CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM Programming & Development (https://www.cfd-online.com/Forums/openfoam-programming-development/)
-   -   New solver: ddtFoam (https://www.cfd-online.com/Forums/openfoam-programming-development/122668-new-solver-ddtfoam.html)

dohnie August 25, 2013 07:30

New solver: ddtFoam
 
2 Attachment(s)
Dear Foamers,
I developed an OpenFOAM solver for the simulation of the deflagration-to-detonation transition in gases.
Some features:
  • accurate shock capturing with a density based solver (largely based on Oliver Borm's development densityBasedTurbo)
  • deflagration modelling via a reaction progress variable (based on XiFoam)
  • auto-ignition sub-grid model separating each computational cell into a "shocked" and an "unshocked" part
  • auto-ignition from tabulated ignition delay times, gained from a detailed chemical mechanism
  • low computational cost due to applicability to coarse grids and usage of pre-tabulated chemistry
Even if you're not interested in detonations/combustion, it might be a nice alternative to existing OpenFOAM solvers when it comes to shock resolution in transonic flows.

The code incl. a short documentation and tutorials can be downloaded from
http://sourceforge.net/projects/ddtfoam/

ChrisA August 27, 2013 13:56

Great work, I haven't had much time to look at it but from first glances the solver looks pretty good. The guys at my lab working on deflagration to detonation will certainly find it interesting. Is there going to be a paper to accompany the solver? I'd be very interested in reading that and seeing some results from the solver.

dohnie August 27, 2013 14:11

Chris,
I have a paper in preparation, but at the moment I can only refer to my thesis (in German, unfortunately):
Effiziente numerische Simulation des Deflagrations-Detonations-Übergangs.pdf
[8.6 MB]

ChrisA August 27, 2013 14:14

Ah, yes, I found the thesis shortly after posting that, unfortunately my German is terrible. However, I do look forward to the other paper you mentioned :)

uwknight October 30, 2013 06:46

Hi dohnie,

Thanks for your sharing this good solver. I am working on this ddt solvers.

Do you HAVE some ideads about how to change the reaction model? Can I use this code to simulate the problem about oblique detonation wave? Could you give me some comments? You comments should be very useful to me. Thanks a lot.

Btw, shall I have your email? You can reply me using priviate message. Thanks.

Best regards,

Yue

dohnie November 11, 2013 15:38

Yue,
sorry for the late reply, I don't work on this topic any longer.
You can use this solver to simulate oblique detonations as well, and of course you can also modify the reaction model.
From your short description I cannot see what kind of problem you really want to simulate, but it seems to me that you are not interested in simulating the deflagration-to-detonation transition. Thus, if you want to investigate only detonations (probably even looking for steady solutions?), you might be better off using simpler solvers. For example you could use a compressible solver (like rhoCentralFoam) and add 1-step Arrhenius kinetics - the publications of Elaine Oran et al. might give you some orientation.

I hope that helps!
Florian

uwknight November 21, 2013 04:39

[QUOTE=dohnie;461587]Yue,
sorry for the late reply, I don't work on this topic any longer.
You can use this solver to simulate oblique detonations as well, and of course you can also modify the reaction model.
From your short description I cannot see what kind of problem you really want to simulate, but it seems to me that you are not interested in simulating the deflagration-to-detonation transition. Thus, if you want to investigate only detonations (probably even looking for steady solutions?), you might be better off using simpler solvers. For example you could use a compressible solver (like rhoCentralFoam) and add 1-step Arrhenius kinetics - the publications of Elaine Oran et al. might give you some orientation.

I hope that helps!
Florian[/QUOTE

Hi Florian,

Thanks for your reply. I would do some works on your mentioned.

Yue

lengshui222 November 28, 2013 03:47

[QUOTE=uwknight;462854]
Quote:

Originally Posted by dohnie (Post 461587)
Yue,
sorry for the late reply, I don't work on this topic any longer.
You can use this solver to simulate oblique detonations as well, and of course you can also modify the reaction model.
From your short description I cannot see what kind of problem you really want to simulate, but it seems to me that you are not interested in simulating the deflagration-to-detonation transition. Thus, if you want to investigate only detonations (probably even looking for steady solutions?), you might be better off using simpler solvers. For example you could use a compressible solver (like rhoCentralFoam) and add 1-step Arrhenius kinetics - the publications of Elaine Oran et al. might give you some orientation.

I hope that helps!
Florian[/QUOTE

Hi Florian,

Thanks for your reply. I would do some works on your mentioned.

