000 4,000 5,000 6,000 0.09 0.10 0.11 0.12 0.13 0.14 0.15 0.16 M od ul us o f E la sti ci ty (k si) wc 1.5 √ ƒc (kcf1.5 . ksi1/2) Limestone Gravel

38 mixtures containing limestone aggregate was slightly higher than the MOE of SCC mixtures containing gravel aggregate, as reported by an earlier study (Mokhtarzadeh and French, 2000); K1 values of 1.0 and 0.95 are proposed for the limestone and gravel aggregates used in this study, respectively. Figure 2-17 shows the measured MOE values of all SCC mixtures versus those predicted by AASHTO LRFD Equation 5.4.2.4-1, using the proposed K1 values (1.0 for limestone mix- tures and 0.95 for gravel mixtures). Figure 2-17 indicates that MOE of SCC mixtures was slightly lower than predicted (a similar observation was reported by Khayat and Mitchell, 2009), which may be attributed to paste-to-coarse aggregate volume, which is higher in SCC than it is in CVC. Therefore, a modifica- tion factor (K2 = 0.96) is proposed for SCC (Ec = 33,000 K1 K2 wc1.5 √fc [ksi]). Tensile Strength Figure 2-18 shows the average measured splitting ten- sile strength for SCC mixtures versus the values predicted by AASHTO LRFD Provision C5.4.2.7 for CVC (f t = 0.23 √fc [ksi]). Figure 2-18 indicates that the splitting tensile strength of SCC y = 0.96x R² = 0.78 2,500 3,000 3,500 4,000 4,500 5,000 5,500 6,000 2,500 3,000 3,500 4,000 4,500 5,000 5,500 6,000 M ea su re d M od ul us o f E la sti ci ty fo r S CC (k si) Predicted Modulus of Elasticity for CVC (ksi) Figure 2-17. AASHTO predicted MOE CVC versus measured MOE for SCC. y = 0.79x R² = 0.60 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.2 0.3 0.4 0.5 0.6 0.7 0.8 M ea su re d A ve ra ge S pl itt in g Te ns ile S tre ng th fo r S CC (k si) Predicted Tensile Strength for CVC (ksi) Figure 2-18. AASHTO predicted tensile strength for CVC versus measured tensile strength for SCC.

39 mixtures was approximately 20% less than that predicted for CVC (an earlier study, Parra, Valcuende, and Benlloch, 2007, reported 18% lower tensile strength). Therefore, a modifica- tion factor of 0.8 is proposed for estimating the splitting tensile strength of SCC (ft = 0.8 × 0.23 √fc [ksi]). Modulus of Rupture (MOR) Figure 2-19 shows the average measured MOR versus the square root of the average compressive strength for SCC mix- tures and the range predicted by AASHTO LRFD Provision C5.4.2.6 for CVC (0.24 √fc to 0.37 √fc [ksi]). Figure 2-19 indi- cates that the MOR of SCC was within the predicted range for CVC but closer to the upper limit (similar results were reported by Mokhtarzadeh and French, 2000). Thus, the AASHTO LRFD provision for CVC could be applied to SCC. Bond Strength Figure 2-20 shows the pull-out bond strength versus the square root of the average compressive strength of SCC and CVC mixtures for 36 vertical deformed reinforcing bars. The SCC y = 0.34x R² = 0.63 y = 0.24x y = 0.37x 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.50 1.75 2.00 2.25 2.50 2.75 3.00 Fl ex ur al S tre ng th (k si) √ ƒc (ksi 1/2) SCC AASHTO Lower Limit For CVC AASHTO Upper Limit For CVC Figure 2-19. Average measured flexural strength of SCC mixtures versus square root of average 28-day compressive strength. SCC y = 2.20x − 2.51 R² = 0.89 CVC y = 2.32x − 2.12 R² = 0.69 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.9 2.0 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3.0 Pu ll- ou t B on d St re ng th (k si) √ fc (ksi1/2) SCC CVC Figure 2-20. Pull-out bond strength versus √fc of SCC and CVC mixtures.

