Wind Shadow of AC36 Boats

[farjoe]

Most normal sailors have a pretty good idea of the area affected by the wind shadow of a normal boat sailing ahead?

But how does it look like for boats going at 4x wind speed?

Does the shadow fan out further when going from hard upwind to when running square?

 

[AlexanderSahlin]

I was going to start a new thread on this subject after viewing the AC-races, and now you have already done that farjoe, what a coincidence!
What we can see from the AC-races is that it seems very hard for the boat behind to pass, and that the distance usually increases during the race independently of which boat is leading.
There are other classes that sail 3 or 4 times the wind-speed. My experience is from skate-sailing, that is about as fast as the AC-75, -some 30 knots upwind and 40 knots downwind with top-speed slightly above 50 knots. What I have observed there is that the downwash behind another skate-sailor lives a very long distance behind that sailor, and is concentrated in a quite narrow path almost straight behind that sailor. However, in skate-sailing it is possible to overtake a sailor in front as long as you have just some small advantage in speed and keep out of that narrow path of downwash.
One difference between skate-sails and the AC-75 is that the skate-sail is just 3.4 m high above the ice, with an aerodynamic force at some 300 N and sails on a 1 km wide course. The AC-75 on the other hand is some 30 m above the water with an aerodynamic force around 30 kN and also sails on a space about 1 km wide. With the high apparent windspeed you get on the AC-75 they can get a substantial span-loading. Obviously, the AC-75 can fly in quite low wind, about 8 knots.
This make me think of the applying simplified wind-turbine theory on the wind-speed leeward of the path of an AC-75.
Assume that the AC-75 has an aerodynamic lift at 30 kN and sweeps over a cross-section area at 30 x 1000 m. This give a disc-loading at 1 N/sqm. If the true windspeed is 8 knots, this disc-loading corresponds to a lift-coefficient at 0.1 (based on swept disc area) and will reduce the windspeed downwind of the leading boats path by some 5 %. The fact that the leading AC-75 sails a little closer to the wind than 90° reduces this a little, but the reduction in windspeed is still in the same ball-park.
I hope that some reader with experience from CFD-analysis of wind-turbine parks can comment on this. Does the turbulent mixing level out the reduction in windspeed after 5 or 30 mast-heights?

 

[BlueBell]

Straight from the racers mouth the dirty air seems to come almost straight back from the leading boat.

 

[farjoe]

Straight from the races mouth the dirty air seems to come almost straight back from the leading boat.

That makes sense for boats going hard on the wind but what about time when the true wind is at 90 degrees to both boats?

Just because the windward boat is going so fast that it is experiencing an apparent wind coming from straight ahead does not mean that the leeward boat is in that same wind shadow. I think!!!

 

[BlueBell]

That makes sense for boats going hard on the wind but what about time when the true wind is at 90 degrees to both boats?

In lighter airs, they're always hard on the apparent wind.

 

[AlexanderSahlin]

That makes sense for boats going hard on the wind but what about time when the true wind is at 90 degrees to both boats?

Just because the windward boat is going so fast that it is experiencing an apparent wind coming from straight ahead does not mean that the leeward boat is in that same wind shadow. I think!!!

As I tried to tell in my previous comment, the downwash and tip-vortex, from a boat sailing 4 times the windspeed with the kind of load you have on an AC-75 in light airs, can live a quite long distance behind that boat and moves only slowly downwind. So you may fall into that wake of low wind quite a long distance after the AC-75 has passed. This probably happened in race 8 when ETNZ fell off the foil after a gybe. When trying to regain speed after the gybe, they luffed to 90° wind-angle and fell into the Italians and their own combined downwash from the previous tack. Have a look at 1.17 in this video:

This can be exploited if the leading boat sails several tacks between the edges of the course, near 90° true wind-angle, so they just avoid their own downwash from the previous tack. Then they will generate a massive field of substantially reduced windspeed that is impossible to pass for their competitor.

 

[tlouth7]

I note that the coverage now includes a wind shadow overlay (only in some shots, typically when one boat is close behind the other), though it is not clear on what basis this graphic is being calculated.

If you imagine a still air mass with the rig being pushed through it at some angle of attack and apparent speed then it is hard to see any reason that the wake should be different on different points of sail. I assume that AWA and therefore sideforce is pretty consistent on the two legs. Of course this wake will be moving down the course at the true windspeed which changes where exactly the rear boat encounters it.

Sorry if my messing around with reference frames confuses things.

