Simulating Flexible filaments with gravity ON

[paulakash24]

Hello,

Hope you are doing well.

I have a doubt regarding the usage of Gravity force. Can we simulate flexible filaments (pinned) and subjected to uniform flow and gravity turned ON in the flow direction ?

I was trying to validate a benchmark case. But the simulation crashes at an initial timestep of 27 saying "position is off entire grid hierarchy". I can see that the simulation runs fine when I mention the filament as RIGID but fails when I make it FLEXIBLE.

There is no change in the source code. I am only giving inputs in the definitions.h and geometry.config file.

I am attaching the definitions.h and geometry.config file for reference.
link

Any help would be highly appreciated.

Thank You
Akash

[ianhinder]

Hi. Have you checked the filament node positions? From the error message, it looks like part of the filament has moved off the computational grid, which is not necessarily indicative of a bug. If this is a benchmark case, it's possible that the case has not been set up correctly; e.g. the Young's modulus of the filament might be set wrong, and it could be too flexible?

I don't know the answer to your question; i.e. whether gravity is compatible with flexible filaments. My guess is that it should work, but someone who knows better should confirm this.

[paulakash24]

Hi,

Thank you for the response.

Yes, it is a benchmark case that I am trying to validate. The E value came out as per my calculation. Assuming I had calculated the E value wrongly, I had run several cases with with large E value but the simulation fails similarly. In fact these cases run fine with gravity turned OFF.

Also, I ran the Turek & Hron (circular cylinder with splitter plate case) that LUMA used as validation with gravity turned ON in the flow direction. It failed too, showing similar error that body is out of computational domain.

[ianhinder]

Have you visualised the simulation, e.g. in paraview, with the bodies, so that you can see what's going on?

[paulakash24]

Hi,

I have visualized but it is hard to understand what is going wrong. For example, I visualized circular cylinder with filament case for which I am attaching the Ux distribution. The gravity is turned ON in the x direction.

One more thing I need to know is that, in definitions.h file, do we give the acceleration due to gravity value in L_GRAVITY_FORCE option?