Numerical simulation of ice accretion on aircraft surfaces necessitates a good prediction of wall friction coefficient and wall heat transfer coefficient. After the icing process begins, surface roughness induces a high increase of friction and heat transfer, but simple Reynolds analogy is no longer valid. An experimental campaign is conducted to provide a database for numerical model development in the simple configuration of a heated flat plate under turbulent cold airflow conditions.The flat plate model is placed in the centre of the test section of a wind tunnel. The test model is designed according to constraints for the identification of friction and heat transfer coefficients. It includes three identical resin plates which are moulded to obtain a specified roughness on the upper surface exposed to the flow. Only the 3rd resin plate is heated on its lower face by an electrical heater connected to a temperature regulator. The evaluation of the friction coefficient is based on velocity profile measurements by laser Doppler Velocimetry. A methodology for heat transfer coefficient identification is defined. It is based on the surface temperature measurement by infrared thermography during thermal transient conditions. The heat flux at the wall is calculated by a thermal numerical model of the resin plate.The experimental campaign begins with validation of the experimental methodologies with the smooth flat plate. The experimental results are compared to reference numerical results and exhibit very good agreement in terms of friction and heat transfer coefficients. From the three test models with roughness, two models are tested under zero pressure gradient flow. In these test conditions, the measurements show an increase of the friction coefficient by a factor 2.0 and 1.9 respectively for models nr. 2 and nr. 3. On the other hand the heat transfer is identical for these 2 models and increases by a factor 1.8.