Aquila Hydro-Glide Foil System with 35% fuel saving

Been through that .. well the 'through' is an euphemism, and i'm still struggeling after a couple years. Doing it for hulls isn't that hard if you have a few weeks at hand; the commercial softs handle that quite well, openFoam-based SimScale - CFD, FEA, and Thermal Simulation in the Cloud | CAE https://www.simscale.com/ should do fine on the cloud and even the stock openFoam distros have tutorials that work more or less. Syms involving foils and other small appendages or small detail on planing surfaces are a whole different story though: Meshing gets very sensitive, input geometry starts to matter a lot (parametric definitions needed), and they get far more unstable than for larger geometries. Calculation times increase at least 10-fold because of the courant number constraints, so that you also have to get into cluster administration (Well unless you can stand to wait weeks for results only to understand one of the 5 tiny errors that blew up the calculation - not exactly a receipe for quick learning). You'll quickly face a lot of theoretical implications, and will have to pick between 5-digit costs on commercial softs, or the catastrophic documentation, 'call-me' partial implementations that lead nowhere if you don't pay for additional consulting and closed source code, and rivaling clans of the openFoam universe.

Maybe it's time for a Google or so to stir up that field with a bit of coherence, ergonomy, and AI-integration.

 

I have that book but haven't finished it yet. And thanks for the link, I've played around with JavaFoil that allows you to simulate very similar to XFoil. The "problem" with both is that there are airfoils in there with ridiculously high L/D ratios, e.g. the NACA 4412 a=5 1MRn, L/D=181. So maybe there are other problems with such foils or you need a database with experimental data.

So yeah I need to read more! But right now I'm reading other books and articles to learn.

I was just curious what people thought about the simple partial lift foil system.

And I agree that there has to be a break even point where the theoretical gains are eaten up by strut drag or added weight or just so small in absolute watts saved it's not worth it.

If I use the above numbers for the NACA4412, to generate 40% lift for a catamaran at 10 knots, 3.5t displacement, 12m LWL, I would need a 0.92m² hydrofoil which takes 389W propulsive power. According to hullcalc.xls that would reduce propulsive power from 6kW to 4kW. So instead of going 9 knots I could almost go 10 knots with 4kW. So even if all these assumptions and calculates were correct I'm basically below this break even point.

Thanks, that made me actually laugh out loud! But I might take a few weeks to do the tutorials for openFoam.

 

Hmm, something just popped into my head: With one or more hydrofoils connecting the hulls below the water line you would also gain stiffness / strength.

Maybe you could build a catamaran or trimaran where the hulls are only connected below the waterline?

 

Probably just the difference between 2D ('infinite span') and a real 3d wing.
Here's a starting point for realistic L/D ratios for small wings: https://www.boatdesign.net/attachments/csyspaperfeb09-beaver-paper-on-moth-pdf.136024/
Then there are also many foil sections to eliminate: Many of the great L/D ones have rather extreme pressure peaks, which could lead to cavitation even at relatively low speeds. Laminar profiles can be problematic in turbulent water and with even slight fouling. And many are just too thin to be build with reasonable effort, and to sustain the loads.

I've found tspeer 's posts around here and his site tremendously helpful, even if much of the info is way over my head.

Drop me a line if you really want to get into such levels of masochism. I'm no crack but at least i have a cluster online and a few templates ... And hell maybe with such enthousiasm you'll be the one who gets these damn sims to work a bit smoother Be warned that "weeks" are pretty much the minimum useful time unit for such work..

 

Thanks, that gives me a good idea what to expect from these simulations. I think for my use case it would only make sense to go for full hydrofoiling at higher speeds with a larger battery, but the numbers in the paper you linked aren't encouraging. I'll keep on reading.

 

I should probably just keep reading and learning but I'm impatient and wanted to post this anyways

This is the NACA 16-1012 foil with the ridiculous L/D of 303 at 1 megareynolds and angle of attack = 0. Interestingly when I change just the number of points the foil is generated from the drag increases from 0.0036 to 0.0075. My guess is there is a numerical instability that pushes the algorithm to assume non-laminar flow. But even that has still a L/D of 147 and a good aspect ratio and drag from struts shouldn't reduce efficiency that much. The pressure peaks Cf seem to be between 1 and -1. Does Cf=-1 mean cavitation?
I'd love to know what is wrong with this.

