Failure for tutorials/perpendicular-flap with solid-calculix and fluid-fake for time-window-size=0.00125

[BenjaminRodenberg]

I ran the perpendicular flap case with the fluid-fake (precice/tutorials#472) participant and solid-calculix. This resulted in the following error:

---[precice]  iteration: 1 of 50 (min 1), time-window: 4000, time: 5 of 5, time-window-size: 0.00125, max-time-step-size: 3.63265e-13, ongoing: yes, time-window-complete: no, 
 initial energy (at start of step) = 0.000000e+00
...
 Using up to 1 cpu(s) for the energy calculation.

 average force= 7.964573
 time avg. forc= 15.686676
 largest residual force= 0.000000 in node 311 and dof 1
 largest increment of disp= 0.000000e+00
 largest correction to disp= 0.000000e+00

 convergence

Adapter writing coupling data...
Writing DISPLACEMENTS coupling data.
Adapter calling advance()...
---[precice] ERROR:  advance() cannot be called with a time step size of 0.

Expected behavior:

The adapter should work for any time step size.

Further information:

1. Using preCICE v3 (minor modification, but this should not be relevant)
2. no problem for fluid-fake + solid-fenics. Therefore, I assume it is an issue of the CalculiX adapter.
3. I found it peculiar that the problem appears only in the very last time step for <max-time value="5.0" />. It does not happen, if I set the <max-time value="1.0" />.
4. ⚠️ In (3.) I forgot to update the max time in flap.inp. One should generally also update the max time in *DYNAMIC, ALPHA=0.0, DIRECT to 1.E-2, 1.0 (second value is max time). However, I do not observe a relevant difference.
5. If I set the <max-time value="6.0" /> (because 5.0 was problematic before), I see no error.

I'm not sure how I acquired this superpower to find strange errors. To me it seems like the problem is very specific for <max-time value="5.0" />.

precice-config.xml (please change extension to .xml when downloading)