Cerfacs Enter the world of high performance ...

NEWS

Medal of the Academy of Air and Space: 3 CERFACS researchers awarded !

Brigitte Yzel |  11 July 2022

The Air and Space Academy awarded Carlos Pérez Arroyo, Gabriel Staffelbach and Jérôme Dombard with the Vermeil medal 2022, to honor the excellence of their work in the realization of the FULLEST project, the first high-fidelity simulation of an aircraft engine. Thanks to PRACE for awarding the access to the Joliot-Curie supercomputer (GENCI hosted at CEA/TGCC) and to DGAC for the funding of the ATOM project (No 2018-39) led by SAFRAN Tech. FULLEST also benefited from developments done in European projects:  EXCELLERAT (H2020 823691) and EPEEC (H2020 801051).  Read more


ROLAND GLOWINSKI CONFERENCE IN PARIS FROM JULY 5th to 7th

CERFACS |  4 July 2022

CERFACS will participate to the Paris meeting dedicated to the memory of Roland Glowinski from July 5th to 7th. Thierry Poinsot will present the advances of HPC in the field of energy and combustion. Roland Glowinski was CERFACS direcfor for 3 years in the 90s and largely shaped present activities of our research center.Read more

All News

RESEARCH PUBLICATIONS

Nadakkal-Appukuttan, S., Riber, E. and Cuenot, B. (2022) Analysis and design of a local time stepping scheme for LES acceleration in reactive and non-reactive flow simulations, Journal of Computational Physics, 470, pp. 111580, doi:10.1016/j.jcp.2022.111580

[pdf] [doi]

@ARTICLE{AR-CFD-22-106, author = {Nadakkal-Appukuttan, S. and Riber, E. and Cuenot, B. }, title = {Analysis and design of a local time stepping scheme for LES acceleration in reactive and non-reactive flow simulations}, year = {2022}, volume = {470}, pages = {111580}, doi = {10.1016/j.jcp.2022.111580}, journal = {Journal of Computational Physics}, abstract = {Explicit time integration based CFD solvers suffer from restriction on the maximum allowable time step computed from the well known Courant-Friedrichs-Lewy (CFL) stability criterion. This restriction poses severe challenge in carrying out large eddy simulation (LES) of reactive and non-reactive flows, where the grid resolution is fine. The challenge of restricted time step is further augmented when dealing with large computational domains that pose a wide disparity in the system time scales. In this study, a numerical methodology is presented based on local time stepping in an overset grid framework. The attainable speedup is found to be a function of the ratio of time steps used in the sub-domains and the ratio of the number of computational degrees of freedom. The method is analyzed using global spectral analysis (GSA) and shows excellent agreement in solution accuracy with the conventional explicit time integration based solver. The impact of local time stepping on the order of accuracy and global conservation properties are also presented. This method is then applied to simulate three flow test cases to demonstrate the ability of the method to reproduce the first and second-order turbulent statistics at reduced computational time.}, pdf = {https://cerfacs.fr/wp-content/uploads/2022/09/CFD_Nadakkal_JCP_AR_CFD_22_106.pdf}}

Suau, A., Staffelbach, G. and Todri-Sanial, A. (2022) qprof: a gprof-inspired quantum profiler, ACM Transactions on Quantum Computing, 1 (1), pp. Article 1, doi:10.1145/3529398

[doi]

@ARTICLE{AR-PA-22-116, author = {Suau, A. and Staffelbach, G. and Todri-Sanial, A. }, title = {qprof: a gprof-inspired quantum profiler}, year = {2022}, number = {1}, volume = {1}, pages = {Article 1}, doi = {10.1145/3529398}, journal = {ACM Transactions on Quantum Computing}, abstract = {We introduce qprof, a new and extensible quantum program profiler able to generate profiling reports of quantum circuits written using various quantum computing frameworks. We describe the internal structure and working of qprof and provide practical examples on quantum circuits with increasing complexity along with benchmarks of the tool execution time on large circuits. This tool will allow researchers to visualise their quantum algorithm implementation in a different and complementary way and reliably localise the bottlenecks for efficient code optimisation.}}

Barleon, N., Cheng, L., Cuenot, B., Vermorel, O. and Bourdon, A. (2022) Investigation of the impact of NRP discharge frequency on the ignition of a lean methane-air mixture using fully coupled plasma-combustion numerical simulations, Proceedings of the Combustion Institute, doi:10.1016/j.proci.2022.07.046

[doi]

@ARTICLE{AR-CFD-22-93, author = {Barleon, N. and Cheng, L. and Cuenot, B. and Vermorel, O. and Bourdon, A. }, title = {Investigation of the impact of NRP discharge frequency on the ignition of a lean methane-air mixture using fully coupled plasma-combustion numerical simulations}, year = {2022}, doi = {10.1016/j.proci.2022.07.046}, journal = {Proceedings of the Combustion Institute}, abstract = {The ignition of an atmospheric pressure laminar premixed methane/air mixture by Nanosecond Repetitively Pulsed (NRP) discharges in a pin-pin configuration is studied using fully coupled plasma-combustion numerical simulations. These simulations are performed using the AVIP code specifically developed for low temperature plasma modeling and coupled to the combustion code AVBP. A reduced chemical scheme for plasma-assisted combustion previously derived and validated is used to investigate the effect of the frequency of NRP discharges and the benefits of their chemical enhancement. It is observed that the induced shock wave produced by strong discharges is of major importance for ignition and can lead to quenching of the ignition kernels through strong induced recirculation of gases. Increasing the frequency of the discharges reduces this effect by depositing less energy at each discharge and accumulating energy more homogeneously between the electrodes, leading to a faster and more stable ignition. The minimum energy necessary to ignite decreases with increasing frequency and at the highest studied frequency (100 kHz) ignition has been achieved with 30% less energy than with a single-pulse discharge.}, keywords = {NRP Discharge, Plasma-assisted combustion, Ignition, Detailed simulations}}