 

Just found I already had this paper so I'm going to attach it here in case anybody else is interested:
Recent Applications of Hydrofoil-Supported-Catamarans by K.G.W. Hoppe

 

Just found an interesting video for "High Performance Foil Assisted Catamarans" by qwestboatsnz. Includes a rendering of how they tilt the hydrofoil and a graph of fuel/nm vs speed. Seems like it's optimized for 30-35 knots.
5 year old video, so it's definitely not new

 

It'd think it's the other way round (software assumes unrealistic laminar flow along the whole chord): the extremely even pressure distribution on the suction side seems to indicate that the turbulent transition is missing.
Yes, results that change drastically with resolution are a clear indication of numerical artifacts - this is why you'd run mesh sensitivity studies in CFD for example.
Guess you should check these figures with an other code or method, not only to confirm the high L/D (147 would already be among the top), but also to make sure that it is not a section that relies on extreme concentrations of suction that would drive you into cavitation way too soon. Who'd want to see his nice wing eaten up by bubbles within minutes ?

btw NACA 16-1012 sounds weird: the 6-digit series usually have numbers starting with 6. It's not a 16-012 either (NACA 16-012 (naca16012-il) http://airfoiltools.com/airfoil/details?airfoil=naca16012-il is symmetric). ... ??

 

The NACA 1-series was their first attempt at designing airfoil sections using an inverse method. The "6" in the second digit indicates the pressure distribution rooftop goes back to 60% chord. In retrospect, they were trying for too much laminar flow and the convex pressure recovery lead to early separation at the trailing edge. The 6-series sections, with their more modest rooftops and linear pressure recovery distributions were more successful. The NACA 1-series has found application, however, in propellers (where maximum Mach number is an issue) and hydrofoils (where cavitation is an issue).

 

It's more or less a "random" example I stumbled on while playing around in javaFoil that had way too good to be true L/D ratio.

The "1" in NACA 16-1012 vs NACA 16-012 stands for the "Design Lift Coefficient=1" in javafoil. So the lift is around 1 even with angle of 0. Definitely an odd bird and probably a numerical fluke.

 

I still don't get it Dejay: Naca 16-012 looks like below in airfoiltools.com and on the UIUC database, that's not the foil in your pressure pic.
If you post point coordinates or the right airfoiltools link for your 'weird' one i'd be curious to throw it into Hanley's Multisurface Aerodynamics or CFD

 

Here are the points for the NACA 16-1012
1.00000000 0.00000000
0.99725609 0.00300943
0.98711502 0.01002075
0.97007336 0.01999901
0.94617998 0.03206509
0.91565237 0.04536635
0.87885246 0.05907058
0.83627428 0.07239325
0.78853126 0.08463441
0.73634070 0.09521406
0.68050536 0.10369937
0.62189281 0.10981991
0.56141368 0.11346917
0.50000000 0.11469463
0.43858546 0.11362833
0.37809432 0.11037820
0.31943292 0.10508162
0.26347698 0.09793027
0.21105984 0.08915736
0.16296239 0.07902746
0.11990486 0.06783022
0.08254062 0.05587934
0.05145280 0.04351839
0.02715582 0.03113663
0.01010826 0.01920475
0.00076516 0.00836899
0.00000000 0.00000000
0.00652597 -0.00452704
0.01894992 -0.00711596
0.03782794 -0.00831975
0.06309117 -0.00858985
0.09447552 -0.00826938
0.13158439 -0.00762430
0.17391495 -0.00685612
0.22087542 -0.00610806
0.27179985 -0.00547009
0.32596219 -0.00498521
0.38259001 -0.00465830
0.44087786 -0.00446740
0.50000000 -0.00437683
0.55912300 -0.00431250
0.61742286 -0.00410836
0.67409953 -0.00361500
0.72838247 -0.00276904
0.77953349 -0.00160263
0.82684838 -0.00024087
0.86965829 0.00111591
0.90733150 0.00222475
0.93927605 0.00284618
0.96494289 0.00280315
0.98382679 0.00205675
0.99545278 0.00082548
1.00000000 0.00000000

I guess it's similar to the "NACA 66-1012". The "design lift coefficient=1" basically pulls up the middle of these two foil and makes the bottom flat.
There is a little bit in the JavaFoil User manual on page 9

But please, don't invest much time in this on my account, since it's really just a random airfoil I played around with. I have yet a ton of reading and learning to do and other tasks before I can seriously approach an addon hydrofoil that might work for a low-powered cruiser powered by solar panels

 

No big deal actually, i was just curious because even your 147 L/D figure was fairly high, and Tom said these 1-series aka 16-series can do well in terms of cavitation avoidance.

