Page 222 - Forensic Structural Engineering Handbook

7.6 CAUSES OF FAILURES

When it is not practical or possible to weigh loads directly, their magnitudes must be
estimated from geometric data collected at the site, or from circumstantial evidence that
may be used to infer the loads that may have been present.
In some cases, the weights on a structure can be estimated with sufficient accuracy
by measuring or estimating the volume of each material in the area of interest and rely-
ing on standard tables 1,2 of unit weights to calculate the weights of each component and
the total weight on a structure. This approach may be appropriate for rough estimates
of weights for early evaluation of the significance of gravity loads to the causes of a
failure.
Dead Loads. The first place to look for estimates of the weight of the structure itself is
the construction documents used to construct the building. When you use this resource,
be certain to assess conditions that may have changed during construction and those
changes that might have been made after the building was finished. It is common for
building alterations to change the weights. Over the life of a building, its roof may be
replaced several times. Since different roofing systems have different unit weights, such
common alterations can increase or decrease weight. Sometimes curtain wall systems
change, or new flooring materials are added. Each of these changes can affect loads on a
building frame.
Even when it can be established that the original construction was not altered, often it
is insufficient to rely on construction documents to estimate volumes of structural materi-
als in a building. Standards of practice for installation of construction materials acknowl-
edge that variations in material quantities are unavoidable, and allow for deviations from
specified design dimensions. For instance, a concrete slab designed to be 4 in thick and
3
built within tolerance can be 4-3/8 in thick. With this acceptable variation from the
intended dimensions, the deadweight of a slab can be more than 9 percent higher than is
normally used for design. Further, deviations exceeding those allowed by normal stan-
dards of care should be investigated whenever gravity loads are suspected to have con-
tributed to a failure. Through variations in care during concrete placement or due to
structural deformations that occur during installation of wet concrete, it is not uncommon
for concrete floor slabs to vary from the specified thickness by 20 percent or more in local
areas—particularly near the middle of slabs cast on potentially flexible structural elements
such as open-web joists.
When appropriate for forensic analyses, samples of each material or building compo-
nent should be taken to determine the actual unit weight. Then, with precise measure-
ments of volume, accurate estimates of total weight are possible. When the greatest
accuracy is required, all materials associated with a failure should be collected and
weighed, if such effort is feasible. Care must be given to include all materials, and only
those materials, affecting the structure at the time of the collapse. Further, to the extent
possible, the location of all materials at the time of the failure should be established and
recorded for analysis.
Live Loads. Loads that are derived from the use of the building normally vary with time.
As such, without an on-site survey, it is difficult to conduct anything but the most rudi-
mentary estimates of the live load on a structure at any particular time. Usually live loads
must be established through a thorough inventory of the debris.
For loads that are derived from solid materials, accurate inventories often are possible.
However, after a major failure it is difficult usually to relate objects to specific locations in
the building. The substantial disruption that a major failure causes obscures much reliable
evidence about the arrangement of building contents. This problem often is complicated
further by rescue operations, which justifiably are not directed toward preservation of evi-
dence for investigations.