Abstract In general, a capsule boundary layer flow is strongly accelerated and has a favorable pressure gradient. Thus, the boundary layer is stable against superimposed freestream disturbances, and modal instabilities can only occur at significantly higher Re/l, which normally cannot be operated in conventional facilities. Capsules goes for transition as far as the boundary layer suffers from a critical amount of disturbance energy in the form of freestream noise, impacts of particles or roughness-induced disturbance energy. Therefore, this thesis aims at a systematic parametric scanning of possible flow parameters influencing the boundary layer transition on a capsule. In the Hypersonic Ludwieg Tube in Braunschweig (HLB), different versions of the Apollo capsule model are operated. Measurements taken with a coated capsule surface provide two sets of IR data. These prove an essential role of distributed surface roughness. The surface roughness characterization confirmed different mean surface roughness heights with respect to the Ra parameter, which is in the range of several micrometers. The noise level of the incoming freestream is assumed to be another important parameter in the transition process. Model instabilities can be successfully detected by surface sensors located close behind a single roughness element. With regard to spatial distribution predicted by stability calculations, the experimental methods can be used to check the spanwise and the streamwise extent of so-called high-speed streaks originating from the vortex system near the disturbance element. Unfortunately, the sampled data do not allow the calculation of amplification factors for the observed instabilities since the sensors also measure too high surface pressure caused by tunnel noise, which is negatively amplified when passing the shock and entering the boundary layer. The examination of deterministically distributed surface roughness takes into account the previous investigations with IR-measurement technique. Moreover, the stochastically distributed roughness mimics the most realistic state of the re-entry capsule in flight. Flush-mounted heat flux sensors identified the transition that only occurs with subcritical ordered roughness elements if they are exposed to sufficient freestream noise. Transition experiments with stochastically distributed roughness help to experimentally substantiate the promising method of predicting the transition for blunt bodies with roughness. These measurements are carried out without any significant influence of tunnel noise as the roughness heights dominated the transition process. The roughness applied on the entire surface shows good agreement with the available correlation. Since the roughness location and spatial extent are varied, the agreement with the correlation decreases, and the limitation of the correlation is demonstrated.
