Ming-Shun Chang 
will realise that almost invariably what you see is a local wind sea 
with a swellunderneath, The wind sea tends to be of relatively short 
wave length. The swell comes from some distant storm ; it has been 
through the filtering effect of distance, so itis a narrow band. Even 
when you cannot see it with your eye the ship usually does see it. The 
local wind sea is usually doing one of two things : it is either growing 
or falling. So what you have is really a dumb-bell spectrum. You have 
a swell here from somewhere else and a wind sea developed locally. 
These are not taken account of in the fully developed spectra on which 
we have been basing much of our work, yet they are very important 
for the naval architect ; we cannot ignore them. 
I was on a ship just two months ago. It was not a normal ship 
but nevertheless it is an interesting case, The average wave height 
on one day was about four feet. There was a swell about 300 feet 
long for our 250 foot ship. We here heading into it. It was a quiet 
day. The surface even glassy. The local wind sea was virtually zero 
- and we slammed about 55 times per hour by count. The next day 
we had a local wind sea about the same height ; a fewwhite caps, not 
much ; a lovely day. The waves were much shorter. It was an abso- 
lutely wonderful day, because the ship was just alive in the water. 
The ship behaved completely differently. 
Over and over again we have used the fully developed spec- 
trum for characterising response when it is just plain wrong. We 
have to look at all of the seas that the ships encounter. 
DISCUSSION 
Michel Huther 
Bureau Verttas 
Paris, France 
I first thank the authors for the very interesting Paper they 
present. 
As the authors noted it, for ship behaviour calculations sea 
states representation by spectra is nowaday commonly used. The 
main problem for naval architects remains the question of the multi- 
directionality of the sea. I shall be pleased to know the opinion of the 
authors upon the two commonly used representations, i.e. the cosine 
1366 