40 trend shown indicates that the pull-out bond strength of SCC was consistently lower than that of CVC (similar results were reported by König et al., 2001 and 2003 and Almeida, Nardin, and Gresce, 2005). Also, analysis of variance (ANOVA) of pull- out test data for the two groups of mixtures confirmed this find- ing at 95% confidence level. Therefore, it appears appropriate to propose a development length modification factor of 1.3 to AASHTO LRFD Bridge Design Specifications Section 5.11.2.1.2 for vertical bars in SCC mixtures. Figure 2-21 shows the ratios of pull-out bond strength to the square root of average compressive strength for 54 horizontal deformed reinforcing bars located at different heights in six wall specimens: two made of SCC with high slump flow, two made of SCC with low slump flow, and two made of CVC. Figure 2-21 indicates no significant difference in the bond strength of hori- zontal bars between low slump flow SCC and CVC mixtures, but shows a slight difference between low slump flow SCC and high slump flow SCC. ANOVA of pull-out test data for the three groups of mixtures confirmed this finding at a 95% confi- dence level. Figure 2-21 also shows a reduction in bond strength as the distance from the bottom of the form increases (top-bar effect), particularly for CVC and low slump flow SCC mixtures, suggesting that the top-bar effect was dependent on the rheo- logical properties of SCC. Therefore, it appears appropriate to propose a development length modification factor of 1.4 in. to AASHTO LRFD Section 5.11.2.1.2 for top horizontal bars with more than 12 in. of fresh SCC cast below regardless of the slump flow. Shear Resistance Figure 2-22 shows the push-off interface shear resistance versus the square root of the average compressive strength for 20 SCC and 12 CVC specimens reinforced with two #3 bars across the shear plane. The developed relationships indicate that the interface shear resistance of SCC was very close to that of CVC; ANOVA results confirmed this find- ing at a 95% confidence level. Figure 2-23 shows a similar interface shear cracking pattern at failure in SCC and CVC specimens. Push-off test results and ANOVA data provided in Appendix E indicate that mixtures containing limestone aggregate exhibited slightly higher interface shear resistance than mixtures containing gravel aggregate. Figure 2-24 shows the measured interface shear resistance of the 20 SCC specimens reinforced with two #3 bars across the shear plane versus that predicted by AASHTO LRFD Sec- tion 5.8.4.1 for CVC with and without the cohesion factor (i.e., c = 0.4 and 0 ksi, respectively). Figure 2-24 indicates that the measured interface shear resistance of SCC was higher than that predicted by AASHTO except for specimens with average compressive strength less than 6 ksi. Therefore, it is proposed that the cohesion factor, c, in the AASHTO LRFD provisions for reinforced normal-weight concrete placed monolithically be 0.0 for SCC with average compressive strength that is less than 6 ksi. Figure 2-25 shows the average of push-off interface shear resistance of two specimens of each of four bridge components made using ready-mixed SCC and without reinforcement across the shear plane. Figure 2-25 indicates that the measured interface shear resistance of SCC was significantly higher than that predicted by AASHTO LRFD for unreinforced normal- weight concrete placed monolithically (c = 0.4 ksi). Figure 2-26 shows the ratio of shear resistance to the square root of average compressive strength for 18 beam specimens made of SCC and CVC with different levels of shear reinforce- ment. It indicates similar shear resistance values for low slump 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 0 8 16 24 32 40 48 B on d St re ng th / √f c (ks i/k si1 /2 ) Distance from the Bottom of the Wall (in.) SCC (high slump flow) SCC (low slump flow) CVC Figure 2-21. Top-bar effect on bond strength of horizontal bars in CVC and SCC mixtures.

