It is common knowledge that one of the main issues of hypersonic flight is the thermal management of the overall vehicle and more specifically the cooling of the engine, since even composite materials can't withstand the large heat load found in a Scramjet combustion chamber. Another critical point is that mixing and combustion should be sufficiently fast in order to avoid long combustion chamber caused by supersonic internal flow and short residence time. Cryogenic fuels are a logical choice but their lack of storability and low density make them second choice compared to liquid hydrocarbons for small vehicle application. Researches are currently conducted in order to determine if the cooling can be achieved by the endothermic thermal decomposition of the fuel itself circulating trough the engine. The other benefit of this decomposition is the expected shift in the fuel mole fraction, from heavy hydrocarbons (with long induction delays), to light species (mainly H 2, CH4 and C2H4). In order to quantify the heat transfer in the cooled structures and the composition of the cracked fuel entering the combustor, an accurate predictive model of the thermal decomposition of the fuel is required. For this purpose, an experimental and modeling study of the thermal decomposition of generic molecules (typically long-chain alkanes and polycyclic non-aromatic complex molecules) that could be good surrogates of real fuels, has been started at the DCPR laboratory. The thermal decomposition of those generic molecules is studied in an isothermal jet-stirred reactor. Products of the thermal decomposition are heavy species like aromatics and poly-aromatics and lighter species like hydrogen, methane, ethylene and other middle-weight alkenes. Those results allowed the validation of the kinetic models that have already been developed. During the same time, MBDA-F launched collaborative project named COMPARER, focusing on system analysis to identify one or two characteristic parameters (able to be measured) needed to understand and control the complex phenomena involved in the presented cooling technology and to evaluate some associated sensors. Laminar flame speed for different fuel composition were computed in order to be able to tune the mixture for the lowest possible ignition delay. If at first glance a high mole fraction of hydrogen seems to be interesting, this generally also lead to a high mole fraction of methane (a molecule well known for its long ignition delay and an un-welcomed product) and high mole fractions of middle-weight species (such as ethylene, acetylene, butadiene, cyclopentadiene...) seems to be more promising from a combustion point of view.