"Boat Speed" vs. SOG calibration challenge

[Snore]

I am with CatNewBee, after wrestling with this issue on my boat this is what I determined.

When on a starboard tack, the boat will “slip” to leeward a smidge. This results in water on the starboard side of the keel traveling faster than the water on the port (leeward) side. So on a starboard tack the water going past the paddle wheel will be greater than the actual speed of the vessel through the water. When on a port tack, the water going past the transducer will be at the same speed as the water going past the rest of the hull.

Hope that all makes sense.

[transmitterdan]

Jr_spyder,

Can you give us a few full examples of wind, STW and SOG? It sounds like the paddle wheel is nonlinear itself or the location causes the water to move slower when heeled. Probably it is a combination of both. Our paddle wheel does something similar to what you describe and the sensor is in about the same place as you describe but we don't have a full keel.

[jr_spyder]

Well, I just wrote a long and thoughtful response to this and when I posted I got an error saying my token had expired and, poof, all that content was lost. Arggggg...

More briefly: My calibration was done right. Out at dawn with no wind, water like glass, no current as evidenced by both the tables, and the fact that buoys were standing still. Drove north, south, north, south (normal directions of the current) at steady 6 kts observing differences in SOG and boat speed and adjusted boat speed to match SOG. This is the right process, and all other evidence/instinct shows it's right. Even today when I'm motoring in a confined area with no current the numbers match.

When sailing the paddle wheel always reports higher than it should, never lower. So while I'd agree that theoretically any cross flow on the paddle would have it go slower, I never see that (although it might be a minuscule difference anyway).

If I'm sailing downwind with a push current I should definitely see SOG higher than boat speed. Sometimes I see them match, but hardly ever is SOG higher. With a very dramatic push current, like in the Cape Cod Canal, SOG definitely goes higher (12 kts!).

My B&G gear is not the race level stuff so I don't have adjustments for heel angle. etc. But you're right, the fact that race boats will have sensors on both the port and starboard side, and the fancy computer analyzes both is evidence that heeling over has an effect on the paddle wheel measurement.

I could just only show SOG on my gauges and not boat speed, and ignore the discrepancy. But my wife is the obsessive trimmer on the boat and likes the numbers. We feel like SOG tells the truth, but boat speed tells us how lucky we are, or not, with current or leeway. Sometimes it matters, like when sailing on the edge of the gulf stream and knowing you're maximizing it's effect on your speed.

I'm enjoying this discussion even though it may not lead to a solution, other than a better understanding. Also good to know other boats have the same issue.

Also curious, as in one post above, if ultrasonic speed sensors make a difference here. But because the flow changes under the boat (like a wing described above) then the type of sensor is probably irrelevant.

Thanks,

JR

[fxykty]

Just a clarification that for myself and 44 cruisingcat, we have catamarans so heel is not an issue. In my case the paddle wheel is located on the centreline of the port hull, just aft of the deepest part of the hull. The daggerboard case is a few metres forward and the skeg and prop are a good 5 metres aft.

As we’re both seeing low and high numbers for speed, it seems we need several calibration points for various speeds?

In any case, who really cares? It’s close enough so that the true wind calculations are correct and we’re not racing so more accuracy doesn’t matter. It does make the maximum speed a little suspect, so we tend to report the maximum SOG.

[jr_spyder]

Transmitterdan, I actually have a detailed log of data that we record every hour on long passages, but it's on the boat and I'm not. I'll examine it more carefully when I'm back on board. I'm just sort of musing over this question while sitting at home wishing I was sailing.

I think it ultimately is a fluid dynamics issue. Since the paddle wheel is located on the hull adjacent to where the leading edge of the keel starts to drop below the hull bottom, the flowing water there has to compress and accelerate to either side of the growing "foil" (keel), and the paddle wheel is right there to pick up this accelerated flow. If the paddle wheel was a foot more forward it might not see this increased flow, and I wouldn't be writing this thread. One would then also theorize that when the boat is heeled this flow acceleration might be greater, hence a faster reading.

Now to disagree with myself - if the flow becomes greater on one side the keel then it is slower on the other. So I should see a different effect on a port tack vs a starboard tack. I've thought this in the past so observed it carefully, but no joy. I have the same results on both tacks.

