The use of composite materials for aeronautical applications has been growing since several years because of the opportunity to produce lightweight structures reducing the fuel bills and emissions. The need for fireproof certification imposes costly and time consuming experiments that might be replaced or complemented in the years to come by numerical calculations. The present work creates a CFD numerical model of a fireproof test. As an example, a composite part (plenum) located in an aircraft APU (auxiliary power unit) which provides power to the aircraft is investigated. A numerical calibration of the flame is conducted according to the fireproof standards. The results of fireproof tests demonstrate a good evaluation of the plenum temperature (discrepancies lower than 19%). The influence of an internal air jet within the studied part is also evaluated observed to evaluate how this could lower the requirements of certification rules. A thermal decrease as high as 38 % is found for a velocity of 10 m/s. Proceedings of the 2 nd IAFSS European Symposium of Fire Safety Science 1. Introduction The use of composite materials for aeronautical applications has been growing since several years because of the opportunity to produce lightweight structures reducing the fuel bills and emissions. The growing use of these materials leads to technical and design challenges to comply with safety standards and certifications, especially when fire safety requirements are concerned. Aircraft parts dedicated to firewall applications or located in a designated fire zone, should meet a fireproof requirement. Therefore the composite parts have to pass fire tests according to ISO 2685 [1] or FAA-AC20-135 (FAR-25) [2] standards. Both standards use an oil burner to heat the part with a minimum temperature of 1100°C for 15 minutes. In this work, a 3D numerical model of a fireproof test using a CFD code is created to investigate the predictivity of a numerical fireproof test. This numerical step is expected to replace experimentation during the development phases of the composite part before the certification test to reduce development cost. This numerical tool would help designers to choose between different composite materials and designs options to avoid critical temperature increases at certain areas and perforation in this composite part during fireproof tests. The second section is dedicated to the presentation of the experimental setup and the third one will present the physical and numerical modelling approaches. In the fourth section the computed temperatures are compared to the experimental ones to validate the presented numerical approach and the results are discussed. The influence of an internal air jet within the studied part is also evaluated The feasibility of replacing a thermal protection by an internal air jet is also presented in this paper as a first design variable case. 2. Experimental setup To be labelled " fireproof " as it is requested in most of the APU (Auxiliary power unit) part specifications and according to the related standards, the concerned part (here a composite plenum) has to resist 15 minutes to a calibrated flame. Criteria to establish the test is passed include no burn through of the part structure, as well as no ignition of the emitted smokes (backside part inner surface self-ignition). This second criteria is here investigated by measuring the part material temperature increase. The Figures 1 and 2 present respectively a picture and an overview of the experimental setup. The composite part is located at 100 mm from the outlet of the cone burner above a vibrating table (sinusoidal vibration of 0.4 mm amplitude and 50 Hz frequency). The oil burner (kerosene-air) operates with a kerosene flow rate of