Duquesne, P., Chanéac, J., Mondin, G. and Dombard, J. (2022) Topology Rule-Based Methodology for Flow Separation Analysis in Turbomachinery, International Journal of Turbomachinery, Propulsion and Power, 7 (3), pp. Article number 21, doi:10.3390/ijtpp7030021

[doi]

@ARTICLE{AR-CFD-22-76, author = {Duquesne, P. and Chanéac , J. and Mondin, G. and Dombard, J. }, title = {Topology Rule-Based Methodology for Flow Separation Analysis in Turbomachinery}, year = {2022}, number = {3}, volume = {7}, pages = {Article number 21}, doi = {10.3390/ijtpp7030021}, journal = {International Journal of Turbomachinery, Propulsion and Power}, abstract = {Boundary-layer flow separation is a common flow feature in many engineering applications. The consequences of flow separation in turbomachinery can be disastrous in terms of performance, stability and noise. In this context, flow separation is particularly difficult to understand because of its three-dimensional and confined aspects. Analyzing the skin friction lines is one key point to understanding and controlling this phenomenon. In the case of separation, the flow at the wall agglutinates around a manifold while the fluid from the boundary layer is ejected toward the flow away from the wall. The analysis of a three-dimensional separation zone based on topology is well addressed for a simple geometry. This paper aims at providing simple rules and methods, with a clear vocabulary based on mathematical background, to conduct a similar analysis with complex turbomachinery geometry (to understand a surface with a high genus). Such an analysis relies on physical principles that help in understanding the mechanisms of flow separation on complex geometries. This paper includes numerous typical turbomachinery surfaces: the stator row, vaneless diffuser, vaned diffuser, axial rotor and shrouded and unshrouded centrifugal impeller. Thanks to surface homeomorphisms, the generic examples presented can easily be converted into realistic shapes. Furthermore, classical turbomachinery problems are also addressed, such as periodicity or rotor clearance. In the last section, the proposed methodology is conducted on a radial diffuser of an industrial compressor. The flow at the wall is extracted from LES computations. This study presents the different closed separation zones in a high-efficiency operating condition}, keywords = {critical point, flow separation, topology rule, turbomachinery}}

Lazzara, M., Chevalier, M., Colombo, M., Garay Garcia, J., Lapeyre, C. and Teste, O. (2022) Surrogate modelling for an aircraft dynamic landing loads simulation using an LSTM AutoEncoder-based dimensionality reduction approach, Aerospace Science and Technology, 126, pp. Article number 107629, doi:10.1016/j.ast.2022.107629

[doi]

@ARTICLE{AR-PA-22-78, author = {Lazzara, M. and Chevalier, M. and Colombo, M. and Garay Garcia, J. and Lapeyre, C. and Teste, O. }, title = {Surrogate modelling for an aircraft dynamic landing loads simulation using an LSTM AutoEncoder-based dimensionality reduction approach}, year = {2022}, volume = {126}, pages = {Article number 107629}, doi = {10.1016/j.ast.2022.107629}, journal = {Aerospace Science and Technology}, abstract = {Surrogate modelling can alleviate the computational burden of design activities as they rely on multiple evaluations of high-fidelity models. However, the learning task can be adversely affected by the high-dimensionality of the system, complex non-linearities and temporal dependencies, leading to an inaccurate surrogate model. In this paper we present an innovative dual-phase Long-Short Term Memory (LSTM) Autoencoder-based surrogate model applied in an industrial context for the prediction of aircraft dynamic landing response over time, conditioned by an exogenous set of design parameters. The LSTM-Autoencoder is adopted as a dimensionality-reduction tool that extracts the temporal features and the nonlinearities of the high-dimensional dynamical system response, and learns a low-dimensional representation of it. Then, a Fully Connected Neural Network is trained to learn the simplified relationship between the input parameters and the reduced representation of the output. For our application, the results demonstrate that our LSTM-AE based model outperforms both Principal Component Analysis and Convolutional-Autoencoder based surrogate models, in predicting the parameters-dependent high-dimensional temporal system response. }}

All publication

JOBS OFFERS

PhD (Cotutelle between Toulouse University and Polytechnique Montréal): Randomized Algorithms for Preconditioning of Nonlinear Inverse Problems

 

PhD (Cotutelle between Toulouse University and Polytechnique Montréal): Randomized Algorithms for Preconditioning of Nonlinear Inverse Problems...Read more


Hydrodynamic modelling at CERFACS

 

Contact: Sophie Ricci Senior Researcher ricci@cerfacs.fr Objective: The project aims at making use of Earth Observation...Read more

ALL OFFERS