|
[Sponsors] |
RMS H-Energy (Particle) Does Not Converge Below 10-6 |
|
LinkBack | Thread Tools | Search this Thread | Display Modes |
October 14, 2017, 03:55 |
RMS H-Energy (Particle) Does Not Converge Below 10-6
|
#1 |
Senior Member
|
Dear all,
Hope all are well. My question may be simple but I am having trouble finding a solution for it so forgive me for asking. I am simulating convective laminar Eulerian-Eulerian homogeneous two phase flow and heat transfer in a pipe. Till a certain Re number my results converge but as I increase the Re number by a certain amount RMS H-Energy (Particle) does not converge below 10-6. It stabilizes to a value just above 10-6. Is a criteria of 10-6 for energy equations too constricted? I have already checked the mesh quality and boundary conditions. I have been changing inlet velocity according to change in concentration of particle as thermophysical properties change with concentration keeping Re number the same. I have been using an advection scheme→high resolution and fluid timescale control→physical timescale→2. MAX H-Energy (Particle) is hovering around 10-5. I read in documentation that "the location of the MAX residual of the momentum equations is the most useful to identify". The thing is I am not interested in residuals of momentum equation. Moreover, how can I locate a node number i.e. location or region in ICEM CFD to determine whether it is far enough from an outlet boundary to have no effect on my solution? My region of interest is entrance region where most of the heat transfer occurs. How can I write the equation residuals out to your RES file, create an isovolume equal to my MAX residual criteria, and verify that the residuals are low enough in my region of interest? I have attached an image of residuals. Hope somebody can shed some light on this. Thanks. |
|
October 14, 2017, 06:14 |
|
#2 |
Super Moderator
Glenn Horrocks
Join Date: Mar 2009
Location: Sydney, Australia
Posts: 17,705
Rep Power: 143 |
Do a simulation where you run it to 10E-5 residuals. If the results are the same as your 10E-6 results then you know converging tighter is not required. This is a sensitivity analysis, and I trust you have also done this for the mesh size and time step size (if transient).
|
|
October 14, 2017, 06:29 |
|
#3 |
Senior Member
|
Yes. Didn't do it for the timestep though as it's a steady state problem. Thanks Glenn. Domain imbalance for H-Energy for particle is -6.5613E-04. Is it sufficient in your opinion? Keep in mind this is not a conjugate heat transfer (CHT) simulation. Also how can I just change criteria for just one fluid (in this case the particle). Attached domain imbalance for the particle. Thanks.
|
|
October 14, 2017, 08:21 |
|
#4 |
Senior Member
|
Error between 10-5 and 10-6 is less than 0.01% Thanks for reminding me the value of sensitivity analysis Glenn. It can save a lot of time.
|
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Energy equation - where is the turbulent kinetic energy? | usv001 | OpenFOAM Programming & Development | 1 | January 25, 2022 15:04 |
UDF for multicomponent particle vaporization | Mohsin | Fluent UDF and Scheme Programming | 17 | January 27, 2021 02:57 |
Blood Damage Modelling via Particle Tracking in a Centrifugal Heart Pump | scatman | CFX | 7 | January 8, 2018 00:59 |
Check particle impaction with User Fortran | Julian K. | CFX | 3 | January 12, 2012 09:46 |
Why FVM for high-Re flows? | Zhong Lei | Main CFD Forum | 23 | May 14, 1999 13:22 |