Hull Asymmetry and Minimum Wave Drag

This is the one: DTMB_1955_T257.pdf

I meant "obscure" as in "not widely known". It was intended as a compliment to you for your vast knowledge of the relevant technical literature.

 

(Bold added by me in the quote above.)

Thanks Leo. I suspected that's what you were doing given your other assumptions.

Another way to calculate the "dynamic" LCB would be to use your equations of equilibrium and determination of F2 and M2 in your report http://www.cyberiad.net/library/pdf/tsl01a.pdf

F1 + F2 = F0
M1 + M2 = M0

F1 = F0 - F2
M1 = M0 - M2

Then X, the location of the dyanmic LCB is given by:

X = M1/F1 = (F0 - F2)/(M0 - M2)

No assumption about the surface is needed for the calculation of X though there are assumptions used in the determination of F2 and M2.

 

In reverse order- Many attempts to analyze groups of tests and extract generalities or produce recursive formula suffer from a common problem- The vessels/models in the data set were designed with opposite goals in "measurement trim". A development class hull is designed to outperform in the common sense of the term; however, a handicap rule such as IMS,IOR,PHRF is designed to have the worst possible measurement score in measurement trim for a given performance in race trim. And it is the measurement trim from which all these data are referenced. This is a pet peeve of mine, so I appologize if the study in question did not make this mistake.

re "optimum". Among the characteristics of hull plating, the two most annoying are cost and weight. The impetus to trim away surface area from a hull is a primary consideration in hull optimisation. I don't see how optimum can be applied to a hull with out addressing this. One could treat it as a force of nature and derive a third nondimensional quantity. You could then look at how CoB varies with Re, Fn, and St (the Stinginess factor.)
I predict that the Stinginess factor would account for most of the variability in CoB. (The slenderness ratio is very closely related to this, perhaps good enough to be a proxy.)

So reason- the desire to reduce costs.
Mechanism- competition in the marketplace.
Basically, I think you are overworking the poor little creature (CoB). It's just one number that can conveniently be drawn on a linesplan. It makes for a nice sanity check at the end of the day.

I offer the following as a proposition-

Proven designs may show a marked correlation regards to CoB; but that in no way implies that boats designed with that CoB are optimal. So when you design a boat and discover the CoB matches good examples of the class you feel good about it. But adjusting a design with poor correlation with respect to CoB towards better correlation is pretty much a shot in the dark as far as optimisation is concerned once all the little details are figured in.

 

The use of the terms "optimum" and "optimal" has clearly caused confusion in this thread. Apparently folks are not understanding that the use here does not mean the "best" hull shape in some global sense, but rather the LCB which provides lowest drag when other hull parameters are held constant. That is the sense it is used in the references of the first post, and which I tried to make clear in subsequent posts.

There has not been any intention at all to suggest that that LCB is the single parameter by which a design should be optimizied.

 

That can already be done in Michlet.
It's a bit non-intuitive, but if you use a negative value for Ntheta, no BL displacement thickness is added, if a positive value of Ntheta is used, the displacement thickness is added before wave resistance is calculated.

This has some effect on small hulls, but it is just a hack, to be honest. That's why I call Michlet a "hydrodynamics workbench" - it allows users to play around with these sort of "enhancements". I might even tell people some of the other undocumented features they can play with one day (if I can remember them myself!)

The slipshod definitions really can be a pain! I analysed about 120 high fidelity velocity profiles to derive a skin-friction formula and some other quantities a few years ago. Some velocity profiles were fairly inaccurate at the outer edge, but many more gave a very clear indication of where the thickness ended.

In any case, I agree that it is best to use the displacement thickness.

 

Thanks, David. It's a good check on other methods I use.
I actually use a couple of methods internally for many quantities that users never see. For example, I also calculate the wave resistance by a pressure integration over the hull but I don't output the results. It makes the codes a bit slower when I forget to comment them out before release, but nobody has complained about speed yet, so it can't be too bad

 

Many famous hydrodynamicists (e.g. Havelock, Wigley, Wu, Maruo and Beck to name a few) have tried including BL displacement thickness on the hull, and some form of trailing extension behind the stern in an attempt to smooth out some of the humps and hollows in the wave resistance curve. None have really been successful. My guess is that they are all inconsistent - they improve agreement with experiments for some hulls, and do worse on others. Of course, in the publishing game, the improvements are published and lauded, while the cases where the theory fails never see light

 

Many designers have tried to find the "optimal" relation between LCB and LCF and come up with nothing convincing. I almost fell asleep during a boring America's Cup meeting where one group explained why LCF should be forward of LCB for "better sea-keeping", and another gave reasons for the exact opposite arrangement

 

The sets of data referenced are not "development" classes or boats designed to a rating rule. The studies are not what you assumed them to have been.

