During a beat against a stiff breeze, at the NZ rally,
Fantail's yard took on a dramatic bend in the centre.
I have long had misgivings about this way of making a yard, and now that I have seen a failure at first hand, I feel emboldened to speak out against the method, of taking a small diameter straight tube and adding a reinforcing tube above it, spaced away in a shallow triangle by piece of sheet or plate.
A great deal of stiffness is gained - but only in the plane of the reinforcement.
A little anecdote: a lifetime ago, I raced dinghies, and on one boat, I made a boom from a piece of wood about 4" deep and 1" thick. As soon as I applied tension to the kicking strap, the boom immediately collapsed sideways, choosing the easiest way to bend. There is an unstable equilibrium position where the load is in line with the strongest, stiffest axis, but if anything at all causes the load to be applied away from this axis, then, sooner or later, dramatic failure must result. This can easily be demonstrated by cutting a profile of such a yard out of stiff card or thin plywood, suspending it from the mid point and progressively loading it at the ends.
PJR gives proportions for a rectangular-section wooden yard of 1.54:1 for the depth/width ratio, and this is clearly safe enough, because it's been used for so long. I would tentatively suggest that a ratio of 1.75:1 or even (perhaps) 2:1 might be tried, but under no circumstances would I want to make a yard with a greater depth/width ratio.
I specified a round tube for Footprint's new yard. A round tube has equal stiffness and strength in all planes and is clearly not prone to instability.
A square tube has equal stiffness when a load is applied in line with its faces or across its diagonal - but, counter-intuitively, when a load is applied across its diagonal, it has only 71% of the strength that it has when a load is applied in line with its edges. So even here, there is a possibility of unexpected failure.
The situation gets worse when the cross section is oval, elliptical or rectangular; but as I said above, a ratio of 1.54:1 has proved acceptable. At any greater ratio - well, I think the risk is unacceptable.
For Tystie's new yard, I will be using a round tubular section. I will be laminating it from carbon fibre braided sleeve, so I have the opportunity to increase the wall thickness in the centre, which I will do. The wall thickness will be the same all round the circumference, so there will be no "weak" axis. However, I will be attaching the halyard by means of a soft rope span to the one-third and two-thirds points along the yard, so that I need not make the yard excessively strong horizontally in order to get sufficient strength and stiffness vertically. This practice has been used for a great many years on the peak halyards of gaff-rigged boats, albeit with wire spans: Dyneema and Spectra are the modern equivalents.