Uncertainty management for computational materials science


Uncertainty management for computational materials science

Under the auspices of the Material Ageing Institute (MAI),the 5th edition of the training course “Uncertainty management for computational materials science” will take place from 23 to 25 May 2018. The course aims at promoting the use of advanced methodologies, existing numerical tools and good practices for quantifying the effects of uncertainties among material researchers-engineers of MAI partners.

The course is well suited to those who confront one (or some) of the following questions in their activities:

  • Which probability distribution describes the best the data acquired from measurements or numerical simulations? How are different quantities of interest correlated?
  • How can one verify the accuracy of an assumption applied to a quantity, e.g. its zero-mean and its normal distribution?
  • How can one calibrate parameters of a numerical model to best fit measured data?
  • Among multiple possible numerical models, which is the best one to be used (according to statistical measures)?
  • Which are input parameters with largest influences on the output of a model? In other words, will the output be changed globally and significantly due to slight variations of a parameter? What are measures of their influences? How can one obtain those measures?
  • What are the best methods and softwares to explore large muti-dimensional data sets? How about a quick introduction and tutorial to those softwares?
  • What is the probability that an output quantity exceeds a prescribed threshold? What is the level of confidence of this measure?
  • How can one model random fields of spatio-temporal quantities?

There wil also be presentations on industrial issues solved with the presented methods. A lab visit will showcase the handling of uncertainties in measurements. In particular, the attendees will have an opportunity to learn about current methods to estimate risk of a complex system such as the emergency power supply of a nuclear power plant.

Subscribe to our newsletter to stay up to date with all of SNETP’s activities and latest news (and/or update your preferences)