theoretical displacement hull shape for min drag

as a thought, is the Series 89 catamaran hull series done by Burkhard Muller-Graf at Berlin Model Basin of any use for your verification?

 

Well, Mr. Willoughby
did that brighten your horizon?





.

Now it really started to go professional here. Thanks mates!

 

I have several series of cats that were in FAST proceedings and I haven't had time to digitise them all yet. I have enough for my resistance and squat predictions.

What I am really waiting on are the results for the DDG61 destroyer hull that is being tested at more than 30 towing tanks around the world. The complete set of data will allow some interesting comparisons of theoretical models and their capabilities, or lack thereof.

I'd also like to see more comparisons of drag predictions using the (IMO dodgy) ITTC line and physics-based methods for skin-friction.

Leo.

 

The series 89 is in the true tradition of German hydrodynamacists...lots of very useful data there.

I presented a paper at a RINA conference in 1995. Was awarded a RINA medal for well...anyway..but we were pipped for the silver medal. It was a paper on just such a subject of skin friction, was areal break through...unfortunately, can't remember the title nor who wrote, would have to dig around for it.

 

Probably Grigson. I know his work pretty well.

 

Form factor?

Let me jump into the loop here, if I may.

We use a couple different in-house implementations of linear wave-making code for slender ship calcs, and I find the one big empty place is the viscous addition of form and the consideration of displacement thickness (i.e., boundary layer). For example, Leo pointed out his concerns when using the small models of the Insel/Molland cat tests. (We also have used these for validation, with the same concerns.) In many cases, the magnitude of the viscous addition of form exceeds the wave-making (as would be the case for most of the Insel/Molland tests below 0.4-ish FN) - so accurate prediction of form factor would seem to be something that gets too little attention.

When model tests are available, you can often derive some sense for the viscous form addition (as was done for the cat tests), then add your wave-making calc. Sometimes, though, you are guessing at a form factor, if the test speeds are too high to determine where wave-making degrades to zero (and then you have to guess-timate if there is any immersed transom pressure drag to contend with). I can get great correlation by fitting it with an arbitrary form factor, but this does not validate the wave-making code, of course.

When doing a prediction for a new ship, you're left determining your own form factor, and any speed dependent correction you might want to use. (We often use Holtrop's 1988 form factor equation with his speed-dependency correction. It's generally a workable solution.)

I'd love to hear your comments on getting the right form factor and viscous addition. Has any one done any work on predicting form factor using the same distributed hull data that goes into the wave-making code? Something like form factor from a sectional area curve rather than simple hull parameters?

Ad hoc: I'd love to read your paper, as well. Thanks.

Very interesting thread...

Don MacPherson
HydroComp

 

I'll also ask a few Prof's here at the Uni i occasionally lecture at too, they are very helpful. They have endless data.....bit of a hidden gold mine really.

Grigson...ahh sounds familiar! Can't recall the title though.....still fair play they got the higher RINA medal than we did.

I've just dug out a short paper called "Short review of semi-displacement series for defining hydrodynamic characteristics of ships (boats)" by Antun Gamulin, have you seen this one?

Trouble is, i have around 2000 papers all stacked away in various files on a range of subjects....I've forgotten what half of them are about now some of them so old, and some highly confidential too, unless i start digging them all up i don't know what I may have lurking on my book shelf..

 

D.MacPherson

Yes, form factor has always been the "problem". as you nicely put it ".. I can get great correlation by fitting it with an arbitrary form factor, but this does not validate the wave-making code, of course..."..

Every systematic series I've ever read all use different 'fudge' factors, which speaks volumes about the difficulty. Perhaps it requires a different approach, since we are all "naval architects/hydrodynamacists" we end up thinking in pretty much the same vein. Perhaps input from say theoretical quantum physics may assist. Their work on Rogue waves, for example, pointed the way forward.

I hate all this management speak (its all bollocks), but can't help saying though, thinking out of the box is required...hence maybe a totally unconventional and different approach to the problem is needed, from an unlikely source.

If you PM me your email address, id be happy to pop over the paper.

