Smeets [1] published in 1988 a new concept for a ram accelerator with guiding tube rails for firing rail stabilized projectiles. This concept replaces Hertzberg et al.'s [2] fin stabilized projectiles accelerated in a cylindrical bore. The rail tube idea offers some advantages, e.g., no sabot is necessary as required for fin guided projectiles, simple projectile geometry, and possibility of varying the inner tube geometry. This principle was tested in 1993 and 1994 in rail tube version I and is now again under investigation since the beginning of 1997 in our RAMAC 30 in version II. In the rail tube concept, circular and finless projectiles are guided in a ram-tube equipped with five inner rails. At the moment we use a ram-section with a length of about 4.8 meters. A conventional powder gun serves as pre-accelerator. In the gun tube with a length of 2.8 meters, projectiles of about 150 grams are accelerated to a muzzle velocity of approximately 1800 m/s which is the initial velocity at the entrance of the ram-section. For successful operating a ram accelerator, the heat release must be limited to avoid thermal choking followed by an unstart. This choking phenomenon will be investigated in detail in this paper from the gasdynamic point of view in order to predict the right mixture for the given flow conditions around the ram projectile. Moreover, to avoid a firing failure, the material point of view must also be considered. Some recent firings have been done using aluminium, titanium and steel as test materials and its behaviour is discussed herein in detail. The first outcome is for example, for a given projectile geometry and a given gas mixture with a steel cowling no ignition occurs, whereas with aluminium or titanium as combustor surface material the ignition starts well followed by a projectile acceleration.