Why is blade tip-to-tower closest approach not always exactly the minimum of the "Blade tip to tower surface" output value?
Clearly the “closest approach” should correspond exactly to the smallest value so far of the tip-to-tower instantaneous value. But sometimes simulations give results like the following, where the closest approach seems not to follow the instantaneous value very closely.
Overview: [black line is blade tip to tower surface, red line is closest approach]
Close-up showing the problem in detail: [same colours]
This is an artefact in the output time series, caused by the calculated values not being exactly in sync with the output values. In fact this value is being calculated correctly. You can confirm this by using a fixed-step integrator and ensuring that the output timestep is the same size as the integrator timestep.