Minimum Wind Polars

[mglonnro]

Greetings!

-- PREAMBLE RANT --

Let me start by saying that I love data and I think the whole sailing market would benefit from getting more of it - especially data that paints realistic pictures of sailing performance, and all of its different aspects.

Anecdotes are cool as well, but sometimes hopelessly unreliable. Sailing magazine reviews bring some sailing facts into the equation, but usually they are valid only for a very limited set of wind conditions - whatever the weather happened to be when they did the test sail.

So, it would be great to see more real life data about everything, partly produced by manufacturers, but mostly by people actually sailing the boats.

-- PREAMBLE RANT ENDS HERE --

In addition to the displacement, sail area, etc figures, the most usually available sailing performance data comes in the form of manufacturer produced (and/or ORC/VPP calculated) polar diagrams.

These polar diagrams show resultant max speeds for different wind conditions. I thought it would be nice to see the inverse, ie what different wind conditions are required to achieve a target speed?

The question I was initially trying to answer was not "how fast will this thing go with wind speed X from angle Y?" but "how much wind did we need to go at least speed S".

Here's an example picture showing our last season data and what the wind conditions were when we were going at least three knots (but less than four):



Looking at the picture, it seems that getting 3+ knots of boatspeed upwind needed at least 6 knots of wind. Beam reaching a bit less (4-5 knots) and broad reaching again at least 6.

Here's another:



For close hauled upwind (port tack) we did 6+ knots with 12 knots of wind. Just a few dots, though, so frequently we didn't go that fast. When reaching, we usually needed at least 10 knots of wind for 6 knots, except for one area around 90-100 where we needed just 8 knots (maybe due to crossover sail hoisted, don't know).

So, wind polars!

I think they show useful information. Do you agree?

(The full blog post I wrote can be found clicking the link in the signature!)

[Tupaia]

As demonstrated by your diagrams the results are "all over the place". This primarily because although the wind strength might be "the same", temperature, pressure, humidity, current and sea state will all be variable within the data.

Just as an example 40knots of wind will have more force in Norway than on the Equator, simply because the air will is denser in higher latitudes.

I generated a stack of polars using OpenCPN in apparently the "same" conditions and they were vastly different. Sail trim will never be identical in similar conditions on different days.

Expecting manufacturers to produce accurate polars is a nice idea but will always be wrong and I suggest that they would be disappointing for buyers.

There are many factors that need to be accounted for when assessing lightwind performance. To achieve the boat must have sufficiently high power (SA) to weight ratio (Disp) fine hulls (>1:12) smooth aerodynamic profile (drag) and be able to point well to windward dagger boards. The faster you go relative to the true wind the further forward will be the apparent wind. In light winds in the right conditions a fast multihull will be sailing close hauled downwind effectively creating it's own wind.

[arsenelupiga]

it is useful but, the main issue with your approach is having speedo calibrated in my opinion. Have found that speedo reading changes even on daily basis, depending maybe on water temperature.... and after couple days one gets growth and this again changes speedo readings. The other issue is wind instrument calibration and precision. And the third issue is wind shear.

i am too lazy to calibrate often & clean so, use different approach.

For given wind speed & wind angle & sea state, boat speed is certain ratio boat speed/apparent wind. Found that with increased wind speed ratio falls, ie doubling wind speed will not double boat speed. Also use ears to listen water flow to determine whether my sail setting is at its best or not. that is interesting topic that always enjoy exploring when sailing albeit with a fat cat.

[Tupaia]

Here are some of my polars light wind different sea state. Different days over a 4 day period so instrument variations, pressure, temperture not accounted for.


You will see that the Bumpy seas has gusting upto 12kts, Smooth seas has occasional gust to 10kts. So taking a slice through all three at 8kts yields considerbly different figures. You will note Bumpy seas is a technical term.


