Flotilla 6.2 Released

This free (64-bit Windows) version of Flotilla is an enhanced thin-ship
code that estimates the total resistance, squat and wave patterns of
monohulls.

Version 6.2 fixes some minor bugs described in the thread for the
previous version. Other changes include better estimates of wave
patterns very close to the hull (e.g. within one beam), some additional
output (e.g. hydrostatics and metacentric heights) and several new
examples.

Latest features, tips, and updated reports can be found at:
www.cyberiad.net/flotilla/flotilla.htm
and
www.cyberiad.net/flotilla/flotilla_mono.htm

INSTRUCTIONS
To install the program, uncompress the zip file into a suitable
directory. To uninstall, just delete the entire sub-directory.

The manual is in the "doc" sub-directory.

 

Pure enjoyment!

Leo, may be you are willing to unveil the source of the spline offsets for the Delft 372. van't Veer's thesis shows only a rough sketch. One would need a lot of creativity to turn this into an offset-table.
Thanks for your gift to the world
Uli

 

Uli:
I took the lines from very much enlarged graphics in some of the
papers I referenced. They are not deadly accurate, but they are
good enough for comparisons. I stress in the manual that the users
primary task is to capture as well as they can the displacement,
wetted area, waterline length etc. I suggest that they should be
within about 1% of the published values to get reasonable results.

Flotilla can produce a table of offsets from the spline files, much
like Michlet does. I left that out of the public release, but I
might include that capability later.

Attached are a spline file for Flotilla and a table of offsets.
I will add these to the catamaran version as an example. There are
many recent experimental results and CFD calculations I can use as
comparisons, so it is an important test case.

Note the "stepped" appearance of the lines near the stern. These are
ugly, but they have a negligible effect on calculations of resistance,
squat and wave elevations. Flotilla (and Michlet) use clever
integrations which smooth out those steps and small bumps. Those
steps in the lines would be very bad news for codes that use
derivatives of hull offsets, for example in surface pressure
calculations, or in the original Michell's integral which was
presented in terms of hull slopes. (For those who care: a "simple"
integration by parts was used to express the various equations
in terms of actual offsets.)

If you can produce smoother lines, then I'll be happy to use them
instead of my rough effort.

Leo.

 

the file _offsets.zip was already included in the flotilla_mono_....._globals_offsets.flo folder.
I converted this file into an excel-file und created a diagram for each station. These sections are very smooth, without any "stepped" appearance. That was the reason why I asked in my previous post for the source of your linesplan, I was impressed by the quality. You seemed to be a good Loftsman!

It is strange that these good splines result in such stepped lines in the on-screen display.

 

It depends on the number of stations and waterlines that you choose at the
top of the offsets.flo file. Some combinations will look smooth, others will give those ugly steps.

 

irritating....
Is it just a display-issue and the internal calculation is using the correct spline, or does it actually affect the basic curve of sectional areas?

 

The program first fits splines through each section, and then fits waterline
splines through interpolated points on the section splines. It's just an
inadequacy of using cubic splines that pass exactly through the supplied
points. I guess that's why NURBS and other splines are used in actual hull
design programs (which Flotilla is not!).

As I said, it makes very little difference in calculations, as long as you have
captured the displacement, wetted area, and BWL reasonably accurately.

When Delftship gets back online properly, I will see if they will modify the
Michlet output option to include a couple of lines that will make creating
offsets easier for some users. Freeship has made a start to that effect, but I
haven't checked if they work properly yet. I know I joke about "Russian
brides", but every time I visit Freeship sites or Ukrainian marine sites I end
up getting spam and weird requests.

Added later: Of course, it is unreasonable to expect simple cubic splines to
approximate rectangular sections very well. That can cause some unusual
oscillations in the lines.

 

Has anyone tried Flotilla with Windows 8.1?
I just got an email from someone who got a "compatibility error" on a laptop running 8.1.
Any advice?

 

Toshiba P75-A7200 laptop with updates to Windows 8.1 (14APR14)
i7-4700 CPU, 8Gig RAM, SSD

Batch file "monoexwp" ran with no errors 0.31718E+02 seconds (wall clock).

Michlet does run oddly in windows 8. The window redraws slowly (e.g. control-V to show sections) and when a calculate drag command (control-R then t) is executed there is no Rt graph, only Rw and Rv but the graph box is scaled for the Rt line. If you issue a "c" instead of a "t" in the second step the total resistance graph is visible.

Windows 8.0 and 8.1 really, really, suck. I am so frustrated with it that I tried to install Windows 7 on this laptop and was ready to pay the 170 us dollars to do it. Unfortunately this laptop did not cooperate.

 

Just some more runs of "monoexwp"

from the test of flotilla 6.1:

0.51780E+02 seconds (wall clock)
i3-2310M 2.1Ghz 4Gb Windows 7 Home Premium
Windows experience index Processor: 6.4, Memory 7.3

and from another run on my desktop computer:

0.64254E+02 seconds (wall clock).
AMD A10-5800k (3.8Ghz - 4.2Ghz turbo), 16Gb 1600Mhz RAM, ssd
Windows experience index 7.3 processor and also 7.3 for memory


Seems almost any reasonable Intel processor beats AMD in single thread math intensive computing. (but watch out and dont get quagmired in that damn Windows 8!)

 

Very much appreciated, John.
I agree that 8.1 is a beast, but maybe the graphics problems in Michlet are unique to your laptop.

Leo.

 

Delft 372

I modeled the hull in Rhino and used the smoothing function to get sections with a continuous curvature. Unfortunately this does not work for the waterlines, they are still slightly wavy. The result is the attached IGES-file. For those people without Rhino I have attached the corresponding file for Delftship. Offset-files for Michlet and Flotilla (.flo is not a valid extension in this forum) are also added.
Hope you have fun
Uli

 

Great work, Uli.
Which version of Delftship did you use? I couldn't load it into version 5.06.186.

 

The free version 6.27.259
The updates come automatically

 

Delft 372

I compared the residuary resistance coefficient to the predicted wave resistance. There is a 33% difference at Fn=0.45.
Did I make a mistake?
Uli