Series 60 was a set of systematic ship-like shapes tested to determine how the hydrodynamic characteristics varied with the hull shape parameters. http://www.dtic.mil/cgi-bin/GetTRDoc?AD=AD419990&Location=U2&doc=GetTRDoc.pdf

The Delft sailing yacht series is a long term effort testing several series of systematic hull shapes based on several different sailing yachts to determine how the hydrodynamic characteristics varied with the hull shape parameters. The results have been used to develop rating rules, but the shapes tested were not intended to be "rule beaters."

I believe Holtrop analysed the results of model tests of many ship hulls and tried to correlate the variation in hydrodynamic characteristics with the hull shape parameters.

I'm not familar with Jensen's research but I believe it also on ship type hulls.

See my comments in post #49 of this thread. Again, "optimum" was used in terms of lowest resistance, not in economic terms.

Altering LCB/CoB location generally has little to negligable effect on total hull surface area if other parameters such as length, displacement, prismatic coefficient, etc are held constant.

I was using "mechanism" in relation to the physics which causes resistance to vary with hull shape variation as characterized by LCB location when other parameters are held constant. Not economics.

There has been considerable research into the variation of resistance with LCB though apparently very little into the underlying mechanisms. LCB is frequently used as one descriptor of how the hull volume is distributed fore/aft. From what I've seen it's more than "a nice sanity check" reducing resistance/drag is of concern and is frequently a fundamental parameter for hull design. Larsson & Raven, in Ship Resistance and Flow which was published in 2010 as one of the volumes in SNAME's redone The Principles of Naval Architecture Series, gives it equal consideration with prismatic coefficient.

 

I know what you are getting at, but I can't see how it is possible to hold all hull parameters constant and only vary the LCB.
The closest I could get was with the (exaggerated) kayak hullforms I posted earlier.

I agree with Phil that you might be asking too much of a single quantity like the LCB (or CoB).

 

I'll start out by noting that the title I gave to this thread is Hull Asymmetry and Minimum Wave Drag, not LCB. I don't think I'm reading too much into LCB because I'm not reading much into it other than as a parameter that's used to characterize hull asymmetry. I'm really interested in hull asymmetry, particlarly volume distribution and it's impact on drag, not LCB per se. Other parameters could and should be also be used to characterize hull asymmetry. LCF is one possiblity but not the only one. But LCB is the one for which data is available. Does anyone know of any data based on other parameters?

Obviously you can't hold all possible parameters to be the same and only vary LCB. Anyone setting up a series of systematic models for tank testing or analysis has to decide what the fundamental set of parameters is and then how the shape will depend on those parameters. Different folks have made different decisions at different times. It seems that LCB has become a standard parameter for hull shape characterization.

I'm also suspicious of the assumption that the effects of LCB variation are largely independent of other parameters. But that calls into question the range of validity of empirical resistance (and other hydrodynamic quantities) methods in general.

 

I agree.

Well perhaps this is also where some of you understanding is blurred.

Asymmetry of a hull refers to the transverse plane, not longitudinal. You do appear to use definitions which are based either in different disciplines and hence cause confusion in the naval architectural/hydrodynamic field of definitions, or not a widely held as a common definition.

So, I’m just wondering have you designed a hull before?

 

I don't think my understanding is blurred at all. I'll refrain from other comments about that statement.

"Symmetric" is not restricted exclusively to transverse symmetry, at least in the small boats I'm interested in. It's also used to describe hulls which are fore-aft symmetric. Leo uses it in that sense in his kayak reports. Also in the in Leo's post from the other thread I referenced he used the term "fore-aft symmetric hull".

Perhaps some folks don't read carefully and/or jump to conclusions too quickly. Here is where I used symmetry in the first two sentences of the first post: Leo explained in another thread that Michell's thin-ship theory predicts minimum wave drag occurs when a hull is fore-aft symmetric. (Post #9 of A few Michlet/Godzilla questions) Fore-aft symmetric hulls have the Longitudinal Center of Buoyancy located at mid-ships.

Interesting that there hasn't been any suggestion in this thread until now that anyone was confused by the use of the term "asymmetric" and related to mean fore-aft asymmetry and not transverse asymmetry. I'll refrain from speculation about why it was suggested now.

Haven't designed a hull for someone else. Don't know what the relevance is for the question here. This thread wasn't intended to be about hull design but about hydrodynamics and hull shape.

 

Correction to previous posts. The authors of Ship Resistance and Flow which was published in 2010 as one of the volumes in SNAME's redone The Principles of Naval Architecture Series are Larsson and Raven, not Paulling. Paulling is the editior of the series. I'll correct the previous posts.

 

Anyone with additional thoughts about the variation of resistance as the volume distribution of a hull is changed fore/aft, and how that the effect of the volume distribution changes vary with Froude numbers?

There have been several comments that it probably isn't due to the boundary layer affecting the waves generated by the hull.

Looking at Leo's charts on the three kayak shapes it appears that differences in viscous drag may be a factor, but not the only factor.

My thought is it has to do with the wave generation itself since that very strongly dependent on the volume distribution and the Froude number. It should be possible to use Michlet to study it by systematically varying hull shape and seeing what happens to the energy going into the waves, provided sinkage and trim are adjusted.