Scale models

[Matt.D]

Do scale models made of high dens. foam give u realistic feed back to how a boat will float and ride in real conditions? Just thought id ask before i waste a week shaping one Any input would be great Thanks Matt

[rasorinc]

I don't think so.....You cannot scale weight to perfection only feet to inches.
Stan

[Matt.D]

Thanks Stan, so how do u go about scaling weight? Does building a larger model help and if so how large? Thanks Matt

[Ilan Voyager]

You can scale weights, it's done all time.
If you scale the boat at 1/10, the surfaces are scaled at 1/10^2=1/100, and the volumes at 1/10^3=1/1000. The scaling of displacement is the same that of the weight.

You need to have the plans and to know the position of the center of buoyancy (if a boat is well trimmed the calculated Centers of Buoyancy and the Gravity are on the same vertical) so you ballast the model adding lead at the exact emplacement of the Center of Gravity.

Example: a motor boat 12m long of 6000 kg, with a CG at 4.8m from the transom will give at 1/10 a model of 1.2m of 6 kg with a CB at 0.48m.

For sure if the CB and the CG were not well calculated the boat won't float in his calculated waterline.

[Matt.D]

awesome!! Thanks mate very handy formual! What if i was designing my own style hull and had no plans?

[eponodyne]

And, at 1:16 scale, a United States penny is about equivalent to 25 Lb. Weston Farmer's excellent book From My Old Boat Shop goes into scale-model hull design at some length.

[daiquiri]

Scaling just weight is ok if you need to test hull drag only, but is not enough if you need to measure angular accellerations around pitch, yaw and roll axis. Which is what you'll need to do if you want to investigate the seaworthiness of your boat.

So you will also need to assure that your model's moments of inertia "I" around x, y, and z axis are scaled properly.

First you need to calculate Ix, Iy, and Iz of the full-scale ship (I'll call it "Ifs" - where "fs" means "full-scale"). If you know the ship's mass Mfs and radiuses of gyration Rx, Ry, Rz, the Ifs' can be calulated like this:

Ifs_x = Mfs * Rfs_x^2
Ifs_y = Mfs * Rfs_y^2
Ifs_z = Mfs * Rfs_z^2

If Lm is model's length (suffix "m" is for "model") and Lfs is full-scale ship's length, let me call "a" the length scale of your model:

a = Lm/Lfs = 1/16 = 0.0625 .

The moments of inertia are scaled as a^5 (fifth power of length scale), so your model's moments around x, y, and z axis will be:

Im_x = Ifs_x * (a^5)
Im_y = Ifs_y * (a^5)
Im_z = Ifs_z * (a^5)

Now you have to arrange a number of small lead weights in your model in such a way to obtain the values of Im's given above.
The total mass of these weights + hull weight must equal the scaled mass of the model, Mm:

Mm = Mfs * a^3

Not an easy job to do, I know. That's why there are people specialized in towing tank model construction.

[marshmat]

And that, daiquiri, is the crux of the matter:
Mass scales with the cube, moments of inertia with the 5th power. (Stability goes with the 4th power.) So trying to get everything proportionally accurate at small scales is virtually impossible; something is always sacrificed. Matching Froude number means that you can't match Reynolds number, etc.

The larger the model, the easier it is to get reasonably close to something representative of the full-scale prototype. Methods for scaling resistance and wavemaking properties are fairly well-developed, but seakeeping and manoeuvring are aspects where even the expensive experts haven't figured it all out yet. (Although, given enough money and sufficient time to build a few models, the pros can get some reasonably good estimates.)

If you don't want to go the expensive tank-test lab route, spending the time to shape an HD foam model is still worthwhile. You can weight it to approximately the correct scale displacement and correct CG, and tow it in a swimming pool or beside a small powerboat (use some scrap lumber to make a boom to keep it clear of the wake). This won't give you good numbers, but it will let you see how the boat responds to waves of different sizes at different speeds- such as: does it have speeds it runs poorly at; does it tend to take spray over the bow; how violent is the motion for a given scale wave height; etc.

[Matt.D]

Thanks for the input guys . How much does a u.s penny weigh as they r a little hard to come accross over here in oz!

daiquiri , must be honest most of what u wrote drowned me!!! I do agree with ur last bit though " not an easy job"

Matt, might do the latter, build a model about 2m long and test it, i live on the water and i have a pool so im half way there!!! HaHaHa!!
Thanks Again
Matt.D

[daiquiri]

Originally Posted by Matt.D

daiquiri , must be honest most of what u wrote drowned me!!! I do agree with ur last bit though " not an easy job"

Matt, it would drown many people who don't have a solid background in physics, don't worry. I'm sorry that I got so mathematical about it.
I just wanted to illustrate a difficulties behind performing a scale model tests, and things got out of control, as usual...

Marshmat did get to the right point - it is difficult to perform a serious towing test at home, without having a good professional or academic structure to rely on.

Now, back to your question, from a more practical point of view...

If you have two or more hulls and you want to know which one is the best for a given speed, you can do a following simple test.
Take two hull models and attach them to a symmetrical balance beam with ropes of equal length. The fulcrum of the balance beam (at the exact center of the beam) is the point at wich the towing force will act.
Then you start towing them at a given speed (calculated so that the Froude number of the model and the full-scale ship is the same). Initially, if the hull forms are not too different, the models will run parallel to each other. But after few moments one of them will start to move forward relative to the other. That's the winner hull for that speed. You can then repeat the process at another speed to see what happens, or you can change the looser hull with another one and repeat the measurement at the same speed.
At the end only one will remain. Kind of Highlander hull...
The balance beam to which the hulls will be attached has to be wide enough to eliminate interference between wave trains from two hulls.

