
[Sponsors] 
Correct method for calculating time step size 

LinkBack  Thread Tools  Search this Thread  Display Modes 
February 9, 2016, 04:49 
Correct method for calculating time step size

#1 
New Member
Aarthy Meena
Join Date: Oct 2015
Location: Trichy
Posts: 29
Rep Power: 9 
Hello Everyone,
I am a newbie to FLUENT & CFD. I wanted to calculate time step size for unsteady flow around a circular cylinder of dia = 0.01m in the Reynolds number of 45 to 50. The working fluid is water with viscosity = 0.001003 kg/ms and density = 998.2 kg/m3. I came across several posts related to this and many methods to calculate the time step size. 1) From Strouhal number (0.2 for circular cylinder approx) Sr (0.2) = (frequency x diameter) / velocity From the above equation find the frequency. The total time period, T = 1/frequency. Hence, time step t = T/25 (approx). 2) From CFL condition t = C_m X x / velocity where, C_m = 1 (approx) and x = minimum cell size. 3)From domain length t = domain length / (20 x velocity) 1) Which of the above method is correct ? 2) How to calculate the suitable time step size for my condition? 3) Please explain. Thanks in advance.
__________________
Aarthy, PG Student, National Institute of Technology, Trichy, India. Last edited by Aarthy Meena; February 9, 2016 at 05:50. 

February 9, 2016, 05:25 

#2 
Senior Member
Filippo Maria Denaro
Join Date: Jul 2010
Posts: 6,266
Rep Power: 67 
the time step is constrained by stability criteria that are based not only on the cfl condition but also on the visicous terms.
Furthermore, each discretization has a specific stability region 

February 9, 2016, 05:28 

#3 
New Member
Aarthy Meena
Join Date: Oct 2015
Location: Trichy
Posts: 29
Rep Power: 9 
Dear Sir,
Can you please explain me the method for my flow scenario around a circular cylinder.
__________________
Aarthy, PG Student, National Institute of Technology, Trichy, India. 

February 9, 2016, 05:36 

#4 
Senior Member
Filippo Maria Denaro
Join Date: Jul 2010
Posts: 6,266
Rep Power: 67 
The numerical stability depends upon the method you use for the integration and the sizes of the mesh (in all directions), does not depend on a specific flow problem.
Assuming you are using nondimensional equations, the velocity magnitude is no more than O(1) so that you can use that as estimation in the (multidimensional) CFL condition 

February 9, 2016, 05:46 

#5 
New Member
Aarthy Meena
Join Date: Oct 2015
Location: Trichy
Posts: 29
Rep Power: 9 
Thank you sir,
Actually am very new to FLUENT and I do not understand your reply. In FLUENT, I am using dimensional form of the governing equations. Kindly please suggest me some material where I can learn about this topic.
__________________
Aarthy, PG Student, National Institute of Technology, Trichy, India. 

February 10, 2016, 04:20 

#6 
New Member
Lucas
Join Date: Jun 2015
Posts: 20
Rep Power: 9 
A theoretical advice:
The CFLnumber and the according "condition" are only part of the stability analysis of numerical schemes. A good introduction (for Finite Difference methods) can be found on: https://en.wikipedia.org/wiki/Von_Ne...ility_analysis Basically the stability depends on the equation you discretise (including all the terms in it  as FMDemaro added), as well as the discretisation scheme you use. As a "rule of thumb": lower order discretisation schemes (e.g.: Upwind schemes) and implicit timedependent solution are very stable, whereas the more accurate methods tend to be unstable. For the Finite Element method, this can be proven analytically for certain cases  An example is also contained in the link. A practical advice: I am not familiar with ANSYS Fluent but I guess, that you will use Finite Volume Method with an implicit solver. The theoretical advice from above might therefore be tough to prove for your problem. I would advise you to have look on the user manual and see, if there is any hint. A cylinder in cross flow is also very well researched and there might be a lot of data and advice for your example already available. Regards, Lucas 

February 10, 2016, 05:46 
Help please!!

#7 
New Member
Ravi
Join Date: Feb 2016
Posts: 1
Rep Power: 0 
I am trying to do a transient analysis of flow through a bifurcation of penstock pipe (wye) using ansys cfx . Here i need to specify the conditions that the valves of the turbine are closed within a 5 seconds of time. I thought of giving pressure inlet and mass flow rate at the outlet. Now i need to vary the mass flow rate at the outlet from 4650kg/s to 0 within a 5 seconds of time and i dont know how to do that in ansys CFX. Can anybody help me please ??


May 7, 2019, 08:56 
question

#8 
Member
Join Date: Sep 2018
Posts: 31
Rep Power: 6 
Dear Aarthy Meena,
did you find a good average to calculate the time step and time step size? 

Tags 
circular cylinder, fluent 14.5, time step size, transient simulations 
Thread Tools  Search this Thread 
Display Modes  


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
AMI speed performance  danny123  OpenFOAM  21  October 24, 2020 05:13 
time step continuity problem in VAWT simulation  lpz_michele  OpenFOAM Running, Solving & CFD  5  February 22, 2018 20:50 
Star cd esice solver error  ernarasimman  STARCD  2  September 12, 2014 01:01 
Micro Scale Pore, icoFoam  gooya_kabir  OpenFOAM Running, Solving & CFD  2  November 2, 2013 14:58 
same geometry,structured and unstructured mesh,different behaviour.  sharonyue  OpenFOAM Running, Solving & CFD  13  January 2, 2013 23:40 