Whole reactor calculation methods

There are three broad categories of deterministic whole reactor calculation methods currently in use (a) diffusion theory (b) spherical harmonics approximations, or Pn methods, and (c) discrete ordinates or Sn methods. Diffusion theory is widely used for reactor core calculations although the faster Pn and Sn methods developed in recent years are now replacing them. 

In recent years, the first order finite difference methods have been superseded by nodal methods. In these the flux in each mesh element, or node, is represented by a set of orthogonal functions, such as Legendre polynomials for each direction, or other types of expansion. Using such flux representations, a more accurate solution can be obtained using a coarser mesh. The matrix equations relating the various components of the flux become more complicated, involving a relationship between components of the flux inside the node and on the surfaces. 

---

A method has also been developed which corrects for mesh size effects and for the difference between diffusion and transport theory calculations, for use in broad-mesh whole reactor diffusion theory calculations (the MONSTRE code [4.37]). 

Structural materials, calculation methods and necessary margins were selected to provide serviceability of the diagrid-pressure pipes system during the whole reactor life. 

---