41 flow SCC, high slump flow SCC, and CVC beams with the same reinforcement level. ANOVA data also indicated that the shear resistance of SCC and CVC mixtures was not sig- nificantly different at various reinforcement levels (similar to earlier results reported by Ebrahimi and Beygi, 2009). Figure 2-26 also indicates higher shear resistance values for all specimens without shear reinforcement and those with two #3 bars at 8 in. than those predicted by AASHTO LRFD Section 5.8.3.3 (sectional design method). These specimens exhibited typical shear cracking and failure with no sig- nificant difference between SCC and CVC beams (see Fig- ures 2-27 and 2-28). Specimens with two #3 bars at 4 in. had reached their ultimate flexure resistance before reach- ing their ultimate shear resistance and exhibited flexural cracking and failure. SCC y = 0.73x − 0.88 R² = 0.72 CVC y = 0.62x − 0.60 R² = 0.50 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 1.9 2.0 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3.0 In te rfa ce S he ar R es ist an ce (k si) √ fc (ksi1/2) SCC CVC Figure 2-22. Interface shear resistance versus √fc of SCC and CVC mixtures. Figure 2-23. Interface shear cracking of SCC and CVC push-off specimens. 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 M ea su re d In te rfa ce S he ar R es ist an ce fo r S CC (k si) Predicted Interface Shear Resistance for CVC (ksi) fc < 6 ksi (c = 0.4 ksi) fc > 6 ksi (c = 0.4 ksi) fc < 6 ksi (c = 0) fc > 6 ksi (c = 0) Figure 2-24. Measured interface shear resistance for SCC versus that predicted by AASHTO LRFD for CVC.

42 2.3.2 Visco-Elastic Properties Drying (Free) Shrinkage Figure 2-29 shows the measured drying shrinkage of SCC mixtures versus that predicted by AASHTO LRFD Equa- tion 5.4.2.3.3-1 for CVC. Figure 2-29 indicates significantly higher shrinkage values for SCC than those predicted for CVC (similar results were reported by Khayat and Mitchell, 2009). Figure 2-29 also indicates that the type of SCM/filler had a large effect on drying shrinkage. Therefore, a modi- fication factor (kp) to AASHTO LRFD Equation 5.4.2.3.3-1 is proposed to consider SCC mixtures with different types of SCM/filler. This SCC powder composition modification factor is estimated at 1.6 for SCC with cement type I/II and 25% Class C fly ash; 1.4 for SCC with cement type I/II and 30% GGBFS; and 1.3 for SCC with cement type I/II and 25% Class F fly ash or 20% Class F fly ash and 15% limestone pow- der. These values were derived from the slopes of the best fit straight lines shown in Figure 2-29. All other parameters of AASHTO LRFD Equation 5.4.2.3.3-1 remain the same (i.e., kp should be 1.0 for CVC). Figure 2-30 shows the average and standard deviation of 56-day drying shrinkage for SCC and CVC mixtures grouped by SCM/filler type. Restrained Shrinkage Figure 2-31 shows the time to cracking versus average stress rate for SCC and CVC mixtures in a restrained condition for up to 28 days. It appears that the time to cracking decreased as the average stress rate increased. About 50% of the mix- tures did not crack during the 28-day test period (these are not shown in Figure 2-31). Figure 2-32 shows the average stress rate and standard deviation of SCC and CVC mixtures grouped by SCM/filler type. Figure 2-32 indicates that the 0.00 0.40 0.80 1.20 1.60 Footing Column Pier Cap Girder In te rfa ce S he ar R es ist an ce (k si) Full-Scale Bridge Component AASHTO LRFD Prediction Figure 2-25. Interface shear resistance of unreinforced SCC push-off specimens. 0.112 0.179 0.216 0.107 0.164 0.190 0.128 0.166 0.194 0.063 0.138 0.212 - 0.050 0.100 0.150 0.200 0.250 None 2#3 bars @8" 2#3 bars @4" Sh ea r R es ist an ce / √f c (ks i/k si1 /2 ) Shear Reinforcement Measured for SCC (high slump flow) Measured for SCC (low slump flow) Measured for CVC Predicted for CVC (per AASHTO LRFD) Figure 2-26. Shear resistance of SCC and CVC mixtures with different reinforcement.

