Roger Long wrote:
Larry wrote:

http://www.thomson.ece.ufl.edu/lightning/

This sailor has done extensive research at ufl.....


Thanks, that was one of the first things I turned up. This quote is
especially applicable to this thread in view of Wilbur's comment that
an ungrounded boat is less likely to be struck:

"While the individual estimates varied widely between surveyors,
there is no support for the argument presented by some sailors that
they should not ground 'their sailboat since it will increase the
chances of it being struck by lightning."

The statistics presented in this article make a good case for just
forgetting about it unless I plan to sail south, which I do hope to
do at some point.

My original post was actually prompted by this and a couple of other
web articles. Here is my situation:

http://home.maine.rr.com/rlma/Keel.jpg

The mast stanchion is essentially equal to a keel stepped mast.
Since my boat was originally a keel / centerboarder, the keel is
large volume and I doubt that Endeavour spent the money for a keel
casting. I'm quite sure the ballast is just stacked lead pigs in
resin. Lighting current going through that stuff would be like a
bomb and the high resistance at the bottom of the main conductor
would create extensive side flashing.

For reasons not evident on the crude drawing, any grounding plates
have to be outboard of the cabin sole. The Thomson paper says not to
let grounding conductors contact the hull but I have no choice if I
am to maintain the maximum radius recommended by other sources. The
reason for overkill on conductor and ground plate size is to
compensate for the tight conductor radius and need to run the
conductors close to the hull skin.

-- Roger Long
The problem is, lightning doesn't like going round corners and *will*
jump back out of the lightning conductor at the bend.

E.M. Thompson states:
At this time the peak lightning current is generated, during the
'return stroke'. Although cresting at ten thousand to hundreds of
thousands of amps, it only lasts for about a millionth of a second.

This 1 us pulse considered in the frequancy domain has a fundamental at
1 MHz and various unspecified (we dont really know the pulse shape)
strong harmonics going *way* up. As a mimimum, the grounding system
*MUST* have a low impedance from DC to 10 MHz which essentially
prohibits anything except large radius bends through less than 90
degrees. Due to the 10 KA current pulse, an extremely high voltage
will be developed accross any bend, so there is an extreme risk of
arcing to the hull if the bent strap is in the bilge. If its touching
the hull, you basically needn't have bothered fitting it from the point
it touches the hull onwards to the grounding plate.

The other nice little problem is that multiple KA currents *will*
straighten out any sharp bends in the conductor carrying them and a
lightning strike consists of multiple strokes so the grounding system
has to survive in usable condition to be effective.

The main lightning conductors need to be joined to the mast support post
at least as high above the grounding plates as the lateral distance
and led in a smoothly swept curve. Unless your pillar is very close to a
bulkhead this may not be practically achivable.

Once you've led the lightning below decks, you are basically screwed
unless you can lead it almost streight down and out. Once you've led it
below the waterline you are totally ****ed unless you can get it out to
large area grounding plates.

From your description, I dont see how anything except boring the top of
the keel downwards and outwards for thick pure copper conducters to
ground plates extending from the sides of the keel to the turn of the
bilge could help much. There are problems with this approach as well,
as the conducters will try to move quite a bit duing the strike and if
undersised and they go open circuit during the strike could cause major
explosive damage to your keel root.