 

EFD test data

Leo, I'm sure you've seen this site, but I'll post it anyway just in case.

www.iihr.uiowa.edu/gothenburg2000/

I've always wanted something like this for smaller ships and all of the other test data that's available in the public domain.

Don MacPherson
HydroComp

 

Thanks, Don. I know the Goth2000 papers well. I love the estimates of surface area by some of the codes - some couldn't even get that right to within 5%!

I addressed the horrible form factor a bit in the first paper in the list on my site at http://www.cyberiad.net/leo.htm

But what I found isn't anywhere near the final word. I would have liked to have done more, but I just couldn't find enough good experimental data and eventually I ran out of time and a supervisor.

All the best,
Leo.

 

Many Thanks to you and Ad Hoc. I am quite a novice here, am given to methods that are sometimes called " Form Finding ", but not so much interested in shapes defined a priori (standard parabolic waterlines/transverse frames or fourth degree buttocks) for hydrodynamics that follow as a consequence. Before I can read the above topics, would appreciate your comments. If the unknown shape is expressed i.e., the offsets are geometrically given in Monge's form y = f(x,z), fluid dynamic and mathematical formulation can be therein incorporated, or so I think, keeping wave drag, skin friction as parameters which can be modelled into a simulation numerical program representing all forces acting on each element and then integrating for full hull having minimum total resistance using a combination of algebraic, pde (or ode). That could be solved later numerically using standard mathematical softwares like Matlab, Maple or Mathematica. As a start we may assume symmetry longitudinally about midship, ignoring transom/bow dissimilarity as a crude starting mathematical model, assume port/starboard mirror imaging and disallowing sudden changes at body line bilge chines.

To illustrate the methods by two examples: a) Finding a stable position of a string of constant linear density hanging between two points, the integral (z ds) potential energy is minimized resulting in a catenary shape. b) Finding the profile of a rocket with minimum aerodynamic drag: Isaac Newton had arrived at the surface of revolution shape by using calculus of variations for the first time. In both cases differential equations describe needed optimal shapes.

Could a reasonably simple program be written defining simplified hull shapes with various parameters for given displacement and speed ? I was expecting F(p,q,r,s,t) = 0 sort of pdes for complicated and odes for simpler cases...

Best Regards

 

I think yes, and I urge you to look at Wehausen's work. G. Weinblum also looked at the problem in the 1950's. As I said, depending on the assumptions you make in your model regarding wave resistance and skin-friction, you might end up with (at best) having to evaluate a number of Mathieu functions. I'm not sure how well Matlab and Mathematica handle those nasty functions for a wide range of parameters. It might be better to get specific C or Fortran code.

In a more general case you will have to solve the problem numerically.

The "optimal" hull forms found in the 1960's were very strange wobbly shapes. One, named Ward's Optimum Symmetric Ship was bundled with Michlet 8.07 as an example. Aslo see:
Ward, Lawrence W., Wave resistance surveys on a ship model of minimum resistance, Webb Inst. of Naval Arch. August 1965, pp. 16. (+ errata).

Experiments on these wobbly hulls showed that their wave resistance at the design speed was very low, as predicted, but the skin-friction was high because of the large surface area, and the form drag was high because of possible early separation.

Other people who have looked at optimal shapes include Ada Gotman of U. Novosibirsk, Russia and Doctors and Day, e.g.
Day, Alexander H. and Doctors, Lawrence J., Resistance optimization of displacement vessels on the basis of principal parameters.

Day, Alexander H. and Doctors, Lawrence J., Design of fast ships for minimal resistance and motions, (submitted to) Sixth Int. Marine Design Conf. IMDC-97, June 23-25, 1997.

I'm not exactly sure how where to find pdes for hull shapes that you will need. Ward and some others used Fourier series if my memory is not completely gone.

Have fun! You have a lot of work ahead of you!
Leo.

 

Padava
Are you familiar with Leo's Michlet/GODZILLA software?

Rick W

 

His question makes clear he is not.

But Rick, are you familiar whith the theories behind?


Your statements in the past made obvious you´re not! Now let the real pro´s discuss the issues, and let´s stay back a while.

Regards
Richard

 

"..Padava
Are you familiar with Leo's Michlet/GODZILLA software?

Rick W.."