CFD Online Logo CFD Online URL
Home > Forums > Main CFD Forum

transition zone

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

LinkBack Thread Tools Display Modes
Old   July 14, 2000, 06:15
Default transition zone
Posts: n/a
Hi, As my earlier message was incomplete, i posted that again. My queries were : (1) Can we predict the heat transfer and pressure drop in the transition region with the help of cfd? (2) In case of standard k-e model with wall function approach, y+ should be between 30-50. In this range, friction factor and heat transfer co-efficient change even for fully developed turbulent flow in a pipe. So how shall we attain the mesh independent solution? Or we have to go on changing the mesh size till we reach to the experimental value.

Please help me to clear my funda. Thanks in advance.

  Reply With Quote

Old   July 15, 2000, 21:07
Default Re: transition zone
John C. Chien
Posts: n/a
(1). The laminar-turbulence transition is a tough research field. It has been simulated since the days of boundary layer/mixing length theory till present time of low Reynolds number models. (2). The wall function treatment replaces the numerical calculation of the near the wall region by a single parameter family of the Law of the Wall, which is also implicitly included in the turbulence model of the outer region. (3). In other words, both regions have the universal Law of the Wall function included in the solutions. The iteration method simply tries to find the proper v* value such that the two Law of the Wall solutions will join properly at the interface, that is the mesh boundary of Y+=30~300. (4). The other way to say is: Plot the velocity solution in semi-log scale, using v* as a parameter (as done in the Law of the wall function). Pick a point in the Y+ scale at a point between Y+=30 and Y+=300, and divide the flow field into two regions. The inner region is represented by the Law of the Wall function, and the outer region is represented by the mesh independent numerical solution of the Law of the Wall function. At the interface point, the sigle parameter is v* from both regions, and the U at the interface, which must be continuous. (5). By the way, the mesh independent numerical solution is for the outer region only. As long as the interface point Y+ is within the range, there is no need the alter the interface location (associated with the Law of the Wall region). And the final Y+ is part of the solution. There is a similar discussion in the previous messages on this subject. Use the search to find it.
  Reply With Quote

Old   July 17, 2000, 05:54
Default Re: transition zone
Posts: n/a
John, Thanks a lot for your nice explanation. But i have still some queries on transition that i'm posting later. I expect your comments on that. dilip
  Reply With Quote


Thread Tools
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 On
Pingbacks are On
Refbacks are On

Similar Threads
Thread Thread Starter Forum Replies Last Post
Fluent3DMeshToFoam simvun OpenFOAM Other Meshers: ICEM, Star, Ansys, Pointwise, GridPro, Ansa, ... 48 May 14, 2012 05:20
Problem in running ICEM grid in Openfoam Tarak OpenFOAM 6 September 9, 2011 17:51
Problem in IMPORT of ICEM input file in FLUENT csvirume FLUENT 2 September 9, 2009 01:08
Error to re-open fluent case file J.Gimbun FLUENT 0 April 27, 2006 08:42
Sliding mesh error Karl Kevala FLUENT 4 February 21, 2001 16:52

All times are GMT -4. The time now is 23:07.