
[Sponsors] 
August 7, 2013, 10:58 

#101  
Member
Join Date: May 2013
Location: Netherlands
Posts: 30
Rep Power: 4 
Quote:
Thank you for the answer. Some people mention that is is also possible to modify the solver of later openfoam versions to run steady state chemistry. For me this is the preferred solution, since we are already running two different OpenFOAM versions (1.6 and 2.1.1), so we like to avoid using a third one. Can anyone tell me which steps are required to change reactingFoam into steadyreactingFoam (or even better, does anyone already have this solver). I already had a quite extensive look on different fora, but I didn't find a clear answer yet ... 

June 6, 2014, 06:53 
reactingFoam to steadyState

#102 
Member
James
Join Date: Jul 2013
Posts: 38
Rep Power: 4 
Hi,
Did you ever get an answer to your question: "Can anyone tell me which steps are required to change reactingFoam into steadyreactingFoam (or even better, does anyone already have this solver)?. " I am also looking to run reactingFOAM in steady state. I am concerned at the moment with a high pressure release, I have switched off combustion and chemistry. What else do I need to modify to run in steady state. Many thanks, James 

June 6, 2014, 10:50 

#103 
Member
Join Date: May 2013
Location: Netherlands
Posts: 30
Rep Power: 4 
Dear James,
I never got a reply to this question, I did some tries with changing to pseudosteady state solvers by changing the ddt schemes, however I was not succesfull with that (see also steady state combustion posts). Later I moved to the FGM based solver libOpenSmoke. This one works very well as steady state combustion solver. Regards 

March 18, 2015, 07:10 

#104 
Member
Join Date: Feb 2015
Posts: 30
Rep Power: 3 
Hi TBO,
I am working on the FGM combustion modeling in OpenFOAM and I find the only one already available by Kroger. I am testing it and I having some problems with the flamelet generator based on Cantera. You say that you moved to the FGM model based on the libOpenSmoke and I also had this idea, but is there an FGM model already available with libOpenSmoke (because I see that it can be used only for non premixed combustion with only the equations for mixture fraction) or did you build on your own? 

March 18, 2015, 10:11 

#105 
Member
Join Date: May 2013
Location: Netherlands
Posts: 30
Rep Power: 4 
Hi Stefano,
I used this model for non premixed combustion, so the libOpenSmoke based solver was suitable for my simulations. Maybe repost the question on the libOpenSmoke forum to see if someone else has experience with using this solver for premixed combustion. Regards, TBO 

March 18, 2015, 10:36 

#106 
Member
Join Date: Feb 2015
Posts: 30
Rep Power: 3 
Ok,
maybe I will ask on the libOpenSmoke forum. Thank you, Stefano 

April 15, 2015, 11:51 

#107 
New Member
Ali Kadar
Join Date: Oct 2014
Location: Delft
Posts: 25
Rep Power: 3 
Hello everyone,
has there been any new development on alternateSteadyReactingFoam which is now compatible only with old versions of OpenFOAM. The transient solver reactingFoam is good but is not very appropriate considering the computational costs. Is there any new development for a steady state combustion solver in OpenFOAM. ? These are the related thread modify reactingFoam in order to make it a steady state solver Steady state chemistry https://github.com/UnofficialExtend...rnateChemistry
__________________
A good solution is one which does justice to the inner nature of the problem Cornelius Lanczos in a letter to Albert Einstein on March 9, 1947 

April 16, 2015, 02:29 

#108  
Member
Join Date: May 2013
Location: Netherlands
Posts: 30
Rep Power: 4 
Quote:
Yes, there are some developments. By default OpenFoam has now the solver LTSReactingFoam, which is a steady state version of reactingFoam (available in OF 2.3). Besides that, also e.g. the solvers based on libOpenSmoke might be worth considering Regards, TBO 

