Speaker
Description
The VKI free-flying ring test case has become an important benchmark for hypersonic shock-body interaction and six-degree-of-freedom free-flight prediction. Several independent CFD rebuilding efforts, including those from DLR, CIRA, Strathclyde, and RTECH, have reproduced the flow physics of the experiment but have consistently predicted a drag coefficient higher than that inferred from the experimental trajectory.
In the present work, simulations performed with the Mistral and Blizzard solvers revisit this discrepancy by examining the sensitivity of the result to the ring geometry and mass properties. A possible inconsistency was identified between the published model dimensions, the reported mass, and the expected mass obtained using the nominal aluminium density. In particular, the published ring thickness of 2.0 mm appears to lead to a mass larger than the reported experimental value.
As a working hypothesis, the ring thickness was reduced from 2.0 mm to 1.8 mm while keeping the rest of the experimental setup unchanged. This 10% reduction yields a substantially improved match with the measured free-flight trajectory and brings the predicted aerodynamic deceleration into close agreement with the experiment. The result suggests that the previously observed drag overprediction may not arise solely from modelling or numerical limitations, but could also be linked to uncertainty in the test-article geometry or mass properties.
This interpretation remains provisional: the proposed thickness correction has not yet been confirmed by VKI, and clarification of the exact manufactured geometry and material properties is still pending. Nevertheless, the study highlights the strong sensitivity of the flying-ring configuration to small geometric variations and underlines the need for precise, traceable mass-property data when using free-flight experiments as validation benchmarks.
The work further demonstrates the ability of the Mistral solver to reproduce complex hypersonic shock-interference dynamics when the reconstructed test-article properties are consistent with the observed motion.