O trabalho analisa as forças atuantes no contraventamento de treliças triangulares de coberturas de madeira por meio de modelos numéricos computacionais. Os modelos foram confeccionados no sistema de análise estrutural Strap e consideram o funcionamento do contraventamento na estrutura tridimensional. Os valores máximos das forças obtidas nos modelos tridimensionais foram comparados aos valores mínimos indicados pelo método de estimativa da NBR 7190 (1997), este, baseado na instabilidade lateral do elemento contraventado. Os valores dos modelos tridimensionais também foram comparados aos métodos do Eurocode 5, da norma sul-africana SABS 0163 e aos valores propostos por Underwood (2000), por apresentarem formulação semelhante ao da NBR 7190 (1997). Os modelos representam galpões usuais na prática, com dimensões variando de 12 a 24 metros de vão; 24 a 96 metros de comprimento; 3, 4 e 6 metros de altura do pilar. Foram analisados para os tipos de treliça Howe e Pratt, para as classes de madeira C30 e C50 e para os tipos de telhas fibrocimento, metálicas e cerâmicas, materiais mais utilizados no Brasil. Os resultados mostram que as forças atuantes no contraventamento dos modelos tridimensionais são superiores aos valores obtidos pelo método da NBR 7190 (1997) e que valores obtidos pelos métodos propostos pelo Eurocode 5 e por Underwood (2000) apresentam-se mais seguros diante do funcionamento global da estrutura. / This work analyzes the forces acting on the bracing bars of triangular timber trusses by computational numerical models. The models are produced in the system of structural analysis Strap and consider the operation of bracing in the three-dimensional structure. The maximum forces obtained in the three-dimensional models are compared to the minimum values indicated by the estimation method of the NBR 7190 (1997), which is based on the lateral instability of the braced element. The values of three-dimensional models are also compared to the methods of Eurocode 5, the South African standard SABS 0163 and the values proposed by Underwood (2000). The models represent sheds with usual dimensions ranging from 12 to 24 meters wide, 24 to 96 feet long, and pillar 3, 4 and 6 meters tall. Are analyzed the Pratt and Howe types of truss, the timber classes C30 and C50 and the types of tiles: cement, metal and ceramics, materials most used in Brazil. The results show that the forces acting on the bracing of the three-dimensional models are greater than those obtained by the NBR 7190 (1997) and values obtained by methods proposed by the Eurocode 5 and Underwood (2000) provides more insurance on the overall functioning structure.

O trabalho analisa as forças atuantes no contraventamento de treliças triangulares de coberturas de madeira por meio de modelos numéricos computacionais. Os modelos foram confeccionados no sistema de análise estrutural Strap e consideram o funcionamento do contraventamento na estrutura tridimensional. Os valores máximos das forças obtidas nos modelos tridimensionais foram comparados aos valores mínimos indicados pelo método de estimativa da NBR 7190 (1997), este, baseado na instabilidade lateral do elemento contraventado. Os valores dos modelos tridimensionais também foram comparados aos métodos do Eurocode 5, da norma sul-africana SABS 0163 e aos valores propostos por Underwood (2000), por apresentarem formulação semelhante ao da NBR 7190 (1997). Os modelos representam galpões usuais na prática, com dimensões variando de 12 a 24 metros de vão; 24 a 96 metros de comprimento; 3, 4 e 6 metros de altura do pilar. Foram analisados para os tipos de treliça Howe e Pratt, para as classes de madeira C30 e C50 e para os tipos de telhas fibrocimento, metálicas e cerâmicas, materiais mais utilizados no Brasil. Os resultados mostram que as forças atuantes no contraventamento dos modelos tridimensionais são superiores aos valores obtidos pelo método da NBR 7190 (1997) e que valores obtidos pelos métodos propostos pelo Eurocode 5 e por Underwood (2000) apresentam-se mais seguros diante do funcionamento global da estrutura. / This work analyzes the forces acting on the bracing bars of triangular timber trusses by computational numerical models. The models are produced in the system of structural analysis Strap and consider the operation of bracing in the three-dimensional structure. The maximum forces obtained in the three-dimensional models are compared to the minimum values indicated by the estimation method of the NBR 7190 (1997), which is based on the lateral instability of the braced element. The values of three-dimensional models are also compared to the methods of Eurocode 5, the South African standard SABS 0163 and the values proposed by Underwood (2000). The models represent sheds with usual dimensions ranging from 12 to 24 meters wide, 24 to 96 feet long, and pillar 3, 4 and 6 meters tall. Are analyzed the Pratt and Howe types of truss, the timber classes C30 and C50 and the types of tiles: cement, metal and ceramics, materials most used in Brazil. The results show that the forces acting on the bracing of the three-dimensional models are greater than those obtained by the NBR 7190 (1997) and values obtained by methods proposed by the Eurocode 5 and Underwood (2000) provides more insurance on the overall functioning structure.

In 2012, falls from elevation in construction industry represented 36% of the total fatalities. The Occupational Safety and Health Administration requires workers to use fall protection systems where workers are 6 feet or more above a lower level. Anchors for fall protection systems attached to roof trusses may cause out-of-plane loading on these structures. Metal plate connected wood trusses (MPCWT) are not designed to carry out-of-plane loads and MPCWT performance under these loads are not evaluated in the design process. The goal of this research is to model and analyze MPCWT assemblies under out-of-plane loads. The rotational stiffness of truss-wall connections, and truss bracing elements are included in the structural component model. Previous experimental data of fall arrest anchor loading were used for model validation. A parametric study considering loading location, joint stiffness and dimension of trusses was conducted. The structural analog of the MPCWT assemblies were found to have first truss deflections within 4% difference, thereby the models were validated. From parametric study results, the load location was not changed the ultimate deflection in the truss assembly by maximum value of 9%. Out-of-plane joint stiffness was the parameter that caused a large difference in the deflection results, when the joists were assumed as either rigid or simple connections. The rotational stiffness of lateral and diagonal bracing should be included as model inputs for the accurate representation of experimental behavior. Truss lengths increased the deflection at the top chord of the first truss in the assembly as truss width increased. / Master of Science