This measurement compares hulls with respect to the wave drag only. It will tell you nothing about viscous drag, and it will not give you any numbers - just the qualitative evaluation of a series of hulls. But it is a simple and quite reliable way to compare different hulls at a given speed. And it can be performed easily, either from shore or from a boat. All you need is a way to accurately measure (and mantain uniform during the test) the towing speed.

[MikeJohns]

Originally Posted by daiquiri

...................
If you have two or more hulls and you want to know which one is the best for a given speed, you can do a following simple test.
Take two hull models and attach them to a symmetrical balance beam with ropes of equal length. The fulcrum of the balance beam (at the exact center of the beam) is the point at wich the towing force will act.
Then you start towing them at a given speed (calculated so that the Froude number of the model and the full-scale ship is the same). Initially, if the hull forms are not too different, the models will run parallel to each other. But after few moments one of them will start to move forward relative to the other. That's the winner hull for that speed. You can then repeat the process at another speed to see what happens, or you can change the looser hull with another one and repeat the measurement at the same speed..................

You are forgetting to disrupt the attached flow, without doing that even this comparison lacks any valididity with 1:1.
We are more interested in the wave generation which is the unpredictable bit. If a sailboat you need to test with heel and leeway as well.

Simulators can give you some idea of the full scale response to simple wave trains. for example 'Seakeeper' and they will be more indicative than a small scale model. Motion can be predicted from the hull coeficients, GM, displacemnt etc.

[Matt.D]

daiquiri , Thanks mate good to hear im not the only one who gets swamped!
U would be the man to ask what the fourmula for displacment is and also the one for calcuating hull speed?????? thanks Matt

[Ilan Voyager]

Matt,

Do not hope to get a lot of useful information, a part basic hydrostatics and a "way" to make rough plans, from a model. If you hope to design a boat and to test it by a model without all the physical means (tank test etc ) and the knowledge implied, you'll go to failure.

Boat design and boat engineering are professions, with years of theoretical studies. When you have your diploma in pocket, you're just a rookie. You'll need a few years more to begin to master your field.

Happily for a lot of self called designers, or architects, the design of a small sail or power boat, if you stay in known forms, is to "high" hydrodynamics, or other fields of the design of ships that cooking is to organic chemistry engineering...

Small boats design may be done by some good and proved empirical approaches that I call "recipes", a very honorable way to design small boats and even not so small, that you have to master. Add some instinct, surely a big bucket of gift, some artistic sense and you can have a good small yacht designer.

You do not need a doctorate to design a good 30 feet cruising sail boat, but you have at least to have a good knowledge of the various fields involved in a such design. That means a good deal of personal work before starting a design, or you'll be simply fooling around and you'll arrive to nothing.

Great naval archis have used or use models. N. Herreshof started with half hulls carved in wood (but he made his studies in the M.I.T. ...), I think that Nigel Irens, who made his studies at Southampton, is always making models in soft wood to appreciate the forms and to have fast first drawings, at the end of the eighties he worked that way.

I do not call myself a designer, but I like better to make a model than to be stacked by the NURBS or struggle with the limited forms generated by most software. But also even simple freewares are useful for repetitive calculations.

So, you have ;

1-To acquire all the basic knowledge. Many excellent books. I mean acquire, not to simply read fast. The simple rule I gave you was the simplest one, that you learn from the very first lessons of geometry.
A revision of high school maths, with a good manual with corrected exercises do not do harm, a manual of strength of materials will give you the necessary knowledge in this field.

2-Navigate. It seems evident, but reading some posts in this forum, I have the feeling that some have never put the feet on a any boat.

3-Study again, making the link between the real world and the book knowledge.

Mean time, you can make models to train you, it may be a good way to understand some notions that seems too abstract in a book.

I conclude this too "scholarish" and "preacherish" post that home made and home tried models can give you some basic infos; how look the model? you can get a set of dimensions and introduce them in a soft, and some hydrostatics data.

Sometimes a model with radio-command can show you that for example that the 24 feet runaboat you modeled is in fact a wanting-be submarine or a extremely wet boat with water spraying on the deck. But no much more.

[daiquiri]

Originally Posted by MikeJohns

You are forgetting to disrupt the attached flow, without doing that even this comparison lacks any valididity with 1:1.

Mike, what do you intend when you say "disrupt the attached flow"?
Do you mean installing vortex generators to create turbulent flow under the hull or else? Actually I forgot to say that laminar flow separations are very unstable and can cause higly oscillating hydrodynamic forces, so it is necessary to induce a turbulent flow under the model hull.

I wonder how will Matt.D react to this complication...

The other thing I didn't say but is understated is that several runs are recommended for each hull comparison, possibly with swapping of their positions with respect to balance beam, to assure that there are no systematic errors due to unbalanced test rig.

[MikeJohns]

Originally Posted by daiquiri

..........Actually I forgot to say that laminar flow separations are very unstable and can cause higly oscillating hydrodynamic forces, so it is necessary to induce a turbulent flow under the model hull.

I wonder how will Matt.D react to this complication...

I figured you had just neglected to mention it.........carry on


PS
Consider If we changed the fuid viscosity in the test tank we could keep both Froude and Reynolds happy.