Impact

The software development performed in this project has impact in two main categories: (i) application areas and (ii) future computing requirement. In application areas, a general coupling library is essential for interfacing specialised application softwares from various strands of research, in order to simulate multi-physics problems. With the ever increasing necessity to simulate complex physics using large-scale scientific computing, it is impractical to redevelop or merge algorithms which simulate various physical phenomena. Such approach is inefficient, and requires exhaustive validation and verification of the new codes. Adopting the coupler as a basis for interfacing existing and established algorithms is a much more robust and efficient approach. Some example applications where the current coupler is expected to be applied are simulations of electro-kinetic flows in microchannels and simulations of drag reduction using surface coatings. The same coupling library can be applied by other researchers, e.g. in coupling Reynolds Averaged Navier-Stokes solvers with higher fidelity methods such as direct numerical simulations, or performing coupled radiation transport and flow simulations in reactor physics. Many of these coupled applications are actively pursued at Imperial College and widely by the UK research community, e..g. members of the UK Turbulence Consortium. In addition, the coupler will create many new opportunities for the scientific community across a broad spectrum of disciplines.

It is important to consider the impact of the coupling library on the HECToR AU footprint. Using the coupler reduces this footprint in two ways: First, the coupler accelerates the solution procedure of multi-physics problems by concurrently running the various software, relative to iterative sequential coupling of applications which is at times adopted in multi-physics problems. Second, interfacing established applications using the coupler library circumvents the need to re-validate and load-balance multi-physics simulations that require more than one application software. This advantage reduced the number of man-months dedicated to this effort, and the HECToR AU requirements for such validation and scalability studies.