Visit our Popular Forums

svrevelations · 07-10-2011, 04:33

Hello Sinbad 7

Tori, I have been following a number of your messages on this forum. Everytime I read some of your messages, I am suprised even amazed at your knowledge depth regarding software and other things.

If you don't have one already, you should really consider creatin a website as there is so much to learn for what or how you do things. Thank you for the many informative messages!

sinbad7 · 07-10-2011, 05:00

MooseMoney..

Thank you for your kind words but I am learning more than my inputs..

I do actually have a 'paralell forum' if I am allowed to tell you:

Quote:

YACHTMARINE USER FORUM - Home

Regds
Tore

PjotrC · 07-10-2011, 13:55

Tore...

The 1.0 knot appearing when there is no polar curve selected is my glitch...

Already fixed it here.

The temporary workaround is to always have some polar curve defined. After all, this is the purpose of the excercise...

You will find an example of the curve definition earlier in this thread.

Good luck,

Piotr

sinbad7 · 08-10-2011, 05:23

Piotr..

I am a novice in this respect so bear over with me...HOW do you create/select the polar curve in OCPN? Mine is presently set for 'opencpn.ini'

Tore

sinbad7 · 08-10-2011, 06:58

Piotr..

I changed the polar curve setting in ocpn.ini to 5.1 but the speed in Route Manager is still shown as 1 kn? I can not see any curves displayed on the map. What am I doing wrong?

Tore

sinbad7 · 08-10-2011, 07:32

Piotr..

Not as easy as that,I understand,you have to create a polar file first and load it into OCPN via the tool box. I gather there are fixed polar profiles for the various boat types? Where do I find a suitable polar file for the sake of this experiment?

Tore

sinbad7 · 08-10-2011, 08:08

Piotr..

I used your example polar file:

Quote:

CM93DisplayOffsetEnable=0
PolarCurve0="00 00.0 4.0 40.0 4.0 60.0 4.0 65.0 4.0 90.0 4.0 130.0 4.0 160.0 4.0 180.0 4.0"
PolarCurve1="10 00.0 1.5 40.0 2.5 60.0 4.0 65.0 4.5 90.0 5.0 130.0 5.5 160.0 5.7 180.0 5.0"
PolarCurve2="32 00.0 0.5 40.0 1.5 60.0 6.5 65.0 7.0 90.0 8.0 130.0 8.5 160.0 8.7 180.0 8.0"
[Settings/GlobalState]
OwnShipLatLon=" 36.487914, 2.139360"
bFollow=0
nColorScheme=1
FrameWinX=1253

and placed it in the OpenCPN 2.5.1006 directory in it's own 'Polar files' sub-directory named Polar-1.pol

I Selected the file in the Toolbox and indicated 'Load polar file'

However,I can still not see any polar curve and the speed is still a constant 1 kn.

I will await your advice..

Tore

PjotrC · 08-10-2011, 13:47

Tore...

Thank you very much for your patience! While this is not overly difficult, it is hard to touch upon every possible case myself, so your testing input is very valuable...

From your message I see that you are not familiar with using OpenCPN configuration file, but let's first discuss the polars.

The experimental version of OCPN that you have installed will handle a single set of Polar Curves at a time. A polar curve is a specification of expected boat speed for given true wind speed and the angle between wind direction and (the opposite of) boat course. One example you quoted is

Code:

10 00.0 1.5 40.0 2.5 60.0 4.0 65.0 4.5 90.0 5.0 130.0 5.5 160.0 5.7 180.0 5.0

This is interpreted as follows:

- for true wind speed of 10 knots (first field of this line)
- the boat speed at 0 degrees to the wind (directly upwind) will be 1.5 knots (second and third field)
- the boat speed at 40 degrees to the wind will be 2.5 knots (fourth and fifth field)
...
- the boat speed at 180 degrees to the wind (directly downwind) will be 5.0 knots

Please keep in mind, that I chose these values to see very clearly the effect of simulation, you are free to chose them according to your boat's characteristics and sailing style.

When predicting boat's speed, OCPN will first select from the available polar curve set (if there is more than one) the curve closest to actual wind condition.
Then it will select the pair (angle, speed) again closest to the actual situation. The speed so obtained will be used for calculation of timing for the next route leg.

You can have as many polar curves in the set, and as many (angle, speed) pairs within each of them as you need to express your simulation needs.

Now, there are two ways to communicate the Polar Curve Set contents to OCPN.

1. Via "opencpn.ini" configuration file. There are three (3) config variables: PolarCurve0, PolarCurve1 and PolarCurve2, which correspond to three (3) consecutive lines of text: 0, 1 and 2. These lines may contain a polar curve for some wind speed value as described above. You may use one, two or three of these variables. You just enter the text as in the example you quoted, perhaps just before the start of section [Settings/GlobalState]. Please have a look again at the earlier posts in this thread to see what is the expected order.

