Graduation Year

2023

Document Type

Dissertation

Degree

Ph.D.

Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Civil and Environmental Engineering

Major Professor

Abla Zayed, Ph.D.

Committee Member

Kyle Riding, Ph.D.

Committee Member

Rajan Sen, Ph.D.

Committee Member

Aydin Sunol, Ph.D.

Committee Member

Abdul Malik, Ph.D.

Keywords

Concrete, Slag blended concrete, Slag cement, Slag characteristics, Heat generation, Sulfate optimization, External sulfate attack, Chloride ingress

Abstract

Performance and durability of slag blended concrete was investigated in this study focusing on slag and cement characteristics. Seven slags with variable Al2O3 content, Al2O3/MgO ratio and fineness were selected for investigation. The eight selected cements included ASTM Type I, Type I(HA), Type II(MH), Type II(MH)(HA), and Type IL(10 and 14). Cementitious materials studied here were subjected to a battery of tests including physical, chemical and mineralogical characterization. Performance and durability testing included heat of hydration measurements, sulfate optimization, external sulfate resistance, adiabatic temperature rise and chloride ingress.

Adiabatic temperature rise was assessed on 60% slag blended concrete, using Type II(MH) and Type IL(10) cements with slags having 8 to 17% Al2O3. Studies revealed that 60% cement replacement with all slags reduced concrete adiabatic temperature rise except the one containing 17% Al2O3. Varying slag Al2O3 content from 8 to 17% resulted in a substantial difference in the adiabatic temperature rise. Meanwhile, slag with low MgO content showed a higher temperature rise during the first day but lower at later ages. Slag Al2O3 content and MgO/Al2O3 ratio had a linear correlation with maximum adiabatic temperature rise rate and hydration parameters (β* and τ*) of blended concrete. Prediction of temperature rise in mass concrete elements showed inclusion of slag reduced both concrete temperature rise and temperature gradients within the elements but this reduction diminished as the member size increased, especially in high Al2O3 slag blended concrete.

Sulfate optimization was performed on a total of 26 blended mixtures of 5 different cements and 6 different slags using isothermal calorimetry at ambient temperature. Research findings showed that slag incorporation resulted in an undersulfated system due to filler effect and additional reactive alumina in slags. Slag Al2O3 and MgO contents were found to significantly affect sulfate demand in blended cementitious systems, where increasing slag Al2O3 content increased optimum sulfate contents at all ages while decreasing slag MgO content increased 1-day optimum sulfate content but not later. Additional sulfate increased ettringite content at early ages and eventually eliminated monosulfoaluminate. Calcite also helped prevent conversion of ettringite to monosulfoaluminate by consuming alumina in forming carboaluminates. Statistical analysis identified that the optimum sulfate content in slag blended cementitious systems was governed by cement Al2O3 content, slag Al2O3 and MgO contents as well as slag mean particle size. Correlation between optimum sulfate content and material characteristics was established for proper sulfate adjustment in slag blended systems.

External sulfate resistance of slag blended systems was investigated. 60% substitution of cement by all slags substantially improved external sulfate resistance irrespective of cement and slag characteristics; however, less improvement could be observed with higher Al2O3 slags. Deterioration of serviceability, such as surface spalling, softening and loss of cohesion, initiated at early age and was the major failure mode for high Al2O3 slag blends with relatively low expansion. Sulfate optimization and addition of limestone were very effective in enhancing sulfate durability performance of high Al2O3 slag mixtures by altering hydration products due to ettringite stabilization and carboaluminate phases’ formation, respectively. To satisfy various performance requirements, different levels of optimum sulfate contents determined from sulfate optimization were required depending on slag Al2O3 content, where higher sulfate demand was indicated for slags of higher Al2O3 contents. Alternatively, limestone incorporation can be coupled with sulfate addition to achieve further performance improvement. A safe threshold of additional sulfate content in slags correlated with cement C3A and slag Al2O3 contents and a relationship was established.

Chloride ingress was studied on 60% slag blended concrete using multiple cements and slags having 8 and 17% Al2O3. Experimental work and analyses included chloride binding capacity measurement, chloride diffusion profile exploration, permeability assessment, diffusion coefficients determination, service life projection and corrosion monitoring of embedded steel reinforcement. Chloride binding tests showed higher binding capacity for slag blended systems, especially for the higher Al2O3 slag blends. Physical binding effect may be primary in high chloride concentration environment. Collected total chloride concentration profiles at 360 days of exposure indicated 60% cement substitution of slags significantly reduced chloride penetration but no major difference was observed between the selected slags. Permeability assessment showed reduced amount of large pores due to incorporation of slags, but the total intruded volume did not show a notable difference. Meanwhile, higher Al2O3 slag resulted in higher early age permeability of slag blended concrete. Although no major difference was observed in apparent diffusion coefficients between slags, lower free chloride concentrations and effective diffusion coefficients in high Al2O3 slag blends, when using IL cement, when chloride binding was considered. Different service life projection methods may give notable difference but generally slag incorporation significantly extended concrete service life. Corrosion of steel bars in plain concrete initiated earlier and developed faster, where the macrocell and microcell corrosion started simultaneously. Slag incorporation slowed corrosion rate but the microcell corrosion may develop ahead of visible macrocell corrosion in slag blended concrete due to lower permeability. Therefore, open circuit potential may be more useful as an indicator of corrosion initiation.

Download

Share

COinS