SMART – Seismic design and best-estimate Methods Assessment for Reinforced concrete buildings subjected to Torsion and non-linear effects

Dedicated website :

Reinforced concrete (RC) structures exhibiting three-dimensional effects such as torsion and non-linear response are a main concern in the field of earthquake research and regulation. Over the last decades, several RC mock-ups have been tested in Western Europe, on simple walls (a dozen campaigns, including SAFE) or on wall and floor structures (CASSBA, CAMUS, etc.). SAFE tests (“Structure Armées Faiblement Elancées”) that lied in pseudo-dynamic tests carried out between 1997 and 1998 at ISPRA JRC, were devoted to the experimental characterization of the dynamic behaviour of low shear span walls [1]. The walls were subjected to a series of earthquake loadings with increasing levels (up to fifteen times the design level). Some well-known phenomena have been confirmed by these tests: failure modes, shape of the nonlinear load/displacement diagrams, allowable distortion described in guidelines. Other phenomena have also been observed: an overestimation of the initial stiffness (maybe due to the boundary conditions), margins dependent on the frequency content of the loadings, limited ductility of the walls or cracking before reaching the maximum bearing capacity. CASSBA tests (1990-1993) [2] and CAMUS (1996-2002) [3] have been performed at CEA Saclay on multi-story mock-ups. These tests have shown the good behaviour of large lightly reinforced wall structures, only minor damages were observed at the design level. These campaigns have both helped to improve the knowledge on RC structural behaviour and provide reference data for model development and validation. It has also helped to introduce special design provisions for similar structural systems in Eurocode 8.

In order to assess the capability of buildings to withstand earthquake loading as well as seismic loads induced by their equipment, a reduced scaled model (1/4th scale) representative of a typical simplified half part of an electrical nuclear building was designed, as part of the SMART 2008 project (Seismic design and best-estimate Methods Assessment for Reinforced concrete buildings subjected to Torsion and non-linear effects) (cf. Figures 1a and 1b) [4]. This specimen was tested between June and October 2008, on the AZALEE shaking table belonging to the TAMARIS experimental facility operated by the Seismic Mechanic Study Laboratory (EMSI) of the Thermal and Mechanical Studies Service (SEMT), as part of the System and Structures Modeling Department (DM2S) of the Nuclear energy Direction (DEN). This project was supported by the French Nuclear Energy and Alternative Energies Commission (CEA) and Electricité de France (EDF). In addition, a blind predictive benchmark started in 2007. The objectives were:

  • to assess the capabilities of conventional design methods for structural dynamic response and floor response spectra calculations;
  • to compare best-estimate methods for structural dynamic response and floor response spectra evaluation including various practices, depending on participants’ experiences.

Results of this benchmark were presented on December 2010 during the 1st SMART workshop held in CEA Saclay.

To keep on improving the knowledge about the seismic behaviour of such reinforced concrete structures, a new research program called SMART 2013 and composed of an experimental campaign and of an international benchmark, was put in place. Regarding the experimental part of the project, the main differences between SMART 2013 and SMART 2008 are the following ones:

  • the input signals come from natural records and are not synthetic (except for low PGA range random signals);
  • high intensity seismic loadings are applied at the beginning of the testing sequence;
  • the effect of an aftershock is considered.

Regarding the SMART2013 benchmark, a specific attention is paid to the characterization of the numerical models developed by the participants and the methodologies used to deal with uncertainties propagation within seismic vulnerability analysis by means of fragility curves. Furthermore, additional data were provided to ensure an efficient control of the boundary conditions such as the time/history quantities at the actuator level.

A workshop gathering the international community working in the field of earthquake engineering will take place at Paris from the 25th of November to the 27th of November 2014. The objectives are to discuss about the best seismic assessment method and floor response spectra computation methodologies. In addition, topics related to the way of modeling uncertainties related to the seismic signal and to the input parameters of the nonlinear constitutive laws will be discussed.


  1. Pegon P, Magonette G, Molina FJ, Verzeletti G, Dyngeland T (1998). Programme SAFE: Rapport du test T5. Technical Note, Joint Research Center.
  2. Coin A. (1993) CASSBA research report. FNB, CEA and MRT. Paris, France.
  3. Coin A. (1998) CAMUS 2000 research report. FNB, CEA and MRT. Paris, France
  4. Lermitte S., Chaudat T., Vandeputte D., Payen T., Viallet E. (2007) SMART 2008 : Evaluation de la fragilité d’un bâtiment de 3 étages présentant des effets tridimensionnels. 7ème Colloque National AFPS 2007 – Ecole Centrale Paris.