CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   FLUENT (https://www.cfd-online.com/Forums/fluent/)
-   -   For Nozzle fluent problem (https://www.cfd-online.com/Forums/fluent/29311-nozzle-fluent-problem.html)

 Jie January 31, 2002 17:38

For Nozzle fluent problem

Hello,

I use Fluent to calculate the nozzle problem with NPR = 20,30,40, 50, 90. My problem is coupled , 2d, implicit, steady problem. Use Spalart-Allmaras as the viscous model.Energy model is enabled.

My question is :

I have calcuate NPR =20 case, Then I took it as the initial field for NPR =30, Take Npr=30 as the initial field for NPR =40 I found that I increase the inlet pressure,but the separete position doesn't change.

Anybody have ideas?

Thanks very much.

 l.g.patil February 1, 2002 07:41

Re: For Nozzle fluent problem

 Jie February 13, 2002 10:53

Re: For Nozzle fluent problem

hello, Sir Sorry to reply to you so late.

My project is about flow separation analysis for rocket nozzles.

I built my model as: inlet : inlet pressure wall farfield: farfield pressure. The model is just like the model from the paper : AIAA 99-2587 The Physical Origins of Side loads in Rocket Nozzles. By M. Onofri and F.Nasuti.

If I take farfield pressure a certain value and increase the inlet pressure, the results are perfect and the separate position will move downstream when the inlet pressure increases. But if I take the inlet pressure as a certain value and decrease the farfield pressure,( theoretically it has the same effects as before),yet I can't move the separation position any more. That't the problem.

For fluent, I used coupled, 2d, implicit , steady model. Take Spalart-Allmaras as the viscous model. and Enable the energy model.

Farfield pressure: Mar#=0.05,Temp=300K etc. If you want to know my project further,please contact with me by jiyao@uncc.edu. Thanks very much.

I will appreciate it, if you can help me in the some problems for this project.

 purushothge December 25, 2011 09:10

HI to all ;
this is my first post in this forum. I am trying to simulate the nozzle flow with plume interaction. My nozzle is having thrust value of 5n and its made in such a way that it is operated in deep space applications.
what are the BC's i have to give for getting the plume exactly ?
because already i have got the converged solution which shows the drastic changes in the temperature profile .
when coming to the mesh , i have made a mesh such that my maximum aspect ratio is 79.82 the total count of 300*100 inside the nozzle and 280*100 on the outer domain.

 duri December 25, 2011 13:49

Quote:
 Originally Posted by purushothge (Post 336969) HI to all ; already i have got the converged solution which shows the drastic changes in the temperature profile .
Is gamma consistent with flow physics?. This could give very different temperature profile.

I am interested to know what conditions do you use for far field boundary and how do you manage continuum condition.

 purushothge December 25, 2011 14:11

Quote:
 Originally Posted by duri (Post 336984) Is gamma consistent with flow physics?. This could give very different temperature profile. I am interested to know what conditions do you use for far field boundary and how do you manage continuum condition.
HI Duri,
Thanks for the earliest reply. taking gamma factor , i have used the values which i got from the NASA CEA program where it gives the gamma value 1.235 wrt that i made my density as 0.587,k=10.942,M=22 and ideal gas.
Indeed i had a doubt that in deep space continuum will be taking no effect and ordinary NS equation is quite suffice to handle this problem ?
though i am trying for that .....
will it yield the proper flow pattern ???

 duri December 26, 2011 05:23

Quote:
 Originally Posted by purushothge (Post 336985) HI Duri, gamma value 1.235 wrt that i made my density as 0.587,k=10.942,M=22 and ideal gas.
I solved similar kind of problem long back may be in 6.2. By changing the flow quantities will not change the gamma. After initialization check the values of gamma in flow field. I think you need to change CP values.
But this will affect the external flow as well. When I did, i remember i added species transport to keep the fluid properties consistent.

Quote:
 Originally Posted by purushothge (Post 336985) Indeed i had a doubt that in deep space continuum will be taking no effect and ordinary NS equation is quite suffice to handle this problem ? though i am trying for that ..... will it yield the proper flow pattern ???
I don't know answer for this. When knudsen number is low you may need to solve kinetic equations instead on NS equtions. I don't know any commercial software that solves boltzmann equation.

 purushothge December 26, 2011 07:03

Dear Duri ,
Thanks for your earnest reply. As you said , i have changed the cp value to. but i made cp as a constant phenomena here and i havent accorded the linearised pattern for cp changes.
I want to stress on your quote that you had attempted the same on "6.2' . I need to know whether you could arrive at the desired pattern ?
more, knudsen number and NS equations arent enough to simute the flow ? Should i go for Boltzmann equation ?

