The designers overview, by Malcolm
SAILROCKET CONTROL PROBLEMS
I developed a computer simulation model of Sailrocket in 2004 before she first sailed in order to evaluate control and response to gusts and pilot inputs. This was used briefly to check out the soft rig behaviour but there were no major control issues with this rig so we didn't dwell on it much. In fact we did experience a rapid spin out into wind on one occasion and thought the rudder system had failed.
As we knew the wing was a very different beast we resurrected the 'sim' and looked at sheeting sequences for building speed , response to gusts and sudden sheeting adjustments. It all looked promising with only very moderate rudder loads of 30-50 kg being needed. The one cause for concern highlighted was the very small rudder angle adjustments needed.
After the first crash at Walvis, aware of this sensitivity issue, we built a smaller secondary rudder to deploy at high speed and focussed on rig adjustments to eliminate the weather helm tendency we thought was causing the spin out. As control seemed pretty good in runs 05 and 06 we thought we were in good shape for a 40 knot run.
In reality we had not understood the full picture.
Chris Hornzee Jones , designer of the Sailrocket wing was first to finally identify the real problem by looking at the issue from Paul's perspective and working back through the observations and sim predictions.
Chris concluded that the boat is directionally unstable and at high speed becomes uncontrollable with realistic pilot inputs and reaction times. The instability is mainly due to the changing orientation of the wing force with apparent wind angle combined with large lateral wing to hydrofoil separation.
The boat response only looked good in the simulation because I had effectively built in an idealised precision 'computer pilot' which was flying the boat like an unstable jet fighter!
What Chris has proposed as a primary solution is a fixed fin or skeg at the back to provide yaw damping and reduce the directional instability. Adding the fixed skeg to the sim has already confirmed the boat will be much more stable and tolerant to sudden rudder inputs.
It remains to design the skeg and any changes to the rudder system in detail, based on achievable human pilot inputs, to provide the margins of control needed to push Sailrocket safely to the next level.
There may of course be other contributing factors such as rigging stretch, ventilation of foils, etc and we intend to investigate these in the coming weeks as well to rule out any more nasty surprises.
Malcolm
I developed a computer simulation model of Sailrocket in 2004 before she first sailed in order to evaluate control and response to gusts and pilot inputs. This was used briefly to check out the soft rig behaviour but there were no major control issues with this rig so we didn't dwell on it much. In fact we did experience a rapid spin out into wind on one occasion and thought the rudder system had failed.
As we knew the wing was a very different beast we resurrected the 'sim' and looked at sheeting sequences for building speed , response to gusts and sudden sheeting adjustments. It all looked promising with only very moderate rudder loads of 30-50 kg being needed. The one cause for concern highlighted was the very small rudder angle adjustments needed.
After the first crash at Walvis, aware of this sensitivity issue, we built a smaller secondary rudder to deploy at high speed and focussed on rig adjustments to eliminate the weather helm tendency we thought was causing the spin out. As control seemed pretty good in runs 05 and 06 we thought we were in good shape for a 40 knot run.
In reality we had not understood the full picture.
Chris Hornzee Jones , designer of the Sailrocket wing was first to finally identify the real problem by looking at the issue from Paul's perspective and working back through the observations and sim predictions.
Chris concluded that the boat is directionally unstable and at high speed becomes uncontrollable with realistic pilot inputs and reaction times. The instability is mainly due to the changing orientation of the wing force with apparent wind angle combined with large lateral wing to hydrofoil separation.
The boat response only looked good in the simulation because I had effectively built in an idealised precision 'computer pilot' which was flying the boat like an unstable jet fighter!
What Chris has proposed as a primary solution is a fixed fin or skeg at the back to provide yaw damping and reduce the directional instability. Adding the fixed skeg to the sim has already confirmed the boat will be much more stable and tolerant to sudden rudder inputs.
It remains to design the skeg and any changes to the rudder system in detail, based on achievable human pilot inputs, to provide the margins of control needed to push Sailrocket safely to the next level.
There may of course be other contributing factors such as rigging stretch, ventilation of foils, etc and we intend to investigate these in the coming weeks as well to rule out any more nasty surprises.
Malcolm
posted by Paul at 2:42 pm
0 Comments:
Post a Comment
<< Home