Tank testing for amateurs

Discussion in 'Boat Design' started by Alwaysthinking., Oct 30, 2019.

  1. Alwaysthinking.
    Joined: Sep 2019
    Posts: 33
    Likes: 1, Points: 8
    Location: Colorado, USA

    Alwaysthinking. Junior Member

    I've toyed with the idea of building a test tank so I can test my pet theories on models before committing to a build. But the conventional method of pulling a model through a tank requires a pretty significant amount of work, money and room. Even the most basic test tank is more of an undertaking than I'm willing to commit to.
    A while back I bought a very precise scale for weighing things like backpack stoves and bike tires. It has a hook on the bottom that you hang stuff on. It will measure the weight of things within a gram or two. I knew this would be a good tool for measuring the resistance of a test model, but I'd still need the tank.
    You don't have to move the model. You can move the water. This is how wind tunnels work. They don't pull a model plane at hundreds of miles an hour. So what's the easiest way to move the water? Nature has already provided the solution in the form of rivers and streams.
    My proposal is to carve a couple models out of blue construction foam and attach them to my scale and put them in a local creek/river. The easiest way would be to just stand in the water and hold the scale, but I suppose my legs would mess up the flow. I'd probably need a stick with a pulley at the bottom and the scale at the top. The line would run from the boat, around the pulley, and up to the scale, with me standing on the bank or on a bridge looking at the readings on the scale. I'm not sure how that would all work, but I'll just have to play around with it. I can use a toy boat to experiment with, before carving my models.
    With regards to my sit on top kayak/paddle board, I'd build two rough models. One your basic swede form and one my updated fish form. I'm hoping the difference will be significant enough to compensate for any irregularities. That is, I think one should be significantly better than the other.
  2. Blueknarr
    Joined: Aug 2017
    Posts: 1,436
    Likes: 406, Points: 83
    Location: Colorado

    Blueknarr Senior Member

    I like your approach.
    The stream presents the problem of exactl reproduction of current to fairly comparable results.

    I suggest a sluce made from plywood or other sheet material that includes;
    A water entry gate
    Level or incline gage
    Flow meter.

    Could be placed in a micro waterfall
  3. bajansailor
    Joined: Oct 2007
    Posts: 3,466
    Likes: 1,434, Points: 113, Legacy Rep: 37
    Location: Barbados

    bajansailor Marine Surveyor

    Thirty years ago I was working in a little design office in England. We had designed a 12 metre fast assymetric hull aluminium catamaran for a customer, and we wanted to see if our resistance and powering calculations were in the right ballpark (as there was a penalty if it didn't manage to do 25 knots). The budget did not allow for professional tank testing to be carried out, so I built a model out of balsa wood, and we carried out some very sophisticated towing using our Boss's motorsailer, motoring along in sheltered water on a very calm day with no wind.
    We used a windsurfer mast to tow the model from amidships on the mother boat, out to one side so it would be in 'clean' water. The towing line went through a pulley on the end of the mast to a simple spring balance (the type that you might use for weighing fish). We towed the model at a range of speeds while filming it on video and measuring the lbs of resistance on the spring balance. We scaled the speed for the model using the formula for Froude Number which is Fr = V/(gL)1/2
    Where V is speed (m/s), g is 9.81 m/s*s, and L is length, all in metric.
    And it all worked surprisingly well - we got good results.
    Until the very last run when the model was running at the equivalent of about 30 knots - something happened and she stuffed her bows in and headed for the bottom. The towing line broke and we never saw her again. :(
    But the system does work, if you can find a suitable boat to tow your model from the side with.
    fallguy likes this.
  4. BlueBell
    Joined: May 2017
    Posts: 2,562
    Likes: 895, Points: 113
    Location: Victoria BC Canada

    BlueBell . . . _ _ _ . . . _ _ _


    But did you make your 25 kt bogie?
    fallguy likes this.
  5. sharpii2
    Joined: May 2004
    Posts: 2,229
    Likes: 310, Points: 83, Legacy Rep: 611
    Location: Michigan, USA

    sharpii2 Senior Member

    You can use the same principle by walking along side a large swimming pool with a tow stick. If you want to do a comparative test, you can make two scale models and have a balanced towing yoke. One model can be your design idea and the other can be of a more conventional shape. They must be of equal displacement, length and beam. The one with the least drag will end up further ahead than the other.
    clmanges and bajansailor like this.
  6. portacruise
    Joined: Jun 2009
    Posts: 1,444
    Likes: 168, Points: 63, Legacy Rep: 218
    Location: USA

    portacruise Senior Member

    I have used both the stream and dead calm lake areas to measure the power draw of my small boats at a target speed of around 3 knots. Both methods may work to give a rough idea when there are big differences in measurement, but neither worked for me when there were small differences in efficiencies.

    I could not find a stream that gave a uniform reproducible flow, even for the same hull to the 10% degree of accuracy that I wanted- as the same hull moved around catching different little eddies on different trials. I don't know what your target speed is, but it might be hard to find a streamflow that is fast enough for your situation, as measurements may not be scalable from lower flows.

    Even with the very best sheltered places and no wind conditions, I could not get reproducible ( within 10%) measurements towing the same hull at my target speed, when running the same trap course in opposite directions between two marker buoys- especially with GPS measurement data. I found a that a tow behind sailboat knot meter seemed to correct for imperceptible water and wind currents at what seemed to be Dead Calm. It gave much better results than a GPS, should you need absolute speed data to find the best test spots. Oops, I see that Sharpii2 side-by-side method might work here if wind were not a big factor. I suppose one would have to tow at the target speed if the relative efficiencies of the model hulls change with speed.