43 (a) Shear cracking in a CVC beam. (b) Shear cracking in a low slump flow SCC beam. (c) Shear cracking in a high slump flow SCC beam. Figure 2-27. Shear cracking of beams with no transverse reinforcement. (a) Shear cracking in a CVC beam. (b) Shear cracking in a low slump flow SCC beam. (c) Shear cracking in a high slump flow SCC beam. Figure 2-28. Shear cracking of beams transversely reinforced with two #3 bars at 8 in. C Ash y = 1.58x R² = 0.45 GGBFS y = 1.36x R² = 0.15 F Ash y = 1.28x R² = 0.40 F Ash+LSP y = 1.29x R² = 0.69 0 100 200 300 400 500 600 700 800 0 200 400 600 800 M ea su re d dr yi n g sh rin ka ge fo r S CC m ix tu re s (μ - st ra in ) Predicted drying shrinkage for CVC (μ-strain) C Ash GGBFS F Ash F Ash+LSP Figure 2-29. AASHTO LRFD predicted drying shrinkage for CVC versus measured drying shrinkage for SCC mixtures.

44 0 100 200 300 400 500 600 700 800 C Fly Ash GGBFS SCM/Filler Type F Fly Ash F Fly Ash + LSP A ve ra ge a n d sta nd ar d de vi at io n of m ea su re d dr yi n g sh rin ka ge a t 5 6 da ys (µ - st ra in ) SCC CVC Figure 2-30. Average and standard deviation of measured 56-day drying shrinkage for CVC and SCC mixtures. 0 7 14 21 28 10 20 30 40 50 60 70 80 90 Ti m e o f C ra ck in g (da y) Average Stress Rate (psi/day) SCC CVC Figure 2-31. Restrained shrinkage test data for SCC and CVC mixtures. 0 10 20 30 40 50 60 70 80 C Fly Ash F Fly Ash GGBFS F Ash + LSP A ve ra ge a nd st an da rd d ev ia tio n of st re ss ra te (ps i/d ay ) SCM/Filler Type SCC CVC Figure 2-32. Average and standard deviation of stress rate for CVC and SCC mixtures with different types of SCM/filler.

45 cracking potential of SCC mixtures was not significantly dif- ferent from that of CVC mixtures with the same SCM (similar results were reported by See and Attiogbe, 2005). However, the type of SCM/filler had a significant effect on the cracking potential of SCC mixtures; mixtures containing Class C fly ash exhibited the highest potential to crack. Figure 2-33 shows the effect of NMSA on cracking potential. It indicates that SCC mixtures containing 3⁄8 in. NMSA exhibited higher potential to crack than mixtures containing ¾ or ½ in. NMSA. Creep Figure 2-34 shows the measured creep coefficients for SCC mixtures at different ages versus the coefficients predicted by AASHTO LRFD provisions for CVC. All SCC mixtures exhibited the same creep trend except those with limestone powder (similar results were reported by Heirman et al., 2008). Figure 2-34 indicates that for all SCC mixtures, except those with limestone powder, the creep coefficient can be accurately predicted using the AASHTO LRFD equation for CVC. How- ever, a 1.2 modification factor, proposed as a multiplier to AASHTO LRFD Equation 5.4.2.3.2-1, would be required for estimating the creep coefficient of SCC mixtures containing 15% limestone powder. Charts showing creep strain versus age of loading for all SCC and CVC mixtures and charts showing the AASHTO LRFD predicted creep strains versus age of load- ing are provided in Appendix E. These data show an agreement between measured and predicted creep curves for all mixtures except for mixture G222S (possibly due to errors in loading specimens and/or recording of test data). 2.3.3 Durability Properties Air Void System Figure 2-35 shows the air void system parameters for the hardened SCC and CVC mixtures. Figure 2-35 indicates that the spacing factor and specific surface of most SCC and CVC mixtures were within the values recommended by PCA (2009) and FHWA (2006). However, the variation in the air content of SCC mixtures was higher than that of CVC mixtures; some values were outside the target range of 6 ± 1.5%. This variation may be caused by variations in the HRWRA dosage used in SCC mixtures. Surface Resistivity Figure 2-36 shows the 28-day surface resistivity of SCC and CVC mixtures. Figure 2-36 indicates that the variation in sur- face resistivity was more related to the SCM/filler type than to the aggregate type (limestone or gravel) or concrete type (SCC 0 10 20 30 40 50 60 70 80 3/4" 1/2" 3/8" A ve ra ge a n d sta nd ar d de vi at io n of st re ss r at e (ps i/d ay ) NMSA Figure 2-33. Average and standard deviation of stress rate for SCC mixtures with different NMSA. SCC without LSP y = 0.98x R² = 0.95 SCC with LSP y = 1.17x R² = 0.93 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 M ea su re d Cr ee p Co ef fic ie nt Predicted Creep Coefficient SCC without LSP SCC with LSP Figure 2-34. Predicted creep coefficient for CVC versus measured creep coefficient for SCC mixtures.

