Author ORCID Identifier
Semester
Fall
Date of Graduation
2024
Document Type
Dissertation
Degree Type
PhD
College
Statler College of Engineering and Mineral Resources
Department
Civil and Environmental Engineering
Committee Chair
Hung-Liang (Roger) Chen
Committee Member
Hota V. GangaRao
Committee Member
Fei Dai
Committee Member
Hailin Li
Committee Member
Terence Musho
Abstract
In high-volume concrete structures, the temperature increment due to hydration can be extremely high. At the same time, the exterior surfaces release heat rapidly, this phenomenon causes large temperature gradients and increases the cracking risk. When the thermal stresses exceed the concrete’s tensile strength, cracking is highly possible. These cracks can reduce the durability of the concrete structures and lead to the corrosion of rebars. Concrete structures with the risk of experiencing thermal cracks are known as “mass concrete”. In RP-312 research projects from WVDOT, finite element models (FEM) were developed to estimate the thermal cracking probability of three different concrete mix designs. Based on the experimentally measured thermal and mechanical properties of three Class M concrete during RP-312, mass concrete definition tables were constructed for three types of pier-stem geometries using the FEM. However, Pier caps and footers can be massive and may have a high probability of early-age cracking. In this study, mass concrete definition tables for four common types of pier-caps with three different mix designs, the 6-bag straight Ordinary Portland Cement (OPC) mix design (Class M Option 1), as well as two thermally friendly mixes with 50% slag replacement (Class M Option 2) or 30% fly ash replacement (Class M Option 3) were developed. The material properties of these mix designs were experimentally measured and used in the thermal stress analysis of the pier caps and footers. Additionally, the mass definition tables were constructed with an insulation value of R = 5 to maximize the size of the non-mass elements. The detailed dimensions of each pier cap type were obtained from the default cross-sections shown in Open Bridge Modeler by Bentley. In West Virginia, three types of pier caps are commonly used in the construction of bridges: hammerhead and pier caps supported on two or three pier stems depending on the number of lanes. The hammerhead pier cap was analyzed for two different lengths, 14-ft (simulating one-lane) and 36-ft (two-lanes). A rectangular pier stem was located at the center of the hammerheads and had a cross-section equal to the bottom of the pier cap to maximize the boundary constraint. For the two-column pier cap, the length was set to be 30-ft (two-lanes) with rectangular columns located 3.3-ft from the end of the pier cap. Similarly, for the three-column pier cap, the length was set to be 40-ft with additional support at the center. Besides, mass concrete definition tables were constructed for four common rectangular footers. Mass concrete definition tables were constructed for rectangular footers with the following dimensions: H (thickness) x 4H (width) x 4H (length), H x 3H x 4H, and H x 4H x 5H. The footer’s templates are for single-column pier stems. An additional case for merged (spread) footers was also analyzed. These footers are commonly H (thickness) x 4H (width) and 6-ft longer than the pier-cap length (i.e., 36-ft length for two-column pier cap). The tables were constructed based on a “worst-case” soil scenario. First, the cracking probability was evaluated for different soil conditions such as dry or wet, and for four types of soils in West Virginia. These cases included steel formwork with and without R = 5 insulation layers. A layer of compacted aggregate was considered as the subbase and analyzed on top of the soil. Furthermore, the tensile stresses developed due to the restrictions from the piles and reinforcements were checked. The mass definition tables were based on the assumptions of the daily ambient temperature variation from 60 °F to 90 °F for summer weather and 30 °F to 60 °F for winter weather conditions. The initial concrete temperature for the summer weather was set to be 75 °F and a placement temperature of 62 °F was considered during the winter conditions. FORTRAN subroutines were developed to perform non-linear thermal stress analysis using ABAQUS during the RP-312 project. The methodology used for creating the pier stem’s mass concrete definition tables was improved to analyze the pier caps and footers with different boundaries. A comparable methodology was implemented for another popular commercially available software, ANSYS Mechanical. A set of subroutines was developed to perform the non-linear transient thermal and viscoelastic stress analysis using ANSYS mechanical considering the degree of hydration-based material properties of early-age concrete. Two separate user material subroutines, USERMATTH for thermal and USERMAT for stress were developed for ANSYS and used in this study. The methodologies were then verified by comparing the temperature and strain measurements of 4-ft cubes against analytical models. The best time to remove the insulated formwork (R=5) was studied using a 14-ft Hammerhead pier cap geometry. Besides, safe construction practices for early insulated formwork removal of different geometries were established. Linear relationships were found between the crack index versus the time of formwork removal for different thicknesses. Besides using R=5 insulation layers, analyses were performed for different geometries with R=2.5 insulation layers as an alternative to increase the maximum allowable sizes. In addition, mass definition tables for non-insulated pier caps and footers were developed for mixes with a water cementitious ratio of 0.42. Furthermore, the thermal shock of hot concrete surface exposed to cold water was extensively analyzed and discussed. The temperature drop of concrete surface after contact with water was measured experimentally and compared to FEM. Using the FEM stress results, an equation was developed to estimate the peak of max principal (tensile) stress. Besides finite element modeling, a multi-component material model was partially developed and validated with experimental measurements. Different cement replacements using GGBFS or Class F fly ash were considered in this model. Different types of GGBFS were hydrated in limewater solution to measure their heat generation separately. The hydration behavior of different GGBFS types was mathematically modeled and incorporated in the multi-component material model to consider the change of degree of hydration based on different GGBFS. Using the material model, the hydration behavior of concrete such as heat of hydration, adiabatic temperature rise, strength, and modulus of elasticity can be evaluated considering the chemical properties of Portland cement, different types of GGBFS, water cementitious ratio, and percentage replacement of GGBFS and Class F fly ash. The material model can potentially improve to provide the thermal and mechanical properties of different concrete mixes for the engineers without the need for time-consuming and expensive experimental measurements. All the analytical models and guidelines developed in this study will enable engineers to take preventative actions to reduce early-age tensile cracking of mass concrete structures.
Recommended Citation
Mardmomen, Seyednavid, "Modeling of Thermal and Mechanical Behavior of Concrete Containing Supplementary Cementitious Materials" (2024). Graduate Theses, Dissertations, and Problem Reports. 12660.
https://researchrepository.wvu.edu/etd/12660