Discussion thread: "A fluid formalism for low-temperature plasma flows dedicated to space propulsion in an unstructured high performance computing solver"

[gbogopolsky]

Hello everyone,

My name is Guillaume Bogopolsky and I am currently in my 2nd year PhD at CERFACS in Toulouse, France. I work with fluid and hybrid simulations of ExB discharges applied to Hall thrusters. We are investigating a methodology to improve the predictivity of simulations for industrial use cases.

In these slides, I present to you the last work realized by the previous PhD student on the subject, V. Joncquières. He developped a fluid formalism for Hall thruster simulation with a 10-moments fluid model, a chemistry, boundary conditions and injection source term based on the Boeuf test case of Landmark 2a (Charoy et. al. 2019) and applied it to an axial-radial HT simulation.
Results show good agreement with literature results and good macroscopic parameters. More work is needed to into account dielectric boundaries and heat flux modeling.

The paper describing this work can be found here.

I will be very happy to discuss this methodology with you,

Guillaume

[eahedo]

HI Guillaume

Nice work.
Can you estimate the computational time of your (z,r) code compared with a similar hybrid code?

And a couple of questions,
How do you manage Bohm condition at sheath edges and losses to walls?
Which value do you use for gamma and how do you justify it?

[gbogopolsky]

Hello Eduardo,

Thank you very much for your interest in this work.
As we solve the full Euler equations for each fluid (ions and electrons) and resolve the sheaths, the CFL condition imposes very small convective time step of the order of 2.5e-12 s. Therefore, I expect the same order of computational time on a 2D case compared with a hybrid code, given the simpler model usually used for electrons.
Without resolving the sheaths, we have a factor 4 compared to PIC on a 3D HET sector.

Walls are considered as metallic, so any plasma hitting a wall is considered lost (with a correction for neutralisation). Dielectric walls are a work in progress.
In this computation, sheaths are resolved following the methodology introduced in Joncquières et. al. (2018) by using a thermal flux condition on the mass and energy conservation equations at the walls.

We use a gamma of 5/3 for a monoatomic ideal gas (Xe here).

Thank you very much for your interest, I am available if you have any more questions.

Guillaume

[pfajardo-uc3m]

Hello Guillaume,

Very interesting work.

I would like to ask you about the gradient reconstructions algorithms that you use in the fluid solver.

[gbogopolsky]

Hello Pablo,

Thank you for your interest in this work.

In the present case, we use a HLLC Riemann solver with MUSCL gradient reconstruction described in Berthon (2006), Alauzet & Loseille (2010) and Toro (2009).

I am available if you have further questions,

Thank you very much,

Guillaume

[eahedo]

Gillaume, Thank you. Best wishes! --------------------- Eduardo Ahedo Professor of Aerospace Engineering Universidad Carlos III de Madrid Avenida Universidad 30, Of. 7.3.H04 28911 Leganés, España +34 91 624 8234 El mar, 22 feb 2022 a las 20:13, Guillaume Bogopolsky (< ***@***.***>) escribió:

…

Hello Eduardo, Thank you very much for your interest in this work. As we solve the full Euler equations for each fluid (ions and electrons) and resolve the sheaths, the CFL condition imposes very small convective time step of the order of 2.5e-12 s. Therefore, I expect the same order of computational time on a 2D case compared with a hybrid code, given the simpler model usually used for electrons. Without resolving the sheaths, we have a factor 4 compared to PIC on a 3D HET sector. Walls are considered as metallic, so any plasma hitting a wall is considered lost (with a correction for neutralisation). Dielectric walls are a work in progress. In this computation, sheaths are resolved following the methodology introduced in Joncquières *et. al.* (2018) <https://doi.org/10.2514/6.2018-4905> by using a thermal flux condition on the mass and energy conservation equations at the walls. We use a gamma of 5/3 for a monoatomic ideal gas (Xe here). Thank you very much for your interest, I am available if you have any more questions. Guillaume — Reply to this email directly, view it on GitHub <#21 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AXZBHFWKFO6I4XZYYJMOOSLU4PN4HANCNFSM5OUYGM5A> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you commented.Message ID: <ep2lab/ExB_workshop_2022_website/repo-discussions/21/comments/2231218@ github.com>