Some questions on polars and upwind/downwind performance

Discussion in 'Sailboats' started by Tedd McHenry, Sep 7, 2021.

  1. philSweet
    Joined: May 2008
    Posts: 2,487
    Likes: 280, Points: 83, Legacy Rep: 1082
    Location: Beaufort, SC and H'ville, NC

    philSweet Senior Member

    If you want to play around with polars, zezo.org provides third party support to Virtual Regatta e-sailing. It has the polars used by the stable of game boats that are quite realistic. This includes Imoca 60, Figaro 3 foilers, supermaxis, Open 40, Minis, and a host of other offshore boats.
     
  2. Tedd McHenry
    Joined: Feb 2020
    Posts: 46
    Likes: 8, Points: 8
    Location: Surrey, BC, Canada

    Tedd McHenry Junior Member

    Thanks, @philSweet . That looks like an interesting site, but I was put off by having to provide a credit card to log on. I like how they display the polars with the lines for maximum VMG.
     
  3. philSweet
    Joined: May 2008
    Posts: 2,487
    Likes: 280, Points: 83, Legacy Rep: 1082
    Location: Beaufort, SC and H'ville, NC

    philSweet Senior Member

    That's weird. I didn't. Maybe you are supposed to be part of the Virtual Regatta crowd. I only found out about the site myself a few days ago.
     
  4. Mikko Brummer
    Joined: May 2006
    Posts: 549
    Likes: 69, Points: 28, Legacy Rep: 258
    Location: Finland

    Mikko Brummer Senior Member

    Where do you find the polars on the site?
     
  5. philSweet
    Joined: May 2008
    Posts: 2,487
    Likes: 280, Points: 83, Legacy Rep: 1082
    Location: Beaufort, SC and H'ville, NC

    philSweet Senior Member

    On the home page, you select a race that is currently open. Each has it's own boat type. Current options are globe forty (open 40 class), mini transat (mini class), around Britain, and around Italy. I just realized that it doesn't actually tell you anywhere which boat type is being competed. That makes it a bit less useful:rolleyes:. But after you select the race, hit GO, and when the chart appears, hit POLARS, and a new window opens with the boat's polars. You can select wind speed in 0.1 knot increments, and select sail options. LS = light jib and light gennaker. C0 is the reacher for each class. HS is heavy air sails. Jib and spinnaker are standard. Some boats have a foils option.

    There are other ways to get at the data. Virtual Regatta has a clunky downloadable tabulated version that can, with some machinations, be loaded into an excel worksheet and graphed.

    As awkward as this all sounds, it is darned hard to find real polar data to use when trying to reverse engineer some of these performance boats, so you take what you can get. For testing VVP modules and checking seaway allowances, they are nice to have.

    The Figaro 3 polars are somewhat suspect at this point. The real boats outperformed the VR boats downwind by a substantial margin. It was obvious from sailing angles and routing choices.
     
    Skyak likes this.
  6. Skyak
    Joined: Jul 2012
    Posts: 1,313
    Likes: 64, Points: 48, Legacy Rep: 152
    Location: United States

    Skyak Senior Member

    I would suggest that polars are not more available and not accurate because there are more significant contributors to to boat speed than wind and angle. The most obvious case would be downwind surfing.
    I think it is time to design a graph that consider more variables.
     
  7. AJB
    Joined: Jul 2021
    Posts: 17
    Likes: 0, Points: 1
    Location: 31 42S 152 04 E

    AJB Junior Member

    Skyak,
    In all reality, the best we can expect is that there might be two sets of polars - one for smooth water, so with no additional resistance from waves upwind; and no surfing.

    And another set for sea state commensurate with wind speed, some fetch assumption and perhaps velocity gradient

    There is an entire universe of wave scenarios for offshore sailing where the sea state has little to do with the existing wind at a particular location.

    Just getting the two modes mentioned above would be a big step forward in general. The ORC stuff is fair way off reality.
     
  8. Skyak
    Joined: Jul 2012
    Posts: 1,313
    Likes: 64, Points: 48, Legacy Rep: 152
    Location: United States

    Skyak Senior Member

    I completely understand that there are lots of variables that come with sea state and I see the value in polar charts.
    On the other hand, polar charts are wrong and the VPP that doesn't consider sea state will lead to the wrong design direction. Look at scow-bows in box rules, the sunfast 3300...the hull shapes would be predicted to lose if you don't consider sea state.
    I can't help thinking that what should be done is to instrument some boats and have machine learning gather insights from some great sailors.
     
  9. AJB
    Joined: Jul 2021
    Posts: 17
    Likes: 0, Points: 1
    Location: 31 42S 152 04 E

    AJB Junior Member

    Well, you can be sure that Raisson and others will have motion sensor data from sailing; surfing and RAW having far more effect on the minis vs. that on the IMOCA 60s.
     
