Rocket booster engines generally must operate at a large variety of atmospheric conditions. To achieve optimal total impulse during engine operation, these engines are normally optimized to be optimally expanded at high altitude, and low ambient pressure. The expansion ratio is constrained by the sea-level operational conditions, for a too high expansion ratio will result in internal flow separation from the nozzle wall. However, during start-up and shutdown, when the engine has not yet achieved nominal nozzle pressure ratio (NPR), between the combustion chamber and the ambient, flow separation still occurs. The unsteadiness and asymmetry of the separation and the resulting internal shock structure, impose significant lateral loads on the nozzle structure. These loads are exacerbated in thrust optimized contour (TOP) nozzles, where the flow transitions between two vastly different configurations, free and restricted shock separation (FSS and RSS), during the start-up or shut-down ramp.

Simulations and small-scale experiments on thick-walled nozzles have provided insight into the physics behind this transition. However, they never consider the effect of the flexibility of the nozzle wall on the flow structure, while in real rocket engines deformation amplitudes of the nozzle wall can reach up to 6% of the nozzle diameter. To investigate the effect of fluid-structure interaction on the magnitude of lateral loads, a test setup was designed at the high-speed laboratory to allow for testing of flexible rocket nozzles. Several tests with stiff and flexible nozzles were performed, taking wall pressure, side load, particle image velocimetry (PIV) and digital image correlation (DIC) measurements. The successful operation of the test setup was validated by comparing stiff nozzle wall pressure measurements to literature, and quantitative velocity field data was obtained using PIV. The PIV data hinted at the existence of a stabilizing recirculation region downstream of the Mach disk, which has been observed in simulations, but has never been experimentally confirmed. The flexible nozzle tests did not result in useful scientific data but did prove that DIC can be used effectively to perform modal analysis on the nozzle lip.