CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > Software User Forums > ANSYS > FLUENT

Unsteady flow

Register Blogs Members List Search Today's Posts Mark Forums Read

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   October 25, 2001, 14:28
Default Unsteady flow
  #1
Ali
Guest
 
Posts: n/a
Dear Fluent Users,

I am running an unsteady supersonic flow directed on the surface. This is the first time I have attempted the unsteady calculation and I am having a considerable difficulties in converging the residuals to the desired level.I have even reduced the time step size as low as 10^-08 and got nowhere with it. Is it the rexation factors or something else that I need to mannipulate to achieve my goal.

What my final goal is to measure the oscillating time period of the blunt body shock on the surface, due to its pressure fluctuation.

Is any body out there who has attempted these kind of CFD models before and know more than I do and one last thing; how do I postprocess the data to simulate this phenomena.

Thank you guys very much in advanced. I am hoping for some excellent answers from you experts.

  Reply With Quote

Old   October 27, 2001, 17:20
Default Re: Unsteady flow
  #2
Axel Rohde
Guest
 
Posts: n/a
Although I am not a Fluent user, your question did strike my interest. Can you be a bit more specific about your problem setup: Is the flow 2-D or 3-D? What is the free-stream Mach number? What is the size of your plate? How does the unsteadiness come in?
  Reply With Quote

Old   October 28, 2001, 16:52
Default Re: Unsteady flow
  #3
reza pegah
Guest
 
Posts: n/a
  Reply With Quote

Old   October 29, 2001, 08:38
Default Re: Unsteady flow
  #4
david
Guest
 
Posts: n/a
A) Have you checked even though ur residuals are not of expected level, still it may represent the right solution? B) for unsteady solution, animation of representative properties would be good to show
  Reply With Quote

Old   October 29, 2001, 12:46
Default Re: Unsteady flow
  #5
Ali
Guest
 
Posts: n/a
Hi Axel

The problem is 2D axisymmetric and the free stream Mach number is around 2. The fluid is ejected from a 1mm nozzle and the plate is infinately long. The flow is impinging on the slot. i.e this is same as the Hartmann-Springer affect. The shock wave on the surface oscillates due to pressure fluctuation on the surface. This phenomena is real and highly validated over the years, but I would like to model this in CFD any thought.

Thank you Ali

  Reply With Quote

Old   October 29, 2001, 12:48
Default Re: Unsteady flow
  #6
Ali
Guest
 
Posts: n/a
Hello David,

Have you tried to show any fluid property animation with your work. If so can you pass on some hint this way please.

Thank you

Ali
  Reply With Quote

Old   October 30, 2001, 08:21
Default Re: Unsteady flow
  #7
Axel Rohde
Guest
 
Posts: n/a
Ali,

Thanks for the description. I am not familiar with this setup, particularly the Hartmann-Springer effect. Do you have a web reference of some sort where I can read up on this? - Also, is this plate horizontal? If the free-stream flow is Mach 2, what is the velocity of the jet (or is the jet velocity at Mach 2)? At what angle does the jet impinge on the plate?

By the way, I publish a shareware program 'MicroTunnel' that deals with 2-D unsteady compressible aerodynamics. Here a link to a transonic flow over a prism (Mach 0.9) with alternating shock shedding,

www.microcfd.com/prism.htm

The program is still in its development stage. I am currently implementing a cut-cell interpolation algorithm for smoother flow over curved surfaces (airfoils). For rectangular geometries, such as horizontal/vertical plates it is well suited, although a 'jet' could only be added by modifying the source code.

Axel
  Reply With Quote

Old   October 30, 2001, 10:40
Default Re: Unsteady flow
  #8
Ali
Guest
 
Posts: n/a
Dear Axel

The jet Mach No. is 2 and the jet is directed perpendicular to the interaction surface. The jet is exausted in to ambient air. The CFD result indicates that pressure absolute rise downstream of the blunt body shock is as much as 4.5 bar/atmosphere.

As far as the Hartmann-Springer effect is conserned,there is very little information available to me, but there are few research institutes that are involed in this line of work. May be you may find some information at the webb sight below.

http://info.fysik.dtu.dk/struktur/morch.html

P.S. What was the time duration of your simulation cycle for the prism ?

Good luck, happy hunting

Ali
  Reply With Quote

Old   October 31, 2001, 09:25
Default Re: Unsteady flow
  #9
Axel Rohde
Guest
 
Posts: n/a
Ali,

Thank you for the link. I now understand the setup of your flow. Your stability problem may have to do with the fact that the surrounding air is stationary. My shareware MicroTunnel evolved from a code that was used to study transient supersonic jets exiting into a subsonic or supersonic free stream. I just started the old Jet program, and a free stream Mach number of M = 0 is not an option (M = 0.1 being the lowest). I will look into that when I get a chance.

About the unsteady simulation ( www.microcfd.com/prism.htm ):

The animated GIF is a closed loop of about 50 frames (48 I think). The unsteadiness is periodic, although the closure at the end of the loop is not 'perfect' as you may have noticed. The first frame of the simulation is shown in,

www.microcfd.com/gallery/prism-trans-m.htm

which also gives you all the test setup data including dimensions. The entire simulation time frame of one 'period' is 5ms (milliseconds), as indicated by the Run Time (0.045s) and Stop Time (0.050s) in the window. The indicated Time Step of 0.10ms is the time interval in between each graphical update, NOT the integration time step (which is about 5 microseconds). Thus each of the 50 frames in the animated GIF is 0.10ms apart.

MicroTunnel marches explicitly in time, and to integrate 1ms takes about 5 minutes on a 1GHz Intel Pentium III (AMD are faster). So the entire simulation (5ms) takes about 25 minutes to compute. Keep in mind that you have 800x600 = 0.48 million cells, and that each pixel on the (800x600) display represents exactly one cell in the Cartesian mesh. I would be curious how long Fluent would take to run this using an inviscid solver at the same resolution.

Axel
  Reply With Quote

Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
how to predict unsteady flow from case definition? Felix Main CFD Forum 6 August 28, 2007 17:40
Unsteady flow in FLUENT Vidya Raja FLUENT 2 November 2, 2005 17:29
Unsteady Boundary Layer Flow Wen Long Main CFD Forum 0 July 30, 2002 00:08
fluid flow fundas ram Main CFD Forum 5 June 17, 2000 22:31
computation about flow around a yawed cone Tylor Xie Main CFD Forum 0 June 9, 1999 08:33


All times are GMT -4. The time now is 20:00.