  10. CT249
    Joined: May 2003
    Posts: 1,327
    Likes: 129, Points: 63, Legacy Rep: 215
    Location: Sydney Australia

    CT249 Senior Member

    But do you know that the polar charts are significantly wrong, in a way that makes them significantly less useful? After all, even if polars under-read or over-read speed and angle, for many purposes all that is relevant is comparative data rather than absolute data.

    Would the Sunfast 3300, for example, be predicted to lose if you don't consider sea state? Why? Because the polar would claim that the 3300 is slower than it actually is? In that case, the Sunfast's ORC rating would probably claim it's slower than it actually is, and it would therefore win lots of ORC races.

    Last time I asked about this sort of stuff, I was told that some top design offices have put top guys on boats and checked the speed against their VPPs and therefore had a pretty good idea of the effect of sea state on their custom VPPs, but that was proprietary. The top guys have plenty of "instrumented boats" with
    "great sailors" who spend lots of their time gathering insights; I can remember years ago when Grant Simmer hopped onto the latest 52' Farr, just about his first question to the AC/Volvo pros on the boat was "when were the instruments last calibrated" - and the answer was within the last four days or so. I can't recall exact numbers but like any good sailor, he was making allowances for sea state and boatspeed and angle all the time.

    And how does one usefully describe sea state for comparison purposes? What chop size? What swell size? What angle? What steepness? What windstrength? What sailing angles? What combinations of all the above? How many boats would have to be put through how many different combinations of conditions? Why?

    Some of my academic friends work in machine learning, and just going from their remarks it seems that any machine learning about all those factors would be extremely difficult.
     
    Doug Halsey likes this.
  11. Skyak
    Joined: Jul 2012
    Posts: 1,313
    Likes: 64, Points: 48, Legacy Rep: 152
    Location: United States

    Skyak Senior Member

    My point was that if you went by the flatwater VPP you wouldn't even build the Sunfast 3300, and yes I think it's rating misses it's ability to keep it's bow up at speed. It has a trick and it wins races when it get to use it's trick. There are also races that don't have significant contributions from sea state, and it looks to me like the "round the cans" boats and teams dominate these -flatwater designs winning flatwater races and classic polar charts more relevant.

    I know instrumented boats are used. The original question was "why is it so hard to get the polar data?" There was a suggestion to derive it from tracker info, and an answer that the result is inaccurate due (most likely) to sea state. That brought my offering -the data is fine, the polar chart is inadequate. I would go further and suggest that manufacturers don't like to provide polar data because they don't want to give customers data that can be argued inaccurate or incomplete. I don't think it is proprietary in that no significant competitor can't calculate it themselves.

    So, if we are done with the posturing, I will continue the thought.

    "And how does one usefully describe sea state for comparison purposes?" From the boats point of view it is just the variables in the average fourier transform. From an energy perspective you likely only need the first two factors and if there isn't some significant depth change the second factor loses significance. Frequency and incidence are tough -navys have wave development models that might offer some data reduction, and skippers themselves avoid bad incidence angles and the natural frequency of the boat. This is the make or break of this idea. There is always a best setting, and the polar chart ONLY indicates the best setting, so I imagine the frequency/incidence factor would often be a boundary that at best indicated the better setting. There would also need to be a confused seas, noise factor which is the inability of the FT to describe the waves for more than a very short time.

    Well, that's more than enough to noodle on for now. On machine learning I would make the case that GPUs can make short work of characterizing sailboat motion and predicting a best path.
     
    Last edited: Sep 28, 2021
  12. Skyak
    Joined: Jul 2012
    Posts: 1,313
    Likes: 64, Points: 48, Legacy Rep: 152
    Location: United States

    Skyak Senior Member

    My thought is that the guys doing the software for the IMOCA autopilot would have the best handle on data and how to use it. Are there rules on what sensor inputs they can use?
     
  13. AJB
    Joined: Jul 2021
    Posts: 17
    Likes: 0, Points: 1
    Location: 31 42S 152 04 E

    AJB Junior Member

    Skyak,
    The short answer is 'No'! One would think that all the accelerations in six axes is enough...?
     

  14. Skyak
    Joined: Jul 2012
    Posts: 1,313
    Likes: 64, Points: 48, Legacy Rep: 152
    Location: United States

    Skyak Senior Member

    So if someone used camera data from the masthead and machine learning to predict wave positions that would be legal to use?
     
Loading...
Forum posts represent the experience, opinion, and view of individual users. Boat Design Net does not necessarily endorse nor share the view of each individual post.
When making potentially dangerous or financial decisions, always employ and consult appropriate professionals. Your circumstances or experience may be different.