In a context where the mass, the cost and the mechanical strength of materials must be jointly optimized, it is necessary to have experimental data quickly available and sufficiently robust to make efficient conception choices. For thermomechanical fatigue, standard tests usually allow comparing material for the same temperature and strain ranges although differences in thermal properties such as conductivity or thermal expansion could make significant deviations when the same thermal flux is applied particularly for structure with forced heat flux operating regimes. A new protocol is then proposed in order to compare the specific resistances of metallic materials against thermomechanical fatigue. It can easily rank materials according to their lifetime under thermomechanical loadings where strain range and temperature amplitude are determined by the heat flux applied on an industrial part. The method is based on a complete numerical analysis to determine experimental loading conditions as a temporal evolution of temperature and mechanical strain representative of thermomechanical loading observed in TMF critical areas for the part. TMF tests on hollow specimens are carried out to rank the materials: temperature and strain amplitude are different for each alloys whereas heat flux is identical. A materials ranking list based on TMF resistance is then determined according to their lifetimes under "heat-flux-controlled" tests. The method is tested for exhaust parts and demonstrates the superiority of some cast irons over others, whereas the intrinsic isotherm mechanical properties suggested an alternative classification. The obtained ranking is confirmed by experimental tests on industrial structures.