    I had an idea about using a dome covered swimming pool with the circulating pump turned off, and a motorized tow rope, but I like Sharpii2 method better, if one can walk or run at Target speed.
    Last edited: Oct 31, 2019
    bajansailor likes this.
  7. Barry
    Joined: Mar 2002
    Posts: 1,813
    Likes: 466, Points: 83, Legacy Rep: 158

    Barry Senior Member

    Is it possible to just run a large pump on a recirc system down a flue so that you could build something like this in a garage
    Pitot tubes to measure flow rate, adjustable input into the flue can then be reset for various values. Certainly some turbulence issues but
    perhaps a series of straighteners to reduce the size of the vortices down stream.
    There was a fellow who wrote an article on the side tow method in Professional Boat Building a few years ago, gave the type of transducers etc and
    even possibly some blue tooth pick up on his phone for recording the data.
    rxcomposite likes this.
  8. bajansailor
    Joined: Oct 2007
    Posts: 3,466
    Likes: 1,434, Points: 113, Legacy Rep: 37
    Location: Barbados

    bajansailor Marine Surveyor

    Yes thanks, the full size version did manage 25 knots - it is a long time ago now, but I think her maximum speed was about 27 knots (she had a pair of 250 hp Mermaid diesels with Hamilton water jets propelling her).
    BlueBell and fallguy like this.
  9. gonzo
    Joined: Aug 2002
    Posts: 16,660
    Likes: 1,619, Points: 123, Legacy Rep: 2031
    Location: Milwaukee, WI

    gonzo Senior Member

    Francis Herreshoff would tow models with a small boat. He was not as interested in total drag as in comparing drag between similar models. He would use a horizontal equal arm scale and see which one pulled more. The key to making the results valid is to pull from the hull models with the same vector the input power is on the full size. A creek or other shallow moving body of water has too much turbulence to generate valid results.
    bajansailor and fallguy like this.
  10. Barry
    Joined: Mar 2002
    Posts: 1,813
    Likes: 466, Points: 83, Legacy Rep: 158

    Barry Senior Member

    The turbulence could be a factor in a creek as the vortices would be large but in say a moving stream in a channel, ( deep enough to mitigate shallow water effects)
    with a series of straighteners upstream of the hull, then the vortices could be made small. Plus the fact that the input conditions of the water would be the same
    Ie a myriad of small vortices, the resultant data could be relatively precise though perhaps lack accuracy.
    So you could come back to several sessions and not have to calibrate each hull everytime
    Alternatively, I suppose, you could have one hull to set up a base point at every session

    I just found this on the net, quite sophisticated, I was thinking more of a series of small
    extremely thin wall tubes, to reduce vortices to small diameters
    rxcomposite and BlueBell like this.
  11. rxcomposite
    Joined: Jan 2005
    Posts: 2,747
    Likes: 600, Points: 113, Legacy Rep: 1110
    Location: Philippines

    rxcomposite Senior Member

    There is one in operation in one of the state university down south. At the bottom is a 6" variable speed axial pump to accelerate the water. The tube goes up and merges to a tank about 24" wide by 12" deep, maybe 36" long. It is a closed loop system. I am thinking of the same but adding some crate or honeycomb to reduce turbulence.

    I will use a digital readout hall effect flow meter to measure speed of water. Relatively cheap.
  12. Doug Halsey
    Joined: Feb 2007
    Posts: 640
    Likes: 212, Points: 53, Legacy Rep: 160
    Location: California, USA

    Doug Halsey Senior Member

    I don't want to put a damper on your enthusiasm, but before you spend too many hours trying to get your test procedures perfect, you should be aware that there will be fundamental problems interpreting the results.

    The problem is that the different components of a hull's drag (skin-friction, wave-making, etc.) scale by different factors, so a shape that is better at a small scale could be worse at full scale. Professionals can (partially) account for this by computing corrections to the various contributors.

    But sometimes, even the best get it horribly wrong. Here's an excerpt of an account of the 12-Meter Mariner:
  13. Earl Boebert
    Joined: Dec 2005
    Posts: 392
    Likes: 62, Points: 28, Legacy Rep: 302
    Location: Albuquerque NM USA

    Earl Boebert Senior Member

    I have always wanted to test various radio sailing hulls (no scale effect involved -- the "model" is 1:1 of the actual radio boat) by putting a model airplane ducted fan on a hull to get a thrust vector independent of the water without the limitations of towing or the confounding effects of water moving in a confined space. If you put the ducted fan on a servo-controlled gimbal you can simulate the effect of leeway and so forth. Just a wild idea :)


    Doug Halsey likes this.
  14. HJS
    Joined: Nov 2008
    Posts: 456
    Likes: 121, Points: 43, Legacy Rep: 288
    Location: 59 45 51 N 019 02 15 E

    HJS Member

    Here comes a proposal that describes an extremely small test device. Good luck.

    Attached Files:

  15. gonzo
    Joined: Aug 2002
    Posts: 16,660
    Likes: 1,619, Points: 123, Legacy Rep: 2031
    Location: Milwaukee, WI

    gonzo Senior Member

    Test tanks with pumps have been around for a long time. The flow speed can be controlled. A creek, on the other hand, will flow at whatever random speed the rain or snow melt dictates. Tank tests need to have control over speed to have meaningful results.
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.