Page 70 - Structural Steel Designers Handbook AISC, AASHTO, AISI, ASTM, and ASCE-07 Design Standards

Brockenbrough_Ch03.qxd 9/29/05 5:05 PM Page 3.2

CONNECTIONS

3.2 CHAPTER THREE

3.1.2 Justifying Connection Design
A proposed connection design method can be justified in three ways: through precedence, testing, or
analysis.
Precedence simply means that there is sufficient historical record of adequate performance of a
connection configuration or an assumption to justify its use. Many valid arguments can be made
against accepting a connection design method based on precedence. Because of the conservatism
built into design loads and load factors and the fact that loads can often redistribute, connections in
service may rarely see their full design loads. Therefore, a history of satisfactory service may not
correlate directly to a safe design. However, some assumptions implicit in AISC, “Manual of Steel
Construction,” 2005 (referred to herein as the AISC Manual), are based largely on precedence. For
instance, the bolts at the supported member of a double-angle connection are typically not designed
to resist any eccentricity, though logically an eccentricity could exist. An argument can be made that
the flexing of the angles relieves the eccentricity, and therefore the bolts do not have to be designed
to resist this rotation. However, the support must now take this neglected moment. The argument can
then be made that the eccentricity is small, and the supporting member probably has some excess
capacity. All of these are qualitative arguments with little analytical basis. The only real justification
that can be found to support this assumption is decades of satisfactory performance. Precedence
should not be overlooked as a valid justification for engineering practices, but it must be used with
caution and must be evaluated whenever paradigm shifts occur in design philosophies, especially
when these shifts involve load determination or resistance factors.
Connection designs or design assumptions can also be justified by testing. This approach has
been used to develop a handful of essentially prescribed connections, the standard single plate shear
connection being the most notable. For many, this approach may be considered the “gold standard”
for justifying a connection design, but it requires a great deal of financial investment, sometimes with
relatively little return, since results are often valid only for a range of strictly defined parameters.
Greater benefits from testing are more often achieved when an analytical model can be found to pre-
dict the results of testing. This analytical model can then be applied to a wider range of conditions.
Often testing is performed to determine the effects of a single limit state. These data are then used
to develop a model for use with more complex conditions.
The final and most common way to justify a connection design is through analysis. Precedence,
testing, and engineering theory and judgment are coalesced to produce a rationale to justify the con-
nection to be used. This is the art of connection design. Simple tests are extrapolated to more com-
plex configurations. Load paths are analyzed and optimized. Assumptions are scrutinized to ensure
their validity. In some cases these procedures are clearly codified. In many others they are not. The
tools are essentially the same as those used in main-member design: statics to satisfy equilibrium,
mechanics of materials to confirm strength and determine load paths, and statistical analysis to deter-
mine reliability. When combined with sound engineering judgment, these tools allow the connection
design engineer to provide safe and economical connections for structural steel.

3.1.3 Choosing Load Distributions—Reconciling As-Built with As-Modeled
As previously stated, the as-built condition seldom re-creates the assumed as-modeled condition
accurately. This fact sets up a paradox for the connection design engineer, who is charged with bringing
the analytical model into existence. The connections must be configured in an attempt to re-create
the assumed behavior, while at the same time recognizing that practical limitations prevent an exact
re-creation.
As an example, consider a 30-ft-long W16 × 30 beam that supports a uniform load of 1.8 kips/ft.
During the design of the beam, the beam ends are assumed to be pinned at the supports. For ease of
erection, the connection design engineer chooses to use extended plates from the webs of the col-
umn to support the beam. This arrangement places a line of bolts 9 in from the center of the column,
and a moment equal to (9 in)(27 kips) = 243 in⋅kips between the bolts and the center of the support.
See Fig. 3.1. The connection design engineer is now faced with the decision as to where to take the

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)
Copyright © 2004 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.