Microscopic and macroscopic modelling: Role of basic data, computer codes and integral experiments

There exist several levels of modelling phenomena observed in nature or in systems built by human enterprise. For systems having safety implications, it is important that their macroscopic behaviour conforms to established norms and regulations. To accomplish this it is often necessary to understand the basic underlying phenomena, their interplay and their overall visible effect on the system, In order to model such behaviour three components are required: (1) the basic data containing and supporting the description of basic phenomena; (2) the computer codes responsible for combining each of these for predicting overall behaviour in a specific context; (3) data from experiments made on well-characterised systems for verifying that the basic data and models used correctly describe and predict the macroscopic phenomena. The OECD/NEA Data Bank, the Radiation Shielding Information Computational Center (RSICC), the IAEA Nuclear Data Section and others play an important role in making these elements available to the community of scientists and engineers. This article describes the idea behind the services provided by these centres and provides examples of the three elements required for system performance design, diagnosis and prognosis. (author)