CFD Online Discussion Forums

CFD Online Discussion Forums (
-   FLUENT (
-   -   About prePDF use ???help me please (

harry q December 7, 2000 22:22

About prePDF use ???help me please
Mr. or Mrs:

I want to simulate natural gas diffusion-type-combustion in a furnace(include mixing, turbulent flow and combustion), must I use the prePDF preprocessor?

Thank u very much for ur helping!

Jin-Wook LEE December 11, 2000 23:05

Re: About prePDF use ???help me please
You can use various combustion model by your choice.

1. finite rate 2. eddy dissipation 3. finite reate / eddy dissipation 4. PDF

You should use prePDF only for PDF combustion model.

Sincerely, Jinwook

harry q December 12, 2000 03:50

Re: About prePDF use ???help me please
Hi I'm harry q.

First of all, thank u for ur answer(combustion models) on cfd-online. the answer is very expert and helps me so much.

I am to use fluent5 to simulate "highly rotate flow and turbulent diffusion combustion" gave me four choices--finite rate,eddy dissipation model,finate rate and dissipation model,PDF model.Coud you tell me which is the best for "highly rotate flow and turbulent diffusion combustion"?

In tutorial guideline(8th), the eddy dissipation model is used. Recently I looked for eddy dissipation model in many books, but find nothing. Can you tell me a book or material involving eddy dissipation model?

I think you are a specialist in CFD area, because your answer is professional version.

Thank you very much!!

Sincerely yours

Harry q

Jin-Wook LEE December 12, 2000 05:13

Re: About prePDF use ???help me please
Dear Harry Q.

I do not understand the meaning of highly ROTATING flow with combustion. I think that you mean highly SWIRLING flow which is usually acompanied with combustion for flame stabilization. If so, it is not dependent on combustion model. It is highly dependent on turbulence model. You might use RSM(Reynolds Stress Model) for highly swirling flow such as cyclone combustor. You may use standard k-epsilon or RNG k-epsilon model for weakly swirling flow, or highly swirling flow but the effect of swirl is confined near the burner region such as small burner compared to large furnace. Computational cost of RSM is very expensive(large CPU time and difficult to converge) and k-epsilon is the cheapest one but not accurate for certain case, especially highly swirling flow.

The selection of combustion model is entirely up to you. But I can give you some basic idea.

Combustion rate is determined by the slower one of 1) turbulent mixing rate and 2) chemical kinetics rate. For high flame temperature, turbulent mixing rate is slow, so that eddy dissipation model based on turbulent mixing rate is good. For low temperature flame combustion, chemical kinetics rate is slow so that finite rate model based on chemical kinetics rate is good. When you can not judge which one is appropriate, you can use 'finite rate / eddy dissipation' model which determines combustion rate as the slower one of two rates. Actually, this is the best approach. How much high or low temperature ? I myself posted the question to CFD-ONLINE but noone answered.

For my opinion(which is entirely based on my experience, not based on the good theory), most combustion system is high enough, e.g., higher than about 1500 Kelvin of flame temperature). For fludized bed combustion, the flame temperature is about 1100-1200 Kelvin, so that many chemical engineers are using chemical kinetics rate(finite rate of Fluent code).

For your case, as a first step, I recommend eddy dissipation model(which is computationally the cheapest one) because in general the flame temperature of gaseous combustion is fairly high. And, as a personal opinion, it is enough for most case.

If you have many species(e.g., greater than 10 species), your best choice is PDF combustion model. But PDF model always overpredict carbon monoxide(CO) because of its limitation of fast chemistry

harry q December 12, 2000 08:02

Re: About prePDF use ???help me please
Dear Jin-Wook LEE:

Thank u for ur answer. Anyway ur English is better than mine. In my mind, you must be a very kind Gentleman. U answered me with such a long email.Can we make friends?

I learnt so much from ur answer.Thank u very much again.

But u know, it is not enough for the report to show how to use fluent, and I have to tell others why I use it(select certain model) like that. when I read undergraduate, my major is HVAC(heating ventilation and air conditioning), So the combustion model is not familar to me. Recently I search the EDDY DISSIPATION model in any books about combustion simulation, but find nothing. My question is "Is EDDY DISSIPATION model popular in CFD area or not? "and "From where can I learn it?"

Tank u sincerely yours, Harry Qiu

Jin-Wook LEE December 13, 2000 06:07

Re: About prePDF use ???help me please
Dear Harry

How can I say that I am a specialisit ? But, I am happy that my answer was helpful to you.

1. EDDY DISSIPATION model is very classical and popular one in CFD area.

2. Because its concept was published long years ago, I can not have the original paper and I can not remember where it is. However, you can find many useful information from following paper.

B.F. Magnussen & B.H. Hjertager, 'On mathematical models of turbulent combustion ........', 16th Symp. (Int'l) on Combustion, 1976

Actually, Fluent's eddy dissipation model is just their's.

3. We can be very good friends, if you want. Just call me Jinwook.

Sincerely, Jinwook

harry q December 13, 2000 09:36

Re: About prePDF use ???help me please

I know u might be from South Korean. I'm very glad to have a oversea friend, let alone we work in the simiar area. Since we are friends now, I think we can talk anything rather than the subject limited in the CFD area. Do u like ur working-area(CFD) in ur deep mind? U know, I don't like it, but I have no other choices.


Jin-Wook LEE December 14, 2000 05:29

Re: About prePDF use ???help me please
Hi, Harry

It seems to be very personal affair. Even though my CFD capability is not so high compared to the many specialists in this forum, I am a CFD consultant in my country. I should like CFD because, except CFD, I can not have any job in which I would be satisifed.

Sometimes, I think that 'which area would be more attractive than CFD in this world ? Because, as you know, there is fluid related phenomenon everywhere in the our life including engineering. And I can explain most phenomenon briefly(roughly) to others. So, it is happy, without considering some problem in the real life.

For technical discussion, we can post our opinion on this forum, and for personal problem like this, I think that e-mail is better.

Sincerely, Jinwook

govinda April 22, 2010 16:58

pdf error
hello ,
this is govi from india
i m solving a non primixed combustion with k epslion standard model with diesel as a fuel . after initialization i m getting a error Pdf_All: Unable to get properties
please help me



All times are GMT -4. The time now is 13:39.