FSI – Flutter in mechanical results

- June 4, 2019 at 9:33 am
  
  Jurjen
  Subscriber
  
  Dear reader,
  
  I am working on two-way FSI simulation of blood vessel.
  In my work I couple transient mechanical and FLUENT for FSI coupling.
  The inlet signal is given by fully developed doppler signal and outlet inserted by a self-programmed windkessel type model.
  
  Since both model converge very well and the coupling does as well i was quite happy. But when looking at the results i noticed a strange 'flutter' (oscillation) like behaviour in the mechanical results (displacements, stresses, strains etc.).
  Since at first i thought this would be a mistake in my outlet pressure code I also monitored the fluid pressure. But these does not seem to have this flutter behaviour.
  
  Therefore my questions:
  - Did someone see/recognizes this flutter like behaviour before?
  - Are there suitable ways to damp or stabalize this behaviour?
  - How could it be that this behaviour is not observed in fluid pressure?
  
  Some illustrations:
  - Fluid pressure movie: https://drive.google.com/file/d/1RxN8lm0rc_qkLapjFWHsj1X9Lqnb4uxA/view?usp=sharing
  - Mises stress movie: https://drive.google.com/file/d/1gOaOsz4UPD2FmcfDodHrR-ECCIYWzjjR/view?usp=sharing
  - Illustration of the max, avg and min values of displacements:
- June 5, 2019 at 5:34 pm
  
  ansysuser
  Subscriber
  
  Hello,
  
  I am doing a similar project. In structural are you using Define By Time or Define By Substeps? I have found this makes a large difference. Also, are you using hyper elastic material with no damping?
- June 6, 2019 at 10:03 am
  
  Jurjen
  Subscriber
  
  Hi,
  
  I am using define by substeps, but restrict it to have only 1 substep per coupling iteration.
  What is according to you experience better to do: substeps or time?
  
  And yes i use hyper elastic neo-hookean material with no damping, do you recommend to put damping on it?
  If yes, do it in system coupling or in mechanical?
  
  Thanks!
- June 7, 2019 at 9:52 pm
  
  ansysuser
  Subscriber
  
  Hello,
  
  Yes, I have found significant difference between using substeps vs. timesteps in FSI. I posted about this but I have not heard a reply from ANSYS yet. I show in this video some of the difference.
  
  https://www.youtube.com/watch?v=mrM70M0WKfk
- June 17, 2019 at 12:46 pm
  
  Jurjen
  Subscriber
  
  So as i understand correctly, in your movie substeps were superior to timesteps?