Home News archive Grant for a PhD in engineering: Efficient parallel optimization algorithms for DNS and LES of turbulent mixing

Grant for a PhD in engineering: Efficient parallel optimization algorithms for DNS and LES of turbulent mixing

Promoters: J. Meyers, S. Vandewalle
Contact: Prof. dr. ir. J. Meyers, Department of Mechanical Engineering, Celestijnenlaan 300A, B3001 Leuven, Belgium. T: +32(0)16 322502. F: +32(0)16 322985. U: http://people.mech.kuleuven.be/~jmeyers/
Apply using the online application tool of Leuven Arenberg Doctoral School. (Applications by Email will not be considered!)

OPTEC (K.U.Leuven Optimization in Engineering Center), and the department of Mechanical Engineering (division TME) offer a grant for candidates who want to pursue a PhD in engineering, and gain teaching experience as a member of the didactic team of the mechanical engineering department.

Research

The candidate aims at a PhD in engineering with research topic: efficient parallel optimization algorithms for transient three-dimensional turbulent-flow simulations (DNS and LES) with applications towards turbulent mixing.

The PhD research focuses on the development of efficient optimization algorithms for parallel computing of complex three-dimensional transient flow simulations, with applications oriented towards turbulent mixing in the presence of passive scalars with and without buoyancy effects. For the flow simulations direct numerical simulations (DNS) and large-eddy simulations (LES) are considered. First of all, the use of algorithms such as BFGS, multiple shooting, multigrid, and various continuation methods are elaborated for large-scale flow optimization on high-performance computing systems. Next to that, the research will concentrate on algorithms which may improve on local optima found using gradient-based optimization techniques (e.g. using a combination with simulated annealing). Finally, the effect of uncertainty, and the presence of background turbulence, on optimization algorithms is investigated.

The research is funded by OPTEC (an excellence center on optimization at K.U.Leuven - www.kuleuven.be/optec/), and is well integrated in OPTEC's research structure, in particular in Working Group 6 on PDE-constrained optimization. The research further closely relates to the fluid-mechanics research at the Mechanical Engineering Department (supervised by Prof. J. Meyers), and the algorithmic research on PDE-optimization at the Computer Science Department (Prof. S. Vandewalle).

Requirements and competencies:

-Master Degree in engineering, applied physics or applied mathematics
-basic knowledge in fluid mechanics and CFD
-good programming skills: knowledge of Fortran, C++
-good English redaction skills for scientific reporting
-good command of the English language in general

Salary and benefits:

The phd grant is for four years. The monthly remuneration is about 1730 € after taxes, social security included.

Application deadline: August 1, 2010
Decision: when a suitable candidate applies
Earliest starting date: July 1, 2010

Research environment:

The Mechanical Engineering department of the K.U.Leuven has an excellent research track in both fluid mechanics in general and turbulent flow optimization. The K.U.Leuven, situated in Leuven near Brussels, is one of the oldest universities in Europe. Its rich history and patrimony provides an inspiring research environment, and this is combined with very high standard accommodation. For its computational research, the K.U.Leuven has a state-of-the-art cluster, 20 TFlops system with 1500 Intel Xeon cores, which is part of the Flemish Super Computing Center.

Leuven is a very agreeable city that is excellently geared to serve almost 30,000 students. The Mechanical Engineering building is situated on its perimeter in a peaceful green park near an ancient Castle. For more information, visit http://www.mech.kuleuven.be/dept/

Newsflash

Johan Suykens has been awarded an ERC Advanced Grant. 

The ERC Project is entitled "A-DATADRIVE-B: Advanced Data-Driven Black-box modelling" and will in the coming 5 years considerably reinforce the research of OPTEC's working group 2 on Data Driven Modelling, which is led by Johan Suykens. More info can be found on
http://www.kuleuven.be/research/erc/suykens.html

Join the OPTEC Info List!