View Full Version : Scale model design testing
06-25-2006, 03:27 PM
I know I've seen an article which described the production of a paper scale model. Built at a specific scale size that mimicked the performance of a finished boat in a tank. The model was varnished or otherwise treated to temporarily stop soaking up water. Pennies were used as balast with a equivalent weight value per penny. Equipment and crew weights could be placed and moved to see the effects on the final trim.
I'm thinking Payson but I've been through every Payson book I have and can't find it. I have an idea for a low power displacement skiff and would like to model it before committing to the full size thing. The method above was said to yield very accurate final results if carefully followed.
06-25-2006, 03:44 PM
What exactly did you want to model?
Extrapolation from scale to real is very accurate in some respects and very "iffy" in others.
06-25-2006, 03:46 PM
Clueless here, but I'd be tempted to ask MMD about it. Because he's a professional designer I'd guess he has considerably more than a clue. I'd also be tempted to contract w/ a designer if the finished boat has significant value or risk factor.
Would paper be rigid enough? And the non-scale water has to have some effect.
06-25-2006, 04:05 PM
Weston Farmer, "From My Old Boat Shop"
Chapters 12, and 13,
Floatational models, part 1 anfd 2
I have often tested panel-construction designs (plywood, aluminum, etc.) for fairness and ease of panel installation by making a cardboard model like the one shown above. I take a copy of the panel layout drawing and glue it to some stiff carboard such as Bristol board using spray-on glue. I then cut out the panels and bulkheads and tape them together with masking tape. Usually I don't care about exterior appearance, only whether the panels lay on properly, so the tape is fine. The one time I did care about the exterior - a planing hull that I wanted to see how the bow spray was deflected - I taped it together outside with low-tack automotive painter's masking tape first, then taped the interior seams, then removed the exterior tape. I painted thick carpenter's glue on all the exterior seams than painted the whole hull with housepaint. I ballasted it with sand in the bathtub until it floated on it's at-rest lines, attached a string bridle to the bulkhead just forward of midships with duct tape (the home hanyman's secret weapon), then took it down to the river for "tank testing". At a fast-moving channel in the river I used a float-type knotmeter to find a spot that was moving at the proper scale speed and sat the boat in there to observe the water flow past the hull. The hull lasted perfectly for the hour or so that I needed it to, and then slowly warped and twisted in the changing humidity over the next few weeks that it sat on top of my file cabinet.
Larger boats with bigger budgets for testing get properly tank-tested at the local university.
06-25-2006, 04:50 PM
MMD, Just what I'd recalled. If memory serves there are some third order relationships such as speed and volume that need to be accounted for to be accurate. That is what I'm looking for.
Canoeyawl, Will try to find a copy locally, I did read some Farmer a few years ago and forgot about him as a source.
PI, Looking at a banks dory basic shape and want to get the engine location/crew weight sorted out for good trim. St Pierre dories are so much bigger that I'm leery of scaling down to a 12' bottom length by using their porportions.
06-25-2006, 06:45 PM
There are a lot of reasons why scaling does not work ---
Viscous effects don't scale unless you use a diferent liquid.
Surface tension effects don't scale.
Structural engineering does not scale - modes of failure change.
Power and time don't scale.
As long as you get the right answers scalling works. When you don't life gets bad real fast.
06-25-2006, 07:57 PM
Agree with all of the above. However this ain't rocket science. I just want to see if a design which hasn't been through a complete set of naval architects calculations will float on its lines. Wasn't planning on a full set of towing tests. I'm too lazy to learn freeship and model building is fun.
06-25-2006, 08:07 PM
There are some things for which that a scale model test can be benneficial. I made a r/c model of my TALLY HO II and found that there was a lack of control at extreme angles of heel. The problem turned out to be that the rudder was not deep enough and therefore was almost totally out of water at the extreme heel. A modification to the length of the rudder fixed the control problem.
Another test I did was centerboard configurations vs balast keel. The original boat had internal balast and a CB, I wanted the keel and only some internal trimming balast. The model sails great with the keel, the addition of a CB resulted in a problem with getting hung in irons when tacking.