(a) Air content versus spacing factor. (b) Air content versus specific surface. 0.00 0.05 0.10 0.15 0.20 0.25 0 1 2 3 4 5 6 7 8 9 Sp ac in g Fa ct or (m m ) Air Content (%) SCC Mixtures CVC Mixtures 0 6 12 18 24 30 36 42 48 54 0 1 2 3 4 5 6 7 8 9 Sp ec ifi c Su rfa ce (m m 2 /m m 3 ) Air Content (%) SCC Mixtures CVC Mixtures Figure 2-35. Air void system parameters for hardened SCC and CVC. 0 3 6 9 12 15 18 21 24 27 Su rf ac e Re sis v ity (k Ω -c m ) Limestone Aggregate Gravel Aggregate SCC + class C fly ash CVC + class F fly ash SCC + class F fly ash SCC + class F fly ash + LSP SCC + GGBFS Figure 2-36. Surface resistivity of SCC and CVC mixtures.

47 or CVC); similar findings have been reported by Tang and Zhu (2007). Mixtures with Class C fly ash exhibited the lowest sur- face resistivity (high penetrability), and mixtures with GGBFS exhibited the highest surface resistivity (low penetrability). In general, SCC mixtures exhibited higher surface resistivity than CVC mixtures with the same type of SCM. Details of surface resistivity measurements for all mixtures are provided in Appendix E. 2.4 Full-Scale Bridge Components Several constructability and structural performance issues associated with using SCC in cast-in-place bridges were inves- tigated by constructing and testing two full-scale bridge com- ponents; details of forming, reinforcing, placing, and testing each component are provided in Appendix F. The findings are summarized in the following sections. 2.4.1 Formwork Pressure Figure 2-37 shows the measured SCC formwork pressure versus time (up to 75 minutes) at three locations in the two pier columns. SCC was placed in each column using a bucket (i.e., discrete placement), which resulted in the steps shown in Figure 2-37. The time interval between successive placements determined the placement rate as column dimensions and bucket size were the same. Figure 2-37 indicates a slight reduc- tion in the maximum pressure with time. Figure 2-38 shows the full hydrostatic pressure and the peak pressure values measured at different heights in the two pier columns constructed using placement rates of 26 ft/hr and 60 ft/hr. Figure 2-38 indicates that the maximum SCC formwork pressure was very close to hydro- static pressure, especially for high placement rates (similar to the findings of the laboratory investigation). A significant reduc- tion in SCC formwork pressure can be obtained by reducing (a) First Column (b) Second Column 0 2 4 6 8 10 12 14 0 15 30 45 60 75 Pr es su re o n Fi rs t C ol um n (ps i) Time (min) At 1 ft At 6 ft At 10 ft 0 2 4 6 8 10 12 14 0 15 30 45 60 75 Pr es su re o n Se co nd C ol um n (ps i) Time (min) At 1 ft At 6 ft At 10 ft Figure 2-37. Measured SCC formwork pressure versus time for the two pier columns.

