Conference unifix::sailing

Gee, I surprised no one has offered any advice yet ...... Oh well.

Clearly, steering failure it one of those extremely nice things to 
avoid, so this is one of those times if it doesn't look much too strong,
it's too weak. I'd get myself a disk sander, a bunch of coarse disks, a
good respirator, disposable coveralls, and start grinding. I'd stop
grinding when I reached the hull laminate. That is, I'd start the
installation all over from the beginning. 

Once you have well-sanded, clean hull laminate exposed, use multiple 
layers of cloth and epoxy resin. Matt isn't as strong as cloth. Orient 
the cloth in various directions to make the strength uniform. After you 
get everything glued together, taper the edges by sanding (be sure to 
make your layers of cloth and resin oversize a bit) to avoid stress 
concentrations. Apply a final coat of epoxy to seal everything. This 
probably won't take all that long (but, as always, double your initial 
cost estimate and triple your initial time estimate). 

I'd say 'have fun', but this isn't likely to be that, especially the 
grinding part.

As I understand it, your pulley installation has the pulleys mounted
against the hull, so that the fiberglass installation is under tension,
i.e. the pulley mounting wants to pull off of the hull when you turn the wheel.

This is inherently a weak system, since the stress distribution at the
attachment point is difficult to predict, and since the materials are stressed
under tension, rather than under compression.

It might be possible to make a much stronger installation by placing
a stout piece of wood transversely between the port and starboard hull
surfaces, and attaching the pulleys using bolts through the wood. The wood
can be glassed against the hull at both ends. This system will be much
stronger, since the pulley force can be taken up by compression against
the OPPOSITE side of the hull, through the wood beam.


        \                                        /
         \                                      /
          \                                    /
           \                                  /
            \                                /
             \                              /
              \ pulley             pulley  /
               \  (O)                (O)  /
                \--|------------------|--/
                 \ |    wood beam     | /
                  \|------------------|/
                   \                  /
                    \                /
                     \              /
                      \            /
                       ------------


--RS

re .2:

Hmmm, I always thought that fiberglass (GRP) was quite strong in 
tension and relatively weak in compression. It seems to me that the most
likely failure would be failure of the bond between the added fiberglass
laminate and the hull. If this bonding area is sufficiently large, the
bond shouldn't fail. After all, the load is spread over several square
inches so that the load per unit area isn't all that high. 

One problem with what is suggested in .2 is that you create a small area 
on each side of the hull that is much stiffer than the surrounding hull. 
Under wave impact, the hull will deflect around the beam and could well 
fatigue and eventually crack. This looks like a good example of the type 
of hard spot that should be avoided. Secondarily, it might not be 
possible to fit a beam across the hull and the beam would have to be 
quite stiff to avoid bending under load. I would guess that the load on 
the steering cable sheaves wouldn't exceed a few hundred pounds at 
worst. The original design, done right, should handle a load like this 
without trouble. 

What did the surveryer suggest (if anything)?

re:               <<< Note 1587.3 by MSCSSE::BERENS "Alan Berens" >>>

>Hmmm, I always thought that fiberglass (GRP) was quite strong in 
>tension and relatively weak in compression. 

This may be true relative to other materials, but most (all?) materials are
stronger in compression than in tension. (Is material X more likely to fail
when you're sitting on it or when you're hanging from it?)
 
>It seems to me that the most
>likely failure would be failure of the bond between the added fiberglass
>laminate and the hull. If this bonding area is sufficiently large, the
>bond shouldn't fail. After all, the load is spread over several square
>inches so that the load per unit area isn't all that high. 

This is not entirely true, and the propensity for failure of this design
has been demonstrated already in this particular instance! If you make it
twice as strong, it will still have less than 100% safety margin. The failure
mode probably consisted of a "tearing" action, with the laminate failing
near the fitting first, and then lifting off the hull, much like a piece
of scotch tape being pulled off a wall. Here's a cross section drawing that
shows how the stress is concentrated:

                                   HH
                         _________ HH ________
      __________________/ ________ HH _______ \__________
      ___________________/ _______ HH ______ \___________
      ____________________/ =============== \____________

        ^ ^            ^ ^
        | |            | |
      not much        stress
    stress here        here

Adding extra area does not help at all until AFTER the high-stress area has
failed! After the initial failure, the high stress region moves outward,
and a lower normal force (due to shallower incidence angle), is applied over
a larger area. You can try this yourself: tape a thumbtack (point up) to your
kitchen counter with a piece of scotch tape. Pull up on the point of the
thumbtack and observe the tape. Does a wider piece of tape secure it better?

>One problem with what is suggested in .2 is that you create a small area 
>on each side of the hull that is much stiffer than the surrounding hull. 
>Under wave impact, the hull will deflect around the beam and could well 
>fatigue and eventually crack. This looks like a good example of the type 
>of hard spot that should be avoided.

