Teaching Aids ...
 (Continued from Page 7)

weak features of their designs. Suc-
cessive refinements usually resulted in
higher load to weight ratios. The
solutions finally arrived at by the
seven teams were quite diverse in
nature and ranged from truss-like
configurations through various plate
solutions to a barrel-vault approach.
Some of these shapes can be seen in
the accompanying photographs.
 The structures were tested to fail-
ure using a steel double-tree system
which can be seen in Figures 4, 5,
and 6. Failure was defined as a com-
plete loss of load-carrying ability, i.e.,
a total collapse. Local failures and
deflection limits were not considered.
Midspan deflection was measured,
however, and varied from about 0.25
inches to 0.70 inches (at approximate
ultimate load) for the various de-
signs.
 The dead weights of the structures
ranged from 0.130 lbs. to 0.459 lbs.,
and the superimposed collapse loads
ranged from 38.2 lbs. to 319.5 lbs.
The load to weight ratios achieved
ranged from 227 to 733 with an
average value for all seven models of
527.
 Most of the failures occurred just
outside of one of the load locations
where the flexural stresses were still
rather high and combined with shear-
ing stresses to maximize on a diag-
onal plane. In only one of the de-
signs, the barrel-vault type, did fail-
ure occur near midspan. These results
tend to substantiate the weakness of
the material in shear. It is interesting
to note that no glue failures occurred
and that when a joint did fail, the
fault lay with the material.
 The assignment was well received
by a large majority of the partici-
pating students and proved very bene-
ficial to the learning process because
of its tangible nature. The student
was dealing with a real structure,
something which he helped to con-
ceive and construct. During the pre-
liminary "design" tests and the final
tests the student observed a great
variety of structural behaviour. He
was able to compare and contrast the
actual behaviour with his own pre-
dictions of strong and weak points in
the design. Some of the observed
behaviour included the elastic insta-
bility of slender compression mem-


Fig. 4 A folded-plate
design, (w cross-section)
under partial load.


9>17M"
 St^*!. ^kj


Fig. 5 -A


semi-elastic arch solution with loading brackets


in place.
in place.


Fig. 6 The barrel-vault model under partial load.


bers, lateral buckling of the entire
structure, plate buckling and shear,
diagonal tension, compression crush-
ing (perpendicular to grain), and
torsional effects due to the concen-
trated line loads and the inconsist-
ency of the material.
 In conclusion it is important to
recognize that this type of problem
can in no way replace the traditional


approach to the study of theoretical
structural behaviour. It can, however,
serve as a very dynamic method of
illustrating the validity or the lack of
validity in applying certain theory to
certain design situations. This type
of problem very readily points up
both the complications and simplifi-
cations involved in the application of
any theory.
 THE FLORIDA ARCHITECT