[en] The Fukushima Daiichi post-accident calculations raised a question of applicability of present severe accident codes on different plant designs and accuracy of computation results. The level of code models precision depends on our understating of complicated severe accident phenomena and accident sequence of events. It is therefore crucial to have a representative experimental database which encompasses data of all relevant phenomena, such as core degradation, vessel and containment failure, hydrogen behavior, etc. The presentation provides a brief summary of Fukushima Daiichi related assessment activities, an overview of existing severe accident experiments and codes, and future plans for detailed model development and validation.(author).