Our deck core samples for testing duplicated the deck core of our test boat, a 32-foot PDQ catamaran. Some other boat cores will be more strongly constructed; some will be lighter.
For each test sample, we hand-laid an 18-inch square section using one layer of 17-ounce biaxial glass cloth on the inside, half-inch end-grain balsa for the core, and two layers of 17-ounce biaxial cloth plus one layer of 6-ounce plain cloth on the deck side. West System 105/205 epoxy (www.westsystem.com) served as resin. We confirmed the soundness of the laminate by testing off-cuts from each sample to destruction.
All testing was based on the force exerted by a quarter-inch, 316 stainless-steel bolt, typical for a winch mounting on this size boat. Clamping force was estimated as 20 times bolt torque in inch-pounds. The standard American National Standards Institute (ANSI) clamping force of a properly tightened bolt is about 1,400 pounds; the proof strength is 1,950 pounds; and the minimum breaking strength is 2,700 pounds. The maximum working load of this bolt under the test boats sheet winch is about 600 pounds, or a 4.5:1 safety factor based on minimum breaking strength.
It is mandatory when mounting through-bolted equipment to cored hulls to seal the core against water entry (see Inside Practical Sailor blog posts Trouble-Free Deck Hardware Installation, Aug. 12, 2013, and Rotten to the Core, Nov. 3, 2015). Water can cause rot in balsa and plywood cores, and delamination in all core types (ice expansion is deadly), leading to structural failure and expensive repairs.
On the majority of test samples, we sealed the core and provided for bolt compression with an epoxy plug. We tested several samples without plugs to establish baseline compression strength.
To determine the baseline compression strength of the balsa sandwich, we tested several washer and backing plate materials by clamping through the laminate with no epoxy plug and with quarter-inch steel plate for support on the topside. Considerable sponginess was evident with plain washers, fender washers, and Starboard. All other materials felt firm when tightened to normal torque (60 inch-pounds is typical for quarter-inch bolts).
We then tested without a topping plate, since most loads are upward pulls. The samples were placed centered over a four-inch wooden ring backed by a quarter-inch steel plate, creating a -inch by 4-inch unsupported annulus. We then gradually tightened the center bolt through the laminate and backing plate or washer candidate material, observing bolt torque at initial inside skin failure (cracking noises) or core sheer failure (silent but visible), 1/8-inch deflection, and failure, failure being the point at which either the bolt failed, the plate failed, or the plate became so distorted that the nut would not turn. Torque was measured with a calibrated load cell.