I come rather late to this interesting discussion.  I have one note of caution to add here.  While a quick reading didn’t pick up an problems with the numbers, an assumption that informs Gravity Groper’s argument requires further examination:

  “this tells us the wall thickness could actually be reduced on the e-glass section and arrive at the same safety factor for each design.”

This claim overlooks the critical issue of fatigue.  In a nut shell, my argument is that we should not apply the same safety factor to glass and carbon composites.

1.       Equal stresses will generate equal deflection but proportionally larger strain in the larger glass section.

2.       Strain is the dominate factor in fatigue analysis.

A rule of thumb is applied in designing boats for longevity.  If the element in question can comfortably survive a million cycles it should suffice.  Carbon epoxy composites are capable of doing this at 60% of the stress level that causes failure.  Glass epoxy composites can only survive a million cycles when the applied stress is less than 40% - and 40% seems an optimistic figure (S-glass epoxy cloth: 20%; glass epoxy uni: almost 40%; curiously, biaxials: roughly 30%).  (Digressing, polyester strain to failure is 1 or 2% depending up the resin system chosen.  Which is why stresses in glass polyester laminates must be kept comfortably below 10% of the strength of glass matrix.)

Unpacking this info somewhat: carbon fails at 2% strain whereas glass and epoxy do so at about 5%.  Given this, at high levels of strain in a glass epoxy matrix, the entire matrix is more highly stressed than would be the case in a carbon matrix.

Returning to the mast case (and as all ready stated): the glass section will be subjected to greater strains than the smaller carbon section, for a given deflection. Rob’s choice of a safety factor of 2.5 (40% stress) for a carbon mast section seems reasonable.  This figure would not be reasonable for the equivalent glass section.  I argue that the safety factor for the glass section ought to be 5 (and certainly no less than 4).

My current state of development has moved to engineering of the structure, notably cross beams with an initial exploration of mast characteristics.  Hence an increased interest in this topic.  A laminate of predominately longitudinal carbon uni with of off axis glass reinforcing is showing some promise (glass is roughly 25% of the laminate).  It should be noted from the preceding that the level of stress in each ply must be examined in a hybrid laminate of this nature.  Also, the buckling characteristics may be different to that of the all carbon laminate and as a result, the rule of thumb, wall thickness no less than 3% of section diameter, should not be assumed.

David