No fluid-structure interaction was observed

Using preCICE
1

Hello
I am learning the partitioned coupling process of the fluid-structure coupling process. By learning the FSI paper on the official website, I can roughly understand the coupling process of fluid and solid:

微信截图_20230605205351
微信截图_20230605205351981×396 66.4 KB

This picture is from this paper: Fluid Structure Interaction Modelling of Flapping Wings

I learned a few key parts:

  1. The fluid solver transfers the force to the solid solver

  2. Solid solver uses force vector to calculate displacement vector

  3. The displacement vector is passed to the fluid solver. If it does not converge, the new displacement vector is used to re-solve the fluid

Then I wanted to use a simple example to understand the fluid-structure coupling process. My idea was to increase the stiffness of the solid-state solver in the vertical-Flap case (by a factor of 1000000, so that the structure is approximately rigid). The degree of freedom in the X direction is not fixed, other Settings are unchanged, so that under the action of the wind, the solid will move under the force of the fluid, by detecting the position of the object, and then calculating the acceleration of the object, it should be proportional.
The next two images are grid files in paraview, and you can see the motion of solids: the wind pushes things horizontally; Solid deformation is so small that it can be ignored

1s
1s1380×790 127 KB

2s
2s1386×782 314 KB

My solid solver is calculix, fluid uses openfoam, step size 0.01, and the observation point of my precice-config file is set to (0,0)

      <watch-point mesh="Solid-Mesh" name="Flap-Tip" coordinate="0;0" />

Down here is the precice-Solid-watchpoint-Flap-Tip file in the solid solver,
precice-Solid-watchpoint-Flap-Tip.log (27.7 KB)

I use displacement-0 in the column, the difference between the next row and the previous row to △x, and take the second derivative with respect to △x to get the acceleration a.

here I find a problem: isn’t displacement in the x direction equal to mass times acceleration? Isn’t the acceleration proportional to the displacement in the x direction? Why is the acceleration in the figure below negative?
In other words, if the acceleration is negative, the force must be negative. Why is the force always positive
acc-x

for-x

This diagram paraview draws the resultant force of the flap path

force-x
force-x1396×829 61.2 KB

2
  • This file is the solution procedure for force and acceleration
    force-acc.zip (47.5 KB)
    The inp file for calculix is as follows
    flap.inp (390 Bytes)
3

No, the sum of forces es equal to mass times acceleration.

Proportional is somehow the wrong word. Displacement and acceleration are related to each other via time derivatives, i.e., a larger acceleration will lead eventually lead to larger displacements.

Do you compute the derivative w.r.t. \Delta x or \Delta t? I’m not sure about the formula you use in order to approximate the derivatives here. Could you also provide a plot of the displacement (of which you compute the derivatives) and the formula you use?