Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

COMSOL AC-DBD Modelling

Please login with a confirmed email address before reporting spam

Hello everyone

I am trying to model a AC-DBD (Alternating Current -Dielectric Barrier Discharge) Plasma Actuator ,using COMSOL.

I am trying to implement the model of Suzen and Hwang(attached) ,whereby the developed model will calculate the body force after calculating the charge density and the electric field.

I am a novice to COMSOL hence i want to know how can i introduce a numerical model in COMSOL?

I have gone through the plasma module and it doesn't have an AC-DBD sub-module under the plasma module. Hence i wonder if i have to use AC-DC package to model the DBD as a capacitor? That seems easy but my question will then be, how do i integrate this with my plasma module in order to study the plasma physics and calculate the body force after i introduce my model (and that too ,How?) .

I will really appreciate the help.

God bless you and best regards
Zaid


1 Reply Last Post 2015年2月9日 GMT-5 19:15

Please login with a confirmed email address before reporting spam

Posted: 10 years ago 2015年2月9日 GMT-5 19:15
Syed,

I currently am in the process of activating our laboratories Plasma module license or I would offer more explicit help. DBD is not a mode of plasma in itself but a type. It transitions through either streamer or diffuse modes (Townsend to corona, spark and then to glow). This has been experimentally reported several years ago with gated CCD's. Your idea is correct. The DBD plasma will be an AC plasma and at the dielectric boundary it would be simple to code the potential delayed charging effect into a boundary force by integrating at that boundary the changing electric field times the maximum charge density.

As for this "body force," I have always confined my studies to pulsed DC driven plasma; however, it is easy to couple the electrohydrodynamic flow due to ionization interacting with rapidly oscillating electric field (~10kHz). Y. B. Suzen et al calculates the body force that your are interested in by "assume[ing] that the Debye thickness is small and the charge on the wall is not large, the distribution of charged species in the domain is governed by the potential caused by the electric charge on the wall and is largely unaffected by the external electric field."

Hence it is determined by the charge density times the changing electric field.
Without this assumption, you might have had to make a variable in your model which is calculated by calling := charge(#) set to equal plasma module abbreviated name. variable name for charge density. space variables for your model (assuming it is x,y) and then achieve the force by integrating the charge density ρ( −∇φ) at the boundary of interest. With the assumption (made by Y. B. Suzen) you should be able to solve the coupled model and output the maximum charge density and then post-process it with the output dE at the boundary (I use 2D cut line and the export function at several time solutions).

Remember, Y. B. Suzen et al states, "Experimental results16,18,19 suggest that this distribution is similar to a half Gaussian distribution given by..." They specify the time changing charge density by merely multiplying the changing electric field (have done it in dc and ac) times a Gaussian distribution and the maximum charge density....

If you do/or-do-not make headway, feel free to post your module. I will be working with dynamic dc pulsed models for micrometer fields of plasma over the next month and would be happy to help if I am able.

The attached file is a "very" simple model of the electrohydrodynamic flow implement in COMSOL which was at least a first step for me. Only takes a few clicks. I also looked at the COMSOL model gallery when first starting on the dc plasma modeling. There is limited help for the plasma module, but the third file (while coupling to reaction kinetics which you do not need) does show the use of two PHYSICS in a dielectric AC plasma (DBD) although at high frequency. These files are able to be downloaded and inspected.

Coupling the physics is not difficult. An example in the model COMSOL gallery for coupling turbulent flow which is solved in one component is coupled to a reaction module simply by inputting the velocity field solution from the first component (comp1.linext1(w)) into the second component (comp2) [the reacting flow module].

The output body force (derived from geometry, applied potential, ion density in plasma module) is intended to be coupled into Navier-Stokes (Y. B. Suzen et al-"A body force vector is calculated from the solutions of these two quantities and incorporated into Navier-Stokes equations to account for the plasma-actuator effects. The model is calibrated against a simple plasma-actuator-driven flow in a quiescent environment."). How that is input is shown in the first attachment although the calculation for the current density is experimentally input as J (mA/cm^2) rather than through a body force.

Examine the gallery link found here:
www.comsol.com/model/dielectric-barrier-discharge-8637


Best,
Justin Pommerenck
Ph.D. Candidate
Oregon State University
Syed, I currently am in the process of activating our laboratories Plasma module license or I would offer more explicit help. DBD is not a mode of plasma in itself but a type. It transitions through either streamer or diffuse modes (Townsend to corona, spark and then to glow). This has been experimentally reported several years ago with gated CCD's. Your idea is correct. The DBD plasma will be an AC plasma and at the dielectric boundary it would be simple to code the potential delayed charging effect into a boundary force by integrating at that boundary the changing electric field times the maximum charge density. As for this "body force," I have always confined my studies to pulsed DC driven plasma; however, it is easy to couple the electrohydrodynamic flow due to ionization interacting with rapidly oscillating electric field (~10kHz). Y. B. Suzen et al calculates the body force that your are interested in by "assume[ing] that the Debye thickness is small and the charge on the wall is not large, the distribution of charged species in the domain is governed by the potential caused by the electric charge on the wall and is largely unaffected by the external electric field." Hence it is determined by the charge density times the changing electric field. Without this assumption, you might have had to make a variable in your model which is calculated by calling := charge(#) set to equal plasma module abbreviated name. variable name for charge density. space variables for your model (assuming it is x,y) and then achieve the force by integrating the charge density ρ( −∇φ) at the boundary of interest. With the assumption (made by Y. B. Suzen) you should be able to solve the coupled model and output the maximum charge density and then post-process it with the output dE at the boundary (I use 2D cut line and the export function at several time solutions). Remember, Y. B. Suzen et al states, "Experimental results16,18,19 suggest that this distribution is similar to a half Gaussian distribution given by..." They specify the time changing charge density by merely multiplying the changing electric field (have done it in dc and ac) times a Gaussian distribution and the maximum charge density.... If you do/or-do-not make headway, feel free to post your module. I will be working with dynamic dc pulsed models for micrometer fields of plasma over the next month and would be happy to help if I am able. The attached file is a "very" simple model of the electrohydrodynamic flow implement in COMSOL which was at least a first step for me. Only takes a few clicks. I also looked at the COMSOL model gallery when first starting on the dc plasma modeling. There is limited help for the plasma module, but the third file (while coupling to reaction kinetics which you do not need) does show the use of two PHYSICS in a dielectric AC plasma (DBD) although at high frequency. These files are able to be downloaded and inspected. Coupling the physics is not difficult. An example in the model COMSOL gallery for coupling turbulent flow which is solved in one component is coupled to a reaction module simply by inputting the velocity field solution from the first component (comp1.linext1(w)) into the second component (comp2) [the reacting flow module]. The output body force (derived from geometry, applied potential, ion density in plasma module) is intended to be coupled into Navier-Stokes (Y. B. Suzen et al-"A body force vector is calculated from the solutions of these two quantities and incorporated into Navier-Stokes equations to account for the plasma-actuator effects. The model is calibrated against a simple plasma-actuator-driven flow in a quiescent environment."). How that is input is shown in the first attachment although the calculation for the current density is experimentally input as J (mA/cm^2) rather than through a body force. Examine the gallery link found here: http://www.comsol.com/model/dielectric-barrier-discharge-8637 Best, Justin Pommerenck Ph.D. Candidate Oregon State University

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.