CFD Online Logo CFD Online URL
Home > Jobs > Job Record #15470

CFD Jobs Database - Job Record #15470

Job Record #15470
TitleMachine learning applied to Computational Fluid Dynamics
LocationFrance, France, Rueil Malmaison
InternationalYes, international applications are welcome
Closure DateMonday, April 01, 2019
Among all CFD methods, the RANS (Reynolds-Averaged Navier-Stokes) approach 
remains the standard in industry and research centers. It allows to model 
flows in complex geometries involving multi-physical phenomena in timescales 
compatible with industrial constraints. This method has however a limited 
accuracy, while industry expects an increasing prectivity. In particular, 
the RANS approach is based on turbulence models to represent the entire 
spectrum of the flow turbulence. These turbulence models rely on strong 
simplifying hypotheses which reduce their ability to account for all 
phenomena and their complex interactions in realistic applications. This 
limits the accuracy of the simulation prediction in terms of flow mixing, 
thus strongly impacting the quality of the results.
The increasing interest for Machine Learning (ML) techniques has recently 
triggered pioneering research showing the potential of ML to improve RANS 
turbulence models based on the knowledge gained from experimental data or 
Direct Numerical Simulations. In this very promising context, this 
internship aims at applying ML algorithms to propose improved RANS models to 
be implement in the CFD code CONVERGE, thus bringing a better predictivity 
of 3D RANS simulations performed at IFPEN (aerodynamic for gas turbines, 
internal combustion engines, etc).
A first step of bibliography will aim at identifying relevant ML algorithms 
and at proposing methods to use them in combination with turbulence models 
in order to extend their domain of application. Tests will then be conducted 
on 1D academic cases, before considering more complex 2D/3D flows.

This internship will offer you:
- a topic applying machine learning to applied physics, in a re-known 
research center
- a supervision by research engineers in CFD and applied mathematics 
- it might be continued by a PhD (in collaboration with Argonne National 
Lab. / University of Illinois at Chicago)

Required skills:
-  Master's degree or third year of engineering school, with a strong 
background either in CFD or machine learning, and the willingness to learn 
the other topic
- Computational and programming skills (LINUX, Python)
- Interest for academic research,  fluent in English 

Internship duration: 5 to 6 months. The internship will take place at IFP 
Energies nouvelles, in Rueil-Malmaison (next to Paris, France). The intern 
will receive a monthly allowance (if he/she does not already earn a salary 

To apply, send your CV and letter of motivation.
Contact Information:
Please mention the CFD Jobs Database, record #15470 when responding to this ad.
NameAdele Poubeau
Email ApplicationYes
Address1-4 avenue du Bois Préau
Rueil Malmaison
Record Data:
Last Modified09:20:52, Friday, November 16, 2018

[Tell a Friend About this Job Advertisement]

Go to top Go to top