Light wind polar bumpy seas
TWA\TWS 0 2 4 6 8 10 12
0 0 0 0 0 0 0 0
70 0 0 0 0 0 7.73 4.68
75 0 0 0 0 7.98 7.5 6.79
80 0 0 0 0 6.81 7.98 7.26
85 0 0 0 5.84 6.31 7.73 8.4
90 0 0 0 0 6.08 7.97 7.97
95 0 0 0 0 5.61 8.65 7.95
100 0 0 0 0 7.01 7.15 7.73
105 0 0 0 0 7.03 6.09 7.5


Light wind polar smooth seas
TWA\TWS 0 2 4 6 8 10 12
0 0 0 0 0 0 0 0
60 0 0 0 4.92 5.86 0 0
65 0 0 0 5.39 6.31 0 0
70 0 0 0 6.09 6.79 0 0
75 0 0 0 7.64 7.48 0 0
80 0 0 5.39 7.5 7.5 0 0
85 0 0 0 7.03 7.5 0 0
90 0 0 0 7.48 7.73 0 0
95 0 0 0 7.73 7.73 0 0
100 0 0 0 7.5 7.73 0 0
105 0 0 0 7.5 0 5.38 0


Light wind polar moderate seas
TWA\TWS 0 2 4 6 8 10 12
0 0 0 0 0 0 0 0
75 0 0 0 0 6.08 5.84 0
80 0 0 0 5.86 6.78 0 0
85 0 0 0 6.32 7.03 6.08 0

[mglonnro]

Originally Posted by Tupaia

As demonstrated by your diagrams the results are "all over the place". This primarily because although the wind strength might be "the same", temperature, pressure, humidity, current and sea state will all be variable within the data.

Just as an example 40knots of wind will have more force in Norway than on the Equator, simply because the air will is denser in higher latitudes.

I generated a stack of polars using OpenCPN in apparently the "same" conditions and they were vastly different. Sail trim will never be identical in similar conditions on different days.

Yes, absolutely! I completely agree that there are a huge number of variables and everything is quite dynamic. That's why this isn't an attempt to model anything (like the manufacturer's polar diagrams do), but rather an attempt to plot (a lot of!) real life data and then have a look at it.

I do see patterns, clusters, shapes already in these plots and that's what I'm looking for.

Expecting manufacturers to produce accurate polars is a nice idea but will always be wrong and I suggest that they would be disappointing for buyers.

Yes, agree here as well. I do expect them to produce some baseline, though, AND several versions of it with different sails. I also expect them to publish the base parameters for the polar models (starting with "what displacement").

What I'm really looking for, as I wrote a few lines earlier, is not to model anything, but to plot a lot of data and see where it takes me.

There are many factors that need to be accounted for when assessing lightwind performance. To achieve the boat must have sufficiently high power (SA) to weight ratio (Disp) fine hulls (>1:12) smooth aerodynamic profile (drag) and be able to point well to windward dagger boards. The faster you go relative to the true wind the further forward will be the apparent wind. In light winds in the right conditions a fast multihull will be sailing close hauled downwind effectively creating it's own wind.

[mglonnro]

Originally Posted by arsenelupiga

it is useful but, the main issue with your approach is having speedo calibrated in my opinion. Have found that speedo reading changes even on daily basis, depending maybe on water temperature.... and after couple days one gets growth and this again changes speedo readings. The other issue is wind instrument calibration and precision. And the third issue is wind shear.

i am too lazy to calibrate often & clean so, use different approach.

For given wind speed & wind angle & sea state, boat speed is certain ratio boat speed/apparent wind. Found that with increased wind speed ratio falls, ie doubling wind speed will not double boat speed. Also use ears to listen water flow to determine whether my sail setting is at its best or not. that is interesting topic that always enjoy exploring when sailing albeit with a fat cat.

Yes, exactly. The calibration bit is a challenge. No immediate solutions there ...

[mglonnro]

Originally Posted by Tupaia

Here are some of my polars light wind different sea state. Different days over a 4 day period so instrument variations, pressure, temperture not accounted for.

You will see that the Bumpy seas has gusting upto 12kts, Smooth seas has occasional gust to 10kts. So taking a slice through all three at 8kts yields considerbly different figures. You will note Bumpy seas is a technical term.

I think that is really interesting data. It gives rough indications of what one might expect with a boat like yours in similar conditions.

Different sailboat models might probably behave quite differently in 'bumpy seas' compared to smooth seas, but this is not something any manufacturer publishes data of.

[Suijin]

This is all familiar territory if you're into advanced racing, where winning and losing can be a matter of fractions of a knot of boat speed and all variables are considered obsessively in the search for more speed. Sail condition and weight are in some respects the primary factors outside of the skill sphere of trim, helming, tactics and maneuvers.

