The present work describes and supports the methodology for the improvement of the wall model devel-oped in AINA and its specific application to the Japanese DEMO Water Cooled Pebbled Bed. The set-upand application of this approach aims to obtain robust models by estimating the behavior of the studiedsystems as accurately as possible. These systems are represented in a simplified way. This requires thecomputation of a 3D radiation transport which has been carried out by means of MCNP6.1, ADVANTG andthermal-hydraulic calculations using ANSYS®Fluent®. Several CFD mesh typologies and discretizationshave also been employed to test the Richardson theorem. In addition, 1D simplified models have alsobeen created and optimized for their usage in AINA code. The temperature distribution also shows goodagreement (within 7%). In some cases the simplified models have not behaved in a conservative man-ner compared with the outcomes obtained for the 3D models. This observed absence of conservatism isintrinsic to the 1D approach. To cope with these effects, scaling functions have been determined as a ratiobetween the most conservative radial temperature distribution – computed by fully detailed 3D CFD –and the 1D simplified model. The scaling functions will be applied to the AINA computed wall tempera-ture distribution. To conclude, the determination and coherence of the result obtained using independenttools and approaches, ANSYS®Fluent®vs AINA thermal-hydraulic routines, lead us to recommend theproposed methodology.