# Full information with illustrations here
https://filesender.renater.fr/?s=download&token=344b64dc-5bd9-4538-a0b4-
bb530453fa95
# Context
A future asset for the low-carbon transition, hydrogen remains a leading
scientific and security challenge. Colorless and odorless, hydrogen leaks
easily, ignites at low concentrations and temperatures, and can lead to the
spread of rapid deflagrations as well as detonations, which are a dangerous type
of supersonic combustion. Understanding the transition mechanisms from
deflagration (slow flame) to detonation (supersonic flame accompanied by a
sudden shock wave) is therefore essential for securing hydrogen production
installations (electrolyzers) as well as for nuclear industry. Indeed, in the
accident scenario of loss of cooling and melting of the core of a reactor, the
oxidation of the uranium rod cladding can lead to the release of hydrogen. It
was the subsequent explosion that led to the loss of containment and the release
of radioactive materials at Fukushima and Three Mile Island. Controlling the
hydrogen risk is therefore one of the major challenges of nuclear safety.
The main mechanism of the deflagration -> detonation transition is the presence
of obstacles along the flame path. These will generate vorticity, which
increases the surface area of the flame and accelerates the reactive wave. When
the number and size of the obstacles is sufficient, a runaway effect and wave
reflections can lead to a shock-chemical reaction coupling: the detonation is
born, propagating at several kilometers per second. Unfortunately, we cannot
avoid that industrial installations are cluttered with obstacles: pipes,
buildings, machines, walkways, structures… and present this type of scenarios.
Conversely, a very heavily congested porous medium can, on the contrary, smother
a flame that is too rapid and allow the reverse transition detonation ->
deflagration, of a less dangerous nature. For example, we see that a detonation
can be attenuated by passing through a porous matrix (see refs. [1,3]), or when
a porous medium is placed along the walls during detonation. propagation in a
tube [4,5]. A crucial safety question then arises: under what circumstances does
an obstacle accelerate or slow down a flame? Can we design porous media that
stop dangerous flames?
# Goals
This thesis work aims to address this question from three angles:
- on the one hand, via the preparation and carrying out of experimental
testson the hydrogen explosion test bench at CEA Saclay (SSEXHY, see Fig. 3).
Among others, these are:
- to explore different geometries for porous media, based on configurable
topologies. For example, we could take inspiration from Triply Periodic Minimal
Surfaces (TPMS) type functions. These porous matrices will then be 3D printed
via metal additive manufacturing (316L stainless steel);
- to prepare the instrumentation of the SSEXHY explosion test bench
equipped with a visualization section via a Schlieren technique coupled with an
ultra-fast camera at several million images per second;
- to post-process the results of shock sensors, pressure sensors and
photomultipliers equipped with an OH* filter.
- on the other hand, via numerical simulations of the DNS or LES type on
search calculation codes: an internal CEA/CNRS code (DRYADS) developed during
the CEA thesis of Luc Lecointre [1], and a CERFACS code (AVBP) well known in the
combustion community. For example, we could be interested in:
- the influence of the geometry of porous obstacles (shape, porosity,
hydraulic diameter, etc.) on the speed of flame propagation and the
deflagration<->detonation transitions;
- the influence of the 2/3D character of the porous materials;
- the proposal of new criteria for choosing the level of mesh refinement
for capturing the phenomena of interest;
- testing new temporal integration schemes for diffusion and reaction
rates;
- to the participation and testing of the porting of the DRYADS code in
C++ on a new HPC platform (samurai), developed by the CMAP laboratory of the
Ecole Polytechnique.
- finally, a theoretical modeling of the problem from the point of view of
volume-averaged equations can be carried out, with the objective of developing
simplified and predictive models on the behavior of porous flame arresters.
This thesis is a continuation of a nearby internship focused on 2D digital
simulations, offered in 2024 and which can be linked to the thesis subject as a
preparatory internship (subject available on the INSTN website or contact the
supervisors for more information). information).
# Work environment
The internship will take place at the CEA center in Saclay within the LE2H
experimental laboratory (formerly LIEFT).
# Skills required or desired
Demonstrated physical acumen, mathematical and modeling skills, knowledge of
numerical schemes/methods, foundation in C++ and Python programming. Knowledge
of compressible fluid mechanics and combustion would be a plus.
# Required profile
- Master 2 or equivalent in fluid mechanics;
- Curiosity about fundamentals, interest in experimental and ideally in
digital simulation;
- Rigorous, involved, passionate.
# References
[1] Lecointre, L. (2022). Hydrogen flame acceleration in non-uniform mixtures.
Doctoral thesis from the University of Paris-Saclay.
[2] Lecointre, L., Vicquelin, R., Kudriakov, S., Studer, E., Tenaud, C. (2022).
High-order numerical scheme for compressible multi-component real gas flows
using an extension of the Roe approximate Riemann solver and specific
Monotonicity-Preserving constraints. Journal of Computational Physics, 450:
110821.
[3] Radulescu, MI, & Maxwell, BM (2011). The mechanism of detonation attenuation
by a porous medium and its subsequent re-initiation. J. Fluid Mech., 667, 96-
134.
[4] Teodorczyk, H., and JHS Lee. (1995) Detonation attenuation by foams and wire
meshes lining the walls, Shock Waves 4:225–236.
[5] Radulescu, R., and JHS Lee. (2002). The failure mechanism of gaseous
detonations: experiments in porous wall tubes, Combust. Flame 131:29–46.
|