48 placement rates, especially for mixtures with high thixotropy and high yield stress. The rheological properties of the SCC mixture used in column fabrication are provided in Appendix F. 2.4.2 Drying Shrinkage Figure 2-39 shows measured drying shrinkage for ready- mixed SCC used in fabricating the full-scale bridge compo- nents versus that predicted by the AASHTO LRFD equation, using the proposed powder composition modification factor for SCC (Section 2.3.2). Figure 2-39 shows that introducing the proposed modification factor improves the predictability of drying shrinkage of SCC for the different curing conditions. 2.4.3 Formed Surface Quality The formed surface quality of the components was evalu- ated according to ACI 347.3R-13 guidance for formed con- 0 1 2 3 4 5 6 7 8 9 10 0 500 1,000 1,500 2,000 2,500 D ist an ce fr om th e bo tto m o f t he c ol um n fo rm (ft ) Pressure (psf) Hydrostatic Pressure Measured Pressure (R=26 ft/hr) Measured Pressure (R=60 ft/hr) Figure 2-38. Measured SCC formwork peak pressure for different placement rates. crete surfaces. Only the surface void ratio criterion was used because it reflects concrete consolidation; not form quality. Table 2-1 lists the maximum surface void diameter, surface void area, and the corresponding surface void ratio class for different formed surfaces of each component. These results indicate that all of the formed surface resembled concrete surface category (CSC) 3 or 4 (categories for exposed sur- faces where visual appearance is important). High slump flow SCC, short free-fall distance, or concrete moved in a bottom- up direction provided the highest surface void ratio classes. 2.4.4 Structural Performance Table 2-2 summarizes the results of four structural tests conducted on the full-scale bridge components made of SCC. In each test, the nominal resistance of the component was predicted by AASHTO LRFD equations for CVC, assuming a resistance factor of 1.0 and using the measured compressive y= 1.39x R² = 0.64 y = 0.99x R² = 0.64 0 100 200 300 400 500 600 700 800 0 100 200 300 400 500 600 700 800 M ea su re d dr yi ng sh rin ka ge (µ - st ra in ) Predicted drying shrinkage (µ-strain) Without Modification With Modification 7-day curing 28-day curing Figure 2-39. Measured versus predicted drying shrinkage for SCC mixtures.

49 Formed SCC Surface Max. Void Diameter (Dmax), in. Surface Void Area (%) Surface Void Ratio Class Footing (short side) 3/8 0.16% SVR3 Footing (long side) 3/8 0.31% SVR3 First Column (at 1.5 ft from the bottom) 3/8 0.14% SVR3 First Column (at 11.3 ft from the bottom) 3/8 0.25% SVR3 Second Column (at 8.5 ft from the bottom) 5/8 0.19% SVR2 Second Column (at 10 ft from the bottom) 5/8 0.31% SVR2 Pier Cap (at the pour line) 5/8 0.68% SVR2 Pier Cap (away from the pour line) 5/8 0.54% SVR2 Top of the Girder (pouring side) 1/2 0.40% SVR3 Bottom of the Girder (pouring side) 1/2 0.51% SVR3 Bottom of the Girder (opposite side) 1/4 0.20% SVR4 Girder End (at the construction joint) 1/4 0.02% SVR4 Table 2-1. Measured surface void size and ratio in the formed SCC surfaces. strength. The predicted nominal resistance was then used to predict the ultimate load, as shown in Table 2-2. All compo- nents showed a higher resistance than predicted and a behavior similar to that expected for components made of CVC. Also, no relative displacement (i.e., slippage) between the SCC top flange and tub section was detected and no cracking or signs of damage were observed at the anchorage zones of the post- tensioned SCC girder. The limited data obtained from these tests suggest that SCC components can be expected to exhibit similar behavior to that of CVC components and the possible applicability of AASHTO LRFD interface shear design and anchorage zone design provisions to SCC components. 2.4.5 Segregation Resistance The stability of SCC used in the construction of full-scale bridge components was evaluated by making several saw cuts at different sections and obtaining specimens for examina- tion. Figure 2-40 shows a mid-section across the width of the footing, Figure 2-41 shows the sections taken at the bottom, middle, and top portions of the two columns of the bridge pier. Figure 2-42 shows a mid-section across the pier cap, and Fig- ures 2-43 and 2-44 show sections taken near the ends of the box girder specimen. For all these sections, the distribution of the coarse aggregate from top to bottom and thickness of the mortar layer at the top of the section were evaluated using the rating criteria of HVSI. Figures 2-40 through 2-44 indi- cate that the SCC mixtures were stable (HVSI 0 or 1), and no signs of segregation, bleeding, or lack of consolidation around the reinforcing bars were observed. Component Test Type Age at Testing (day) Actual Compressive Strength (ksi) Predicted Ultimate Load (kip) Cracking Load (kip) Applied Load (kip) Pier Cap Strut-and-Tie Resistance 33 7.5 380.0 150 451.0* Column Flexural Resistance 103 8.5 94.5 40 101.5* Box Girder Flexural Resistance 39 8.2 297.0 200 301.5* Shear Resistance 39 8.2 353.0 250 400.4* * Loading was stopped to maintain specimen stability and integrity for further testing. Table 2-2. Results of testing full-scale bridge components. Figure 2-40. Saw cut at the middle of the bridge pier footing.

