Add time derivative Jacobian tests

[AntoniaBerger]

No description provided.

[jbreue16]

The time derivative Jacobian tests did not catch a missing alpha factor in the CSTR, which was fixed in #310

We could extend the time-derivative Jacobian tests to also check for alpha.
It might seem trivial since we would check for $\alpha \cdot J$ instead of just $J$, but since this already lead to an error we've decided to catch this.
We cannot use AD for this check since AD units use the same addTimeDerivative function for that part of the Jacobian.
So we need to adapt the FD Jacobian check for $\alpha != 1.0$

Ive also cleaned up some repetitive code in the CSTR multiplyDerivativeJacobian function, where we now use the addTimerDerivative function

[jbreue16]

I realized that the multiplyWithDerivativeJacobian functions dont call the addTimederivativeJacobian function because we dont want to use (sparse) matrix vector multiplication here since the time derivative jacobian is even more sparse (almost a diagonal matrix). So we should probably revert the changes of commit 69422bb, even though its probably not performance relevant for the CSTR.

We need a whole different, dedicated test for addTimeDerivativeJacobian

[AntoniaBerger]

Why not modify the old multiplyWithDerivativeJacobian function and use the matrix from addTimeDerivativeJacobian?

[jbreue16]

Why not modify the old multiplyWithDerivativeJacobian function and use the matrix from addTimeDerivativeJacobian?

because of efficieny: we dont want to use (sparse) matrix vector multiplication here since the time derivative jacobian is even more sparse (almost a diagonal matrix)

[AntoniaBerger]

Due to more time-sensitive tasks, we will not be working on this issue in the near future and have therefore decided to close this PR without a commit.
To keep track of this problem see issue #330