This can be a potential problem, but it isn't likely to be in this area of
the boat, since the waves don't generally hit hardest from this quarter. The
hull is already fairly stiff in this area, and there may even be an existing
bulkhead that can be used in lieu of the beam..

>Secondarily, it might not be 
>possible to fit a beam across the hull and the beam would have to be 
>quite stiff to avoid bending under load. I would guess that the load on 
>the steering cable sheaves wouldn't exceed a few hundred pounds at 
>worst. The original design, done right, should handle a load like this 
>without trouble. 

I would guess the worst case loads to be at least ten times higher, since
these sheaves will take much of the shock load when your rudder hits a
wave or whale or submerged rock. The beam is only stressed from the ends, and
most of the force runs parallel to the beam, so there will be little deflection.

>What did the surveyer suggest (if anything)?

I'm not a surveyor (or a boat "expert" of any kind), so if his advice differs
from mine, take his!

--RS

re .4:

There's materials and then there's how the material is used. Using a 
rope for a chair (compression) isn't likely to be very satisfactory. 
Using the same rope to hang yourself is likely to work very well.

:-)

re .4:

You (my humorous last reply aside) raise some valid points. However .....

Assume for the moment that the material bonding the sheave to the hull 
is totally inelastic. Then the load will be distributed over the entire 
area. As the material becomes more and more elastic, the load will 
concentrate at the edges as you describe (the tape in your tape and 
thumbtack analogy is probably quite elastic). 

Obviously, no material is totally inelastic. But, it would seem to me, 
if the GRP bonding the sheave to the hull is sufficiently thick and
stiff (heck, use some carbon fiber tows -- not too expensive), then the
load will in fact be adequately spread over the attachment area and the
bonding will be sufficiently strong. Bashing the rudder into something
like a whale or granite ledge is likely to bend the rudder stock or
break the rudder. The steering system will fail at the weakest point,
which may well not be the attachment of the sheaves to the hull -- it
might be steering cable, the internal parts of the wheel pedestal, the
quadrant, etc. 

There is another solution ..... how 'bout a tiller instead of a wheel? 
(Sorry, that's already been thrashed in another note.)

    The surveyor recommended "beefing up the quadrants with additional
    layers of fiberglass. My hull has a canoe stern that does not allow 
    a board to cross in the way described. Worse, the Pullies operates in the
    horizontal plane with their backs facing outward toward the sea,
    parallel to the long axis (stem to stern) of the boat. 
    
    
    You do bring up an interesting point: my aft section looks something
    like this from the top:       ST=steering, dots are cable routes.
    
                      |        |  Cockpit    |        |
    		      |        |             |        |    
                      \        |    .ST.     |        /
    		       \       |.          . |       /
    			\      ---------------      /
    			 \  . 			.  /
    		           .P                    P.
    			     .                 .  <-outside hull shape
                                .          .
                                     .Q.    <- quadrant
                                   
    What if I moved my pulleys to the walls of my cockpit? I would then
    be compressing. I would also generate a much smaller angle that the
    cables would make to turn the quadrant. Would that make a difference?
    Should I beef up the quadrants as mounted or investigate mounting them
    in the wall of the cockpit somehow???
    
                                    
    
    
    

Relocating the sheaves sounds like an excellent idea.

You can get by with a steeper angle to the quadrant, provided that there is
always some cable laying along the sheave portion of the quadrant. You don't
want the cable continuously flexing at the same point, or it will likely suffer
a fatigue failure at that point.
    
Relocating the sheaves is a good idea, if you can find a reasonable way to
do it. There are a number of ways you can use some extra piece(s) of wood to
help relocate the sheaves. Here are a few possibilities:

   plywood laid horizontally, bolted to the cockpit floor, and glassed into
   the canoe stern. (Cut holes into it to clear obstructions or allow access
   to gear.)

   plywood or planks bolted to the cockpit sides and extending (slightly)
   beyond the aft edge of the cockpit.

   plywood or planks laid transversely across the aft wall of the cockpit,
   possibly glassed to the hull.

   thick pieces of wood against the cockpit sidewalls to move the sheaves
   further outboard.

I think you will find it much easier to complete a wood-based installation to
high quality standards. Fiberglass work is obnoxious and messy even under good
working conditions. Working inside your canoe stern will be very difficult. 

--RS

Fiberglass work is much more safe and pleasant if you use epoxy resin. 
Epoxy does not give off any fumes to speak of when curing and is easier to 
mix (in my opinion). I've worked with epoxy in confined spaces (my canoe 
stern) without problem. I wouldn't even consider using polyester. 

    Looks like the consensus is to move the pulley's inboard for
    compression loading rather than leave them in place and fiberglass
    reinforce. I'll be back down at the boat in two weeks and will begin
    seriously looking at how to do it.
    
    Thanks for the advice. I ALWAYS appreciate it!
    
    Robert