REVIEW: Full Scale Measurements on a Hydrofoil International Moth

I am always a sucker for the science of mothology. You can imagine my excitement this week when: “Full Scale Measurements on a Hydrofoil International Moth” by Bill Beaver and John Zseleczky was released this week. Published moth tow tank data! And lots of it. They have been so busy, so much data, so many runs. Hull resistance! Foil L/Ds! Air resistance. I doubt there has ever been any moth data of this magnitude ever released.

Paper available here, presumably by permission of author....

I have been casually working on a hydrofoil simulator. The goal of this is to determine for a given foil configuration:

• Take off speed
• Time to take off
• Top speed.

Beaver and Zseleczky’s research info is invaluable to my research and I certainly make it publicly available when it is done.

I was so excited by this report I wanted to digested it piece by piece. What started as my notes based on this report became a bit of a review. My notes are collated below. N4rkla says I am being too harsh, I at no time mean to offensive, but I do highlight some issues. I would welcome discussion with the authors or others if interested (I am but a lowly naval architect).

I will produce a couple of additional graphs from the raw data this weekend (with hope) and publish them here next week.

----------------------

Pg 3: At the light displacement, over half the wetted

area is retained, despite shedding ¾ of the weight.

I believe wetted surface area should optimally decrease linearly with displacement. This suggests a new hull design - not a circle section, not a rectangle, triangle maybe. I hope it’s a really cool, funky cross section. I will investigate.

Pg 4: Figure 6 – Collapsed Resistance Data

I think there may be a problem because vertical axis is coupled to horizontal axis by V. I don’t think it is necessary to “normalized by…speed squared” (pg 3). Full scale results shouldn’t need to be nondimensionalized. And horizontal axis isn’t nondimed anyway.

Pg 4: This effect was not investigated as

the constraint of having adequate angle of attack on the

foils limits the usefulness of dramatic trim changes.

Is it possible to constrain pitch in a tow tank?

Probably not, but chaning static trim would be more interesting anyway and easily achieved. This would investigate the benefits sitting aft for more aoa on foils.

Pg: 4 Figure 7 – Pitch and Heave Data

Do you think hull lift = drag?

I think so so an optimal foiler hull may have no rise.

Figure 8 –Hungry Beaver at 13fps Design Displacement

Transom separation does not appear very well developed despite ~7kts boat speed.

Pg 5: limiting the test configurations to a … zero yaw conditions

A couple of yaw measurements would be good to confirm suspicions of http://giovannigaleotti.blogspot.com/2009/02/ask-pilot-about-take-off.html

Pg 5: These components are labeled “Vendor 1” and “Vendor 2”

Is this bladerider and fastacraft respectively?

Can’t see what the JZ foil is.

Pg 5

3) Lift and drag of the T foil were again measured at zero

roll, zero yaw, 20 fps, and 18 inches immersion, but the

flap angle was varied for a given pitch angle to achieve a

constant lift of 180 lbs. This highlights the effect of the

flap on the system efficiency and provides some guidance

how best to set up the boat.

From this description presumably he would have needed to do a few short runs, playing with the flap angle until the desired lift is achieved. Hopefuly there is a recording of these “rejected” run data for other flap settings that resulted in more, or less lift. There doesn’t appear to be enuf data in the appendix for this to be included.

Pg 8

The wave and spray drag

was then calculated as:

RtWave&Spray = 2 * Rtfaired 12”immersion – Rtfaired 24”immersion

The wave and spray drag was converted into a coefficient

according to the Hoerner formulation where t is the foil

thickness:

CdW&S = RtWave&Spray/(0.5 * ρ * V2*t2)

Smart! But again I woulnd’t nondimensionalize it. Especially not by foil thickness that varies between the different models tested. Also it would also be interesting to confirm that this spray resistance is proportional to V^2 by testing alteast 2 more speeds.

Pg 8 Figure 17 – Calculated Wave and Spray Drag

Coefficient for Struts

A definition of Hoerner Guidance would be good.

