While I was writing that last post I had to stop and convince myself that a course optimized to a point dead upwind would be the same as a course optimized to a point that was, say 25 degrees off the wind. After drawing a few vector diagrams, I managed to convince myself that the optimization would give the same result to either course. (Note well: convincing myself is NOT the same as a proof. I could be wrong!)

Here is my logic... While sailing up wind I can break my velocity vector into two component vectors: one straight upwind, and the other perpendicular to that. Any useful optimization program must maximize the upwind velocity vector and minimize the cross wind vector. No matter where my destination is in the upwind zone (between the tack angles) the vector perpendicular to the wind is taking me away from where I want to go, so minimizing that is always best. QED. (Maybe)

And I definitely hear you about avoiding beating! We just had an unusually delightful close hauled sail from Bar Harbor to Gloucester, but that is definitely the exception the the rule! I'd rather wait a week, or sail an extra 1000 miles than spend 3 days beating.

Bill Kinney

SM160, *Harmonie*

Gloucester, Mass

---In amelyachtowners@..., <svperegrinus@...> wrote :

Bill,

Thanks for the detailed explanation. My only remaining question is if under "W", the autopilot is trying to optimize VMG to an Apparent Wind Angle AWA of zero, or is it trying to optimize to a shallower angle, say, 10° or 20°. It seems to me that optimized VMG is not the same to a destination that is {0°, 10°, 20° or 30°} AWA of one's current position.

When under autopilot, we sail in "Wind" mode most of the time by far, so we'll try in the weeks and months ahead to see what we can learn of the VMG algorithm the B&G autopilot uses. The only problem is that we try and avoid beating like the plague, so it may take a while to accumulate observations on VMG-optimization.

Cheerio,

Peregrinus

SM2K N. 350

at anchor, Thermopylae