sailing efficiency meter

Recently I was given a mast head unit consisting of a wind direction indicator and a wind speed sensor.

Due to the need to take into account mast rotation and the fact that standard equipment with this option is, for me, hard to get and expensive, I decided to design and build my own.

The software and hardware is almost ready and I am considering if I should add a "sailing efficiency" option or not.

The need for this option comes about from the fact that many times one may feel he is sailing fast but when another boat zips past one realises that his speed is not so good after all.

Of course it is very difficult to always sail against competition and therefore the idea came up to display the ratio of COG against apparent speed weighed by a factor which reflects wind direction.

Does this idea have any merits or is it complete nonsense?

All suggestions will be appreciated.

regards

 

Sailing efficiency would probably be a function of true wind speed and your sailing speed. Problem is measuring the true wind speed while you're moving and the angle to the wind...

Be interesting to see what you used... I'm buzy designing a whole set of instruments for my cat.

 

My boat speed comes from a GPS and interfaced using RS232 and NMEA commands. From this, True wind and angle can be resolved.

For hardware, I am using a micro controller and a 160x128 lcd display board.

 

If you want to measure the efficiency of sailing through the water (as opposed to navigating), then you might consider the basic equations for sailing performance. Steady-state speed through the water is

Vb = Vt * sin(gamma - beta)/sin(beta)
Vb = boat speed through the water
gamma = course relative to the true wind (gamma=0 is head to wind)
beta = apparent wind angle, measured between course through the water and apparent wind (not quite the same as what's measured by the masthead unit)

So a key efficiency measure is the boat-speed/wind-speed ratio, Vb/Vt

Beta itself is one of the most important measures - the smaller beta is, the faster the boat is going. The problem is the measured apparent wind bounces around a great deal due to turbulence and movement of the mast-head in a seaway, making it difficult to perceive small differences. Also, the apparent wind angle used in the performance equation is the sum of the apparent wind angle relative to the centerline of the boat (measured by the masthead unit) and the leeway angle. It's rare to have a boat that measures leeway, so it's hard to get the true value of beta.

The COG element you mention includes current, so now it's becoming a measure of navigational performance as much as sailing performance. Have picked the right water to sail in, as well as sailing through the water as quickly as possible? Again, the lack of leeway complicates the issue, since leeway can look like an apparent current.

The most important navigational measure is, of course, velocity made good toward the goal. However, when going upwind, you need to us Vmg relative to the wind, not toward the mark, because any location along a line perpendicular to the true wind is equal sailing distance. But dividing by the true wind speed yields a nondimensional Vmg ratio that is an indicator of sailing performance. There may be a way of factoring in the current, too.

 

There already exists a set of sailing instruments that give VMG, which is "velocity made good to windward". The maths is definitely NOT rocket science but understanding algebra helps.

VMG is used when racing to the windward mark, which when you round the bottom mark should be directly to windward. Good racing sailors check this as they round the bottom mark then tack to the most favoured for performance (bearing in mind close competitors, perceived advantage of wind shift - observed and trend in wind)...

The idea is to maximise VMG, and at that point your sail handling and boat capabilities will limit the maximum value BUT IT DOES NOT MEAN POINTING HIGHER than the competition... You will be surprised at the advantage of easing off a little to gain better boat-speed whilst watching VMG...

 

As you say, the true sailing efficiency of the boat is Vb/Vt but since I have no idea of how to measure the speed of the current, COG will have to do. In any case, if you have ended up sailing in an area where you have a big opposing current, you may realise earlier that you are in the sh%^, if you see your efficiency going down.

The effect of the mast head movements on the actual measured values is also a concern and is as yet not catered for. This is also true for most instruments I have seen. In the future, I may try to tackle this issue using a low cost accelerometer chip which are starting to appear on the market.

VMG is of course another most important efficiency measure but to me it is more geared towards reporting your best efficient angle than to your best possible boat speed.

 

Thimk Guys, For sailors racing the key is VMG (velocity made good to windward) all other stuff -

Dealing with tidal and current course/direction is "good seamanship" and local knowledge - that will maximise your performance over ground.

The comparative advantage against another vessel in a race is for the tactics guy who KNOWS the racing rules and stuff to "force their right of way to advantage...

Performance improvement is relative so if your tactician & weather and current guy says hold this course then sail trim and watching an accelerometer and sensitive speed/log will be ALL that is needed - Many VMG instruments will also work in other points to the wind and give a plus or minus as to change in vessel velocity (acceleration)....

Don't get all confused with issues not relevant to sailing/hull performance (sail tuning)....

 

Real Data

Perhaps it would help to see what we're talking about. Here are some data I collected last Saturday during a beat up Sinclair Inlet in 14 - 16 kt of wind.

It ain't a pretty picture.

 

You guys are sailing a mono, I presume, To assess the performance of your boat you need to be able to sail consistently, so I would reckon at holding a series of tacks, alternating port & starboard tacks and each set to the same relative wind angle, and use your "paddlewheel" (not gps) to evaluate the best sail setting (max velocity and hold it) for a period of say 60 to 120 seconds or so after everything is "optimised" - record apparent wind angle (should be constant), apparent wind velocity - expect some variation from "puffs and lulls" - don't forget to trim sails constantly if wind velocity changes, and for skipper to steer to hold constant apparent wind angle. also record VMG and actual hull speed through the water

Eventually you will have a database of apparent wind speed/apparent wind direction and hull speed and have just the numbers for true wind speed / angle and then you should be able to go to the boat and say wind speed =x & velocity = Y therefore we should do z knots here and a knots there etc etc... any variation needs to be evaluated as to why - what was different in our settings etc

The object always being to drive your boat as fast as possible THROUGH the water... Tides and currents are a local knowledge thing, and performance evaluation of the boat is a through the water thing.

