Time evolution

tchitchas · July 6, 2018, 17:10

Hello everyone, I'm having a question regarding time integration.

I developed a code for unsteady flows using FVM, however, my problem evolves in time in a strange way (depends on the Reynolds). In my code, the flow takes almost twice to enter in steady state, when compared to the same problem executed in another code. After moving to the steady state, the results are the same when compared to the benchmark.

For temporal discretization, I used the fully implicit method. To approximate the velocity values on the faces of the control volume I used linear interpolation. Could this be the problem? Since this method of interpolation does not work well for higher Reynolds?

Ps: For Re<200 the temporal evolution occurred in an expected way.

Thank you in advance.

Hamidzoka · July 8, 2018, 07:23

It is not a serious problem If the only issue is the convergence speed!
Generally rate of convergence depends on many things from selected algorithm (SIMPLE, PISO,...) to relaxation factors, selected method for numerical solution of equations ( ADI,...).
you had better to use staggered grids for mementum equations in which velocities are calculated at faces of the element. for scalar equations like energy linear averaging the velocity field sounds ok.
Note that numerical codes usually becomes unstable as Re increases. relaxation factors can fix it to some extends. If Re reaches the turbulent flow margins you should add a proper turbulent model to find a stable convergence.

Regards

July 6, 2018, 17:10	Time evolution	#1
tchitchas New Member Daniel Gustavo Benvenutti Join Date: Jun 2017 Posts: 5 Rep Power: 8	Hello everyone, I'm having a question regarding time integration. I developed a code for unsteady flows using FVM, however, my problem evolves in time in a strange way (depends on the Reynolds). In my code, the flow takes almost twice to enter in steady state, when compared to the same problem executed in another code. After moving to the steady state, the results are the same when compared to the benchmark. For temporal discretization, I used the fully implicit method. To approximate the velocity values on the faces of the control volume I used linear interpolation. Could this be the problem? Since this method of interpolation does not work well for higher Reynolds? Ps: For Re<200 the temporal evolution occurred in an expected way. Thank you in advance.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Elastic or Plastic Deformations in OpenFOAM	NickolasPl	OpenFOAM Pre-Processing	36	September 23, 2023 08:22
bash script for pseudo-parallel usage of reconstructPar	kwardle	OpenFOAM Post-Processing	41	August 23, 2023 02:48
AMI speed performance	danny123	OpenFOAM	21	October 24, 2020 04:13
Floating point exception error	lpz_michele	OpenFOAM Running, Solving & CFD	53	October 19, 2015 02:50
simpleFoam error - "Floating point exception"	mbcx4jc2	OpenFOAM Running, Solving & CFD	12	August 4, 2015 02:20

July 8, 2018, 07:23		#2
Hamidzoka Senior Member Hamid Zoka Join Date: Nov 2009 Posts: 282 Rep Power: 18	It is not a serious problem If the only issue is the convergence speed! Generally rate of convergence depends on many things from selected algorithm (SIMPLE, PISO,...) to relaxation factors, selected method for numerical solution of equations ( ADI,...). you had better to use staggered grids for mementum equations in which velocities are calculated at faces of the element. for scalar equations like energy linear averaging the velocity field sounds ok. Note that numerical codes usually becomes unstable as Re increases. relaxation factors can fix it to some extends. If Re reaches the turbulent flow margins you should add a proper turbulent model to find a stable convergence. Regards