 

[AlexanderSahlin]

I note that the coverage now includes a wind shadow overlay (only in some shots, typically when one boat is close behind the other), though it is not clear on what basis this graphic is being calculated.

If you imagine a still air mass with the rig being pushed through it at some angle of attack and apparent speed then it is hard to see any reason that the wake should be different on different points of sail. I assume that AWA and therefore sideforce is pretty consistent on the two legs. Of course this wake will be moving down the course at the true windspeed which changes where exactly the rear boat encounters it.

Sorry if my messing around with reference frames confuses things.

The AC-75 rigs are fairly aerodynamically clean, so it is reasonable to assume that they give the air, that flows past the sails, a change in momentum perpendicular to the apparent wind-direction and in the opposite direction to the aerodynamic lift. When the AC-75 is travelling several times the windspeed at a true wind-angle a little more than 90°, so the apparent wind-direction is perpendicular to the true wind, this change in momentum will result in just a reduction of the windspeed in the path that the AC-75 rig has passed. This zone with reduced windspeed will travel downwind a little slower than the true windspeed with a tip-vortex on top. If applying simple lifting-line theory on an AC-75 sailing with an apparent windspeed at 25 knots in 8 knots true wind, I estimate that reduction in true wind to some 1/3 of the true wind.
If the AC-75 on the other hand is sailing upwind close to the wind at a small true wind-angle the change in momentum for the wind that flows past the rig will mostly result in a change in direction of the true wind. This effect can in fact be experienced in classes that sail at lower speed and close to the wind. If you have a field of boats sailing on starboard tack and you sail through that field on port tack, this will give you a favourable wind-shift behind the starboard-tack boats.

 

[tspeer]

The wake trails at the apparent wind angle. This is due to the wake being convected with the true wind while simultaneously being generated at the boat's velocity.

Fast boats are sensitive to small changes in the true wind speed. If the boat is going 3 times wind speed and encounters a 1 kt loss of true wind due to sailing in another boat's wake, that is a 3 kt loss of boat speed!

The deflection of the air (wake wash) behind the rig extends laterally over a larger width than the reduced velocity due to the viscous wake. The wake wash constitutes a header wind shift for the following boat. This is what makes it difficult to remain on the hip of a boat to leeward.

If you want to get a sense of how far back the wake wash extends, take a look at this photo by Daniel Forster of the Volvo fleet leaving Cape Town in a fog bank:

 

[Mikko Brummer]

I added a simulation in Youtube around the topic:

Upwind, there is not that much difference to the sub-wind speed case, even if the bad air zone is wider and lasts longer behind the boat. Downwind, however, the bad air zone is totally different: When foiling at several times the wind speed, the wind shadow is left in the wake as a persisting wind shift, extending nearly right behind the boat and for a long time. There is no effect to the side or to downwind of the boat, as would be in the case of a traditional yacht, sailing at sub-wind speeds. You cannot speak at all of a wind shadow in the traditional meaning.

In both cases, the influence of the bad air on the wind speed is small, while the influence of the direction of the wind is significant, nearly 30 degrees.

The performance data courtesy of AC36 telemetry https://ac36.herokuapp.com/map. The simulation was performed by WB-Sails on Dassault Systèmes XFlow XFlow | High Fidelity CFD - Dassault Systèmes® https://www.3ds.com/products-services/simulia/products/xflow/.
Comments/critics/interpretations/explanations are most welcome

 

[AlexanderSahlin]

Nice simulations indeed. I am surprised that the reduction in wind-speed was only 3% at 45° true wind-angle. I would have expected some 20% reduction in the upwind case.
Did you also try 100° true wind-angle? There one can expect much more reduction in wind-speed as the major effect.
How fast is the wake-wash zone propagating downwind in your simulations?

 

[Mikko Brummer]

I am surprised that the reduction in wind-speed was only 3% at 45° true wind-angle. I would have expected some 20% reduction in the upwind case.

Well it is, or nearly, 16% closer by and 8% further away. My reference velocity was 5.144 m/s, or 10 kn. In the video, I think the lower scale was at 4.9 kn, so all the blue is not really the lowest value there. Attached probing in the velo field shows how you are right.

 

[Mikko Brummer]

Did you also try 100° true wind-angle? There one can expect much more reduction in wind-speed as the major effect.

Not yet - I would like to do a quick bear away - jibe into one's own bad air, which was seen to be a bad manoeuvre in the races, and also an attempt of "hooking" to the leeward aft corner of the opponent before the start - a standard match racing manoeuvre that never worked in this cup.