Time Derivative Kernels for Array Variables

[zachmprince]

Reason

This issue is an extension of the closed issue #6881. Currently, there is no Array Kernel version of TimeKernel or TimeDerivative. Having this would allow the time derivative residual contribution of multiple variables to be computed in the same kernel.

Design

The design would mirror that of TimeKernel and TimeDerivative. One important distinction would be to allow the time derivative of the array variable components to be coupled, i.e.:
$(\mathbf{T}\dot{\vec{u}}i = \sum{j=1}^{N}T_{ij}\dot{u}_j$.
Where $N$ is the number of components in the array variable and $\mathbf{T}$ can be a scalar, a diagonal matrix, or a full matrix. The implementation would be similar to ArrayReaction and ArrayDiffusion.

Impact

This implementation would reduce the number of kernels generated with systems of equations with many coupled variables, namely in radiation transport.