2. Via an external file, selected by a button control in the Toolbox/Settings panel. The file then should contain just the numerical data, e.g. just the text

Code:

00 00.0 4.0 40.0 4.0 60.0 4.0 65.0 4.0 90.0 4.0 130.0 4.0 160.0 4.0 180.0 4.0
10 00.0 1.5 40.0 2.5 60.0 4.0 65.0 4.5 90.0 5.0 130.0 5.5 160.0 5.7 180.0 5.0
32 00.0 0.5 40.0 1.5 60.0 6.5 65.0 7.0 90.0 8.0 130.0 8.5 160.0 8.7 180.0 8.0

which contains the 3 curves for ranges 0-4, 5-21, 21+ wind speeds (or so)

If you wonder about the meaning of boat speed at 0 knots of wind, the answer is: motoring sometimes is part of planning, sometimes not.

You can make such Polars Set as simple or as elaborate as you need. Basically, just one line

Code:

10 10.0 6.0

should replace the Planned Speed of 6.0 knots for all possible conditions... (the search algorith will always stop at 6.0 knots, I hope).

You can directly use polars included in popular Weather Routing programs (see Bluewater Racing docs and kit).

The 1.0 value appearing now means "no speed", because there is no polar, but a fix is underway. I just feel it is not worth posting a new kit just for this now.

Good luck and thanks for trying this! Keep going!

Piotr

PS.

Maybe some more remarks on polars, etc.

I recommend trying out Bluewater Racing software, and reading their manual.

Besides Bluewater Racing I looked at, and I am sometimes using in real routing cases, the "zezo.org" web page and the "VRTool" software, but:

- for zezo I do not see how to supply my own polar set, and in daily life I am not really interested in Volvo Ocean Race type of performance
- for VRTool I am definitely not smart enough to use it efficiently :-(
- Bluewater Racing is very intuitive, albeit it is sometimes over-optimistic about boat performance. However, very, very useful overall. I hope it will be developed in the right direction.

One can use Bluewater Racing to view (and edit) graphically the polars.

Bluewater Racing do use cube-spline interpolation on individual polar curves. I just choose the nearest value, because I think it easier and more natural to define my curve this way. This is one source of differences in simulations between us.

I attach sample polar files for a pure sailing scenario (no motoring, 0 speed upwind) and a cruising scenario (motoring allowed, no speed 0 ever occurring). Please appreciate the difference between a Sail Race Game and real cruising planning: in a racing/gaming environment it just does not happen, that the course is directly into the wind, while in a cruising scenario it is quite often the case - at a price that has to be considered and accepted. It is not easy to program the strategy for all situations, without requiring a lot of options, user interaction etc., what I want to avoid. I tend now to use the "cruising" (i.e. motoring allowed) polars with a speed visibly indicating that it is a motoring speed (e.g. speed 5.0 - under sail, speed 5.01 - under power).

I do not intend to write a GUI editor for polars for OCPN - I think there is no need for it, and those who need to see the graphs drawn as color lines can very well use Bluewater Racing... there is a lot of other useful things to implement...

sinbad7 · 08-10-2011, 15:47

Piotr..

Now this looks a little better...thanks for your very comprehensive explanation,much appreciated. I will try to digest...

Tore

jonasaberg · 09-10-2011, 03:39

About 10 years ago a PhD student presented an Excel sheet calculation as a part of a paper on yacht performance calculations at the Chesapeake Bay Symposium.
It calculated polars based on known algorithms, as for example those in this book.
Amazon.com: Principles of Yacht Design (9780071487696): Lars Larsson, Rolf Eliasson: Books

Unfortunately the Excel sheet is outdated, it was made for Excel2000 and would not work in current Excel versions. However, if anyone is interested I might be able to dig it up.

Also there is this software:
Techsail - Sailing Yacht Performance Analysis
But the software has not been developed much during a couple of years, and you need to shell up $50 which is questionable.

A third option is to find a ORC sheet for your yacht. These can be found on the internet and is a good starting point for any standard yacht that has been used for racing according to the ORC data sheet (Formerly IMS) based on MHS.
Example link:
Mätbrev 2011 - SeglarFörbundet

Suggestions:
1. It would be good if the ORC data sheet data could be used straight off. (as an option)
Open one of the pdf sheets in the link above to find the time allowances data, they state the time needed to sail one NM, so the polar data is presented a bit differently and needs to be recalculated.