Weight, particularly in light winds, is a HUGE factor. In a light wind race you reduce as much weight as you possibly can and the difference on a performance boat can be enormous.

I think the scatter in your plots is due primarily to sea state, air density, and perhaps displacement depending on whether that changes much from data point to data point.

[thinwater]

Sorry, I disagree.


Polars should reflect the BEST you can do in a given set of conditions, not what happened with a full cruising load and poor sail trim during a gust when the boat had not yet accelerated. 8-18 knots to reach 6 knots? That's historic data, but history of what? We don't know. If we don't know, it isn't usable data.



I've put together polars before, and given a set sea state, ANY racer will expect to be much closer to his targets than that. If you are off by a fraction of a knot you find the reason. If you are off by 2 knots... well, you just aren't in the range of speeds discussed.


Polars are only going to be data if the conditions are stated and limited so that another person can reproduce them. That is the nature of science.

[mglonnro]

Originally Posted by Suijin

I think the scatter in your plots is due primarily to sea state, air density, and perhaps displacement depending on whether that changes much from data point to data point.

I think momentum as well? Wind is really turbulent but the boat is constantly accelerating/decelerating towards an equilibrium that already happened...

[mglonnro]

Originally Posted by thinwater

Sorry, I disagree.

No need to be sorry! Where would the world be without disagreement

Polars should reflect the BEST you can do in a given set of conditions, not what happened with a full cruising load and poor sail trim during a gust when the boat had not yet accelerated. 8-18 knots to reach 6 knots? That's historic data, but history of what? We don't know. If we don't know, it isn't usable data.

Ah, that's a good observation. I was actually just trying to find the minimum by looking for lightest winds where the scatter turns dense enough to not be just a few outliers.

Why did we also do a lot of 6 knots with double the wind speed? That's a good follow-up question that I don't knot the answer to. Acceleration, deceleration, sea state, temperature, a lot of other variables are in the data, though, so it might become more evident when looking at those.

Sail trim is unfortunately not in the data, so in addition to sea state, I assume a lot of the worse data points are due to poorly trimmed sails!

I've put together polars before, and given a set sea state, ANY racer will expect to be much closer to his targets than that. If you are off by a fraction of a knot you find the reason. If you are off by 2 knots... well, you just aren't in the range of speeds discussed.

I think that is a different use case-- being close to your maximum polar speed as much as possible.

My polar data attempts to show: at what minimum wind speed does the boat regularly move at least X knots. And this IS cruising mode including everything that goes with that.

If I tidied up the chart a bit, made some educated means and drew lines instead of sampled dots and it would show about 90-94% of Hanse's published polars (them having more downwind sail area).

Polars are only going to be data if the conditions are stated and limited so that another person can reproduce them. That is the nature of science.

They buy a Hanse 388 with the same options as we have, pack it with six similarly sized people, a dog, our food and drink (some wine, not too much of it!), and then they take a time machine to go sailing in the Baltic Sea last summer.

Easy!

[billknny]

Yes, polar data is typically all over the place. But that is kind of the point of collecting it.

Under any set of conditions it is really helpful to know the FASTEST the boat can go so you know if you are sailing the boat to its full potential.

A great deal of the scatter comes from sailing in unsteady winds, so applying a lot of filtering to the data is important. Looking for places where the TRUE wind speed AND wind angle AND STW have been (reasonably) stable for several sequential data points will help a lot in generating data that is consistent and makes sense..

For those of us who do long passages, accurate polar data is critical to feed into a weather routing program. Without accurate polar data, it is very much Garbage IN-Garbage Out. Especially important are the data points around close hauled and downwind. REALLY knowing the wind angles at various wind speeds that give you your best VMG is important. Unless you are REALLY good with LOTS of racing experience doing it by "feel" is very unlikely to get you upwind as quickly as you can go.

Looking at a whole season's worth of data at once, is to me, not useful. I develop my polars a bit at a time, collecting data with identical sail plans, and sea states. An hour's data is usually enough to add a couple data points to my polar plots. I have a ketch, so between reefing the three white sails, a downwind rig, and a mizzen staysail I probably have a 100 possible sail combinations, most of which we'd never use, but there are still a LOT that are useful and practical at some wind speed and angle. It makes for a complex graph. Without some way of keeping track of the performance of the various combinations, I'd just be guessing what was best to use.

