Below you find the complete list of Tier-1-projects since the start of the regular project application programme.

Robustness and Sensitivity of Pareto-optimal Selection Designs under Various Types of Candidate Withdrawal

Date: 01.11.2016
  • Promotor(s): Wilfried De Corte
  • Institution(s): UGent
  • Domain(s): Psychology
The validity/diversity dilemma constitutes the biggest threat to the practice of personnel selection because applicant pools are increasingly heterogeneous and the most valid pre-employment tests often show substantial mean group difference. To address the issue, methods have been proposed for designing selection systems that are Pareto-optimal and, hence, are expected to offer the best possible trade-off between the goals of selection quality and diversity. However, it is currently uncertain whether these systems live up to their promise in actual selection applications where selected candidates may refuse job offers. Using simulation methods, the present project aims to resolve this uncertainty.

xPic3D : Next generation Particle-in-Cell algorithms for exascale HPC architectures

Date: 01.11.2016
  • Promotor(s): Giovanni Lapenta
  • Institution(s): KU Leuven
  • Domain(s): Astronomy and astrophysics
The next generation of supercomputers, composed of millions of processors, will be online in the 2020-2025 timeframe. Such machines will use a mix of multiple technologies, including the classic CPUs and new accelerators (MICs and/or GPUs), to achieve extreme computing power. The CmPA of KU Leuven is targeting such systems to perform detailed simulations of plasmas, for applications in space physics, space weather, electric propulsion and plasma confinement. However, the codes that we use today for our simulations are not fully adapted to the next generation of supercomputers. We are currently testing the next generation of codes that will allow to make full use of the upcoming architectures. We have developed a Particle-in-Cell code that shows perfect weak and strong scaling in small systems. In this project we will go one step further, testing the code in middle range systems and targeting real physical applications. Our goal is to demonstrate the scalability of the next generation of Particle-in-Cell algorithms and gather data for the projection of the performances on large-scale systems.

Not for publication

Date: 01.11.2016
  • Promotor(s): Kurt Lejaeghere
  • Institution(s): UGent
  • Domain(s): Physics
/

High-throughput screening of an unknown quaternary crystal space

Date: 01.11.2016
  • Promotor(s): Michael Sluydts
  • Institution(s): UGent
  • Domain(s): Technology , Physics
While many materials are known, many more are waiting to be discovered. One way to find them is massive screening using quantum simulations, which is capable of predicting the stability of hypothetical materials. Following up on previous projects where existing families of quaternary crystals were significantly expanded, here a completely hypothetical new crystal structure is explored. In this way the database of quaternary materials will be significantly diversified opening the way to advanced data mining techniques.

The role of humidity in the CO gas sensing mechanism of the clean and impurity modified ZnO surfaces

Date: 01.11.2016
  • Promotor(s): Nasrin Sarmadian
  • Institution(s): UAntwerp
  • Domain(s): Physics
Nanosized chemical sensors hold great potential for applications where conventional analytical systems are prohibitively expensive or impractical. Optimization of the sensor performance requires a detailed understanding of the underlying microscopic mechanisms responsible for its selectivity and sensitivity. In the current proposal, we use density functional theory (DFT) to investigate the local changes in the structure and electronic properties of ZnO due to the adsorption of the CO molecule on the defect-free and impurity-doped ZnO in presence of humidity. We calculate the surface energy, adsorption energy of the CO molecule and the charge transfer between the molecule and the ZnO (10-10) surface.