Reactive flow cylinder code runs on six explosives were made with rate constants varying from 0.03 to 70 mus(-1). Six unconfined/steel sets of original ANFO and dynamite data are presented. A means of comparing confinement effects both at constant radius and at constant detonation velocity is presented. Calculations show two qualitatively different modes of behavior. For U-s/C-0 greater than or equal to 1.2, where U-s is the detonation velocity and C-0 the zero-pressure sound speed in steel, we find a sharp shock wave in the metal. The shock passes through the steel and the outer wall has a velocity jump-off. For U-s/C-0 less than or equal to 1.04, we find a pressure gradient that moves at the detonation velocity. A precursor pulse drives in the explosive ahead of the detonation front. The outer wall begins to move outward at the same time the shock arrives in the explosive, and the outer wall slowly and continuously increases in velocity. The U-s/C-0 greater than or equal to 1.2 cylinders saturate in detonation velocity for thick walls but the U-s/C-0 much less than 1.04 case does not. The unconfined cylinder shows an edge lag in the front that approximately equals the reaction zone length, but the highly confined detonation front is straight and contains no reaction zone information. The wall thickness divided by the reaction zone length yields a dimensionless wall thickness, which allows comparison of explosives with different detonation rates. Even so, a rate effect is found in the detonation velocities, which amounts to the inverse 0.15-0.5 power.