Pg 8 Figure 18 – Spray Generation off the Vendor2

Daggerboard T-Foil at 20 fps

An argument for a foil with a sharp leading edge. Just ensure you sail with 0 leeway angle.

Future thesis – is it better to sail with 0 leeway? (Sideforce balanced by lifter).

Pg 8• The wave and spray drag of the struts is some 30%

higher than would be anticipated based on simple

calculation from Hoerner.

Who cares. What about the component of spray to total drag vs ride height.

Page 9: Figure 20

Dimensionalised Drag! Yay!

Have you subtracted strut drag from these values?

Plotting c/h rather than h/c is annoying and makes making a conclusion of drag @ h=inf dangerous from that graph.

Also as h -> inf. You would expect a curve plateauing, not a linear regression. Perhaps this would be apparent in a plot of h/c but I can’t tell without regraphing it.

Page 9 CdWave = RtWave/(0.5* ρ*V2*S) = C^L2 * c/h * CDH/CLh^2

I think this needs to be fleshed out some more. Give more complete reference to the Hoerner formulation.

Pg: 9 The lower proportion of induced drag on the rudder T-foil is

probably an indication that the rudder lifting foil is too

large, or that the assumed rudder loading is not

representative of what the rudder actually encounters.

Lower drag = bad? Not in my book. As you say, this needs to be investigated further.

Pg: 11 Figure 24 – Impact on flap angle

As lift remains constant in this graph (180LB) why publish L/D ratio, much better just to plot drag. These graphs are essentially showing 1/D. Low is bad, high is good. Easier to comprehend without this inverse relationship.

Pg: 11 Most sailors

obsess about the surface finish of their foils and the

antidotal evidence from these tests indicate that they are

right in doing so.

Use moody chart to quantify this

Pg 13: The initial test results on the HB daggerboard foil indicate that a permeable hinge joint which allows pressure relief across the foil may well be a greater liability than a larger impermeable hinge gap.

I wouldn’t discount the faired hinge yet. If you still have the facilities how about putting setting a flap angle then applying a thin pass of tape over the hinge to make it impermeable. This way your tests of the hinge are constrained to one foil assembly.

Pg 13: The scatter in the data was

assumed to be caused by large scale turbulence and

separation off of the hull, appendages and helmsman.

Could also have been latent turbulence in the tow tank – air would take much longer to settle than water.

Pg: 14 It isinteresting to see that the highest aerodynamic drag was

measured with the hull upright.

Yes, but “sideforce” increases with heel. Loading up centerboard = drag, not to mention this side force reduces VMG. Vector sum perpendicular to the true wind direction is more appropriate for upwind case, however I can be convinced that another angle (not true wind direction) is more appropriate.

Pg: 14“winds eye view”

“Profile area”?

Why are tramps not shown?

The drag was tested with tramps.

Pg: 15 Higher is better. The hydrodynamic drag on the foils

decreases with reduced immersion.

Only for zero leeway. Leeway angle will influence this conclusion greatly.

Pg15: Existing rudder lifting foils appear too large.

Experiments with smaller foils may yet lead to

improved performance.

Don’t believe this report supports this finding. May be the case but a force balance is required.

General Comment

I have an issue with the non-dimensionalizing. Do you use the thickness, chord or area? Given neither area nor thickness is consistant between foil models you shouldn’t use either parameter to non-dim it. There is no need to reduce lift and drag to coefficients for comparison.

Much better to use outright drag for comparison, as he has done on page 10, figure 23, however I cannot see what AOA he is testing these at. I suspect 0 degrees and then 180lb provided with flap. However as shown in figure 24, optimal drag setting appears to be with zero flap angle.

This really makes me think that the main foil global angle should be set at the start of the day when you decide what you want the takeoff speed to be. Light days, lots of global angle. Heavy days, not much. Requires simple and rapid adjustment of CB rake.