1832
Proceedings of the 18
th
International Conference on Soil Mechanics and Geotechnical Engineering, Paris 2013
Proceedings of the 18
th
International Conference on Soil Mechanics and Geotechnical Engineering, Paris 2013
methods, but this influence often has a lesser effect on resulting
designs.
Designers can also influence variability and
uncertainty in construction, by developing “constructable”
designs, as well as by requiring and/or engaging in effective
QC/QA. However, the influence is again generally smaller than
what can be achieved through effective site characterization.
3 COMPARISON OF TRADITIONAL & LRFD PRACTICE
While
traditional “allowable stress design” (ASD) and LRFD
practices seek to account for variability and uncertainty
introduced by all three sources, they do so differently.
Differences among traditional geotechnical practice, current
AASHTO specifications, and the new MoDOT guidelines arise
primarily from differences in how these sources are addressed.
3.1
Traditional ASD Practice
Traditional practice for geotechnical site characterization, in
terms of the specific types of measurements made and the
quantity of such measurements, is largely dictated by the
judgment of the designer. In establishing the scope of site
characterization activities, designers generally consider (often
local) standards of practice for structures of similar complexity
and importance as well as general characteristics about the site.
The actual site characterization activities undertaken are also
subject to the
designer’s ability to “sell” the importance of the
activities to those that are paying for the characterization. This
task is often challenging because it can be difficult to quantify
the potential value of site characterization activities in ways that
are meaningful to those outside the profession.
Importantly, traditional geotechnical practice also provides
some flexibility in selection of appropriate values for the factor
of safety to be used in design. In selecting a specific value for a
specific project, designers generally consider the importance
and complexity of the structure, the complexity of the site, and
the appropriateness of site characterization that has been
performed. Thus, there is an implicit link between the quality
and rigor of the site characterization and the safety margins that
are employed in design. This link is clearly subjective, which
introduces the potential for inconsistent application and
inconsistent reliabilities for resulting designs. The subjectivity
may also expose designers to substantial risk, since it can be
difficult to justify specific design postures when performance
does not meet expectations (e.g. if a problem occurs, one can
often easily argue that site characterization was insufficient or
that sufficient margins of safety were not used).
3.2
AASHTO LRFD Practice
Design according to the AASHTO LRFD specifications
largely follows traditional practice, but with two important
distinctions. First, the AASHTO LRFD code explicitly
establishes minimum standards for the quantity and type of site
characterization that must be performed in order for the
standards to be used. These minimum requirements enhance the
designer’s ability to “sell” site characterization and
provide
some minimum level of confidence in the design parameters.
However, the requirements also pose challenges for some
regions of practice where traditional site characterization
practices do not work well. Secondly, the AASHTO code
stipulates fixed values for the margin of safety, via resistance
factors. Table 1 shows a listing of resistance factors for side
resistance of drilled shafts from the AASHTO LRFD
specification (AASHTO, 2010). These
“method specific”
resistance factors are
“lumped”
factors in that they account for
all three sources of variability and uncertainty collectively.
Fixing the magnitude of resistance factors results in more
consistent designs and likely produces the intended effect of
achieving more consistent reliability compared to ASD practice.
However, fixing the magnitude of resistance factors also
eliminates the flexibility provided in ASD to select appropriate
safety margins and limits the capability to improve design
efficiency through improved site characterization. Conducting
more tests, or higher quality tests, to improve confidence in
design parameters does not allow one to use more advantageous
resistance factors. Improving the scope or quality of site
characterization may have a second order effect of changing
predictions of nominal capacity, but it does not allow designers
to exploit the improved confidence in design parameters.
Table 1. Resistance factors from AASHTO LRFD Specifications for
side resistance of drilled shafts (AASHTO, 2010).
Soil/Rock
Type
Design Method
Resistance
Factor,
Clay
O’Neill & Reese (1999)
0.45
Sand
O’Neill & Reese (1999)
0.55
IGM
O’Neill & Reese (1999)
0.60
Rock
Horvath and Kenney (1979)
0.55
O’Neill & Reese (1999)
0.55
Carter & Kulhawy (1988)
0.50
The AASHTO code may reduce the risk to geotechnical
designers in the sense that they may have stronger defense
against litigation as long as the minimum requirements are
satisfied. However, the code also requires designers to employ
judgment to expand the site characterization where conditions
warrant so the practical truth for this is at least debatable.
3.3
MoDOT LRFD Practice
The MoDOT design guidelines seek to address limitations in
traditional ASD and AASHTO LRFD practices by linking the
resistance factors used with the quality of site characterization
performed, and simultaneously improving consistency by
restricting this link so that the target reliability is more
consistently achieved.
The “link” in this case is formed by
implementing resistance factors that depend on the variability
and uncertainty in design input parameters, which in turn
depends on the quality of the site characterization conducted. In
implementing this link, the guidelines also provide designers
with practical means to estimate the potential value of site
investigation activities on a project specific basis so that more
effective site characterization decisions can be made.
4 GUIDING PRINCIPLES
The principal objective for development of the MoDOT design
guidelines was to provide procedures that would save agency
funds by more precisely and consistently achieving target
probabilities of failure in design (i.e. applying appropriate
conservatism for the variability and uncertainty present for each
specific project). The primary means for improving the
precision of the procedures is by considering the variability and
uncertainty in design input parameters separately from the
variability and uncertainty in design and construction methods.
The predominant cost savings are expected to be savings in
construction costs rather than savings in site characterization
costs. However, it was recognized that conducting advanced or
extensive site characterization to reduce variability and
uncertainty in design parameters is not always justified and will
not always produce net cost savings. The overall intent was
therefore to provide the agency with practical procedures to
identify conditions where more extensive investigations are
likely to produce cost savings, considering the costs for site
investigation, costs for construction, as well as potential future
costs for maintenance and repair.
Conscious effort was also made to avoid overly prescriptive
provisions. Rather, the intent was to provide methods that
inform the judgment of the designer about the value of
