• 16th April 2010 - By greg
    Side view showing the water line

    Side view showing the water line, old rudder (grey), new rudder (black), keel, drive leg and new fresh water ballast at the stern

    A way back a million years ago (it seems!) when Ken and I first started to plan the build of WiTHiN, Ken made a scale model based on the panel patterns generated by Stuart Bloomfield (the naval architect). Ken’s model was great, but I totally destroyed it the other day when I attempted to float it in my bath tub to see what effect increasing surface area under the water line would do to weather cocking. I had taped the hull up with duct tape, but the model fell off the edge of the tube into the water overnight and Ken’s awesome model was destroyed. My sincerest apologies to Clives little boy Shackleton who loved to play with the model!
    Foam model with the small rudder on

    Foam model with the small rudder on

    I was able to salvage it by pouring 2 part foam into the soggy cardboard form and then sanding down the foam. I was able to get it pretty close to the shape of the hull. Good enough for this test. I added a little rudder and keel – both scaled to size and weight. Then I attached a string to a push pin and adjusted it’s location on the top deck such that pulling it made the model track square through the water in the bath tub – just like WiTHiN tracks in the water when pushed by wind.

    Model in the tub with pull string, large rudder and 60 lbs of ballast

    Model in the tub with pull string, large rudder and 60 lbs of additional ballast at the stern

    When I exchange the rudder with a larger one that is close to the scale size of the new revised rudder, the model immediately starts to turn bow forward by about 10 to 15 degrees and then it picks up speed in that direction.

    Floating in the tub with small rudder

    Floating in the tub with small rudder

    When I add about 60 lbs (to-scale) of weight to the very stern directly over the rudder, it does the same thing, but turns to about 30 degrees, picks up speed and hits the side of the tub. The reason weight at the stern is more effective in moving that center of lateral resistance back is because when the model heels (rolls), the lateral area of the hull that is under the water stays consistent, whereas the lateral area of the rudder and keel actually decrease. See the drawing below and compare it to the drawing at the top of this page:

    WiTHiN rolled to 45 degrees. Not the lateral area of the rudder and keel actually get smaller

    WiTHiN rolled to 45 degrees. Note the lateral area of the rudder and keel actually get smaller

    It was obvious that adding the weight at the stern was more effective in making the bow turn downwind than a larger rudder. However, neither approach resulted in the model turning all the way (90 degrees) to face down wind. Rick thought that a stick pushing rather than a strong pulling was a more accurate way of simulating wind, so I used a pencil on the pushpin and centered it on the top of the gunwale. When I pushed directly down the length of the tub, the model eventually turned to about 45 to 50 degrees. When I add a bit of rudder turn, the model easily turned all the way to 90 degrees to face down wind.

  • 4 Comments to “Model testing”

    • Steve Retz on April 16, 2010

      Sounds like you’ve got it! COOL BEANS!!! Good Job!

    • Nick Hein on April 16, 2010

      I just love it when a systematic, scientific approach solves a problem – so you can move on to the next one.

    • Bruce Bolster on April 16, 2010

      Sounds like you have a solution in hand! Now that I see how far forward the keel is in relation to the wind-induced centre of effort, I understand why the boat tends to weathercock so much. The upside is it will be easy to maintain an upwind heading if desired since the propeller is so far forward in relation to the rudder. If the larger rudder doesn’t solve the problem, you could consider a simple modification to the keel portion of the hull aft of the ballast keel – just make it pointy instead of round in cross section, and deeper close to the rudder itself, to add some lateral resistance. This would add minimal drag or weight, and would require less in the way of rudder correction to maintain a downwind course. With a fixed skeg such as this, the boat would tend to track more strongly, but would be a little slower to turn in calm conditions.

    • Alex on April 22, 2010

      Have you considered that with scale testing in fluids velocity doesn’t scale the same way as model size? When working with scales in fluids you may find the model performs differently than your full sized boat and cocks differently that the full sized will.

    Leave a Reply