CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CD-adapco (http://www.cfd-online.com/Forums/cd-adapco/)
-   -   Urgent help needed with spray injection (http://www.cfd-online.com/Forums/cd-adapco/56691-urgent-help-needed-spray-injection.html)

nick June 27, 2008 20:53

Urgent help needed with spray injection
 
Hi, i am using a lagrangian multi-phase injection of dedecane into a chamber. i have set up the model using the explicitly defined parcel injection, and set up the velocity, flowrate, injection timing, parcel/s etc and the rotation axis about the z axis in cylindrical coordinates. I double checked these inputs with someone else and they seem correct. i have defined the injection points at the top of the chamber, on a symmetrical plane (its a half a chamber model) with a vector pointing in the -z direction.

when i run the simulation, my concentration of C12H26 stays at zero, and the O2 and N2 values for air also stay constant.

i was wondering if any of you would be able to help me out, its quite urgent that i get this working soon!! regards nick

usker June 28, 2008 07:35

Re: Urgent help needed with spray injection
 
Did you set C12H26 evaporates to scalar? You have to set O2 and N2 fractions in additional scalars.

nick June 29, 2008 01:22

Re: Urgent help needed with spray injection
 
I do not have the dodecane(L) (from the database) to evaporate to scalar.

and do you mean setting the initialisation O2 and N2 mass fractions? are they not set when i select air as the user defined material?

nick June 29, 2008 01:27

Re: Urgent help needed with spray injection
 
i have tried setting the initialization mass fractions to 0.233 and 0.767, but it still has no effect on the C12H26 concentration. its as if the injection process is not even starting

usker June 29, 2008 05:40

Re: Urgent help needed with spray injection
 
You have to select dodecane vapor as a scalar, and evaporate your dodecane liquid to that scalar. Just go to additional scalars and set dodecane as scalar 1, then assign specific heat from chemkin database, it will be dodecane vapor. Then set scalars N2 and O2. In initialization set the mass fractions for N2 and O2. In spray model select dodecale (liquid) and evaporate to scalar 1.

nick June 29, 2008 07:04

Re: Urgent help needed with spray injection
 
I am unable to set the dodecane to evaporate, the option to do so is unselectable. any idea why this may be?

nick June 29, 2008 07:21

Re: Urgent help needed with spray injection
 
ok, can i just say, you are awesome!! i just had to turn on mass transfer and i now have ignition, modelling soot and all. thank you very much for your help!! you shall be thanked in my thesis!

nick June 29, 2008 08:15

Re: Urgent help needed with spray injection
 
ok, sorry, premature celebrations there. it now has C12H26 in the residuals, and it seems my chemical reaction scheme is working as i am getting H2O and CO2 produced, but my temperature is dropping towards the temperature of the spray, even though at 967K it should be igniting. can anyone see why i might not be getting any temperature increase? also, it says i have 0 pressure at my monitoring point.

peculiar

usker June 29, 2008 09:58

Re: Urgent help needed with spray injection
 
I would strongly recommend that you first do some background reading such as Star-CD Methodology, User Guide and do some exercises from Tutorial. Then, think about for a while and try to understand what exactly you are trying to do.

"Engage the brain before you put the mouth into gear".

nick June 29, 2008 10:04

Re: Urgent help needed with spray injection
 
I have read over the methodology, and also followed some of the tutorials. From what i can see, i have everything set up correctly. I see it as a quite simple model and I cannot understand why the simulation is following my reaction scheme, but the dodecane is not combusting in conditions where it should. i have also followed a methodology for similar setup, which is pretty much exactly what i have done.

To me, from what i have read the simulation should be working correctly. Can you see something i can't?

nick June 29, 2008 12:53

Re: Urgent help needed with spray injection
 
Is it possible for the injected fuel to ignite due to the high temp and pressure, or do i have to set ignition cells? i was hoping for it to ignite itself, if thats possible, which ignition model do you recommend

Andrea June 29, 2008 13:56

Re: Urgent help needed with spray injection
 
Hi nick
before anything else, what combustion model are you using? And what ignition model?
If you use shell model you haven't to set the igniont cell, you just have to set some parameters that defines the "ignition speed", the speed of the pre-flame reactions.
Write down everything...
best regards,
Andrea

nick June 29, 2008 14:13

Re: Urgent help needed with spray injection
 
I'm pretty new to Star-CD, so bear with me. I'm using the chemical kinetics model, and i was trying to use the numerical ignition method. I haven't used ignition before, but i've tried reading over the manuals, and tried a few things, but my temperature is always stuffed, and pressure reading 0 for my monitored point. my chemical reaction seems to be working, but i just need the combustion part.


Andrea June 29, 2008 18:20

Re: Urgent help needed with spray injection
 
ok, about the pressure reading "0": is it relative pressure I suppose, it's right? Try to open a .pst file inside pro-Star and type after the command "load" try to use "getc,p,abso". Look at the pressure, that's the absolute pressure, try to find if the monitored cell (and the others) have the right pressure value.
About the ignition: Note that the TIME or TIME-STEP option in the ignition model refer only to the time where is applied your ignition model, so they defines the time where the ignition CAN occour... You can choose as starting time the time where the injection start and as end time a time very very distant from the end of your simulation... One more thing: how can you say that the chemical reaction is working ok if you don't get an ignition? If you get CO2 and H20 the ignition is done, the combustion is taking place and the temperature must rise...but if the combustion is little and the CO2 and H20 created is also little, the rise in temperature is very small and so you don't see it. Is the combustion you get a REAL combustion or it is a small creation of CO2 and H20 at low reaction rate? Because if is it you have to speed up the ignition process and the combustion rate: lower your ignition temperature and takes higher reaction rate in the combustion mechanism.
Personally it's a long time that I don't use kinetics combustion model: I prefer the EBU-LATCT combustion model (Magnussen model). YOu can use it with Shell autoignition model (able to predict double delay combustion) and you can turn on NOx model and SOOT formation.

Andrea June 29, 2008 18:48

Re: Urgent help needed with spray injection
 
I've read also you other post on the forum: it's normal that the gas temperature drop down inside the spray...the liquid fuel is evaporating and so it takes heat from the gas phase inside the cylinder. Where the combustion occour, in spite of the drop, the temperature will rise after some reaction, because the release of energy.
You must analyze the results of you simulation: does the combustion starts or not? In other words...you get CO2 and H20? If yes, ok, speed up the reaction rate and the ingnition rate, you have to tune your model but it's all ok...when you'll get enough CO2 and H2O produced you'll se the rise of temperature.
If not, the combustion doesn't start and so the problem is probably the ignition model. Follow my previous post.
Best regards,
Andrea


All times are GMT -4. The time now is 13:42.