Michlet and immersed transoms

Hi,

Michlet estimates the additional drag produced by the immersed transom as a "hydrostatic drag". The accuracy of the drag calculation is most important when you optimize a hull shape.

This is a typical result of Godzilla optimization of the total drag in wide range of speed, with the length and displacement restricted, max beam and min GMT limited.
You can see the big, square, immersed transom. Its area is the biggest section area of the hull.

However, the drag calculated by Michlet is actually pretty small, and the hydrostatic component at small speed is about a half of the skin friction.

As I know, the form drag of the cut-ended body is 2-3 times bigger than the sharp-ended one, so the 50% increase of drag by the transom seems to be unrealistic.
In fact, I've never seen such hulls in real ships.

So, the question appears: under what conditions the Michlet hydrostatic drag is a good estimation of the drag produced by the immersed transom? Especially, for low speed. And, how to optimize the transomed hulls in the wide range of speed?

regards

krzys

 

Broli Taut! (Is that Polish, or some other mysterious language?)

I think you might be expecting too much from Michlet to optimize a
transom stern vessel at low Froude numbers. That is a very difficult
flow regime for all codes, particularly if the transom is not running
fully dry.

Form drag for transom stern hulls is a controversial issue. If you
take the point of view that form drag is mostly due to BL separation,
then (IMO) the concept of form drag for the dry portion of transom
sterns at high Froude numbers is meaningless. I am not alone in
having this view: See, for example:
Min, Keh-Sik and Kang, Seon-Hyung,
"Systematic study on the hull form design and resistance prediction
of displacement-type super high-speed ships",
J. Marine Science and Tech., Vol. 3, 1998, pp. 63--75.

(I prefer that view to the model of Molland et al who use form factors
of the order of 0.3 to 0.5 for fairly fine hulls with smallish transoms.)

I have extended this thinking to form drag estimates of pointed and
cruiser sterns. If the form drag is due to boundary layer separation,
then that should only occur for the wet portion of the hull near the
stern. (Thus, in my more recent models there is a Froude number
dependence as well as a Reynolds number dependence).

I have a much better (free) program called Flotilla that you will soon
be able to use. See:
www.cyberiad.net/flotilla/flotilla.htm
(I hope to release it here on boatdesgn.net by Monday.)

Transoms are handled better in Flotilla because I am able to use
some near-field computations, such as the height of the wave at the
transom. That can't be done in Michlet because it is purely a far-
field code.

I am now finishing off the (60+ page) manual and documentation. (I
hate writing manuals, so I'm glad you pulled me away from that dull
task with your questions!)

I am also working on making an easy method to input hulls using the
Michlet Export Option in Delftship and Freeship. Step-by-step
instructions for how to do that are available now at:
www.cyberiad.net/flotilla/flotilla_mono_ds.htm

I don't know if I will have time to put that feature into version 6.1.
I should probably consult Martin at Delftship and Viktor at Freeship
before committing myself to a fairly weak solution.

Your original original question was:
"...under what conditions the Michlet hydrostatic drag is a good
estimation of the drag produced by the immersed transom? Especially,
for low speed. And, how to optimize the transomed hulls in the wide
range of speed?"

1. "Hydrostatic drag" is due only to the loss of hydrostatic pressure on the
transom. It is not the complete story: there is wave-breaking and spray
and much more to consider.

2. You could use the Godzilla option of optimising for more than one design speed.

3. Or you could try OpenFoam.

Good luck!
Leo.

 

Your version of Michlet/Godzilla appears to have more freedom of shape than mine. What hull series did you use - series 8?

There is no way to get an aft flaring waterline with my version 807mt with any published series. It looks like F1 can vary along z, which would be very handy for transom designs.

 

I have allowed some greater freedom in shape since 8.07 but I have
always felt a bit uncomfortable about it. As always, constraints and
practical limitations need to be carefully applied to the problem
or some very strange shapes can emerge. If I allowed complete
freedom of shapes then many "designs" would end up looking like
Ward's Optimum Symmetric Ships, or even weirder ones that look like
double-headed marlins, with a huge spike at the bow and another at
the stern.

I'm not sure that Michlet is the right tool for fine "transom design".
As I mentioned in my previous post, Michlet lacks some capabilities (like
near-field calculations and wave profiles) that can have a large effect
on the hydrodynamics.

Leo.

 

Michlet version 9.30 and series 20. The result obtained after 60 000 iterations.

Leo, great news about Flotilla! I am waiting...

I realize, that the estimation of form drag is difficult. I hoped, that there are experimental results that help to estimate the range of speed and transom size (especially depth), where the Michlet predictions are (more or less) accurate.

For example, let's analyze two hulls of the same dimensions and displacement - they are actually Mini 650 boats. BTW, both are developable, designed to be built of plywood.

The first hull is a typical planing one, with a sharp bow and wide, flat transom. The second one is quite different, it resembles old working boats. Note the deep, triangular transom.

Michlet predicts almost the same drag for both hulls at wide range of speed!
Of course, the dynamic lift can reduce the drag of the first hull at high speed. At lower speed, the form drag of the second hull seems to be underestimated. So, I expect that the second hull will be worse. But, will it actually? And how much worse?...

regards

krzys

 

Your wait is over
http://www.boatdesign.net/forums/design-software/flotilla-6-1-released-49750.html#post677563

Forget it: you would be wasting your time.
Your main problem is that you are using Michlet for hulls that are definitely
not thin.

I'm not sure what is best to use: if Flotilla can't help (and I doubt it
will if Fr < 0.35) then maybe OpenFOAM, or maybe something else somebody
else here can suggest.

Good luck!