Add a derivative of the mixture fraction or progress variable over time

[czc-zju]

How to add a derivative of the mixture fraction or progress variable over time

[esclapez]

I see at least two ways to get those: either combining the time derivatives of mass fractions, already available in the advance function or computing mixture fraction and progress variable on both the old and new states, and getting the time derivative, in the WritePlotFile function for instance.

[czc-zju]

@esclapez My purpose is to get the flame displacement velocity, I need to output the Fm species diffusion fluxes or the divergence of the Fm , the calculation equation of Fm is as follows, but in PeleLMeX, how should I output Fm or Vm

[drummerdoc]

In “advance” you want to compute (Dn + Dnp1)/2 on the last SDC iteration.

[czc-zju]

@esclapez @drummerdoc I've been learning PeleLMeX software for three months now and I'm still trying to learn it. Your contribution is really great and helped me a lot, but I still don't know how to get these parameters ( (Dn + Dnp1)/2 ) directly. I sincerely hope that you can help me, or you can share some tutorials that can be referenced, thank you very much

[esclapez]

@czc-zju getting the displacement speed, and I'm guessing in the future its different components (S_r, S_{d,t}, S_{d,n},...) is a relatively heavy post-processing which could be part of one of PeleAnalysis set of tools (evaluating the terms post-mortem, less consistent with the actual advance of PeleLMeX, but easier) or can be added as an evaluate derived, in which case it can be very close to the actual terms used in PeleLMeX advance.
How familiar are you with the code itself at this point ?

[czc-zju]

Thank you for your reply!!!

It can be seen that sd =(diffusion term +the source term)/(density and modulus length), according to diffusion coefficients of fuel and reaction rate of fuel, I have calculated it in yt, but the result of the diffusion term of fuel is not accurate. So I want to output the diffusion term of fuel directly from PeleLMeX, it should be D(ϕ) in the following formula: [[ϕt=F≡A(ϕ)+D(ϕ)+R(ϕ)]].(https://pelelm.readthedocs.io/en/latest/Model.html#misdc-correction-equations)
I've been studying your commit log these days, it's very helpful to me, but I haven't been able to modify the source code accurately, I'm very sorry.
Thank you again

[drummerdoc]

Yeah, it seems that what you want to compute here really is at the core of what the code computes to advance the solution. There are lots of intermediate data structures in the code that you can take advantage of, but it really requires a thorough understanding, not only of the algorithm, but of how to manipulate and reward/write AMReX data structures. We could post a bunch of code of suggestions here, but it might be more productive to set up a meeting to discuss some of the nuances required. In the best of cases, we can figure something out and not only makes it available to all, but improve the documentation for all as well. Please feel free to reach out to me at Marc.Day@nrel.gov to see if we can schedule some time at a whiteboard to talk some of this through.