Reducing the number of state variables

Try to limit the number of state variables in your model as this has a direct influence on the CPU time. For an “easy” model (Adams methods are used) there is roughly a linear variation, but for stiff models (BDF methods are used) a Jacobian matrix must be evaluated. The size of the Jacobian matrix is equal to the square of the number of state variables in your system: 200 state variables in a stiff model, for example, would be quite a high number.

Figure 499: State variables

In order to limit this number, you might, for example, replace several inertias with one inertia when possible.