Page 262 - Forensic Structural Engineering Handbook

8.2 CAUSES OF FAILURES

phase communications, and construction defects. Although some of these terms have well
established legal connotations and implications, there is no attempt to conform to those in
the following.

DESIGN ERRORS AND OMISSIONS

Range of Design Errors
Design errors have been known to lead to failures. These failures span a wide range of out-
comes from a failure to function as intended to outright collapses, often fatal. Although
some errors are attributable to negligence, the more interesting ones are caused by our
imperfect knowledge or appreciation of the intricacies of a structural system, the effects of
some of the loads it is subjected to, the limitations of various analysis method used in
design, and the dramatic effects that small details can sometimes have on an entire struc-
ture. It is from these errors that structural engineers have learned, and continue to do so, and
from which have evolved safer and more rational practices.
It is difficult to develop a clear or complete categorization of design errors, mainly
because most failures involve more than one error, and arguably because single errors often
do not lead to failure. Usually, factors of safety and the general resilience or redundancy of
structures provide sufficient margins to offset these. Generally though, design errors fall
into one or more of the following categories:

Structural Systems or Concepts, and Load Paths. There must be a continuous and sensi-
ble load path from the point of application of a load all the way to the foundation. Although
this seems evident, load path issues do arise when the design analyses and detailing lose
sight of the overall structural concept and become inconsistent with it. They also arise when
the structural concept itself is inconsistent with the loads it needs to resolve and resist.
Incorrect Loads or Incorrect Assessment of Load Effects. Determining the correct
design loads is not just a question of extracting values from a building code. It is a critical
element of the design. Even building code load values have and continue to evolve as our
understanding of natural events, such as winds and earthquakes, and their effects improves
and as the statistical foundation for their frequency and magnitude are enhanced.

Structural Analyses and Calculations. With the prevalence of terms such as computer-
aided engineering, it is understandable if people think that computers can do all the work.
They can, but they cannot do the thinking. The adage “garbage in, garbage out” remains
true, except it occurs faster and faster. Structural analysis, whether computer-aided or not,
is fundamentally a mathematical representation of physics. It remains a designer’s respon-
sibility to ensure that the complete structural behavior of a given system is captured, and to
ensure that the analyses results are properly interpreted. The problem is actually exacer-
bated when engineers fail to appreciate the underlying assumptions and limits of applica-
bility of the analysis software they sometimes use as a design aid. For example, most
“simple” structural analysis approaches assume slabs to behave as rigid diaphragms under
lateral loads. Unless the slabs have sufficient rigidity relative to shear walls and frames to
genuinely qualify as rigid, such analyses will yield unrealistic load distributions, resulting
in some of the lateral frame members to be substantially under-designed.
Design Detailing. How much detail must be shown on construction documents? The sim-
ple answer is: as much as needed to clearly describe the structure. Standard details are often