Menu
▾
▴
Re: [sparta-users] [EXTERNAL] Sparta Clarifications
|
From: Gallis, M. A <ma...@sa...> - 2018-08-14 05:26:55
|
Lorenzo - Emit/Surf: I am still somewhat confused on how to properly emit particles from a surface then, should I use create particles instead? I'd personally be inclined to specify Np, but from the way it is worded in the documentation it is unclear to me if I can just straight-up use the total number of particles I am injected divided by fnum multiplied by the time step. (apologies if this isn't clear enough) You can create a second mixture with the desired number density (ns) for the surface emission. Based on this ns (particles/m3) you can calculate the flux (Eq 4.22 in Bird’s 1994 monograph). Create_particles will create particles based on the atmospheric density at the altitude you as simulating flight. - Would an idea like this idea hypothetically work? I.e. would a modification to the react*.cpp group be self consistent if implemented correctly, or there are other files I need to be wary of? It could work but you have to be careful when you are using cross sections. The DSMC algorithm needs probabilities. To turn a cross section to a probability you need to divide by the total cross section. Most likely you are thinking of using the total cross section of the VHS/VSS model. Is that compatible. Is there a chance you may end up with physically unrealistic probabilities (P > 1). - In react_tce, is pre_etrans the total (relative to each other) energy of the particles colliding? Correct - What does react_bird feed to react_tce (or vice-versa) React_bird inclide tce and the QK models and allows a hybrid of the two. - How come in react_tce, ionization and dissociation are blank? They are not, I am not sure what you mean. Thanks again, Best regards, Lorenzo -- Lorenzo Limonta Ph.D. candidate Aeronautics & Astronautics Stanford University https://sess.stanford.edu/lorenzo http://www.sjsu.edu/ae/faculty_staff/lorenzo_limonta/ https://italiansociety.stanford.edu/ ________________________________ From: Gallis, Michail A <ma...@sa...> Sent: Monday, August 6, 2018 13:03 To: Lorenzo Limonta; spa...@li... Subject: Re: [EXTERNAL] [sparta-users] Sparta Clarifications Lorenzo 1st TOPIC: Emit/surf The mixture nrho has to do with the atmospheric density at the altitude you are looking at. It has nothing to do with the rate the surface is emitting. 2nd TOPIC: Chemistry We also have reactions like O + O -> O + O+ + e- and O + e- -> O+ + e- These are very common reactions in atmospheric chemistry. Which are not recombination reactions. Is there another kind of reaction you need ? 3rd TOPIC: Ambipolar approximation A fully kinetic (free e-) would require timesteps that do not allow e- to move more than a cell per move. That would have limited to the timestep to ps at best. Using the ambipolar approximation, you can have timesteps that match the heavy particle motion at the expense of loosing charge separation. DSNC has similar to CFD restrictions: the mean transit time. Effectively you should not allow a molecule to travel distances greater that a cell width per time step. Otherwise you artificially increase transport. (high viscosity, conductivity etc). From: Lorenzo Limonta <lli...@st...> Date: Friday, August 3, 2018 at 4:10 PM To: "spa...@li..." <spa...@li...> Subject: [EXTERNAL] [sparta-users] Sparta Clarifications Good morning, I am Lorenzo Limonta, a Ph.D. working on meteor ablation. I have been using SPARTA for the past few weeks and have been thoroughly impressed with the software and the documentation. Nonetheless, I have a few questions regarding the software functionalities as I am trying to integrate it with the ablation code I have been developing as part of my studies. 1st TOPIC: Emit/surf I have a conundrum on how to correctly specify this command for the case I am looking at. The meteor is continuously emitting 8.9e21 molecules/sec from a surface area of pi*10^(-4), would the correct mixture nrho be simply 8.91e21/(Volume of meteoroid)? i.e. (abbreviated to relevant commands) global fnum 6.089e16 weight cell radius species air.species N2 O2 etc.. mixture iron Fe nrho 8.e21/2.7e-7 vstream 0 0 0 temp 2860 group patch surf id 1:25 fix emit/surf iron patch normal yes perspecies yes 2nd TOPIC: Chemistry Looking at the documentation, under the ambipolar approximation, ionization reactions between neutrals appear to be restricted only to recombination cases (i.e. R1 + R2 -> R1R2+ + e). Could you point me to a resource that explain why is that the case? To make sure I don't make any mistakes when adding chemical reaction, the units of the coefficients used for the Arrhenius equation in the companion file are A: m^3/molecules, whereas Ea is equivalently expressed in J/molecules, correct? Do you recommend any repository for ionization reaction coefficients (beside obviously looking at primary sources)? 3rd TOPIC: Ambipolar approximation In the documentation it is claimed that the ambipolar approximation is "a computationally efficient way to model low-density plasmas" meanwhile in the ambipolar example, you use a time step of 1e-7 while asserting that is too large of a time step for a 42.5 km/s flow. What would you recommend to properly capture similarly fast flows? Is there an equivalent condition for DSMC simulations to the CFL condition in traditional CFD simulations? 4th TOPIC: Write out adapting grid Currently, is it possible to write out the grid every N steps similarly to what is done for the dump file? From the dump image I would assume so, but the documentation for the write_grid command doesn't mention it. Thanks for all the help, Best regards, Lorenzo -- Lorenzo Limonta Ph.D. candidate Aeronautics & Astronautics Stanford University https://sess.stanford.edu/lorenzo http://www.sjsu.edu/ae/faculty_staff/lorenzo_limonta/ https://italiansociety.stanford.edu/ |
