BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Dr. Stéphane Roux: "From tomography to motion ... and from motion
  to tomography"
DTSTART:20220608T170000
DTEND:20220608T180000
DTSTAMP:20260924T085515Z
UID:71e2e48ba9074f6496a37c8b253800b74983211de8660f425ebbdd9f
CATEGORIES:Conferences - Seminars
DESCRIPTION:Dr. Stéphane Roux\, CNRS\nAbstract. Materials science and now
  solid mechanics are revolutionised by the use of tomography. It provides 
 a complete microscale 3D image of both microstructure and kinematics in a 
 non-intrusive way. To reveal the mechanical properties of materials or str
 uctures\, a series of tomographic images acquired in situ all along a mech
 anical test appears to be the ideal way to measure 3D displacement fields 
 based on non-rigid tomographic image registration (called Digital Volume C
 orrelation\, or DVC\, in solid mechanics) from which the parameters of its
  constitutive law can be calibrated. Tomographic reconstruction\, registra
 tion\, and identification of the mechanical behaviour are three inverse pr
 oblems\, which can be chained or fused optimally with respect to noise (i.
 e. showing the least sensitivity). Integrating all those operations into a
  single procedure has the additional benefit of cutting down the needed ac
 quisition time by a significant factor (from several days to a few minutes
 ). The principle of such an approach is first to perform a plain tomograph
 y of a sample (potentially with no applied load) and then\, under increasi
 ng load\, evaluate the constitutive parameters from a series of radiograph
 s or projections. In this way\, motion (and its mechanical justification) 
 could be obtained efficiently at high temporal resolution.\nYet\, a first 
 static reconstructed image is still needed in such a procedure. Thus\, a n
 atural question is whether this first step is dispensable. This opens the 
 way to “dynamic tomography”\, whereby a series of radiographs (not lon
 ger than for a usual tomography) of a non-rigid body is used to reconstruc
 t both the body and its motion at each instant of time. An application to 
 medical imaging (chest tomography of a breathing patient) will be shown.\n
 \nBiography. Stéphane Roux is a CNRS Research Professor at the Ecole Norm
 ale Supérieure Paris-Saclay. After graduating from Ecole Polytechnique\, 
 and Ecole Nationale des Ponts et Chaussées\, he earned a Ph.D. in 1990 in
  Statistical Physics\, before joining the CNRS as a Research Professor in 
 1991 at Ecole Supérieure de Physique et Chimie Industrielles de Paris. Fr
 om 1998 to 2007\, he was the Director of a Joint Research Unit between Sai
 nt-Gobain (a worldwide company in material manufacturing and distribution)
  and the CNRS. He has worked at ENS Paris-Saclay since then. There\, his r
 esearch interests are in Digital Image Correlation\, Stereo-vision\, and T
 omography at the interface between experimental solid mechanics and modell
 ing. He received the CNRS Silver Medal in 2006\, and the Prix Jaffé from 
 the French Academy of Science in 2019.
LOCATION:https://epfl.zoom.us/j/66247740858
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR