This paper describes the results of numerical analysis of a hydraulic structure subjected to seismic load. The paper is divided into two main parts. First, a numerical procedure for the description of mechanical effects of alkaliaggregate reaction in concrete structures is briefly outlined. Subsequently, a single unit of the Beauharnois power plant, situated in Quebec (Canada), is analysed in the context of seismic excitation. The numerical simulations are aimed at assessing the dynamic stability of this unit and investigating the evolution of damage associated with inception of macrocracks.Résumé : Cet article décrit les résultats de l'analyse numérique d'une structure hydraulique sujette à une charge séismique. L'article est divisé en deux parties principales. En premier lieu, une procédure numérique pour la description des effets mécaniques de la réaction alcali-granulat dans les structures en béton est exposée brièvement. Par la suite, un simple groupe de la centrale de Beauharnois, situé dans la province de Québec (Canada), est analysé dans le contexte d'une excitation séismique. Les simulations numériques ont pour buts d'évaluer la stabilité dynamique de ce groupe et d'étudier l'évolution des dommages associés à la création de macro-craques.Mots clés : réaction alcali-granulat, plasticité, analyse séismique, structure hydraulique. [Traduit par la Rédaction] 338Parvini et al.

No abstract

The seismic stability of the facade brick-masonry walls of the machinery building of the Beauharnois powerhouse near Montreal, Quebec, Canada were investigated numerically by use of non-linear models and applying experimental methods on site and on the IZIIS' seismic shake-table. The dynamic properties of the machinery building were obtained by ambient vibration measurements. Based on these results, a model of a representative part of the building, consisting of steel frames and brick masonry wall, was designed and constructed to the reduced scale at the IZIIS' Dynamic Testing Laboratory and then tested on the two-component shake-table. The geometry of the original structure was completely scaled to 1/3, consisting of many realistically simulated details such us: brick layers, steel columns, openings, window frames, steel connectors between brick layers, number of layers, brick dimensions, etc. The material used for the model was: original steel for the frame structure and bricks of reduced mechanical properties for the masonry wall, close to the similitude requirements according to the Backingham's theorem, valuable for adequate artificial-mass simulation model as well as true replica simulation model. More than 50 seismic tests were performed considering the design earthquake Nahanni NWT, H1, with a time scaling factor of 3 1/2 , and acceleration scaling factor 1, according to the model design rules. The intensity of the applied input earthquake excitation was from 0.05 to 1.2 g. The design peak acceleration of Nahanni earthquake was 0.2 g. The cracks development was stated at 0.7 g input acceleration. These were concentrated around the openings. No collapse happened even under the strongest earthquake input. The numerical part of this paper deals with formulation/application of the critical plane approach to seismic analysis of masonry structures. Starting with the constituents, i.e. mortar and bricks, the macroscopic strength properties of masonry were established based on numerical homogenization. Generally, based on all the performed experimental 123 422 Bull Earthquake Eng (2010) 8:421-450 tests, considering some simplifications and assumptions in the constructing details, as well as in the design of the model, the global conclusion is that the existing wall is very well incorporated in the steel structure of the powerhouse. The complementary stiffness of the steel frame and the brick masonry wall produces interactive deformation of the system. Only local cracking and relative displacement between the wall and the steel frames could be expected in the case of a strong earthquake.

No abstract

No abstract