SID VISCOUS - Longitudinal Wave Cuts

Hi Leo, you seem to have a very productive phase these days, with high output. What makes you work so much, did you ran out of movies?

You might remember that I tried to compare the results from Michlet to the test data for a surface piercing NACA foil, as communicated by Kuhn & Scragg. The wave resistance was measured 1991 at the David Taylor Model Basin using the longitudinal cut method of Eggers, Sharma and Ward. I do not have the original 1967 paper of Eggers, but in a later paper he confines his method to the wave resistance for non-breaking waves. The wave resistance determined by the wave cut method yielded lower values than the thin ship theory used by Kuhn&Scragg and also compared to Michlet and compared to the residual resistance measured in parallel at the DTMB. I could imagine that the breaking mechanism dissipates energy and the wave elevation measured further away from the hull gives resistance values that are too low.

When you are applying your wave cut method to a theoretical wave pattern, as determined by thin ship theory, you work in a theoretical environment, that is not influenced by wave braking. But what does it say about the usefulness of a wave cut method to determine the wave resistance in a towing tank?

Regards
Uli

 

I am often slipshod in my terminology. To be consistent I
should call the resistance obtained from wave cuts as "wave
pattern resistance", Rwp. This does not include non-linear
effects such as wave-breaking. And it does not include viscous
effects along the ship's track.

There are a couple of different methods for measuring Rwp,
the most popular being the transverse cut method (which requires
two cuts perpendicular to the ship's track) and the longitudinal
cut method which requires only one cut for monohulls. There is
also the XY-method which uses one transverse and one longitudinal
cut, but I don't know if that is in use these days.

There are great problems measuring Rwp accurately in towing
tanks using longitudinal cuts. For accurate results, long cuts are
needed, and that is difficult without wide tanks where wall
reflections can be avoided. There are corrections available for
wall reflections, but it usually best to avoid them.

Transverse cuts also have their own problems, notably that they
cut across the viscous wake.

Finite depth effects can also increase the difficulties.
For example, as the vessel approaches the critical Froude
number, Fh=1, the wavelength of the transverse waves increases.
Close to Fh = 1, it might not be possible for a steady state
to be achieved.

You are correct that the wave pattern resistance will be generally
be less than than the residuary resistance, say, Rr = Rt - Rittc57.
I guess that you need to be aware of what is being measured and
reported, and make sure you are comparing apples with apples.

I could be even more pessimistic and say that the whole idea of
wave pattern resistance in a viscous fluid is not perfectly
defined. What are to we do when, a long way from the ship, there are
no measurable waves because viscosity has damped them out of
existence?

And, yes, we have run out of movies and are now listening to audio
books: hence the productivity increase.

 

Thanks for the clear words, this is in line with Egger's paper.
If I understand viscous resistance Rv = skin friction + pressure drag caused by boundary layer separation, then
Wave resistance Rw = Rt - Rv
In this case Rwp should be less than Rw

If this is true and I am interested in Rw, then the wave cut results from the towing tank are not very helpful. The computed results from Michelet should be closer to reality. Alternatively I could use the measured total resistance and subtract a realistic estimate (not ITTC57!) of Rv. How to get this realistic estimate of Rv is then another story.

Can you agree?

 

Thanks for the Eggers paper, Uli. I was using the earlier paper
with Sharma and Ward. There is a more recent study by Lalli et al:
"Longitudinal cut method revisited: a survey of main error sources",
J. Ship Research, vol 44, 2000, pp 120-139.

I doubt that SID will be of use to you: Michlet or Ulimich should
be fine for preliminary design purposes. SID might be of use for
those who want to try to simulate a towing tank, however, the
water will have to be pretty calm. It might be best to use multiple
probes at different distances from the track and average the results
in some sense.
How to estimate form drag (to get Rv from Rf) is a problem for most
people and I don't have any advice, except to look at previous work
with similar hulls.

I also don't have much of an idea of what frequencies people would use
to record data. 10Hz? 100 Hz? (Ad Hoc suggested to me that there would
be little use in logging data at 100Hz, and that 10Hz would probably
be quite reasonable.)

There are some esoteric situations (e.g. yawed ACV or SES) where wave
resistance is very difficult to estimate by "standard" numerical
methods, and it is easier to calculate the wave pattern, and then
use that to get the wave resistance.

 

I attended a presentation today during which Michlet was praised and then SID VISCOUS was described as a replacement for Michlet which includes viscous effects. That seems to be at variance with Leo's explainations of SID in this thread.

 

Thanks for the heads up, David!
SID (which stands for Ship IDentificatiation) is nothing even remotely like
Michlet.
It is better than Godzilla for estimating the size, location, principal dimensions,
speed, and direction of ships given a wave pattern. But that is probably the
least used feature of Michlet.
Hope you enjoy the rest of the talks at IBEX.
Leo.

 

Hi Leo,

Neat work. Get your commercial head on, couple it to a Synthetic Aperture Radar and go submarine hunting! Or does it require the vessel to be surface piercing?


Love the name btw.

 

Thanks PI, but I am sure that towing tanks and various miltiaries
have far more capable programs. This version of SID is the simplest
embodiment of some very well-known techniques, although I have
included some unique features that are hidden to users at the moment
because they aren't relevant to the wave resistance problem.

Submarines do not have to be surface-piercing, but they must make
waves at the surface. Similar techniques to those used in SID could
be used for deeply-submerged subs when they make "internal waves"
but it's not something I'm interested in at the moment.