[jr_spyder]

I tend to agree "who really cares." But my geeky mind likes these kinds of puzzles.

[skipper53]

I have a Datamarine speed instrument. Always reads about .7 mph slower than the GPS calculated SOG. Decided to use the boat speed instrument only to make sail trim adjustments. With several GPS instruments onboard, and the observed accuracy of all of them in terms of position plotting, I have given up on trying to reconcile the two sources and just blame the difference on an old, tired paddle wheel.

[jr_spyder]

And now for some hilarity on this subject:

Last June we did the Marion-Bermuda race which is a Corinthian event for non-professional cruisers. It is decidedly not the Newport-Bermuda, but is pretty much the same route (just longer, with a 20 mile beat down Buzzards Bay before turning left to Bermuda.) Anyway, as a race you can never use your motor. We had a couple days of very light air so had to learn a lot about sailing slowly. We also had to amuse ourselves.

This question of SOG vs boat speed was a popular subject. We decided we needed a more traditional measure of a knot as a different data point and had details in a sailing book on board. It turns out our boat is almost exactly one knot (47.2 feet) long. So my wife sat on the bow seat and would drop a piece of pasta in the water, and my daughter hung off the back rail watching for it to pass. We used a stop watch and timed this very scientific procedure, then did the heavy math converting the time to speed. We ran the test 10 times. Now the process had many flaws, including the fact that we were generally laughing so hard it was difficult to repeat each "run" with much accuracy. Nonetheless 9 out 10 runs the data showed the boat's paddle wheel was reporting faster than the timed pasta experiment. So there you have it!

Maybe we're getting too carried away with this....

[YellowcakeMan]

I have had the same situation with my B&G system and I'm reasonable tech-savvy. My best hypothesis is that the water flow over the paddle wheel is sufficiently non-linear such that a linear adjustment is simply insufficient for all wind/speed/heel combinations. I have been playing with this for two seasons, and have come to the conclusion that I need to re-calibrate when wind/speed/heel changes substantially. I believe it's also a case of having electronics that are far more capable than the accuracy of the relatively coarse instrument in the water. That is, garbage in = garbage out, regardless of the capability of the electronics. This doesn't solve your problem, but maybe it will give you a little comfort that you're not alone.

[lateral]

I have two DST 800, one either side for each tack.
They are a crappy mechanism fraught with multiple short comings.
Waste of time getting anything meaningful from them that doesn't need to be "adjusted".
My B&G don't do non-linear either.

[KiwiKen]

Originally Posted by StuM

Where is the paddle wheel in relation to the keel and propellor? in a location where the propellor when motoring creates some sort of eddy effect, slowing down the water flow at that point?

As described, it suggest that you are getting flow over the paddlewheel impeded when motoring, and you are calibrating based on that water flow. When you re not motoring, the flow over the paddlewheel is not impeded.

A Curved surface can cause loequacal waterflow acceleration, just like air flowing over the curved upper surface of a Wing is locally accelerated ( this is what generates lift)
Paddle wheel transducers should if possible be located where the waterflow is the same as the boat speed and not distorted by local accelerations caused by the curvature of the hull surface.

[CatNewBee]

Well I guess you have very sensitive instrumentation.

Since Albert Einstein, we know E=mc², and in a accelerated system
there will be relativistic effects, especially if you sale near speed of light.

(this was for the geeky explanation... now try too beat this)

[MarkSF]

Another error I see is related to cleanliness of the bottom. If I were to do the calibration the day after the bottom is cleaned, after a couple of months the DST800 will read a knot low. This must be the slime growth disturbing the flow near to the hull.

[markpierce]

On my previous boat (Bluewater Blackwatch) noticed minor difference (about a tenth of a knot) in boat speed depending which windward tack. ... Recommend you rely on GPS readings as that is measuring real progress.

[donradcliffe]

The non linearity of the paddle wheel is because the boundary layer thickness varies with velocity, not because Airmar are crap. If the system does not have a correction factor, i calibrate at the typical upwind boatspeed, because that is what most helmsmen use to determine optimum VMG upwind.