|
[Sponsors] |
How can we choose the residuals for the transient flow? should 1e-3 enough or not? sh |
|
LinkBack | Thread Tools | Search this Thread | Display Modes |
May 12, 2020, 20:27 |
How can we choose the residuals for the transient flow? should 1e-3 enough or not? sh
|
#1 |
Senior Member
Join Date: Dec 2017
Posts: 388
Rep Power: 9 |
How can we choose the residuals for the transient flow? should 1e-3 enough or not? should we choose 1e-6? I see different people have different ideas. what is the difference between the absolute convergence criterion and relative convergence criterion?
|
|
May 12, 2020, 22:00 |
|
#2 |
New Member
Alex
Join Date: May 2020
Posts: 21
Rep Power: 6 |
Case in case for transient flows, in some type of transient flows the residuals never get small enough,for example :flow cross the cylinder
|
|
May 13, 2020, 12:13 |
What is the reason? How can I solve it? If I only set it as 1e-3, whether the result
|
#3 |
Senior Member
Join Date: Dec 2017
Posts: 388
Rep Power: 9 |
||
May 14, 2020, 04:34 |
Residuals
|
#4 |
Senior Member
|
You can reduce the residuals but do note that the reduction in residuals do not directly translate to more accurate solutions. Therefore, you should not worry that much about residuals. Use monitors. Read Chapter 2, especially page 8 from the
https://books.google.nl/books?id=uSM...torial&f=false
__________________
Regards, Vinerm PM to be used if and only if you do not want something to be shared publicly. PM is considered to be of the least priority. |
|
May 14, 2020, 11:55 |
I cannot understand its explanation in the book? Could you explain it easily for me w
|
#5 | |
Senior Member
Join Date: Dec 2017
Posts: 388
Rep Power: 9 |
Quote:
|
||
May 14, 2020, 12:12 |
Residual and Accuracy
|
#6 |
Senior Member
|
You don't need to understand much from the book. Don't start reading whole of it, only the content within the box on page 8. Rest is not important as of now. What you need to understand is that residuals are not always good indicators of the accuracy. All it talks about is that there are two systems of linear equations and both have same exact solution, u1 = 1, u2 = 2. However, when another solution is assumed, i.e., u1 = -0.95 and u2 = -1, it shows large residual for second case, which is good but small residual for first system. Though the solution is pretty bad yet the residual is small. In other words, small residuals don't translate to more accurate solutions.
No, you cannot download the book.
__________________
Regards, Vinerm PM to be used if and only if you do not want something to be shared publicly. PM is considered to be of the least priority. |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
[TUTORIAL] Run fluent on distributed memory with 2 windows 7 64 bit machines | ghost82 | FLUENT | 54 | February 9, 2022 04:32 |
Transient solution residuals | lyro | FLUENT | 7 | May 5, 2020 14:09 |
Global residuals in transient case | mx | Main CFD Forum | 0 | August 6, 2007 08:12 |
transient run and residuals | Mark | Main CFD Forum | 4 | April 21, 2004 11:12 |
transient simulation: natural convection problem? | Basics | CFX | 3 | September 25, 2002 10:42 |