The project aims at studying the effects of surface roughness on the mean and turbulent properties across a high Reynolds number supersonic turbulent boundary layer. High-speed aircrafts travelling at supersonic speeds are vulnerable to the onset of surface roughness. They are notorious for increasing skin friction drag and convective heat transfer. Studying these flows is critical in several aerospace applications. The majority of research to predict and understand these flows have been limited to experiments. These experiments have laid the foundation into providing a fundamental understanding of the physics of supersonic boundary layers affected by surface roughness. However, the database is not very comprehensive and little is known about the effects of different roughness geometries along with varying Mach and Reynolds number. This work proposes to tackle these problems and bridge the gap currently present in understanding these flows through a numerical characterization. This is done through large scale Direct Numerical Simulations which was historically improbable to perform for such high Reynolds number flows due to massive computational requirements. This enormous task will be addressed through the usage of a robust in-house solver developed in the research group. This solver is capable of utilising the potential of the powerful Graphics processing unit (GPU) technology, which, through its efficient computing capabilities, makes it more advanced than the present state of the art procedures. Surface roughness is currently an inevitable problem due to its inherent presence owing to manufacturing defects. These simulations will, therefore, give us a deeper understanding of the working of supersonic boundary layers which will certainly help build efficient future high-speed vehicles.
If the objectives laid out for this project are achieved, it is possible to foresee that the results will have a significant impact bringing contributions to the understanding of unsolved key issues, including:
i) characterization of turbulence alteration induced by the surface roughness geometry and importance of compressibility effects related to the generation of shock and expansion waves.
ii) verification of Townsend's outer layer similarity hypothesis under the various flow conditions investigated. This is a fundamental research question, with strong impact in terms of potential engineering applications, providing the possibility of using DNS data to estimate drag variation in operating conditions.
A parallel goal of the present project is to make the database available to the scientific community through the Web. Each user will have the possibility to access and explore the data and to use them interactively. This is a particularly important aspect of the project, considering that many research groups do not have the numerical expertise and the possibility to access the computational resources needed to perform this kind of simulations.