"Steve Lusardi" wrote

(some really, really, bad structural advice)

It's important in engineering something like this to consider the inevitible
flexing that takes place with concentrated loads. Discontinuities and
changes in materials, expecially of different stiffnesses, are trouble. I
can't think of anything much worse than introducing an interruption of the
inner skin. Any micro cracking around the aluminum in this arrangement will
let water into the core.

This is something that needs to be designed for deflection rather than
strength. Think of a too tight jackstand pushing in the side of your hull.
It's in no danger of breaking but you wouldn't want your autopilot ram to be
doing that.

I would not want you to build something this critical according the these
ideas without further advice from someone familiar with fiberglasss boat
design who can look at the actual installation but I think a solution should
include elements like these:

Drill the smallest holes in the inner core that will let you remove core
with an "L" shaped piece of wire chucked into a drill and a shop vac. Pack
these voids with chopped mat and resin to tie the skins together. Do this
for about a foot around the attachment point and then put a layer or two of
glass over the whole area.

Much depends on how close existing stiffening such as bulkheads are to the
ram attachment point. If the hull is fairly stiff due to other structure, a
large pad for the ram may be suffiicient. Find out the tensile strength of
the epoxy bond to the hull and divide by about 4. Divide the maximum force
the ram can exert by this number and you will need that many square inches
of pad for bonding.

You need to give some careful thought to the rudder stops. Will the ram
limits now become the rudder stops? If so, the rudder could put still
larger loads on the ram, especially in a grounding. If the rudder can hit
the stops, they need to be able to accept the full strength of the ram in
case the autopilot gets into a mode where it thinks it needs to be hard
over. This could be through a control failure.

If the area of hull where the ram is going to go is unsupported for a long
distance, consider a web frame of glass covered plywood a few inches deep on
each side of the ram running up and down as far as possible. The ends
should land on other structure or be gradually tapered. Put one of these on
each side of the attachement point and then mount the ram on a header
running between them.

Here is a way you can evaluate your hull stiffness without engineering:

Install the ram against a temporary mount glued to the hull. Rig up a way
to pump oil into it with a hand pump, (a high pressure grease gun might do)
and a gauge. Piston area times PSI - force. Fasten a timber from the rail
to the ground and rig a dial gauge on it touching the hull. Record some
force and deflection readings and plot them on a graph. Run a line out from
zero through the points untill it reaches 1500 pounds, or whatever maximum
ram force is, (remember that rudder force could be greater if the ram
becomes the rudder stop). If predicted deflection is fairly minimal, you
probably don't need much additional stiffening. This sounds like an
involved operation but hiring and engineer can be expensive too.

Whatever you do with a point load like this, think about keeping all
structural members and transitions tapered. Visualize the whole thing as if
it was flexed and distorted way out of shape and look for areas where things
would bend sharply or loads would be concentrated. Designing things to flex
properly is often more important than strength.

--
Roger Long