The scale model was also a good actual test that my displacement, balast and righting calculations were accurate.
Bottom line,the model was well worth the time and effort and also a lot of fun to sail.:)
06-25-2006, 08:18 PM
Tom Lathrop did something like this while designing "Liz", his Bluejacket 24 and he may chime in here any minute now. Rick
06-26-2006, 09:33 AM
Rather than a throw-away paper model, why not a nicer one you can keep?
06-26-2006, 09:45 AM
The paper model is not a mad idea to confirm the marks.
Remember that the masses scale as the cube of the linear scale ratio - so if you ar building a 1:10 model you need 1:1,000 weights.
The University of Michigan had a 640 ft long testing tank for hull models. Some of the pre WWII cup defender designs were tested there. The US Navy has a testing tank, and so did/does the Stevens Institute. There may be others. Model testing must be good for something.
06-26-2006, 12:15 PM
06-26-2006, 12:52 PM
You can use scale models for a number of things.
The simplest is a planking layout model. It will enable you to figure the shapes of the planks. The larger the scale, the more accurate the planking shapes. This type of model is more than 100-years old.
A second model type used by Naval Architects is used to optimize appendages. These are self-powered and help ensure rudder response is adequate. Models of this type are anywhere from 8-ft to 40-ft long.
A third model sometimes used is a scale propeller model. These are more sensitive than the above two examples to scale effects. !/4 scale is about the absolute smallest that gives usable data.
Models are used to estimate sea keeping in wave tanks.
Models are used to estimate performance in Ice too.
Lastly there are hydrodynamic models, commonly used in a towing basin. If scale is too small, these models will misdirect you due to Reynolds Number problems. However if made big enough they work quite well. In fact they are what we (Naval Architects) use to verify our computational dynamic calculations. These Models start at about 6-ft and go to over 20-ft.
A towing basin is not the only way to run these hydrodynamic models but they tend to have the best instrumentation and are generally the most repeatable. You can easily measure straight-ahead drag between two models with a 'yoke tow'. These things have been around a long time. They are simply a pivot pin with a swept back arm on both sides. A known model is placed on one side and towed at a scale speed. The new model is towed from the opposing arm. Depending on how the arm rotates the relative performance can be calculated using simple trigonometry. Models for this test start at about 6-ft.
As pointed out earlier, modeling of hydrodynamics can be misleading even for the professional (remember Courages?). This is especially true for small-scale models. So just be aware.
06-26-2006, 04:48 PM
P.I. That is the relationship I was looking for. I remembered there was a cube involved but not where. Thanks.
06-28-2006, 01:35 PM
I'm thinking of making a model of my boat to test her lateral stability with different loadings . A scale of 1 1/2 in.=1 ft. would yield a 3 ft. model , which would fit comfortably in my small tub .Is this big enough to be useful ? I know the waterline of the boat with 2 crew and all equipment and fuel ,so weighting the model down to that point would be my start ,I guess . It seems to me that knowing the exact waterline of the boat ( I've floated her ) ,at an accurately known weight , gives me a big head start on the problem .I know the seat ht. of the passengers ,the height of the outboard , cabin , and cargo area .When placing the representative weights at the center of the mass of these items , would the vertical distances above the waterline be measured at the scale of the model ?
The point would be to determine a sensible maximum load ,and to Identify a safe maximum tilt at that load ,and at a few lesser loads . Do you think I could get useful data ? The "angle of no return" could be determined by simply rotating the model with my hands . I'd get a direct feel of the boats resistance to heeling at various angles from which I'd determine a safe operating range .I'll install an incline measure ,like sailboats have , at least as I get a feel for my boat .
This is a shallow Vbottom outboard . will the boat in motion be stiffer than the static model ,as I've read?
Option B is to present an NA with the plans and commission the mathematical solutions to my questions , which I'm not dead set against,but it may be unnecessary if the model can yield good info .
Powered by vBulletin® Version 4.2.1 Copyright © 2013 vBulletin Solutions, Inc. All rights reserved.