[Montanan]

My polars vary A LOT, depending on how much beer has been brought on board, how many bottles of beer have been taken down by the crew and skipper, how much effort has been taken to establish towards optimizing trim, reefing / unreefing, and choice of sails, and of course sea conditions and current speed and direction all come into play, etc.

When we be out cruising there is very little time spent monitoring wind speed, subtle shifts in direction of wind or boat, adjusting trim, or VMG.

I suppose one could diagnose my boat and its crew as exhibiting bi-polarism, or bi-polar disordered states of sailing.

Bipolar disorder is a mental disorder that causes unusual shifts in mood, energy, activity levels, concentration, and the ability to carry out day-to-day tasks.

There are three types of bipolar disorder. All three types involve clear changes in mood, energy, and activity levels. These moods range from periods of extremely “up,” elated, irritable, or energized behavior (known as manic episodes) to very “down,” sad, indifferent.

[mglonnro]

Originally Posted by billknny

Yes, polar data is typically all over the place. But that is kind of the point of collecting it.

Under any set of conditions it is really helpful to know the FASTEST the boat can go so you know if you are sailing the boat to its full potential.

Yes, sure! What I've done is basically the same thing, though, just plotted wind speed on the circles instead of boatspeed.

A great deal of the scatter comes from sailing in unsteady winds, so applying a lot of filtering to the data is important. Looking for places where the TRUE wind speed AND wind angle AND STW have been (reasonably) stable for several sequential data points will help a lot in generating data that is consistent and makes sense..

For those of us who do long passages, accurate polar data is critical to feed into a weather routing program. Without accurate polar data, it is very much Garbage IN-Garbage Out. Especially important are the data points around close hauled and downwind. REALLY knowing the wind angles at various wind speeds that give you your best VMG is important. Unless you are REALLY good with LOTS of racing experience doing it by "feel" is very unlikely to get you upwind as quickly as you can go.

Looking at a whole season's worth of data at once, is to me, not useful. I develop my polars a bit at a time, collecting data with identical sail plans, and sea states. An hour's data is usually enough to add a couple data points to my polar plots. I have a ketch, so between reefing the three white sails, a downwind rig, and a mizzen staysail I probably have a 100 possible sail combinations, most of which we'd never use, but there are still a LOT that are useful and practical at some wind speed and angle. It makes for a complex graph. Without some way of keeping track of the performance of the various combinations, I'd just be guessing what was best to use.

If you are looking for just maximum speed, then yes, of course the whole season's data won't make much sense. If you are, on the other hand, looking to visualize what the season's sailing looked like -- how much of the time did you spend at only 90% of planned polar speeds, how much time even less than that, then it - in my opinion - does make sense.

Maximum is maximum. Minimum is minimum. Average is ... well yes. And the whole shebang is the whole shebang!

Here's a version of the 6+ knots visual where:

- speed trap is now 6.0 - 6.8 knots (6.0 = blue color -> gradient 6.8 = yellow)
- average line is average of the lowest 60% wind speeds when speed was between 6.0 - 6.3 knots.

[mglonnro]

Originally Posted by Montanan

My polars vary A LOT, depending on how much beer has been brought on board, how many bottles of beer have been taken down by the crew and skipper, how much effort has been taken to establish towards optimizing trim, reefing / unreefing, and choice of sails, and of course sea conditions and current speed and direction all come into play, etc.

When we be out cruising there is very little time spent monitoring wind speed, subtle shifts in direction of wind or boat, adjusting trim, or VMG.

I suppose one could diagnose my boat and its crew as exhibiting bi-polarism, or bi-polar disordered states of sailing.

Bipolar disorder is a mental disorder that causes unusual shifts in mood, energy, activity levels, concentration, and the ability to carry out day-to-day tasks.

There are three types of bipolar disorder. All three types involve clear changes in mood, energy, and activity levels. These moods range from periods of extremely “up,” elated, irritable, or energized behavior (known as manic episodes) to very “down,” sad, indifferent.

I'm diagnosing that this sounds like good sailing!