Yue

Hi Yue,

I want to work on this ddt solvers, but when I am installing it, it shows some mistake like "expected unqualified-id before '.' token". Have you met the similar mistake when you were installing the ddt solver? Can you show me the detailed steps of installing this solver, if you remember the installing steps? Thank you very much.

Best regards,

Zhong

cfdsat February 4, 2014 07:00

Hi Dohnie
 
You have used two different solver one based on densityBasedTurbo solver package and other based on sonicFoam. What is your experience regarding which one is best for solving compressible and DDT and detonation problems.

I have got your PhD thesis, since it is written in German, I am not able to understand it. Could you please provide me a link to your papers if published ?

dohnie February 11, 2014 16:18

The density based solver is definitely better for simulating DDT and detonation.
I use the pressure based solver only to get an initial solution for unsteady problems where I start in quiescent or very low-speed (Mach number < 0.1) flow, because the density based solver tends to get unstable at very low Mach numbers.

I have prepared a journal paper which is currently in the review process. I'll let you know once it has been accepted.

Regards,
Florian

cfdsat February 12, 2014 06:03

Hi Dohnie

Thanks for the comments. Which openfoam version did you run this solver ?

Can you already send me the pdf file of the publication you have submitted ?

dohnie June 2, 2014 05:56

Paper published
 
Dear all,
a paper describing the features of ddtFoam has now been published:

http://www.hindawi.com/journals/jc/2014/686347/

Florian Ettner, Klaus G. Vollmer, and Thomas Sattelmayer, “Numerical Simulation of the Deflagration-to-Detonation Transition in Inhomogeneous Mixtures,” Journal of Combustion, vol. 2014, Article ID 686347, 14 pages, 2014. doi:10.1155/2014/686347

anh065 June 5, 2014 10:08

Great. Thanks, Florian, for sharing this. I would want to apply this code to problems with non-confined domains and wonder what types of boundary condition should be applied to the open (non-reflective) boundaries?

dohnie June 7, 2014 15:42

Dear Anh,
you should get the best result by using the waveTransmissive BC.
However, this will require some extra programming work. The standard wave transmissive BC in OpenFOAM is based on the assumption of a constant ratio of specific heats (gamma) and a flux field which is called phi in most OpenFOAM solvers. In ddtFoam, where the fluxes are computed using a Riemann solver, this flux corresponds to Riemann.rhoUFlux() divided by Riemann.rhoFlux().
I hope that helps!
Florian

panda1100 June 11, 2014 03:21

Dear dohnie,

Thank you for sharing this great work:)

Is there any information about how to change H2+air to Diesel mixture?
And If its possible, Can you share your utility codes?
for example,
  • loolupIgn_06.m
  • automaticSetFields_tfc_zu_Grad3s_03.c etc
I want to try this solver with another type of gas.
(I tried your tutorial and that works pretty good.)

Any information you can provide me would be greatly appreciated.
I wish I could read German...
panda1100

dohnie June 18, 2014 04:05

Dear Yoshiaki,
in principle you just have to replace the tabulated ignition delay times in the respective case. However, you should be aware that the heat release of long-chained hydrocarbons is not as instantaneous as the heat release of hydrogen.
See, for example:
O. Colin, A. Pires da Cruz and S. Jay: "Detailed Chemistry-Based Auto-Ignition Model Including Low Temperature Phenomena Applied to 3-D Engine Calculations", Proceedings of the Combustion Institute, Vol.30, pp.2649-2656 (2005)

Send me your e-mail address in a PM and I can send you the files you requested.

Regards,
Florian

panda1100 June 19, 2014 00:12

Dear Florian,

Thank you so much for your advice and your help.
I sent you PM with my e-mail address .

Thank you again for sparing your time for me.

Best regards,
Yoshiaki

blais.bruno June 19, 2014 08:41

Great!
 
Looks both very good and very interesting!

Thanks for posting it online this way... :)

dowlee September 1, 2014 11:26

Dear Florian Ettner
Thanks for your efforts on developing solver DDTFoam. I am very interested to this solver. And at present, I am working on a project concerning detonation combustion. In my case, the combustible mixture close to the end wall of the resonators is heated up to a temperature that creates ignition automatically. And after a short while, the flame front reaches from behind the reflected shock wave and combines with it forming a plannar detonation wave. My question is whether this solver is suitable for automatic ignition and detonation. If not, what should be modified in this solver. I am looking forward to your reply.
Thanks very much
Dowlee

dohnie September 8, 2014 12:57

Dear Dowlee,
I don't understand from your description HOW the mixture close to the end wall is heated up, but I assume it is done by a shock, as you mention a reflected shock wave later on.
I think the solver should work fine for your problem!


All times are GMT -4. The time now is 19:39.