Page 306 - Practical Design Ships and Floating Structures
P. 306
28 1
reference point reference point
main ontoo on
part part
initial design z
Optimized design
d., averaged downtime Pd = 64%
k, averaged downtime Pd = 27%
free variables
I all I
designs
pontoon breadthlheight ratio B~/T, displacement V 50.000m'
pontoon centre line draught dp
VCBIL of column from waterline &< pontoon length
pontoon centre line shift A+
Figure 5. Geometric properties of initial design and the semisubmersible optimized with respect
to downtime due to excessive accelerations at the reference point.
_._..
A,, /A',,, ratio of cross section areas of pontoon main part A,, and central part A, ,
B, / T, ratio of width B, and height T,, of pontoon cross section (all parts),
d, draught of pontoon centreline (all parts),
4,' normalized vertical centre of buoyancy of column 4,' = VCB / L, , measured from
waterline.
AYP shift of pontoon centreline with respect to column centreline in direction of y-axis;
breadth is increased when Ay, is positive.
Fig. 5 presents the main geometric properties of initial and optimized semisubmersible design,
respectively. The increase in pontoon centreline draught d, decreases heave exciting forces. This ect
is amplified by shifting displacement from the central pontoon part to the pontoon main part, i.e.,
A,,,,, /A,,, is increased, and the outward shift of the pontoon centreline by Ayp =1:86m. Due to a
decrease of 4,' a pronounced shoulder in the profile of the column is developed. Heave added mass
and damping of the new configuration are adopted by reducing the B,, /T,, -ratio. The expected
downtime Pd is decreased considerably for all wave directions. The optimisation process extends the
region of feasible sea states especially where high probabilities of occurrence are present (Fig. 6).
The decrease in acceleration levels is reflected by a significant improvement of motion behaviour. Fig.
7 shows the response amplitude operators of surge ( s, ), heave (s?), roll (s, ) and pitch ( s5 ) motion.
The wave heading is 120 deg, which corresponds to the most probable direction of sea states (B
=240deg). In all cases and all ranges of wave frequencies lower motion amplitudes are achieved. Note
that the amplitudes of heave motion are overestimated in the range of the resonance frequency and
underestimated for the cancellation frequency. This is of course due to the lack of viscous effects in the
hydrodynamic analysis. Indeed, the excessive heave resonance motions provide a positive side effect,