real-gas EOS with 1-D flame simulations

[junfeng bai]

Hello everyone, 

I am working on real gas effects on transport properties and 1-D flame simulations. 

I got "TypeError: ThermoPhase object is not an IdealGasPhase" error using  R-K EoS and got ''NotImplementedError thrown by ThermoPhase::critTemperature:Not implemented'' error using high pressure transport model. 

Based on previous discussions on this topic in our group, I understand that current code base in 1-D flame simulations need to be modified to combine real gas EoS. 

To my understanding, the main changes should be 

(a) partial molar enthalpy in Redlich-Kwong model should be called to calculate h_k in energy equations.

(b) Critical temperature calculation in Redlich-Kwong model should be called to be used in high pressure transport model. 

(c) Onedim.py should be modified to allow real gas phase in ThermoPhase.

For (a), I think  partial molar enthalpy   should be easily incorporated in governing energy equations just like how activity concentration was  incorporated.

So I am curious whether there were any other difficulties in this modification since previous discussions on this topic were last for three years and  real gas EoS  still could not used in 1-D flame simulations in Cantera?

I am glad to contribute some effort on this part if it is not beyond my abilities. Could anyone explain my confusions and give me some advice? Thank you very much!

Sincerely,

Junfeng 

[Mohammed Niyasdeen]

Hi, 

Did you check with the latest Cantera 2.6.0 release? 

I haven't tried the latest version but would like to know if the 1D flame has the real-gas EOS in the latest version of Cantera. 

[Z. Weng]

Hello,

I am also interested in this topic. I have checked the source code before but not have a clear understanding on how the governing equations are solved. Any explaination on the numerical scheme or the source of the ideal gas based flame solver will be appreciated.

Best,

Zifeng

[Bryan Weber]

Hi all,

This feature is not available in Cantera yet. It is a quite complex problem to solve, there are several threads here on the Users' Group about the topic if you're interested in previous attempts.

In any case, the solution code for the 1-D solver is in src/oneD/Sim1D.cpp, specifically the solve method kicks off here: https://github.com/Cantera/cantera/blob/48457cc2ba49d75230103b0f113166f873969660/src/oneD/Sim1D.cpp#L307

Conceptually, a 1-D simulations is built from a series of Domain1D class instances that represent different types of flow (inlet, outlet, free flow, surface, etc.). The main flow domain is usually a StFlow instance (src/oneD/StFlow.cpp) which implements an eval method that actually does the calculations.

The numerical scheme is quite similar to the one used in Chemkin. It is a hybrid transient/steady-state Newton solver which is solved on a spatial grid, and it supports grid refinement based on the slope and curvature of the solution variables. Most of the numerical code is in the src/oneD/MultiJac.cpp and src/oneD/MultiNewton.cpp files.

A few years ago, an undergraduate student wrote up a series of posts partially about this hybrid solver, you can read them here: https://cantera.org/blog/gsoc-2020-intro

Hope that helps!
Bryan

[Bryan Weber]

Hi all,

To follow up, this is probably the most detailed thread on the group about real gas EOS in 1-D simulations: https://groups.google.com/g/cantera-users/c/96myuNgXN-o/m/2fLCCkGgBwAJ

Best,
Bryan

[Ingmar Schoegl]

In addition to user group threads, there also is a pull request on the GitHub site, see https://github.com/Cantera/cantera/pull/1079. I don’t know about the current status, as there hasn’t been any recent work other than trying to keep it up to date with the latest version of the source code.

-ingmar-

[宋业城]

Hi, I'm having trouble with the calculation of a one-dimensional flame for a real gas.I was very inspired after seeing your post, after trying, I still can't solve the problem of 'ThermoPhase object is not an IdealGasPhase', may I ask how you fix Onedim.py?

在2021年11月2日星期二 UTC+8 20:53:26<baijunf...@gmail.com> 写道：

[Bryan Weber]

Hi,

This is still not possible, please see the pull request that Ingmar linked to, which is currently undergoing review.

Best,

Bryan