theoretical displacement hull shape for min drag

I shouldn't have used 5% from memory. Most are a bit better than that.

From Gothenburg 2000...
Code: S/L^2
Measured: 0.149
FinFlo: 0.1558
MGShip: 0.1485
CFDSHIP: 0.1550
ICARE: 0.1486
UNCLE: 0.1520
CFX: 0.1517

Mean: 0.1519

And, yes, most of my work with Tuck was concerned with the wave-making of thin ships and moving (over-)pressure distributions.

Leo.

 

To me the most interesting aspect of arriving at the lowest drag hull for given speed and displacement constraints alone is that the wave drag ends up being very low.

Although the evolved hull shape is highly dependent on the wave drag, its actual value is almost negligible in determining the total drag of the vessel. Hence the accuracy of determining the friction component plays a more significant role in the total drag than wave drag.

It is enlightening to watch the evolution of a hull or hulls during an optimising run in Michlet. It inevitably makes sense when you see the output and reason it out but the result is sometimes surprising. Those so enlightened realise the benefit of analytical methods over empirical methods. Extrapolating from existing forms simply does not match what analytical based approach offers.

Maybe one day there will be sound analytical methods for determining the viscous component of hull drag and some will find smart ways to lower it. For now I am happy that Michlet provides an accurate estimation of wave drag from fundamental physics and GODZILLA has the power to minimise the overall hull drag; albeit relient on the in-built empirical based determination of viscous drag.

Rick W

 

Leo

As far as I am aware Molland et al, described 2 methods for determining form factor, One being the raising the transom clear of the water, as you noted, but is based on Prohaska’s work. They didn’t actually suggest doing this. He indicates that it is only suitable for conventional ships below 0.4Fn, and hence inappropriate for Higher Fn numbers. Also that it introduces more complexities.

Ahh, ok, so Michlet is estimating the Cw. And how is this verified to obtain a degree of accuracy or convergence?

Yes, we use the ITTC line.

We have sometimes, when time and money allows, measure the sinkage of the waterline around the model/ship…these in reality have only been done when funded by EU projects, since we are not really interested in this, only the end result. However we did do so for a client where wash was a prime objective in securing the contract.

 

Rick

"..Extrapolating from existing forms simply does not match what analytical based approach offers..."

and what analytical methods do you use to validate and verify this?

 

I tend to treat the whole thing as some sort of weird computer game with artificial life-forms. For me the most interesting point was Godzilla's "discovery" of the hulls and diamond arrangement that almost completely eliminated waves at one speed:
http://www.cyberiad.net/waketet.htm

That sort of thing used to really bug Tuck until he could prove mathematically why it should be so.

I don't think that we ever satisfactorily explained the strange "optimal" pressure distributions shown in:
"Free-surface pressure distributions with minimum wave resistance"
http://www.cyberiad.net/library/pdf/tl01.pdf

At low speeds they are a bit like the weird and wobbly Ward Optimum Sym. Ship Shapes, but acting on top of the water. At high speed there is definitely something like hydrofoils at the ends.

That's the great thing about wave-making theories. They might not always give perfect real-world answers, but they can provide interesting insights. (Or to use Don's analogy, they might not get you all the way to Rome from London, but they can get you as far as the middle of the English Channel.)

Leo.

 

From my own experience I am heading up the Thames at least in calm water. May be in the Channel when the wind driven waves need to be accounted for.

Actually in sheltered water with short fetch distances the windage is much more significant than the wind waves.

Rick W

 

I just noticed that they tried the bow down test for only one hull.

I still find the use of form factors as large as those shown in Molland et al (i.e. from 1.2 to 1.7) very worrying.

By running the code and looking at the results. You might change your mind about declaring Michlet "Dangerous" after you see what it does and under what assumptions.

Is that because it is easy to use on a calculator? Or because it is still the accepted "standard"?

Cheers,
Leo.

 

Leo

I've read their papers quoting as high as 1.45, but not 1.7. Which paper shows this as a result?
He has said though that the 1.45, whilst still being on the high side and is still unclear why it is high, does however, suggesting it may be owing to using different towing points compared to the original experiments.

Any software is "dangerous" until it is fully verified. To suggest otherwise is fool hardy. It may be ok to suggest as such when the only thing to loose is face amongst ones peers. But try telling that to a client when their 40m, 400 passenger, 40 knot, $10million ferry only goes 35knots, and the only explanation one can give is, well, the software has yet to be fully verified but we're told it's close enough. That's all.

Until another "standard" comes along that is proven to be more effective and consistent and verifiably so, then we shall continue to use the ITTC.

 

I will need to sleep on the hydrofoil pressure distribution and its significance.

I had never even thought of hovercraft making waves. I might drag out one of the models I have.

Rick W

 

Report 71, 1994

Monohulls: 1.23 - 1.45
Cats: 1.40 - 2.18

Fair enough. But so strongly criticising other people's attempts without knowing what they are actually doing is a bit over the top.

