Earthquake engineering of nonstandard concrete structures: from the assessment of spatial variability of earthquake signals to numerical modelling of structural response
Event details
| Date | 16.12.2016 |
| Hour | 12:15 › 13:15 |
| Speaker | Prof. Dr Frédéric Dufour, Professor at 3SR laboratory and Vice President for research of Grenoble Institute of Technology, Grenoble, France |
| Location | |
| Category | Conferences - Seminars |
Over the last 30 to 40 years, modern design codes have been developed worldwide to increase the safety of civil engineering structures against an earthquake. For practical/economic reasons they mainly focus on classical structures to be built with restrictive hypothesis to allow simple engineering analysis. However, historical monuments and large structures for energy production (dams, nuclear power plants) need a specific attention of the scientific community.
Large structures for energy production have a huge interfacial surface with the ground. Therefore, the hypothesis of having a homogeneous loading must be questioned. Several sensor networks have been set in the world to analyze the amplitude and phase changes in space due to complex wave propagation in the underground. More recently, a dam has been instrumented with about 20 velocimeters to measure the spatial variability both under ambient noise and seismic events. The effect of topography on the phase and amplitude heterogeneities is analyzed with the help of spectral element models. The objective is to better identify the real loading felt by large structures and to identify whether the spatial variability is more or less critical for the structural safety.
Besides, for ancient/historical structures classical engineering methods are not valid. Thus, one needs to use modern finite element codes to assess the earthquake safety of such structures. Dynamic analyses have the potential to evaluate accurately local information (damage, crack opening, rebar yielding, etc.) although the computational time may be a drawback. Therefore, simplified numerical methods may be used for specific structures. For instance, the number of degrees of freedom of the model can largely be reduced by applying beam kinematic for some structural elements. This yields to the development of multifiber beam finite elements. Recently warping has been added to those models to account for shear deformation in the damage of concrete and yielding of rebars.
Finally, for the purpose of a probabilistic risk analysis in engineering design, one must rely on an efficient intensity measure to estimate the structural response. For instance, the well-known PGA accounts only for the signal measure without taking into account the modal analysis of the structure. Thus, it cannot be reliable for any structures since it does not account for resonance. Besides, by definition the spectral acceleration (SPa) is the best intensity measure for a single degree of freedom system with a linear behavior. However, real structures may undergo nonlinear behavior under intense earthquake. Recently, a new intensity measure has been developed accounting for the fundamental frequency of the structure and arbitrarily its reduction upon structural damage. This IM called ASA40 has been proved to be simple and efficient, although in some specific case a time-frequency analysis remain necessary to evaluate structural damage.
Bio : Prof. Frédéric Dufour obtained his PhD from the University of Nantes and Ecole Centrale of Nantes in 2002. He is now professor at Grenoble Institute of Technology in the 3SR laboratory and Vice President for Research of the Grenoble Institute of Technology. He is the head of the PERENITI Chair funded by EDF on the reliability of large structures. His research interests comprise the numerical modelling of complex fluids and the modelling of concrete structures with non-local approaches. Prof Dufour is the co-author of about 30 peer review journal papers with a h-index of 12 and 400 citations.
Practical information
- General public
- Free
Organizer
- Prof. Dr Brice Lecampion & Prof. Dr Katrin Beyer
Contact
- Prof. Dr Katrin Beyer