 purushothge December 30, 2011 01:45

Dear DURI,
I have tried to attain the nozzle simulation for vacuum conditon. This follows the conventional under expandd pattern of plume. I wonder that it would happen in deep space !
More, I just came across the CHAPTER 18 of rocket propulsion elements by sutton , in which he clearly gave the illustrative picture of vacuum expansion condition.
The pattern which i obtained is using the boundary conditions LEFT AND TOP EXTREME DOMAIN as PRESSURE INLET WITH 100 Pa and FOR THE RIGHT EXTREME DOMAIN I USD THE PRESSURE OUTLET WITH 0 GAUGE PRESSURE.
WHEN I USED THE PRESSURE OUTLET CONDITION FOR ALL THE ABOVE MENTIONED DOMAINS I GOT A PATTERN WHICH IS GETTING WIDENED DOWNSTREAM OF THE DOMAIN.
WHICH CONDITON SHOULD I USE FOR THAT ?
Will you please help me to sort this out ? I will upload both of the pictures in my next post as i have been running the same for other case .
THANK YOU
PURUSHOTHMAN.N

 duri December 31, 2011 07:28

2 Attachment(s)
I tried to reproduce you problem with two different nozzle to ambient pressure ratio (1000 and 10000). Exhaust plume i got is almost like high altitude plumes. Check the attachments. The boundary conditions i used are pressure inlet at left and pressure outlet at top and right. Second order with k-epsilon turbulence model. Pressure ratio at nozzle inlet is 2.

Later i realized that domain is not sufficient enough to solve this problem. Nevertheless, it shows its possible to solve these kind of problems.

 purushothge January 1, 2012 22:03

Dear DURI,
I want to thank you for your kind help. As you said , Yo have attempted for NPR 100-1000. have used multi species concept ?
In my case, nozzle inlet pressure is 7 bar and expanding to the vacuum . I will be pleased if you send me your mail address, so that i can send my case and data file for your perusal.
My mail id is : purushothge@gmail.com.
Thanking you
PURUSHOTHAMAN.N

 purushothge January 1, 2012 22:17

1 Attachment(s)
Later i realized that domain is not sufficient enough to solve this problem. Nevertheless, it shows its possible to solve these kind of problems.[/QUOTE]
Here , i have attached my velocity plot in which i have used ideal gas alone as a fluid . should i change it to real case ??

 flashkicker9 January 9, 2012 17:51

compressible converging nozzle

hi, everybody i am new to the gambit and fluent software, and i needed help solving a converging nozzle problem,i am not getting the convergence in iterations, its actually showing lot of divergence during iteration process,
and it ends with warning like reverse flow at outlet and turbulent flow at outlet,
i have selected viscous flow to be -spalart allamaras and the courant number as 5.
plz plz plz i need the solution urgent and asap,
million thanks
flash
http://http://s1187.photobucket.com/.../flashkicker9/

above is the image of my model, grey part shows pressure inlet and pink shows the pressure far field. plz help!!!

 duri January 10, 2012 03:24

Your image is not available. I hope you haven't used pressure outlet boundary. Post your problem clearly.

 flashkicker9 January 10, 2012 15:04

1 Attachment(s)
i have not used pressure outlet boundary.
i have used solver as density based and fluid as Air(property-ideal gas).
As the iteration progresses the reverse flow and turbulent flow error is shown, after 80-90 iterations it shows divergence and accordingly it reduces the courant number from .5 to 5e-5 for each iterations and it ends with error that solution is diverging.
below is image of my model- the white region is wall, the pink region is pressure far field, and grey(ish) region is pressure inlet.
i need to find the flow characteristics of compressible flow in nozzle.

 duri January 10, 2012 15:24

You could do this as Axi-symmetry case it saves lot of time and effort.
Either axi-symmetry or 3D. Use pressure outlet at down stream exit. If flow at that region is subsonic then choose appropriate exit pressure. Keep low Courant number initially (< 1). Extend the inlet for 2 or 3 cells to have constant area near inlet. Area change near the inlet sometime behaves badly (I found solution usually starts to converge and suddenly diverge after some iterations).

 flashkicker9 January 10, 2012 15:33

even its same in my case it converges at the beginning and later on its diverging, i wanna know why it is diverging and i tried in many ways but still its diverging. and could you plz explain to me how to extend the inlet for 2-3 cells?
what you mean by pressure outlet at downstream exit?

 flashkicker9 January 11, 2012 14:44

Quote:
 Originally Posted by duri (Post 338686) You could do this as Axi-symmetry case it saves lot of time and effort. Either axi-symmetry or 3D. Use pressure outlet at down stream exit. If flow at that region is subsonic then choose appropriate exit pressure. Keep low Courant number initially (< 1). Extend the inlet for 2 or 3 cells to have constant area near inlet. Area change near the inlet sometime behaves badly (I found solution usually starts to converge and suddenly diverge after some iterations).
thank you for the help but please could you tell me what condition to take for compressible flow?

 All times are GMT -4. The time now is 06:12.