Flotilla 6.2 Released

xuanyz,
here is a question for you.
Is there a single Chinese character for "symmetry"? Is it a symmetric character?
What about "anti-symmetry"?

 

I just realized I may have misled you a bit.
The exact hull condition for waveplot is only at the
undisturbed free-surface. It does not mean the wave along the hull is
exact all the way to the top of a wave, or all the way down to a
trough.

 

thank you very much
I have another question about the theories and application conditions on Michlet Flotilla and ZGREEN.
In other words, which theory dose Michlet use ? thin boat theory? only use for far field behind the hull ?
what about flotilla ? Does it use linear theory or nolinear theory? can it be used next to the hull? Does this theory think about bubbles flow around the hull?
what about ZGREEN ? I mean the theory different from the theories used in Michlet and Flotilla.

 

Michlet is based on thin-ship theory. Wave patterns are only accurate behind
the hull (say > 1 boat length).

Flotilla is an enhanced thin-ship code using linear wave theory.
It cannot handle unsteady bubbles, like in air-lubrication applications.

ZGREEN is a linear wave code. Because it uses isolated sources and sinks,
the concept of "thin-ship" is meaningless.

SID Viscous is another code I have released. See:
http://www.boatdesign.net/forums/design-software/sid-viscous-longitudinal-wave-cuts-44698.html
It can estimate the wave resistance from a given longitudinal wave cut. It
does not depend on thin-ship theory. The wave cut can be created by a
code for displacement ships or hovercraft, or you can use measured wave
elevations along a longitudinal cut.

 

In chinese symmetry means "对称" ，such as the single Chinese character "林"，from the left side to right or from the right side to the left ,the single Chinese character,"林",looks the same.
another example "日"，from the bottom to top or from the top to bottom ,it looks the same.

 

thank you very much.
what do you mean about "enhanced thin-ship code using linear wave theory"?
can I understand it as "Higher order linear theory"? second order?

I also want to ask you some theory or software about bubbles flow aourd a hull? Do you know it?
thanks a lot

 

No, it is NOT a higher order linear theory.
I have made some small modifications to thin-ship theory to better
approximate waves close to the hull, and also to the flow behind a transom
stern. I (and my former colleague E.O. Tuck) did not like the methods used
by some other researchers, like adding a hollow at atmospheric pressure
behind the transom. We found a much simpler way of approximating the
flow and its effect on resistance.

When I get time (maybe in version 6.4) I will also include some additional
calculations to work with the viscous wake behind the hull. That is one
application of the perturbation velocities, especially the transverse velocity.
That velocity (shown in the 2nd graphic I attached earlier) can be thought of
as a "tearing" velocity, pulling the viscous wake apart sideways along the ship's track.
I can't think of many applications except remote ship detection yet.

No, sorry.
Maybe you could try OpenFOAM or some other CFD code.

 

than you very much ,I learned a lot tonight . Best wishes for you .
thank you

 

In Flotilla 7.x I have abandoned the crude form factor approach because it is
not suitable for all hull shapes.
I am now working on estimating the viscous pressure resistance which
(hopefully) should have wider applicability.

 

Leo, I was out on the lake - blue skies, warm air, steady strong winds - ideal sailing weather - so it took 3 days till I saw your post.

I am currently going through all the Delft-hulls and it is impossible to come up with a meaningful estimate for the form factor. I was hoping to replace my BL-simulation (costly computer clock-time) by the simple product of friction coefficient*formfactor, but I will scrap this idea now.
I can show the development of the skin-friction coefficient along a real yacht-hull compared to that along a flat plate of the same length. The characteristics are totally different. A streamlined body has a much higher cf at the forebody and a much lower cf on the tail, compared to the flat plate. A form factor approach might be possible for commercial ships (e.g. tankers) with a long parallel middle-body, but not for yachts or other ships with appreciable waterline curvature.
And this is just the skin friction coefficient! As you say, on top of that, one needs an estimate for the viscous pressure resistance!
I can't count any more the different methods that I tried and how often I went to bed frustrated at the end of the night.
Good luck to all the brave men, who try to simulate this!
Uli

 

Uli:
I share your frustration. I have also tried many dozens of different
schemes. I am now on my last attempt, one which was inspired by a
discussion I had with Prof. Bill Webster (from the Uni of California)
about 20 years ago. His advice about form factors was to only consider
the aft half of the hull.