April 17, 2015, 02:47 

#109 
New Member
Ali Kadar
Join Date: Oct 2014
Location: Delft
Posts: 25
Rep Power: 3 
Thank you for pointing that. I did test runs using the LTSreactingFoam, but unfortunately it fails(after 27 time iterations) with the following error(attached).
It runs perfectly ok with reactingFoam and I get good results. I could not understand the parameters in the PIMPLE Algorithm too. What are alphaTemp, rDeltaTSmoothingCoeff, rDeltaTDampingCoeff. and why there is no relaxation ? I think I will have to look at the code to understand what they are doing and what is the idea of local time stepping(LTS). I know that if you want to go fast in time you have to solve more pressurevelocity couplings per iteration i.e. nOuterCorrectors > 1 with underrelaxation(idea behind SIMPLE). But they are not doing that!!. Whats the idea behind LTS ? Any ideas would be really helpful !! Code:
PIMPLE { momentumPredictor no; nOuterCorrectors 1; nCorrectors 1; nNonOrthogonalCorrectors 0; maxDeltaT 1e2; maxCo 1; alphaTemp 0.05; rDeltaTSmoothingCoeff 1; rDeltaTDampingCoeff 1; } relaxationFactors { fields { } equations { ".*" 1; } } Code:
Time = 26 Time scales min/max: Flow = 1.28808e05, 0.01 Temperature = 6.78274e06, 1e+300 Overall = 6.78274e06, 0.01 diagonal: Solving for rho, Initial residual = 0, Final residual = 0, No Iterations 0 DILUPBiCG: Solving for O2, Initial residual = 0.075075, Final residual = 0.00549963, No Iterations 4 DILUPBiCG: Solving for H2O, Initial residual = 0.963064, Final residual = 0.075786, No Iterations 3 DILUPBiCG: Solving for CH4, Initial residual = 0.978202, Final residual = 0.0737773, No Iterations 3 DILUPBiCG: Solving for CO2, Initial residual = 0.962777, Final residual = 0.0756163, No Iterations 3 DILUPBiCG: Solving for h, Initial residual = 0.996642, Final residual = 0.0833314, No Iterations 2 min/max(T) = 297.502, 2074.17 DICPCG: Solving for p, Initial residual = 0.846321, Final residual = 9.13219e07, No Iterations 226 diagonal: Solving for rho, Initial residual = 0, Final residual = 0, No Iterations 0 time step continuity errors : sum local = 1.14244e05, global = 5.85091e08, cumulative = 1.33362e08 DILUPBiCG: Solving for epsilon, Initial residual = 0.000435871, Final residual = 2.96746e07, No Iterations 1 bounding epsilon, min: 45699 max: 4.85026e+08 average: 332153 DILUPBiCG: Solving for k, Initial residual = 8.21421e09, Final residual = 8.21421e09, No Iterations 0 ExecutionTime = 32.79 s ClockTime = 34 s cellSource volumeTemperature output: average(sampledSurface) for T = 1285.89 faceSource faceTemperature output: average(outlet) for T = 1113.5 Time = 27 Time scales min/max: Flow = 6.88187e06, 0.01 Temperature = 6.73481e+07, 1e+300 Overall = 6.88187e06, 0.01 diagonal: Solving for rho, Initial residual = 0, Final residual = 0, No Iterations 0 #0 Foam::error::printStack(Foam::Ostream&) in "/opt/openfoam230/platforms/linux64GccDPOpt/lib/libOpenFOAM.so" #1 Foam::sigFpe::sigHandler(int) in "/opt/openfoam230/platforms/linux64GccDPOpt/lib/libOpenFOAM.so" #2 in "/lib/x86_64linuxgnu/libc.so.6" #3 Foam::EulerImplicit<Foam::chemistryModel<Foam::psiChemistryModel, Foam::sutherlandTransport<Foam::species::thermo<Foam::janafThermo<Foam::perfectGas<Foam::specie> >, Foam::sensibleEnthalpy> > > >::solve(Foam::Field<double>&, double&, double&, double&, double&) const in "/opt/openfoam230/platforms/linux64GccDPOpt/lib/libchemistryModel.so" #4 double Foam::chemistryModel<Foam::psiChemistryModel, Foam::sutherlandTransport<Foam::species::thermo<Foam::janafThermo<Foam::perfectGas<Foam::specie> >, Foam::sensibleEnthalpy> > >::solve<Foam::Field<double> >(Foam::Field<double> const&) in "/opt/openfoam230/platforms/linux64GccDPOpt/lib/libchemistryModel.so" #5 Foam::combustionModels::laminar<Foam::combustionModels::psiChemistryCombustion>::correct() in "/opt/openfoam230/platforms/linux64GccDPOpt/lib/libcombustionModels.so" #6 Foam::combustionModels::PaSR<Foam::combustionModels::psiChemistryCombustion>::correct() in "/opt/openfoam230/platforms/linux64GccDPOpt/lib/libcombustionModels.so" #7 at /home/opencfd/OpenFOAM/OpenFOAM2.3.0/src/OpenFOAM/lnInclude/autoPtrI.H:174 #8 __libc_start_main in "/lib/x86_64linuxgnu/libc.so.6" #9 in "/opt/openfoam230/platforms/linux64GccDPOpt/bin/LTSReactingFoam" Floating point exception (core dumped)
__________________
A good solution is one which does justice to the inner nature of the problem Cornelius Lanczos in a letter to Albert Einstein on March 9, 1947 

September 20, 2015, 15:13 

#110  
Super Moderator
Bruno Santos
Join Date: Mar 2009
Location: Lisbon, Portugal
Posts: 9,507
Blog Entries: 36
Rep Power: 96 
Greetings to all!
@Ali: Quote:
As far as I can understand (I am not an expert on this), the LTS modelling strategy uses an adaptive time step depending on the Courant Number in each cell (or small region of cells) and then uses a smoothing/damping strategy for making the flow compatible between regions that have different time steps. That's why the parameters "rDeltaTSmoothingCoeff, rDeltaTDampingCoeff" are needed, so that it can sortof have a fastforward feature for the regions that require a small local time step, when compared with other regions. This strategy is sometimes preferable to using explicit steadystate modelling, because this method still preserves some time accuracy, which usually is not preserved in steadystate modelling... and time accuracy is important for certain simulations, such as chemical reactions, heat transfer and multiphase flows. Best regards, Bruno
__________________
I'll be at OFW11 in Portugal 

Thread Tools  
Display Modes  

