Safety in nonlinear reinforced concrete design
Author
Olesen, Daniel Vestergaard
Term
4. term
Education
Publication year
2020
Pages
123
Abstract
The thesis investigates how the choice of material strength parameters (mean, characteristic, and design values) influences predicted failure modes and safety in nonlinear reinforced concrete design. Two simply supported beams—a rectangular section and a T-section—are analyzed with ATENA 2D, and four safety formats are evaluated: the partial safety factor method, a global resistance factor, ECOV (estimate of coefficient of variation for resistance), and a full probabilistic method. The probabilistic approach randomizes material resistances using SARA coupled with ATENA to estimate resistance distributions and compare reliability. The study explores whether varying initial material resistance values can lead to different failure modes and provides a reliability comparison across safety formats. Uncertainty is noted in the T-beam results, and definitive conclusions depend on the complete analysis presented in the thesis.
Specialet undersøger, hvordan valg af materialestyrkeparametre (middel-, karakteristiske og dimensionsværdier) påvirker den forudsagte brudmåde og sikkerhed i ikke-lineær dimensionering af armeret beton. To simpelt understøttede bjælker—en rektangulær og en med T-tværsnit—er analyseret i ATENA 2D, og fire sikkerhedsformater vurderes: partialkoefficientmetoden, global modstandskoefficient, ECOV (estimat af variationskoefficient for modstand) og en fuldt probabilistisk metode. Den probabilistiske tilgang randomiserer materialemodstande via SARA koblet til ATENA for at estimere modstandsfordelinger og sammenligne pålidelighed. Arbejdet belyser, om forskellige startværdier for materialemodstand kan føre til forskellige brudmåder, og indeholder en sammenligning af pålidelighed mellem sikkerhedsformater. Der rapporteres usikkerhed i resultaterne for T-bjælken, og endelige konklusioner afhænger af den fulde analyse i specialet.
[This apstract has been generated with the help of AI directly from the project full text]