Weather Routing

[rgleason]

Shipahoy

Stavanger, we took the fast ferry from there a long time ago and listened to school kids singing during the trip. It was a wonderful experience.

I am looking at the TWS-0-6-power polar and it doesn't look right to me (and it may be wrong in the distribution as well).
I see you have just one polar in the boat.xml which is a good simple way to isolate problems.

No Windage: If your boat can power at 6 knots with no wind, the polar would be a simple one, basically six knots at all points of TWA. This would be one polar.

Upwind: If your boat is hampered going upwind by varying amounts of windage, then the speed would be slower, say 4 knots. There would be multiple polars for various TWS.

Downwind: If your boat is assisted by wind going down wind by varying amounts of windage, then the speed would be faster, say 6.5 or 7 knots for example. There would be multiple polars for various TWS.

I created this 0-6 knot Power polar as an example for sailboat and powerboat users. it could be improved. For different TWS (true wind speeds) there will be different polars which express the effect of windage on the boat headed at different true wind angles.

I think this example power polar could be improved. My suggestion is to "edit" the polar so that it works the way you want it to work and represents your boats power characteristics! Then rename it for your boat.

We are pleased you are using weather_routing. When you develop a power polar of your own using the Polar "Edit" feature, that you like and represents your boat, please attach it to this thread, we may use it in the examples instead!

Thanks. Rick

[did-g]

Hi,


It seems to be a bug..
I can reproduce it, sort of.

[rgleason]

Originally Posted by did-g

Hi,


It seems to be a bug..
I can reproduce it, sort of.

did-g How?

[did-g]

I did a routage and got weird steps 0.8 rather than around 5 knots ,

[seandepagnier]

Originally Posted by shipahoy

I get a bit confused by my weather routing results. The boat speed is much higher than the one I specified in the polar. See the example below. This is a simple polar for motoring with max speed of 7 knots, but I get similar results with the sail polar.
I get a boat speed of 10 knots when I was expecting 4 knots. See the attached screenshots.

Something is very strange. No other users have reported this that I know of, and the code is untouched for years.

Can you check for inconsistencies? Try different polars, etc? If currents are enabled they can increase boat speed, or else maybe it is a strange bug.

[rgleason]

I think it is working correctly. I have added some polar data to the power polars for higher winds. and run a routing across the atlantic, assuming there was unlimited fuel in the world in a very unrealistic simulation, that would create environmental degradation. Then using just climatology and the one polar file, run a routing.
Inspecting the boat speed and route, it looks fine to me.

[shipahoy]

Originally Posted by rgleason

I think it is working correctly.
....
Inspecting the boat speed and route, it looks fine to me.

I expect that I have done something wrong with my settings, but I cannot see what the problem is...

Can you check the maximum boat speed of the route?

The first strange thing I discovered was that I got maximum boat speed of around 12 knots when routing with the included Bavaria 37 Cruiser polar in light winds where I was expecting no more than 7 to 8 knots. The average boat speed where of reasonable value.

[shipahoy]

I tried now on another PC, and made a polar with just 5 knots for all wind speeds and angles.
When sailing downwind I got 5 knots average and max as expected, but when sailing into the wind the average was 8 knots and maximum 16 knots.
If I change the tacking time from 60 to 0 or if I change timestep from 15 min to 1h, everything is OK.

[rgleason]

shipahoy


Here is the max speed.

[rgleason]

Originally Posted by shipahoy

I tried now on another PC, and made a polar with just 5 knots for all wind speeds and angles.
When sailing downwind I got 5 knots average and max as expected, but when sailing into the wind the average was 8 knots and maximum 16 knots.
If I change the tacking time from 60 to 0 or if I change timestep from 15 min to 1h, everything is OK.

Please attach this Polar to this thread.
How are you doing this? Using Climatology or a grib?

[seandepagnier]

Originally Posted by shipahoy

I tried now on another PC, and made a polar with just 5 knots for all wind speeds and angles.
When sailing downwind I got 5 knots average and max as expected, but when sailing into the wind the average was 8 knots and maximum 16 knots.
If I change the tacking time from 60 to 0 or if I change timestep from 15 min to 1h, everything is OK.

Look for the weatherroutingconfiguration.xml file. In the same folder the boat.xml. If you can attach this and a grib file you used then we can reproduce it.

Very helpful to note the tacking time, it may be a bug there, for now I suggest using 0.

[shipahoy]

I was not able to reproduce the error with tacking time set to 0.
Attached is the simplest example I could think of that shows the error. In this example I am using Climatology and going downwind. In the previous tests I used grib-files.

[shipahoy]

Should it be += for the time penalty in line 820 in the following file?
https://github.com/seandepagnier/wea...c/RouteMap.cpp

/* did we tack thru the wind? apply penalty */
bool tacked = false;
if(parent_heading*H < 0 && fabs(parent_heading - H) < 180) {
timeseconds -= configuration.TackingTime;
tacked = true;
}

[ozolli]

Originally Posted by shipahoy

Should it be += for the time penalty in line 820 in the following file?

Maybe you can test with negative tacking time.

Edit: Does not work; It's not saved when negative.

[did-g]

Originally Posted by shipahoy

Should it be += for the time penalty in line 820 in the following file?
https://github.com/seandepagnier/wea...c/RouteMap.cpp

/* did we tack thru the wind? apply penalty */
bool tacked = false;
if(parent_heading*H < 0 && fabs(parent_heading - H) < 180) {
timeseconds -= configuration.TackingTime;
tacked = true;
}

Hi the formula doesn't seem to work anyway, in your example and mine there's no tacking.