It has been very quiet on this channel recently, but things have been happening behind the scenes. The working group of Rob Prince, Chris Gallienne, David Tyler, and myself have spent a lot of time and many emails discussing what equipment to buy, and concluded that ultrasonic speed detection was the most accurate. Consequently we have purchased a DX900 boat speed, leeway, and water temp sensor, from Airmar, and an LC7 wind and direction sensor from LJ Capteurs.

The boat speed is said to be accurate to 0.01 knots, and the wind speed to 0.25 knots. Wind direction accuracy is to 1.5 degrees, which is the best we could find within the budget. We have developed an instrument mounting "rig" which can be moved from boat to boat and which has the boat speed sensor in a faired strut or "paddle" and the wind sensor on a mast, both of which can be "tacked" fairly easily - we hope.

We have bought an HP X2 hybrid tablet/laptop for its extended battery capacity, and because we wanted a windows based system to use the free polar diagram software "Polauto".

To get the data into the computer requires a multiplexer to receive the NMEA 0183 signals and combine them for feeding to the computer. Here we hit a snag because the first unit we tried, a Quark A031, we could not get to function properly. After Chris Gallienne and I both tried and failed, we sent it back, and we are now getting a unit from V-Boat, a Swedish company, which we hope will arrive this week.

This has meant we have missed the opportunity to measure Weaverbird and KRKA in Falmouth which was scheduled for this week. We will reschedule when we have the system up and running.

It may have been quiet on the surface, but we have been paddling like crazy underneath.

Arne,

Correct on both counts. The ultrasonic sensors are more accurate and have better resolution than paddlewheels, which can only register one pulse per revolution. On a fin keel boat, the best position is 300 - 600mm forward of the keel. Ultrasonic sensors do need bubble-free waterflow, and this gives the best chance of achieving that.

Arne Kverneland wrote:

Better disregard GPS, but keep measuring the leeway angle

Arne

It is available at a price within the budget agreed by the JRA committee, so I'm inclined to think it's the way to go. It can be retrofitted into some of the skin fittings used by some existing paddlewheel logs, but not all. It does mean that a prospective test boat owner must be prepared to fit a compatible skin fitting, as this is the only sensible way to install it with an accuracy of one degree. Without that accuracy, reliable leeway data cannot be gathered. 

I think Eric has an interesting point here. In fact, to get the true wind from the apparent wind you need the speed and direction of movement of the boat RELATIVE to the WIND, so you should ideally take into account leeway, and current, although you can argue the current is already taken care of in the apparent wind speed and direction.

The apparent wind direction is, of course, measured relative to the yacht's centreline, which is the HEADING, but this is not necessarily (or usually) the direction of the boat's motion relative to the water, thanks to leeway.

The simplest and traditional approach as suggested by Paul is to measure apparent wind speed and direction, and boat speed, and to ignore leeway, or make a correction for an assumed leeway, which is what we did in India.

In theory, Eric's approach could potentially be more accurate, as it takes leeway into account more or less instantaneously, and leeway is really tricky to measure on board because the angles are so small, and there is so much noise from the boat's motions.

The problem with Eric's approach is that it brings the current into the mix, and it is difficult to take it out again, because you can't tell which part of the COG difference from the heading is due to leeway, and which part due to current. It also requires accurate and simultaneous compass readings

At the moment, we are taking the simple approach, and aiming to record apparent wind speed and direction, and boat speed, as accurately as we can. We are considering ways to measure leeway as well, but it is tricky and we may just make the crude assumption that the boats being tested are all roughly similar types which can be expected to make similar leeway. We might use GPS to try and observe some typical leeway figures when we manage to find some slack water. Then we can correct the results for assumed leeway, or not, as we choose.

We have calculated that provided the wind speed is greater than the boat speed, the change in the true wind angle for leeway of 5 degrees (on the large side) is less than 2.5 degrees, and if the wind speed is twice the boat speed the error is less than 1.53 degrees, so the error created by ignoring leeway is not huge, and about the same as the accuracy of the best wind instruments (we can afford) which is +/- 1.5 degrees.

I could go on... but I won't!

Bonjour

I don't understand it the same way. In modern navigation and autopilot systems the true instantaneous true wind (speed and direction) is computed out of the instantaneous apparent wind provided by the anenometer and wind direction indicator and the boat speed provided by the GPS to have a very accurate true wind data. On the most sofisticated systems they use a strapped  down  laser inertial platform (as on the airliners) to correct instaneous airspeed values from the mast movements.

Remark : In case of strong tide stream it may induce a small error on the true wind value.

Then the polar is computed out of the boat speed (over watter) measured by a speedometer;  the heading, measured by a compass (usually fluxgate), and the true wind data.

Eric