It's a very versatile beast. I've seen NASA papers where it is used (after correcting for compressibility effects) to predict skin-friction on re-entry vehicles.

 

Ahh..that’s the only paper I don’t have no.71. So can’t comment further on that. Do they state why they get higher coef’s and how they arrived at them?

As for verification and the likes.
You have slowly given details of the “caveats” and hence limitations of the programs. Even to the extent where you say you have no way of knowing if the results are a validation of anything, and that you’re not validating them anyway, as it is not your field, to mention a few. In light of this and other comments, outside of the world of pure research and theoretical field, which you reside in, once you cross over into using such software for real hulls used in anger all those caveats come into play, rather ‘seriously’ too I is have noted, hence your comment. This is not to say it is not a useful tool, as it is, especially in the theoretical fields for others. It also one hopes, shall eventually lead to furthering the science in such areas, as this is always welcomed.

But it is just a tool and not an absolute. When someone who is not a professional naval architect uses this ‘tool’ to supplement their gaps in knowledge, or inexperienced novices or students etc, it becomes dangerous. Owing to the fact that there is just unconditional acceptance of the results without any hint of “is this real and believable” when there are so many caveats to start with. (Unless of course one stays within the cevats -always!). If it were me, I would not be so laissez-faire about how it is used so wantonly; since it reflects on you early on in your career. But that is my own personal opinion.

In doing my own PhD, the ideas, language and words I use, and have to use, are totally different to the “real world” in which I make a living. The two are more often than not incompatible.

If any of the naval architects I’ve had under me and trained up came to me with colour plots of wave trains and ‘optimal’ hulls and suggested we redesign the hull as to the output, I would have to seriously question their analytical ability in the professional world. One must always ask questions “why” and “how” and obtain independent verification as much as possible. When designing vessels that cost millions of dollars, that I do, I’m not interested in personal feelings, just quantitative results and a professional attitude that demonstrates the ability to critique results, regardless where they come from to justify their bold statements. I’ve worked with some highly intelligent naval architects; some just could not work in the professional field, owing to being constantly questioned, they preferred the esoteric field…and so left. Sad.

It’s a simple as that.

To paraphrase Don, I would rather a naval architect I have trained get to Rome methodically then get to Paris in a fast and brilliant unique way, but get lost afterwards or stuck in never ending traffic and never seen again.

 

The caveats and limitations of thin-ship theory are very well known. I assumed you knew the limitations of the program which is why there was some confusion.

Here's the first two main paras of the manual just so we are on the same page

MICHLET IS RESEARCH CODE.
Please check all estimates generated by Michlet against
experimental results before committing any time or funds
to your project ...

and

Michlet is a computer workbench that can be used for investigations into some aspects of ship hydrodynamics. Although it is not a ship design program, Michlet can be used for preliminary design work such as estimating the resistance, wave elevation patterns and bottom pressure signatures of monohulls, multihulls and submarines. GODZILLA, the optimisation module of Michlet, uses artificial life algorithms to search for hulls of minimum resistance, or for hulls with other specific characteristics.

I agree with you that some people will stray outside the caveats despite warnings, but that's their look-out IMO. If someone blows themselves up because they designed a pressure mine using Michlet without knowing about what they are doing, I can hardly be held responsible. You might as well blame Mr Simpson for making his well-known integration method public.

As to the large form factors in Molland et al, in their favour they do stress that the factors "may not necessarily be used directly for design or resistance scaling purposes". Part of the reason (for cats) could be because of wave-breaking between the demihulls, or viscous interference.
For thin demihulls, where skin-friction is the main component of drag, the magnitude of the factors is particularly worrying. On the other hand, the models are only 1.6m long, so maybe they should only be used as a rough indication of what might happen at full scale. Like some mathematical models, I guess ;-)

It will be interesting to see what comes out of the DTMB5415 destroyer model tests around the world. At least they are being performed on larger models.

Leo.

 

And I'm just happy to fly them to Paris safely, and let them find their own way from there.

 

"..I agree with you that some people will stray outside the caveats despite warnings... I can hardly be held responsible.."
With all the caveats stated, that is fair comment. So long as the users has the intelligence to understand the limitations and caveats, he shouldn't blow himself up!


Ahh..the effects of wave breaking and spray et al...that is a perennial problem that affects the Cw, along with far filed waves, not to mention transom sterns. Although the viscous effects are generally more pronounced when hull separation (for catamarans) is low.

Still, without work like yours and others, the science and understanding wont progress...keep plugging away.

 

There are all sorts of messy issues when the demihull separation is low, including single breaking bore waves at some speeds in shallow water which I most definitely do not model yet.

Thanks.
I agree that the 15 or so comparisons of Michlet predictions with experiments you have seen is still not enough to verify the code yet. The real problem for me is getting hold of good data (with error bands) for a wide range of hull types and then collating the results. Maybe some other people could submit comparisons in return for the free code.

Cheers,
Leo.