Prediction of thermal loads within cavities in space turbine, has as been a challenging task  in aspects of achieving accurate and reasonable estimations that are crucial for design concepts. The difficulty lies within the turbulent flow and its thermal interaction with the structure inside such section. It does not exist a method that works perfectly for prediction of thermal loads in any cavity and the taken approach to perform this kind of analysis has been differently chosen. The objectives of this work have been to improve methods for assessment of thermal loads in space turbines, especially calculation of the heat transfer coefficient and bulk temperature. As the thesis was conducted at GKN Aerospace Sweden, Trollhättan, one of theirs demonstrator turbine was chosen for the study case. Its first stage rotor blade and the nearby cavity were the main research regions. The flow can enter and exit the cavity through one slot and is characterized with a very low axial speed. For the studied regions, the wall surface has been subdivided into a number of segments. With prescribed wall temperatures and use of computational fluid dynamics (CFD) to compute the wall heat flux at the sections, the heat transfer coefficient and bulk temperatures were determined in three different ways. One of them was based on evaluating one single CFD result and derive the thermal loads from formulas. The others used by point-plotting approach, whereas one of them focused on formulating a model that describes the thermal interaction between the section walls. The results demonstrate that this model was able to predict a section's wall heat flux as a function of the wall temperatures in fair agreement with CFD results for a range of temperature variations. Further more, some of the predicted heat transfer coefficient at a section shows to be highly sensitive to the prescribed wall temperatures in the cavity and rotor blade.