Sailboat data acquisition

Some of the newer boat instrument packages have the ability to store and later download data. My older Raymarine Series 60 instruments lack this, but I was able to put together a home made system.

The heart of the system is a little netbook computer I bought used of EBay for $52 US. People are dumping these old XP era computers, but they run just fine with Ubuntu Linux. I purchased a little 3 axis accelerometer board for $10 to provide the boat heel and pitch, plus another $10 for the serial to USB interface. Data collection and writing to disk via a small Python program, which also allows manual input of a number of sail parameters.

The most expensive part of the project was trying to get at the SeaTalk data from the ship's instruments. Raymarine used to make a SeaTalk to NMEA 0183 converter, but it is long out of production. I ended up having to do a couple of conversions in sequence to get to NMEA 0183.

I did a short test sail on Thursday. Now I have a new problem - making sense of the thousands of data points I just collected...

More details at: https://sites.google.com/site/sailcfd/home/calibration

 

Learn multi-variable stochastic theory....

Any data you collected is in the past and only tenuously has any relevance with the future...BTW...welcome to my world....

 

Knot meter bias

One of the first things to pop out of the data is how far off the knot meter gets once the boat is heeled. The plot shows two days of similar data collected on both port and starboard tacks. The numbers are about 10% high on port tacks and 10% low on stbd compared with the CFD model expectations.

This is likely due to the knot meter impeller being slightly off center line to port ahead of the keel. On port tacks the keel must generate lift towards the port side, which means low pressure/higher velocity on the port side. Everything reverses on the stbd tack.

You would think I would known all about this after sailing the boat for 3 1/2 years, but no...

 

Funny isn't it, I had a boat which was better upwind on port than starboard. In the end I found the cause - the c/board case was slightly off vertical......

Not by much either maybe 0.5 - 0.75 deg but it made a difference. That took a while to spot, as the best test was hard racing at close quarters. As I was getting decentish results at the time, it was not obvious. Even with the boat upside down and board extended, it was not easy to see.

Since then I have also become fastidious in correcting rudder fixings (if I have to work anywhere near them) on dinghies, as very often the manufacturer has not got the thing aligned!.

 

This is of real interest to me.

Data acquisition on a scale model is one of my projects.
eg
http://www.boatdesign.net/forums/on...home-made-wind-direction-indicator-47750.html

For a model, it would be great to acquire the sail angles, wind velocity automatically, and a whole bunch of other figures.

I will do a google search, but if you have any hints on the best place to source the components, I would be glad of any suggestions.

 

How big is your model? I ask because things get very different below about 1 meter. The boat ends up in a surface region with a large wind speed gradient. Surfaces can also fall below the critical Reynolds number, putting the air flow in the laminar regime.

 

Ardunio

 

Yes indeed. I got up to speed on too small scale models pretty early . This one is 1:5 scale, and is nearly 2 meters long. Hopefully it will present some useful data

I'm glad you said that - I started a thread
http://www.boatdesign.net/forums/on...data-acquisition-thread-51395.html#post704260

the other day and zoomed down to the Arduino platform pretty quickly.

I am investigating pressure sensors at

http://www.boatdesign.net/forums/on...g-pressure-diffferences-47511.html#post704265

 

This is fascinating. I never suspected such an effect.

It begs the question - if you could 'cant' your daggerboard by only 1 degree, not a technically tricky problem - you have quite an advantage ?

 

It may be a downwash effect with the rudder taking more load in 'clean' water? The foils were reasonable section, I'd made new ones and saved over 5.5KG weight. Foil area was OK, perhaps a tiny bit small on the board inlight air. That particular boat (and class) does not make much leeway.

 

Update on procedure to collect sailing data

What I'm finding is that you need a period of 4-5 minutes with one sail trim and one wind angle to get single measurement, and even then there is a lot of variation. The procedure I'm using now is:

1) Set the sails to the desired position/shape and leave them alone
2) Set the autopilot to hold one relative wind angle
3) Collect data for 4-5 minutes
4) Change the autopilot wind angle setting five degrees and repeat
5) Do this at 3-4 wind angles
6) Tack and do it all over again on the opposite tack.

I do the same thing on multiple days to get different average wind speeds. The attached plots show that after about three days of collections the data starts to fill in the curve. The vertical bars show one standard deviation in one second data for each of the 4-5 minute periods.

The data collection hardware is detailed at: https://sites.google.com/site/sailcfd/home/calibration

 

That's a good thing to know.

I had visions of having to talk someone with a similar yacht into trying to match a course, and measuring the distance made good of each boat.

I have memories of very, very small sail adjustments making very distinct differences in performance of a boat.

 

Could you tell me the mechanism for measuring the wind angle ?

I am having trouble finding small enough gear to do this for scale models.

The best solution I have been given is
http://www.eagletreesystems.com/inde...&product_id=54

but they emailed saying that they dont have anything to do apparent wind angle

 

I can see how minaturising a wind gauge is problematic.

Have you considered a separate base station with a full size wind recorder?
provided the real time clocks were in sync it would give a fair indication

 

I think the shore station is probably best, unless you went with an acoustic/doppler wind speed/direction instrument. Any small mechanical instrument would end up in the laminar regime, and would be difficult to interpret.