Deck sweeping sails and effective aspect ratio

Discussion in 'Hydrodynamics and Aerodynamics' started by Will Fraser, Mar 31, 2015.

  1. Ben G
    Joined: Jun 2013
    Posts: 49
    Likes: 0, Points: 6, Legacy Rep: 10
    Location: Brisbane

    Ben G Junior Member

    Here are some images. From memory, going 10 knots upwind in 15 knots breeze, 45 deg TWA / ~27 deg AWA

    The idea was to add an aerodynamic 'lever vang' which would continue the lifting aerofoil all the way down the mast to deck level. See comparison. Chord was about 300mm at boom level.
    The difference in force vectors was very little for this modification. Mostly the aero force increased but the drive component didn't really change
     

    Attached Files:

  2. Ben G
    Joined: Jun 2013
    Posts: 49
    Likes: 0, Points: 6, Legacy Rep: 10
    Location: Brisbane

    Ben G Junior Member

    force vectors for 27degree AWA
     

    Attached Files:

  3. Will Fraser
    Joined: Feb 2014
    Posts: 168
    Likes: 17, Points: 18, Legacy Rep: 11
    Location: South Africa

    Will Fraser Senior Member

    Very interesting Ben, did you ever apply it in practice?

    The increase in leeway force should not be disregarded, especially if the additional force is applied low down where it has a small contribution to heel. The benefit lies under the water, where the keel/dagger board is forced to produce more lift. Since any increase in keel hydro drag associated with the increase in lift is small compared to the hull drag, the net result is an increase in hydro L/D.

    This phenomenon is used to good effect by glider pilots who carry huge amounts of water ballast to simultaneously increase speed and glide ratio, i.e. L/D.
     
  4. Will Fraser
    Joined: Feb 2014
    Posts: 168
    Likes: 17, Points: 18, Legacy Rep: 11
    Location: South Africa

    Will Fraser Senior Member

    I have been doing some more simulations on a sailing kayak concept model. As a control case I used the Hobie sail kit (20 sq-ft, AR = 3.5) mounted as low as possible.

    For the modified, flush-footed sail I kept the chord distribution the same (except in the foot area, obviously) as well as the camber, area and twist.

    In both cases heel is at 6deg.

    The airflow has a little bit of velocity gradient but no twist. I have an idea of how to model the twist and will hopefully re-evaluate these models with it later.

    Apparent wind is 18.8kts at 30.6deg, equivalent to doing 4.9kts in 15kts true wind with 3.8deg leeway. These odd values were chosen to correspond to an equivalent case on the spreadsheet for comparison.

    The flush footed sail showed persistent trailing edge flow separation at deck level. Non of the "usual" aerodynamic fixes seemed to solve this (leading edge vortex generators, root fillet etc.), but I did eventually find a solution and the results showed corresponding improvement.

    With the modified sail's AR equal to the original, the drive was up by 16% while the heel went down 28%.

    I gradually decreased the modified sail's AR (from 3.5 to 1.5) until the sail L/D as well as the drive corresponded with the stock Hobie sail. By this time the modified sail's heeling moment was 55% of the original despite the tack being 2" higher.

    While these results are not conclusive, I think it is safe to say that:
    - results obtained on monohulls, especially when heeled, are not representative of the possible benefits on more appropriate vessel platforms.
    - sheeting angle on the modified sail is generally 4deg less for maximum drive compared to the original sail, and might hint at wing-sails being more appropriate for the application
    - factors that play a role in obtaining best results include sail size and position relative the the endplate in use as well as flow over (and under) the endplate.

    [​IMG]

    [​IMG]

    [​IMG]

    [​IMG]
     
    Last edited: Apr 29, 2015
  5. daiquiri
    Joined: May 2004
    Posts: 5,373
    Likes: 247, Points: 73, Legacy Rep: 3380
    Location: Italy (Garda Lake) and Croatia (Istria)

    daiquiri Engineering and Design

  6. Ben G
    Joined: Jun 2013
    Posts: 49
    Likes: 0, Points: 6, Legacy Rep: 10
    Location: Brisbane

    Ben G Junior Member

    I partly applied it... in that I now have a lever vang that is more structurally ideal and simpler rather than getting fancy with aero details. It is still somewhat aerodynamic but more to reduce drag than increase lift.
    The main benefit of the lever is keeping weight forward during tacks and gybes. I understand what you're getting at with the increased side force, but the (vector) angles are getting pretty close, especially once you consider leeway, which is not in the diagram above. If there's any error in the computation there's a somewhat equal chance of increasing either drag or thrust, as far as I could tell
     

    Attached Files:

  7. Ben G
    Joined: Jun 2013
    Posts: 49
    Likes: 0, Points: 6, Legacy Rep: 10
    Location: Brisbane

    Ben G Junior Member

    PS Please appreciate how the foot of the jib is mostly in contact with the deck, and the leeward side of the jib follows the curve of the topsides ;)

    I would like to close the gap near the clew of the jib, but I suspect the sailmaker will advise me to concentrate on where I'm steering instead
     
  8. Erwan
    Joined: Oct 2005
    Posts: 385
    Likes: 17, Points: 18, Legacy Rep: 97
    Location: France

    Erwan Senior Member

    Thank you Will for posting your research.

    I would have one question, or a suggestion:
    On your last picture we can see a hudge amount of blue area leeward and some green yellow almost orange windward.
    Did you try to quantify the lift on the leeward trampoline of your trimaran (assuming "airproof" tramp) if you consider the blue pressure fields on it

    Does it contribute to increase the righting moment ?
    In other words does aero lift on the leeward tramp alleviate the leeward hull ?
    Symetrically does windward pressure of the sail creates a downward force on the windward trampoline?