2. In general, It would be good if it was allowed to use more than three wind speeds.

3. I respectfully beg to differ that it is not useful with having the tool integrated in the live version and in a version that optimizes the route with respect to changing weather over the time of the route. (I hope I did not misunderstand this ....have not had opportunity to try the beta...sorry)
I have used a home-hack version of a wind prediction system for a few years and it is a fantastic tool to pick when to sail and how to sail to get best use of the wind. It would be even greater to be able to download predictions under way to plan how to avoid foul weather.
On our yacht we have effectively used it to sail on a reach for 200NM instead of finding ourself sailing into the wind on the open sea. It can be a very useful tool, and I expect it would be the same on the high seas.

/Jonas

PjotrC · 09-10-2011, 04:04

Quote:

Originally Posted by jonasaberg

2. In general, It would be good if it was allowed to use more than three wind speeds.

3. I respectfully beg to differ that it is not useful with having the tool integrated in the live version and in a version that optimizes the route with respect to changing weather over the time of the route.

Jonas,

ad. 2. You can have as many speeds as you wish in the file load option. The .ini version is a simplified, fall-back option. It can easily be extended.

ad. 3. My thinking is mostly along the non-connected cruiser line. Before leaving the harbour I use all the data available, and all the tools that process this to decide on rough strategy. Once offshore, when no new data is available, I think it is not useful to re-optimize automatically without current inputs. I just revert to traditional thinking then. Having said that, I prefer to have my nav program lightweight and use excellent external tools, when available. However, I still keep the last GRIB as a reference and I like to view it in sync with time progress. Not a very strong feeling, though...

for those cruising in Internet-connected waters, it may look differently, indeed.

Cheers,

Piotr

dacust · 09-10-2011, 06:44

Quote:

Originally Posted by jonasaberg

Unfortunately the Excel sheet is outdated, it was made for Excel2000 and would not work in current Excel versions. However, if anyone is interested I might be able to dig it up.

I have some older versions of Excel at work. I bet one is old enough to open back that far.

Quote:

Originally Posted by jonasaberg

1. It would be good if the ORC data sheet data could be used straight off. (as an option)
Open one of the pdf sheets in the link above to find the time allowances data, they state the time needed to sail one NM, so the polar data is presented a bit differently and needs to be recalculated.

...

Unless I am mis-reading, they also have a straight table for wind vs. speed.

Here's one I copied.

Code:

Wind Velocity   6 kt    8 kt   10 kt   12 kt   14 kt   16 kt   20 kt
  Beat Angles   45.7°   43.4°   42.1°   41.3°   40.6°   40.7°   41.6°
     Beat VMG   3.49    4.24    4.67    4.96    5.14    5.22    5.15
           52°  5.48    6.47    7.09    7.40    7.57    7.67    7.68
           60   5.87    6.85    7.47    7.73    7.89    7.99    8.05
           75°  6.19    7.16    7.75    8.13    8.32    8.45    8.59
           90°  6.22    7.31    7.90    8.16    8.49    8.79    9.03
          110°  6.06    7.26    8.04    8.41    8.85    9.19    9.49
          120°  5.89    7.11    7.98    8.58    8.98    9.24    9.89
          135°  5.28    6.51    7.57    8.16    8.68    9.21   10.34
          150°  4.39    5.60    6.73    7.54    8.07    8.55    9.55
      Run VMG   3.80    4.85    5.83    6.60    7.23    7.73    8.60
  Gybe Angles  141.6°  146.8°  150.5°  156.7°  163.8°  166.9°  167.3°

-dan

jonasaberg · 09-10-2011, 09:02

Dan,
Ooops, sorry, embarrassing.
I did not read the new version of the sheet before writing...in the old IMS it was not.

Anyway, one idea would be to use this format since it makes it easy to cut and paste existing data.

I will try to dive into my old backups.
Also, I am kind of surprised that none has written an opensource VPP - velocity Predicition Program based on the open information and data.

After writing my post this morning I opened the Larson book I referred to in my earlier post and all the information is quite clearly stated to write a piece of software that would produce polars given the measurements of the yacht and its sails.

The algorithm is based on an article from 1980 by a guy called Hazen. It includes the equations, the data for different kinds of sails and an algorithm to follow.

What Hazen seem to have done was to summarize experimental data and gather them together with the necessary equations and an algorithm.

If anyone is interested in programming it could possibly help out with writing up the equations... and the framework of the algorithm.

/J

champ8242 · 09-10-2011, 09:13

Quote:

Originally Posted by jonasaberg

I will try to dive into my old backups.
Also, I am kind of surprised that none has written an opensource VPP - velocity Predicition Program based on the open information and data.

After writing my post this morning I opened the Larson book I referred to in my earlier post and all the information is quite clearly stated to write a piece of software that would produce polars given the measurements of the yacht and its sails.

This is the sheet you are referring to.

Bart