Page 258 - Forensic Structural Engineering Handbook

7.42 CAUSES OF FAILURES

severe enough to cause significant distress in conventional buildings, its manifestation is
readily visible.

Determination of Settlement Loads
As with self-straining loads, loads due to settlement are the result of internal restraint of
movement, rather than from an externally applied load. Structures that are fully pinned
theoretically can be subjected to settlement without developing stresses. However, since
few practical structures are without restraint, structures generally become stressed when
settlement occurs.
The analysis of settlement effects is best begun with a survey of the building to measure
its distortion. Care must be taken to obtain accurate data on relevant structural elements.
Data points must be selected to be representative of the movements of the structure. Normal
deformations due to beam flexure, for instance, must be excluded from settlement defor-
mation analyses. Data collected on architectural features that are not locked fully with the
structural system, such as wood trim or column boxing, may not reflect structural move-
ments accurately.
Useful places to acquire data are the tops of column baseplates, because these ele-
ments usually are installed at elevations close to those specified. Differences in eleva-
tions of the tops of structural steel near columns on any level of the building usually are
reliable indications of post-construction settlement amounts. When the floor system is
concrete slab, analysis of the patterns of elevation differences, as measured on the top
of slabs next to columns over several floor levels, can reveal settlement amounts. When
performing such surveys, care must be taken to account for elevation differences intended
by the design.
The investigation must establish whether the original, as-built positions can be deter-
mined with sufficient certainty. Normal construction tolerances allow floor surfaces, for
instance, to vary from specified elevations. Data should be assessed with consideration
to the possibility that data points were not precisely at specified elevations, but within
tolerances. Furthermore, it is common for buildings to be erected out of level or out of
plumb.
It is also possible the differential settlement begins during construction. As such, sub-
sequent construction sometimes masks settlement that occurred previously, by progres-
sively “re-leveling” at each new construction phase. For example, foundation settlement
during construction of a concrete building is more easily detected at the base of the build-
ing, where the structural elements experienced the full amount of construction phase set-
tlement, than at the top of the building.
If an as-built survey was performed at the time of construction, it can be used to com-
pare to the data collected during the investigation. The investigator must assess the preci-
sion with which field surveys are performed so that measurement errors do not unduly
affect judgment about settlement. When high precision is required, such as when move-
ments are to be monitored over time, special survey instruments and procedures are needed
to achieve the required precision.
Sometimes useful information can be obtained by observing the architectural and struc-
tural features of the building. For instance, mortar joints in masonry construction usually
are constructed relatively straight and level.
Once the amount of settlement is known, the load effects on the structure must be
determined by analysis. Reliable models must represent the behavior of the structure as
it is deformed. Using these reliable models, forces in structural elements can be evalu-
ated by imposing the measured support movements. Care must be taken to ensure that the
model adequately represents connection stiffness and the potential for bearing points or