AR3-GP wrote: ↑06 Sep 2024, 22:33
Interesting. It would probably be useful if the model motion systems were more powerful, and they were allowed to run them at higher transition velocities (with a suitably robust model). You might discover some transient flow instabilities that you were not aware of. Given that the real car is fitted with pressure sensors and push rod load cells, I wonder what would be visible in that data. Do those sensors have a fast enough response time to capture such effects?
Sensors are not an issue, you can set sampling rate far above 0.001s, ie 1000Hz, but you don't really need this even if you want to capture transient effects.
Like I said, transition from state to state takes a lot less time than actually spending time in any said state on track. On top of that - if you have a separation that causes a significant drop in load at a static state, it won't matter when it happens. If you don't have any separation in static state in WT, it's very unlikely that you had it during transition. Once separated, the flow needs a lot of encouragment to reattach, like introducing some suction on the wall (in our case, bodywork) and as we know - that's very illegal in F1
This is why bouncing was and is such a big issue, you can have a non-aero related disturbance cause a vertical movement of the car that leads to critical separation and this then kickstarts the porpoising phenomena. The core of this issue is hysteresis between aero and suspension reaction, but also the fact that detached flow won't reattach at exactly the same geometric conditions of initial detachment - you have to get geometry well inside the "safe" zone for that.
Fluids, like everything else, follow the path of least resistance. If it takes more energy to detach, they stay attached. When you included the losses, the energy requirement of reattachment is higher than that of initial detachment (simplified explanation to the best of my own understanding)