Sailing Performance Trials - 2018

This page contains information about and access to the results of the JRA Sailing Performance measurement project. To see the history of this project, please look at the Performance Measurement thread on the Technical Forum or the article(s) in the Magazine.

So far we have run trials on four boats. Details of these boats can be found on the page "Details of Tested Boats" and here

Results comparing the performance of the four boats tested so far can be found on the "4 boat comparison polars" page

The raw data from the trials can be found here.

There are separate folders for each boat tested.
The data is available in the following formats
1/ as .txt files containing the NMEA sentences from the wind and water sensors recorded during the trials
2/ as .csv files containing the boat speed and apparent wind speed and direction every second

In the data folders are separate files for each test run and for port and starboard tacks. There are up to 8 runs on each tack for each boat. To simplify processing the text files to produce polars in Polauto, the data files for each run on each tack on each boat have been combined into a single file called "Name of Boat P (or S) combined". In combining these files any NMEA sentences before the first boat speed sentence have been deleted to prevent wind data from one file being used with boat speed data from another file. The deleted data would not be used anyway as there is not boatspeed value for it to be used with.

Polar Diagrams using the Median Boat Speed  and True Wind Speed and Direction can be seen on the page "Polars (Median Boat Speeds)". These were created by Robert Self from the .csv raw data files.

Polar diagrams using the Maximum speeds recorded in the raw data using Apparent Wind speed and direction which have been created by Polauto7 can be seen on the page "Polars (Apparent Wind)" and more are stored here.

There is a key to the colours used on the Apparent Wind polar diagrams on th2 Polars (Apparent Wind) page and also here
In the data files on Dropbox there are separate polar folders for each boat tested. In some there are additional folders for trial polars that were produced while testing various filter options in Polauto. Ignore those unless you are interested in the effects of different filters. 

In the main folder for each boat there are polars produced from the unfiltered raw data, and another set of polars produced using my selected set of filter options, and a correction factor to adjust the boat speeds for waterline length and for sail area to displacement ratio. These filtered and corrected files contain the words "JRA filters & corrn" in their file name.

I have concluded that the polars need editing to smooth the rather spiky curves for each wind speed, and to remove data for 2kt and 4kt windspeeds, which produce very erratic results and are of little relevance. This editing and smoothing is a work in progress, but an early example for Poppy can be seen here.

This editing is done following a set of rules as follows:

1/ When filling in dips or removing spikes in the polar the boat speeds would be interpolated in a linear fashion both along the windspeed curve and across the set of windspeed curves to create a fair curve between the points where the dip or spike starts and finishes.

2/ Where there are only a few points for a wind speed, they would be deleted.

3/ That data for apparent windspeeds of 2 and 4kts would be deleted as the accuracy is very dubious.

4/ Where necessary the fairing would tend to make the curves conform to the expected shape of sailing polar curves, that is to say, boat speed falling off when too close to the wind, and increasing on a close fetch and reach, and decreasing slightly near a dead run.

In the process of running these trials and creating these polars we have discovered some limitations in our equipment that we plan to rectify soon.

We have discovered that the Vyacht multiplexer we have been using to combine the NMEA sentences from the wind and water sensors has been limiting the frequency at which the boat speed number was being transmitted, meaning that it is changing far less frequently than the wind data. We plan to buy a better multiplexer that will give us more control over this by prioritising the NMEA sentences we are most interested in.

We have also discovered that the process in Polauto which is supposed to generate True Wind Speed and Direction has been giving erroneous results. We have not yet been able to persuade the author to correct this, which is why at the moment we are only showing polar diagrams using Apparent Wind.

If you want to compare several polars at the same time, I recommend using Faststone Photo Viewer which is available to download free here It allows you to view up to four images at the same time.

       " ...there is nothing - absolutely nothing - half so much worth doing as simply messing about in junk-rigged boats" 
                                                               - the Chinese Water Rat

                                                              Site contents © the Junk Rig Association and/or individual authors

Powered by Wild Apricot Membership Software