50 Location First Column Second Column Bottom Middle Top Figure 2-41. Saw cuts at the bottom, middle, and top of the two pier columns.

51 Figure 2-42. Saw cut at the middle of bridge pier cap. Figure 2-43. Saw cut in the bridge box girder specimen. Figure 2-44. Saw cut in the bridge box girder specimen showing top and bottom corners. Limited mechanical vibration of the concrete surface prior to placing the next lift is a common practice used to avoid the formation of pour lines. To investigate the SR of SCC mixtures handled in this way, three 6 in. × 12 in. cylinders were made using low slump flow SCC. SCC in one cylinder was not vibrated, but SCC in the other two cylinders was subjected to low (2 sec) and moderate vibration (8 sec). The three cylinders were saw cut after hardening to examine the distribution of coarse aggregate. Figure 2-45 shows that the resistance to segregation was not affected by the level of vibration (HVSI of the three cylinders was 1).

52 2.4.6 Air Void System Table 2-3 lists the air content measurements for the ready- mixed fresh SCC at the plant and at the job site after adding an additional dosage of HRWRA (the target air content was 6±1.5%). The air void system parameters for the hardened concrete were measured on cylinders and cores extracted from No vibration Low vibration (2 sec) Moderate vibration (8 sec) Figure 2-45. Saw cuts of three cylinders with different levels of vibration. the full-scale components. These data indicated a significant difference between the air content in fresh concrete at the plant and that at job site (more than 1.5%) due to time, transporta- tion, and addition of HRWRA. A comparison of the air void system parameters for the cylinders and cores indicates no sig- nificant difference, suggesting that placement method did not have a significant effect on the air void system in SCC. Component Fresh SCC Placement method Hardened SCC % Air content at plant* Dosage of HRWRA added at jobsite (oz/cwt) % Air content at jobsite Measured on 4 x 8 in. cylinders Measured on 3 in. diameter cores % Air content Spacing factor (mm) Specific surface (mm2/ mm3) % Air content Spacing factor (mm) Specific surface (mm2/ mm3) Footing 4.5 2.3 2.5 Truck Chute 4.1 0.17 32.8 2.9 0.23 22.5 First Column 4.5 2.0 4.0 Bucket and Tremie Pipe 5.7 0.17 26.4 4.2 0.18 31.0 Second Column 6.2 1.0 4.0 Bucket (free fall) 4.9 0.15 27.2 5.2 0.15 32.7 Pier Cap 6.4 4.0 4.0 1/2 cy Bucket 4.3 0.22 24.3 – – – Box Girder 6.0 1.5 4.5 Pumping (2 and 3 in. hose) 4.0 0.16 34.0 5.2 0.14 29.0 Top Flange 4.5 1.5 3.5 Pumping (3 in. hose) 5.0 0.13 33.1 4.6 0.17 31.9 AVERAGE 5.35 2.05 3.75 – 4.7 0.17 29.6 4.4 0.18 29.4 * Measured by plant technician Table 2-3. Air void system in fresh and hardened SCC used in full-scale specimens.