Necessity of TangDis fallback in ConstDisVelAccPress predictor and importance of predicting velocities and accelerations in the static time integration

[danielwolff1]

While implementing a new predictor, I was wondering about the necessity of having a TangDis fallback predictor when employing a predictor of type ConstDisVelAccPress.

The TangDis fallback predictor is only triggered in this line which according to the coverage report basically never happens. But it has a lot of implications, also for other classes, i.e., the different kinds of time integrations: The return values of the time integrations' methods predict_const_acc() and predict_const_vel_consist_acc() have been changed from void to bool, even though they almost always return true.
The only exception to this is the time integration class for static simulations Statics. Here, the return values indicate whether enough history information is available to reconstruct meaningful velocity and acceleration vectors and then the TangDis fallback predictor is only used if this is not the case.
However, this brings up the question of how meaningful the definition of global velocity and acceleration vectors in a static simulation is, which is why I wanted to ask here, whether anyone knows about a case where vel_n, acc_n, vel_np, or acc_np are meaningfully used in stationary simulations?

If not, I would propose to remove the TangDis fallback predictor from the ConstDisVelAccPress class and change the return values of the predict_const_acc() and predict_const_vel_consist_acc() methods in the time integration classes back to void which would then again be consistent with the signature of predict_const_dis_consist_vel_acc().