    Hope it's not irrelevant

    Regards

    EK
     
  9. Will Fraser
    Joined: Feb 2014
    Posts: 168
    Likes: 17, Points: 18, Legacy Rep: 11
    Location: South Africa

    Will Fraser Senior Member

    Erwan, I never isolated the loads and moments of the tramps, but the overall moment of the entire boat and crew is indeed 30% less than that of the sail for the AR = 2.2 case that is pictured above. (The moment was measured at water level).

    When reducing the AR to 1.5, the overall moment is 42% less than that of the sail, most likely because of the long sail foot now spreading the differential pressure over a greater extent of the tramp. And yes, your assumption of an "airproof" tramp is correct.

    I have corrected the figure in the post that compares the low AR sail heeling moment with that of the original. It is in fact the overall heeling moment that is 34% of the original case, not the moment of the sail itself.
     
  10. Will Fraser
    Joined: Feb 2014
    Posts: 168
    Likes: 17, Points: 18, Legacy Rep: 11
    Location: South Africa

    Will Fraser Senior Member

    I managed to get the wind gradient and twist modelled and have re-tested all my previous models for the little kayak trimaran. I have also tested a host of variations on the flush-decked sail.

    The stock sail was modelled with its tack 12" off the deck. The flush-deck sail was modelled at the same height i.e. the end-plate deck was added at 12", so that both scenarios get the same amount of wind. The flush-deck sail was also tested at a reduced aspect ratio, but with the tack still at the same height.

    The flush-deck sail shows trailing edge flow separation at the foot despite the decreased AoA of the local apparent wind, and none of the solutions that solved this issue in the older tests seemed to work this time around.

    Despite the problematic separation bubble, the flush-deck sail still showed 34% less heeling moment for a given amount of drive compared to the stock sail. The overall heeling moment of the boat was reduced a further 7% due to the significant righting moment of the deck.

    The reduced AR flush-deck sail managed to attain 90% of the drive of the stock sail at only 51% of the heeling moment.

    There is still a lot left to explain, especially why the flow at the foot separates at such small angles of attack and why none of the traditional boundary layer control solutions seem to have any effect. The learning curve continues...

    The stock sail, AR = 3.5:
    [​IMG]


    Flush-deck sail, AR = 2.2
    [​IMG]

    [​IMG]
     
  11. Erwan
    Joined: Oct 2005
    Posts: 385
    Likes: 17, Points: 18, Legacy Rep: 97
    Location: France

    Erwan Senior Member

    Thanks for sharing Will,
    So with deck as an end-plate, you achieve 34% less heeling moment for the same driving force. That's quite an improvement, usually in sailing people use to deal with the 1% performance pick-up
    It could be interesting to get an idea of the difference in driving force for the same righting moment.
    And the righting moment from pressure fields on the tramp is around 7%, that's significant too.

    Please, which software are you using for these simulations?

    Regards

    EK
     
  12. TeddyDiver
    Joined: Dec 2007
    Posts: 2,578
    Likes: 120, Points: 73, Legacy Rep: 1650
    Location: Finland/Norway

    TeddyDiver Gollywobbler

    Looks like stalling... but there's not visible if the vortex comes from the pressure side? but flow separation shows..
     
  13. Will Fraser
    Joined: Feb 2014
    Posts: 168
    Likes: 17, Points: 18, Legacy Rep: 11
    Location: South Africa

    Will Fraser Senior Member

    I use Solidworks with its built-in Flow Simulation add-on.

    Note that the figures quoted are specific to this peculiar configuration. In my earlier tests (with no wind gradient) I noted that there is much less separated flow if the end-plate deck is raised above the actual deck and air is allowed to flow more freely between the two decks. I assume it has something to do with the air pressure as it exits at the rear of the end-plate deck.
    There was also a notable improvement when I changed the structure connecting the end-plate deck to the amas from individual fore- and aft struts to a closed, inclined side plate as seen the the pics above. A further small increase came from the diagonal cut-away at the leading edge of this side plate.

    I mention these modifications because their effects were not readily foreseen but were stumbled upon during an exploratory "what-if" approach. Many other modifications that I tried had detrimental effects and were discarded. Other hull and rig combinations might call for a very different overall solution.

    I hear what you are saying about the 1% being significant in the real world, but I suspect that is partly because it normally involves an already very optimised racing design within the constraints of some class rule or other.

    To answer your question about the amount of drive for equal heeling moment, I must choose between a few different approaches:
    -reduce the AoA on the stock sail,
    -increase the AR of the flush-deck sail
    -increase the area of the flush-deck sail
    Any preference?
     
  14. Will Fraser
    Joined: Feb 2014
    Posts: 168
    Likes: 17, Points: 18, Legacy Rep: 11
    Location: South Africa

    Will Fraser Senior Member

    I have just completed an analysis with some changes to the deck surface properties that suggests that the separated flow was perhaps nothing more than interference drag of a tired boundary layer.

    With reduced roughness on the deck, I was able to sheet the sail in an additional 2deg with only the tiniest bit of separation starting to form. The heeling moment of the flush-deck sail is now almost as much as that of the stock sail, the drive is 28% more.
     

  15. daiquiri
    Joined: May 2004
    Posts: 5,373
    Likes: 247, Points: 73, Legacy Rep: 3380
    Location: Italy (Garda Lake) and Croatia (Istria)

    daiquiri Engineering and Design

    You have basically re-discovered that sails with high aspect ratio (with the base plate, which doubles the AR) stalls at smaller angle of attack than a sail with a smaller AR (the stock sail).
     
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.