|
[Sponsors] |
February 19, 2018, 11:20 |
Flutter: convergence of aerodynamic work
|
#1 |
Senior Member
M
Join Date: Dec 2017
Posts: 698
Rep Power: 12 |
Hi all. I had some questions recently on my way to setting up my simulation and yet there is another obstacle. Thanks for the help to get here anyway.
For my flutter setup, I am using a mesh with 1 turbine blade and a long outlet. The aerodynamic work I try to extract is the area integral of the pressure at the blade surface integrated over a vibration cycle. I can monitor this value with a monitor point, see pictures to the right. After a few vibrations I expect this value to converge and extract this as result. Currently I am using 80 steps to resolve a vibration cycle which should be sufficiently small (?). The current model can only represent an interblade phase angle of 0, meaning all blades are vibrating in phase and periodic BC are applicable. The specific problem here is, I have two identical setups, one time a tip gap of about 2mm is included, the other case is w/o tip gap. I added two images of solver manager monitor prints, one for each case. The tip_solver picture shows good convergence for both residuals (below 1E-3) and aerodynamic work. The other file shows how aerodynamic becomes somewhat stable and then starts fluctuating and looks like its diverging, same goes for the efficiency plots. The residuals are higher but dont seem to change -> flat lines, which I find suspective. I made sure to do a steady state simulation for each geometry and extract inlet profiles from the respective run to use them as boundary profile. Same goes for the outlet. What I tried to do? I figured it must be a setup issue, I compared both .out files and they are identical, despite another mesh is used (of course) and I had to use another mesh displacement for the shroud (unspecified for noTip, surface of revolution for Tip). I do not see an obvious difference in setups that could cause it. I looked up where those larger residuals are occuring in the domain, they are located right @ the inlet. I have no idea why, but the rest of the domain is well within desired range. My intention is, that the residuals are not causing this fluctuation. I am happy for any suggestion. I need to find a consistent setup which I can use for all geometries (other tip gap sizes) and interblade phase angles which means I need to model more passages. This fluctuation looks like some strange behavior from the solver rather than a wrong setup to me. |
|
February 20, 2018, 06:40 |
|
#2 |
Super Moderator
Glenn Horrocks
Join Date: Mar 2009
Location: Sydney, Australia
Posts: 17,841
Rep Power: 144 |
Before you start looking at your results too hard, have you done the normal checks for simulation accuracy? See FAQ: https://www.cfd-online.com/Wiki/Ansy..._inaccurate.3F
The most frequently overlooked part of these checks is sensitivity checks on mesh size, convergence criteria and time step size. You need to check these are OK before you can trust your results.
__________________
Note: I do not answer CFD questions by PM. CFD questions should be posted on the forum. |
|
Thread Tools | Search this Thread |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Convergence in steady state simulations vs transient ones | cardioCFD | CFX | 5 | January 21, 2018 11:59 |
How to determine the convergence of solutions | cfx_ws1992 | CFX | 6 | November 9, 2017 17:12 |
Does the Case Termination Control based on coefficient convergence work? | ronithstanly | OpenFOAM Running, Solving & CFD | 0 | May 1, 2017 05:27 |
determining flutter index corresponding to flutter speed and flutter index variati | varghese9 | Main CFD Forum | 2 | January 29, 2015 23:38 |
Why FVM for high-Re flows? | Zhong Lei | Main CFD Forum | 23 | May 14, 1999 14:22 |