| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 2.26 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
Esta tese apresenta uma investigação de paredes LSF (Light Steel Frame) portantes em condições de incêndio, sendo possível determinar a temperatura crítica. Estas paredes LSF possuem utilização crescente na construção de edifícios sustentáveis. O estudo se baseia no comparativo entre os resultados experimentais de um ensaio à escala real e modelo numérico 3D, na tentativa de replicar os desempenhos do modelo à escala real feito experimentalmente e o modelo numérico 3D desenvolvido durante o estudo.
A análise segue o procedimento feito experimentalmente, com três corpos de prova, sendo um deles em escala menor de 3x1 m, para calcular a capacidade portante à temperatura ambiente, sequencialmente dois corpos de prova em escala maior de 3x3 m, um deles testado a 84% da carga portante e outro a 27% da carga portante. São realizadas quatro simulações sequências para simulação 3D. A primeira simulação considera cargas unitárias a fim de determinar os modos de instabilidade e as cargas críticas. A segunda simulação utiliza uma análise não linear geométrica e material com imperfeição geométrica (GMNIA), com o intuito de determinar a capacidade portante da estrutura à temperatura ambiente. A terceira simulação faz uma análise térmica transiente não linear material para determinação da temperatura ao longo do tempo e por fim, a quarta simulação faz uma análise termomecânica a carga constante com o objetivo de averiguar a temperatura crítica da estrutura e o tempo de resistência ao fogo.
O modelo é avaliado com base nos resultados experimentais, fazendo um comparativo em todos as variáveis apresentadas e seguindo os passos feitos durante o processo, o erro quadrático médio (RMSE) é calculado para averiguar a diferença entre as temperaturas medidas e as temperaturas determinadas numericamente. No caso da temperatura, é assumindo uma boa aproximação sempre que o RMSE seja inferior a 100 ᵒC, durante o tempo de exposição ao fogo da parede LSF.
Os resultados mostram um comparativo seguro entre os testes experimentais e as simulações computacionais, demonstrando não somente a capacidade portante, o desenvolvimento da temperatura e o valor da temperatura crítica, assegurando a confiabilidade das simulações computacionais, sendo uma solução alternativa para ensaios experimentais dispendiosos, possibilitando o teste de soluções alternativas para paredes LSF.
É apresentada uma nova proposta para determinar a confiabilidade e semelhança entre simulações numéricas e testes experimentais, testando novos formatos e confecções diferentes de paredes LSF.
This thesis presents an investigation of load-bearing LSF (Light Steel Frame) walls under fire conditions, allowing for the determination of critical temperature. These LSF walls are increasingly used in the construction of sustainable buildings. The study is based on a comparison between experimental results from a full-scale test and a 3D numerical model, in an attempt to replicate the performance of both the experimentally conducted full-scale model and the 3D numerical model developed during the study. The analysis follows the experimental procedure, with three test specimens, one of them at a smaller scale of 3x1 m to calculate the load-bearing capacity at room temperature, and sequentially, two larger-scale specimens of 3x3 m, one tested at 84% of the load capacity and the other at 27% of the load capacity. Four sequential simulations are performed for the 3D simulation. The first simulation considers unit loads tio determine instability modes ando critical loads. The escondi simulation uses a geometria and material non-linear análises with geometric imperfection (GMNIA) tão determine the load-bearing capacito of the structure ato room temperature. The third simulation conducts a transient thermal non-linear material analysis to determine temperature over time, and finally, the fourth simulation performs a thermomechanical analysis under constant load to verify the structure's critical temperature and fire resistance time. The model is evaluated based on the experimental results, comparing all the presented variables. Following the steps taken during the process, the root mean square error (RMSE) is calculated to verify the difference between the measured temperatures and those determined numerically. For temperature, a good approximation is assumed whenever the RMSE is below 100°C during the LSF wall's fire exposure time. The results show a reliable comparison between experimental tests and computational simulations, demonstrating not only the load-bearing capacity, temperature development, and critical temperature value but also ensuring the reliability of the computational simulations. This offers an alternative solution to expensive experimental tests, enabling the testing of alternative solutions for LSF walls. A new proposal is presented to determine the reliability and similarity between numerical simulations and experimental tests, testing new formats and different configurations of LSF walls.
This thesis presents an investigation of load-bearing LSF (Light Steel Frame) walls under fire conditions, allowing for the determination of critical temperature. These LSF walls are increasingly used in the construction of sustainable buildings. The study is based on a comparison between experimental results from a full-scale test and a 3D numerical model, in an attempt to replicate the performance of both the experimentally conducted full-scale model and the 3D numerical model developed during the study. The analysis follows the experimental procedure, with three test specimens, one of them at a smaller scale of 3x1 m to calculate the load-bearing capacity at room temperature, and sequentially, two larger-scale specimens of 3x3 m, one tested at 84% of the load capacity and the other at 27% of the load capacity. Four sequential simulations are performed for the 3D simulation. The first simulation considers unit loads tio determine instability modes ando critical loads. The escondi simulation uses a geometria and material non-linear análises with geometric imperfection (GMNIA) tão determine the load-bearing capacito of the structure ato room temperature. The third simulation conducts a transient thermal non-linear material analysis to determine temperature over time, and finally, the fourth simulation performs a thermomechanical analysis under constant load to verify the structure's critical temperature and fire resistance time. The model is evaluated based on the experimental results, comparing all the presented variables. Following the steps taken during the process, the root mean square error (RMSE) is calculated to verify the difference between the measured temperatures and those determined numerically. For temperature, a good approximation is assumed whenever the RMSE is below 100°C during the LSF wall's fire exposure time. The results show a reliable comparison between experimental tests and computational simulations, demonstrating not only the load-bearing capacity, temperature development, and critical temperature value but also ensuring the reliability of the computational simulations. This offers an alternative solution to expensive experimental tests, enabling the testing of alternative solutions for LSF walls. A new proposal is presented to determine the reliability and similarity between numerical simulations and experimental tests, testing new formats and different configurations of LSF walls.
Description
Mestrado de dupla diplomação com a UTFPR, Universidade Tecnológica Federal do Paraná
Keywords
Paredes LSF Resistência ao fogo Simulações numéricas