Do this for as many different wind velocities as possible as behaviour is different when in light airs and everywhere in between up to "lets stay ashore it may damage the rig etc... also sail on as many different wind bearings start with using max VMG to get best into the wind tacking angle (expect different for port & starboard) then work back say every 10 or 15 degrees. that will get you the usable performance data to optimise and tune your sailing performance.

If your boat is a cat - learn to sail... otherwise find out what your optimal designed hull speed is and work up to that - you cannot expect above that without regularly breaking stuff....

The idea is to see some consistent relationship with apparent wind angle / velocity and boatspeed... That is your baseline. maintain that at least in racing... and know you can do no better so lots of handicap wins should be seen/achieved..

Good sailing

 

tspeer,

Are you still working on this project? Maybe plot the data on a polar chart so it would be more like what people are used to seeing?

-jim lee

 

Oh, yeah, I'm still working on it. And will be for the next couple of decades!

I plotted the time history to show what the variability looks like in real data. Attached are the same data in polar form. I've normalized by the true wind speed to show boatspeed/windspeed ratios.

When I first plotted up these data, I was pretty discouraged. I've since had a chance to see how a very sophisticated team collects their performance data. Guess what - theirs are very scattered, too. Maybe not as bad as mine, but still quite scattered. It's just the nature of the thing.

 

Hi Tom
Wouldn't you get a closer grouping if you strictly sail on compass instead of on the wind ? Also how is it going to look if your boat is steered by a wind vein ?

Sorry, not suggesting you don't have a steady hand, but one does not react to and changes as instuments do.

If you calculate the ave of the gross, and disregard a certain deviation off that you should get a pretty accurate average.

I hope you don't shoot this bad when the pirates... Look, only 3 or 4 on the 90 deg

 

I don't have a wind vane, but I will be installing an autopilot and hooking it up to the instruments. So I guess I'll find out what things look like then.

I was mainly sailing by visual references, and not chasing either the compass course or the apparent wind that much. It turns out there's a lot of dynamic variation in the wind readings just from the motion of the boat. And this is a trimaran, so it doesn't move as much in heel as would a monohull.

True. I was always humbled when sailing the F24 upwind in steady conditions, and saw the speed drop by half a knot when I punched off the autopilot.

I'm also discovering that one just about needs to have a person dedicated to acquiring the data. So it's not very practical to try to sail for data with a crew of two. You need someone steering, someone handling the sails, and someone on the computer. But with an autopilot, it would become manageable with two.

There's a better way than simple averaging. I plan to develop a Kalman filter algorithm that I can run on the laptop. A Kalman filter provides a structure for handling everything you know about estimating the state of a dynamic system. It handles the fact that you know something about what to expect (a VPP), and the fact that the VPP is not perfect. It handles the fact that you have measurements of various kinds, and that the measurements have errors and random noise, too. For a linear system driven by Gaussian white noise, the Kalman filter is the optimal least-squares estimate of the system state.

There are five steps to the full-blown, time-varying Kalman filter. 1. Use the VPP to predict the measurements you're going to take, based on the previous estimate of the system state. 2. Use the VPP to predict the growth in the uncertainty in the estimates (covariance matrix; basically the standard deviation of the estimated state vector). 3. Compute an optimal weighting between the prediction and the measurements (Kalman gain matrix). 4. Update the predicted state estimate with the Kalman gain matrix times the difference between the actual and predicted measurements. 5. Update the uncertainty in the state estimates. As a practical matter, the filter is typically run in a simulation until the Kalman gain stops changing with time. Then fixed values for the gains are used, and only steps 1 and 4 need be implemented in real time.

I'm also looking into inexpensive instrumentation for measuring the boat motion and the position of various items. Did you know the Wiimote has a three-axis accelerometer sampled at 100 hz, an IR camera with 1024x768 resolution and hardware blob tracking, and Bluetooth communication? And up to 4 can be used simultaneously. With a few IR LEDs on the boom, tiller, and even sails, one could do some awesome things. Three widely separated Wiimotes can give you three-axis angular acceleration that can be integrated to get angular rates. Tracking LEDs can give you boom position, tiller position, mast rotation. You could put reflective tape on control lines and illuminate with IR to measure their position.

Now to find a good way to measure leeway angle...

 

Probably the easiest of them all to do judging by what you wrote there.

Angles are measured with a Clinometer. If you buy one they are probably going to rip you for it.

All you need is a pot (variable resistor like a volume control) with a weight that centers it down. If the boat heels you get a deviation to one side and that is easy to measure

I've had the same problem with sailing my little tri, if you don't concentrate all the time you drop off on speed.

What exactly are you looking at measuring ? I was under the impression just speed vs wind angle / aparent wind angle would be it.

Using a wind vein should have your chart almost dead on all the time if it works properly.

 

Oh, I was going to mention. You do get 360 deg pots, they can turn through 360 deg, mostly used for wind direction.

It is only the monohulls that would require the 360 deg pots for a clinometer