Page 546 - Forensic Structural Engineering Handbook

14.36 MATERIAL-SPECIFIC FORENSIC ANALYSES

Over subsequent winters, snow had accumulated on the roof, although not up to the
25 psf design load. In 1989 flooding from broken pipes indicated a problem, Al Bassett,
P.E., S.E., was called and during his inspection not only got to see the fractured lower few
laminations but witnessed the explosion accompanying the fracture of the next lamination.
He had a contractor immediately construct shoring towers to prevent total collapse. There
was no snow on the roof. His investigation indicated that this girder was overstressed with
dead loads. The allowable flexure stress was reduced by lamination grade adjustments and
a size factor of .805 based on research since the beam was originally designed. Since his
investigation, the allowable stress for such large glulam beams would be reduced by the
volume factor of 0.64 replacing the size factor. This is a good example of fiber damage
accumulation progressing until failure. This large glulam beam failed under lesser load than
the beam had resisted previously.

Case Study 10. LaGrande Middle School Gymnasium, LaGrande, Oregon
This case study illustrates impact effects and resulting fiber damage.
A light snow had accumulated on the roof of the school gymnasium when one of the
glulam roof beams collapsed. During the initial site inspection, Donald Neal, P.E., S.E.,
noticed wood putty filling a vertical scratch in the beam in the area of fracture. Further
investigation determined that during construction, the architect had directed the contractor
to cosmetically apply the wood putty. A concrete tilt-up wall panel had fallen after being
released from the crane and the corner hit the side of the recently placed and anchored glu-
lam beam prior to purlin and sheathing placement.
This presented the author, working with Mr. Neal, an almost storybook problem for ana-
lyzing the energy and resulting forces and fiber damage from the impact. The spring con-
stant for the beam was easy to determine accurately without complications of other attached
members and sheathing which also were not there to absorb some of the impact energy. The
panel for which the weight could be accurately determined started in a stationary position
without rotation restriction at the base nor an initial velocity from a moving crane. So the
resulting damage was not just on the side with the scratch but from flexural tension dam-
age at the far side and lower corner of the beam. This beam accumulated fiber damage from
one impact that led to its failure 5 years later. Since the snow was light, the reduced effec-
tive section from the initial fiber damage may have created continued fiber damage accu-
mulated with dead load that could have led to eventual failure without the light snow. The
theoretical damage was just below the proportional limit where it would be visible, thus
making the beam susceptible to the failure that occurred. After the failure with hindsight
and time allotted for such an analysis it was a problem with a solution that matched what
had happened. The question is whether we would have advised the architect to do exactly
what he did after seeing a small scratch on the beam without performing the appropriate
analysis at the time of construction.

Case Study 11. William Shore Memorial Pool, Port Angeles, Washington
This case study illustrates inadequately sized tapered glulam beams, dead load in excess of
design, accumulation of fiber damage, tension lamination fracture, and field installation of
additional laminations.
After being called to investigate horizontal splits of glulam roof beams covering a pool
facility, the author was not too concerned, considering the possible variable moisture con-
ditions in a swimming pool enclosure. The first observations confirmed the low signifi-
cance of the minor splitting and checking that had occurred. However, in the process of