“…In addition, experimental measurements in passive safety systems have often been limited by the high cost, and technical difficulties in achieving the required spatial resolutions needed for detailed understanding, associated with the large scale containments and extended flow loops normally involved (Chu et al, 1997;Rouge, 1997;Theofanus et al, 1997). Numerical modelling has also been mainly restricted to onedimensional, best-estimate thermal hydraulic system codes (Park et al, 2016;Hu et al, 2018;Valincius et al, 2018) that, in adopting a one-dimensional approach, are not equipped to deal with the intrinsic three-dimensional, multi-scale nature of most natural circulation flow patterns.…”