[Public-List] chain plate spacers

Mon Mar 23 15:31:44 PDT 2009

Gordon et. al. Albergers,

Since properly connecting the rig to the boat is A), so important, and B),
something I need to do soon, I am going to jump on my friction vs. shear
soapbox again hoping others might confirm the logic.  I have attached a copy
of an epistle I wrote back in 2004 (please note disclaimer) below.  

The basic argument is, by putting enough tension (tightening) on the bolts,
the chain plate is forced against the knee/bulkhead with enough force that
it will not slide in the vertical plane (direction of load).  This
"friction" connection is quite strong and is the design used on most bolted
connections for steel building structures.    

The problem with shear a connection is it relies only on the strength of the
weakest connection material over less than half of the surface area of the
bolts.  This makes the shear strength of the bolts irrelevant because the
shear strength of the wood thru which the bolts go is less; the bolts would
elongate the holes in the wood long before the bolt would break from shear.
Also, with a shear connection, only bolts in exact alignment with the holes
in the chain plate would achieve their full shear strength.  

With a friction connection every bolt applies friction so their strength is
cumulative.  I checked the tensile strength of a 1/4" 304 SS bolt and it is
about 3,700 pounds.  Assuming a 1.3 safety factor (see below) and a 6,900
pound breaking strength of a 1/4" SS wire mainstay, sufficient friction
strength could be achieved with 3 - 1/4" bolts if each were tightened to 80%
of maximum load.  So, it appears 5/16" bolts are only necessary if one wants
to move up to 9/32" mainstay wire.  In fact, larger holes would reduce the
area of the chain plate and possibly degrade the ability of the plate to
distribute the needed compression load across sufficient area (the plate
could bend at the narrowest part beside each hole).  

Another issue is the compressive strength of the bulkhead (marine plywood)
and knees (hardwood).  This could be field tested by torquing down a bolt
(does anyone know how to convert tensile strength to bolt torque) with a
couple of layers of wide washers on each side and observe if the wood
crushes.  To me, this is perhaps the most important test since several
owners have mentioned water leaks and water damaged bulkheads/knees would be
the weak link in any connection detail.

Does this stuff make any sense or is my version of reality drifting yet
further away from the norm?         

Best - Roger 148

Aug, 2004
Albergers,

RE the thread on reinforcing the knees, I checked with the NA who surveyed
my boat and learned FG tape has a strength in shear of 8,000 pounds per
square inch.  The breaking strength of the 1/4" lowers is 4,700 pounds.  To
insure the stays break first (comforting thought), the chain plates should
hold 1.3 times the stay breaking load and the knees 1.5 times, or 7,000
pounds.  The load on the knees is transferred to the inside surface of the
hull (which, like the main bulkhead, is a good, strong diaphragm) via the
vertical FG tape on each side of the knee.  Assuming the knee is 8" high (I
haven't measured) or a total length of 16", the thickness of the tape should
be a minimum of 1/16" (7,000 pounds / 16 inches = 440 pounds/inch; 8,000
pounds / 440 = 1/18" tape thickness).

I plan to drill a little hole in the tape to check thickness but suspect it
is greater than 1/16."  So, the knees are probably adequate but, for the
belt-and-suspender folks, easily reinforced by simply adding new layers of
tape to the existing.   

RE the chain plate size, I started looking into the shear strength of
stainless steel and learned what a megapasquale is, and also learned its
strength goes down drastically over 400 degrees so, let's hope global
warming doesn't catch up with us too quickly. Then I realized an easy way to
insure adequate chain plate strength is confirm that the minimum cross
sectional area above or beside the turnbuckle fastener pin is at least 1.3+
times the cross sectional area of the SS stay the chain plate supports.
Since the area of the main stays is 0.20 sq. in. and the lowers is 0.11 sq.
in., chances are the chain plates are more than adequate.

Our structural engineers tell me the majority of structural failures occur
at the mechanical connections so, I suspect the weak link is the chain plate
connection at the knees or bulkhead. Unless a friction connection is
employed, the entire load will be placed on the upper part of the bolt holes
in the knees and main bulkhead (shear connection).  A friction connection
involves compressing the material between the chain plate and backer plate
(or washers) sufficiently so the friction between the different surfaces
prevents them from "slipping" (like Chinese handcuffs).

The Gougeon Brothers (West System) suggest friction connections under load
(just about everything on a sailboat) be "bonded" by adding a layer of high
compression, adhesive epoxy (West makes a slick powder additive) between the
surfaces to be connected.  This insures friction across the entire surface
area, not just the "high points."  They also point out that the bond must be
rigid.  If the bond is soft and flexible (5200?), the load will cause
movement which will degrade the friction into a shear load on the bolts.
Once this occurs, the bolt shafts work against the holes resulting in
substantial reduction of load capacity and leaks (like the toe rail at the
genoa track?).  

The wood knees on #148 appear to have the grain oriented vertically so the
upward load from the stays runs parallel to the grain.  No big deal with a
friction connection but real a problem with a shear connection because wood
is much weaker parallel to the grain than perpendicular to it. The need for
a good friction connection argues for backer plates as large as the chain
plate and, perhaps even increasing the size of the chain plate.

Assuming the above actually makes sense (time for disclaimer; the author is
not a structural engineer and all said above could be total BS; so, rely on
it under pearl of tumbling mast), my plan to insure stays are properly
fastened is to check thickness and quality of the fiberglass connection
between the knees / bulkhead and the hull.  Then, make backer plates about
the same size as the chain plates and fasten them with epoxy and a bunch of
tension on the bolts.  My main bulkhead is in good shape but I have heard
some are rotted, particularly if water seal between chain plate and deck has
not been maintained.

Hope this helps sort out the issues, it did for me; but, then, I find it so
easy to agree with myself.

As always (jealous of those with boats in the water),

Roger

KSBA 
________________________________

Architects/Planners/InteriorDesigners/ProjectManagers

3441 Butler Street
Pittsburgh, PA 15201
N 40° 27.8344'  W79° 57.9831'

412-252-1500 ext.101 
412-779-5101 cell 
412-252-1510 fax
r.kingsland at ksba.com
www.ksba.com

-----Original Message-----
From: public-list-bounces at lists.alberg30.org
[mailto:public-list-bounces at lists.alberg30.org] On Behalf Of gordon white
Sent: Monday, March 23, 2009 3:04 PM
To: public-list at lists.alberg30.org
Subject: Re: [Public-List] chain plate spacers

Washer-type spacers are a bad thing, if I understand the description as to
how they are used. They apply a bending moment to the bolt rather than
straight shear. Reminds me of a vintage racer whose rear axle was a couple
of inches wider than  his suspension setup. He used spacers much as you
describe to bolt the axle into the car. He was running on a race track just
ahead of me when one of the bolts broke, letting the car's chassis dig into
the track. The car flipped over and killed him.
Not a good thing.

But if you put a shim the full width of the chainplate you can mostly
correct the situation, even though I would not recommend it if it could be
avoided. Certainly up-size the bolts.

- Gordon White
Brigadoon II
_______________________________________________
These businesses support your Association:
http://www.alberg30.org/store/A30supporters.html
Please support them.
_______________________________________________
Public-List mailing list
Public-List at lists.alberg30.org
http://lists.alberg30.org/listinfo.cgi/public-list-alberg30.org

 1237847504.0