Like you, I also tip my hat to those who attempt complicated schemes to
estimate viscous pressure resistance. It is an awful problem, and made
worse by having to consider hull slopes, which can be very sensitive to
the accuracy of the offsets. That's why I prefer other methods that by-
pass pressure integrations!

Good to hear that you are getting some good sailing weather. It is
foul here, but it means I got to stay indoors and watch Germany v Brazil!

 

Uli:
I'm not sure if I sent you the paper on this topic a while ago.
Kouh, Jen-Shiang, Chen, Yen-Jen, and Chau, Shiu-Wu,
"Numerical study on scale effect of form factor",
Ocean Engineering, 36, 2009, pp. 403-413.

It compares scale effects on viscous pressure resistance of 5 quite
different hulls, namely a tanker, destroyer, fishing vessel, Wigley
hull and a torpedo. IMO it's the most important paper on the
topic that I have found recently. Email me if you want a copy.

 

Leo:
I don't have this paper but I am surely interested.

I am trying to show you where I stand at the moment with the following attachments and would appreciate your comment.
As I already said, the friction coefficient along a sailing yacht is quite different from that on a flat plate. The diagram "Sysser24" shows the comparison. The first B.L.-trip on the model is 0.02 meters behind the bow and the second trip is located 0.35 meters behind the bow. None of the curves for the flat plate is similar to the flow along the hull.
Surprisingly, if I use Hoerner's equation for the form factor I get an estimate for the Delft-hulls that is close within 8% of the measured values. This is by no means sufficient for accurate power- predictions, but not as bad as one would expect from the diagram "Sysser24".
I have modified Hoerner's equation slightly, interestingly you also proposed to use the length of the aft-body in your post #41. The equations are shown in the attachment "FormFactor" (equations are difficult to type in the forum).

PS.: once every 4 years during the world cup I develop an interest in soccer. While my program is running on one computer I watch the game on the other one. Usually my program is running much longer than the game, so extra times and penalty kicks are O.K.

 

Uli:
We seem to be trying very similar methods! I have thought about using
Hoerner's formula too, but I decided to do that after I exhausted some
other possibilities, e.g. Stratford's BL separation criteria.

I am also looking at the section area (and girth) as a basis for BL
separation and flow curvature effects because I think (or hope!) they
will be numerically more stable than the more usual pressure
integration using offsets. I have read several papers that mention the
difficulties CFD and other techniques have, especially near the bow.
I often check my wave resistance calculations using a pressure
integration of the (far-field) linearised pressure. In 95% of the
cases it matches to within 1% using Michell's integral, but every so
often the offsets are not smooth enough and there can be differences
of up to 20% between the two methods. Viscous pressure calculations
can vary by even more than that.

A difference of 8% in viscous drag might seem large to you, but I
regard that as a very reasonable start for yacht hulls at relatively
low Froude numbers. As you appreciate, it's a very difficult problem
and one where the experimental scatter could be much greater
than 8%. I'm not sure if that's true for the Delft hulls because I
haven't had time to trawl through the data to see if they have
conducted multiple runs.

I am using Wigley hulls for my calculations. They are not quite as
curvy as yacht hulls, but at least the experiments (for L/B=8 and 10)
cover a very wide Froude number range. When I have something
reasonable I will include it in the next release of Flotilla. I am
just trying to find a more generally applicable form factor method
than the crude one in version 6.x.

P.S. Like you I get interested in football every 4 years. I doubt I
will ever see a game as hilarious as the Brazil-Germany game again.
Shame. I could watch those sort of matches every 2 years

 

If you have time, can you explain the definition of your friction coefficient and how you got the value for the Sysser 24 hull?