Here's the 3Dfoil output for a quasi-infinite span at 18kn in saltwater. (Span : 100m mirrored, meaning the LD is that of a 200m wing, and lift ist for one half of this wing. Chord = 1m):

Coefficients @ AOA = 0.00 Deg.
Lift Coefficient: 0.821
Side Force Coefficient: 1.01E-16
Vortex/Total Drag Coefficients: 0.00111/0.00565
Lift-to-Drag Ratio: 145.36 -- That's pretty close to your 'resonable' figure isn't it ?
Moment Coefficient (about CG): -0.422
Roll Coefficient (about CG): -40.81
Yaw Coefficient (about CG): 0.0858
Forces/Moments @ AOA = 0.00 Deg.
Lift Force is:
Drag Force is:
Moment (about CG)
Minimum Sink Rate
Total Area is: 100.0
3.61E+06 N (8.12E+05 lbs)
24,839.54 N (5,584.43 lbs)
is: -1.85E+06 N-m (-1.37E+06 lbs-ft)
is: 0.0111 m/s (0.0365 ft/s)
m^2 (1,076.39 ft^2)

_________________________________
Coefficients @ AOA = 4.00 Deg.
Lift Coefficient: 1.25
Side Force Coefficient: 1.53E-16
Vortex/Total Drag Coefficients: 0.00286/0.0113
Lift-to-Drag Ratio: 110.47
Moment Coefficient (about CG): -0.53
Roll Coefficient (about CG): -62.03
Yaw Coefficient (about CG): 0.212
Forces/Moments @ AOA = 4.00 Deg.
Lift Force is:
Drag Force is:
Moment (about CG)
Minimum Sink Rate
Total Area is: 100.0
5.50E+06 N (1.24E+06 lbs)
49,807.66 N (11,197.77 lbs)
is: -2.33E+06 N-m (-1.72E+06 lbs-ft)
is: 0.0119 m/s (0.0389 ft/s)
m^2 (1,076.39 ft^2

_______________________________
if you go down to a reasonable span (5m mirrored for 1m chord as an example), things don't look quite as sexy, but not bad either:
Coefficients @ AOA = 0.00 Deg.
Lift Coefficient: 0.67
Side Force Coefficient: 8.18E-17
Vortex/Total Drag Coefficients: 0.0144/0.019
Lift-to-Drag Ratio: 35.31
Moment Coefficient (about CG): -0.378
Roll Coefficient (about CG): -1.57
Yaw Coefficient (about CG): 0.0416
Forces/Moments @ AOA = 0.00 Deg.
Lift Force is:
Drag Force is:
Moment (about CG)
Minimum Sink Rate
Total Area is: 5.00
1.47E+05 N (33,109.57 lbs)
4,170.39 N (937.59 lbs)
is: -83,096.45 N-m (-61,291.85 lbs-ft)
is: 0.227 m/s (0.744 ft/s)
m^2 (53.82 ft^2)

________________________________________
Coefficients @ AOA = 4.00 Deg.
Lift Coefficient: 1.01
Side Force Coefficient: 1.24E-16
Vortex/Total Drag Coefficients: 0.0328/0.0371
Lift-to-Drag Ratio: 27.31
Moment Coefficient (about CG): -0.463
Roll Coefficient (about CG): -2.35
Yaw Coefficient (about CG): 0.0931
Forces/Moments @ AOA = 4.00 Deg.
Lift Force is:
Drag Force is:
Moment (about CG)
Minimum Sink Rate
Total Area is: 5.00
2.23E+05 N (50,025.36 lbs)
8,146.72 N (1,831.55 lbs)
is: -1.02E+05 N-m (-75,052.53 lbs-ft)
is: 0.238 m/s (0.782 ft/s)
m^2 (53.82 ft^2)

 

Oh thanks, that is interesting! I didn't expect going to aspect ratio of 5 would be this "bad". But I guess you get both less lift and more drag. I only considered the reduced lift which isn't so bad at AR 5.

So that is probably the reason why hydrofoils only make sense for fast speeds, because at faster speeds you need less foil area which allows you to use higher aspect ratios leading to higher efficiency.

With an L/D of 90 you need about 0.9m² of foil area to lift a 4t boat out of the water at 20 knots with 4.5kW. Or 0.4m² at 30 knots with 6.7kW. Seems insanely attractive for a power boat, but unfortunately not possible with solar power alone.