Loading...
Loading...

Go to the content (press return)

Advanced numerical algorithms for the improvement of the efficiency applied to wind-energy and solar thermal: tackling with new computational arquitectures

Total activity: 7
Type of activity
Competitive project
Funding entity
AGENCIA ESTATAL DE INVESTIGACION
Acronym
ANUMESOL
Funding entity code
ENE2017-88697-R
Amount
151.250,00 €
Start date
2018-01-01
End date
2020-12-31
Keywords
CFD, energía eólica, heat transfer, heteregenous computing, high performance computing, métodos numéricos, numerical methods, procesadores heterogeneos, solar thermal, solar térmico, turbulence, turbulencia, wind energy
Abstract
The objectives of this project are oriented towards the third challenge in the Spanish Strategy of Science, Technology and Innovation, i.e.
safe, efficient and clean energy. Within this framework, we propose to work on the development of advanced supercomputing algorithms.
They should be stable and reliable, working with a minimal amount of empirical information, and focused on two important renewable
energy fields: solar thermal plants and wind-energy. These fields involve many different physical mechanisms combining a broad range of
physical and temporal scales (multiphysics-multiscale modeling).
The project consists on two interrelated main lines of research. The first one is a basic research line focused on the development and
improvement of numerical techniques for fluid mechanics in turbulent regime and heat transfer. The key elements to improve their
accuracy are threefold: the computing capacity of modern (i) high performance computers (HPC) based on hybrid architectures (e.g,
CPU+GPU), the improvement of our (ii) numerical techniques and the reliability of (iii) turbulence models. In this regard, this research
project interlaces these three pillars with the aim to tackle the problem of turbulence and heat transfer in incoming HPC systems that
demand new algorithmic approaches with much higher arithmetic intensity. This central idea of the project will be present in the first seven
objectives. Namely, (i) the development of highly-parallel Poisson solvers (O2) and numerical algorithms for the radiative transport
equation (O3) that can run efficiently on any kind of modern supercomputer (O1) will constitute the computational base of the project.
Moreover, (ii) the development of numerical algorithms for complex geometries will also be addressed in the same vein. This includes the
construction of high-order formulations for unstructured meshes with high arithmetic intensity and low memory consumption (O4). In doing
so, the basic conservation principles of the governing equations need to be exactly conserved at discrete level; therefore, this type of
discretization are the most appropriate for turbulent flows. This should form a solid numerical basis for testing new turbulence models for
the subgrid heat flux (O7) and wall-models for turbulent boundary layers (O6) at high Reynolds numbers. The latter will be developed on
the basis of new implicit-explicit time-integration schemes with high arithmetic intensity (O5) with the aim to significantly increase the timestep
in the near-wall regions. Hence, near-wall regions will be integrated implicitly whereas the rest of the domain will remain solved
explicitly.
The second line of research is applied to the above-mentioned renewable energy fields: solar thermal plants and wind energy. This line
follows the supercomputing trends of the Scientific Excellence Challenges within the Horizon 2020 framework. Moreover, it constitutes a
great tool for the priority challenges in renewable energies in Social Challenges of the Horizon 2020. This line consists on two topics of
research; namely, multi-physics and multi-scale simulations of high-temperature solar thermal systems (O8) and wind energy (O9). The
project focuses on the detailed analysis of relevant elements of these systems with a minimal amount of empirical information.

Participants

Scientific and technological production

1 to 7 of 7 results