Page 384 - Forensic Structural Engineering Handbook

STEEL STRUCTURES 11.15

13
a role. These properties are defined by ASTM Specifications such as E8 and A370. The
significance of these properties to failure is discussed later in this chapter; but, in general,
since design against failure requires knowledge of the maximum load which any part of a
structure must be able to bear without either excessive local or general yielding or separa-
tion of members through ductile fracture, they are critical properties. The significance of
the ductility-based parameters, total elongation and reduction of area, is not that designers
directly utilize these properties, but rather, they design structures to undergo general yield-
ing and to demonstrate that the materials are being used in the ductile regime. Since all
structures, especially welded structures exhibit local “hot spot” stresses that can be at the
yield strength of the material, the ability to plastically deform, so as to relax or redistribute
these local stresses, is essential to the practical utilization of the measured yield and tensile
strengths. Normal design practice is based on utilizing only some fraction of the yield and
tensile strengths of the materials, typically on the order of 70 percent of the yield strength
and/or 50 percent of the tensile strength, thus leaving a reserve “safety factor” beyond
which may result from typically conservative sizing of members. However, potential prob-
lems exist when the tensile or other mechanical properties of steels are determined accord-
ing to the requirements of ASTM Specification A6 or A20, which is the normal practice.
These two specifications, for materials for general construction and for pressure vessel
materials, respectively, also define the manner in which tension and other kinds of tests
shall be made, where tests to measure these properties will be taken, and how many tests
shall be run. In attempting to use these measurements to determine what the mechanical
properties of a structural material are at a potential or actual failure location, it is important
to realize that the metallurgical structure and therefore strength, ductility, and toughness of
a steel will vary with material thickness, location within a specific member, and even loca-
tion within the original ingot (if ingot cast) from which it was produced. These variations
are not usually revealed when performing standard tests taken from standard locations with
specimens in standard orientations and, thus, it may be essential to perform tests at the loca-
tions and in the orientations which, as much as possible, match those of the actual or poten-
tial failure location in service. Deviations from standard test values determined from such
nonstandard tests should not be taken as a failure of the material to meet “specifications”
unless modified test procedures were requested and agreed to by the material supplier; but
the deviations can reveal why a failure occurred or could occur even though it would oth-
erwise seem unlikely. In addition to the discussion on the resistance of structural steel
members to failure, it should be noted that most failures of steel structures are caused by or
within the connections that join the individual members together.
At the beginning of the 20th century a common method of joining structural steel mem-
bers together was the use of rivets. By the mid-20th century, structural bolts and welding
were the primary methods used to connect structural steel members. Welding, in particu-
lar, has created extreme challenges to the designer of steel structures to understand the
behavior of structural connections in addition to the performance of the structural steel
members themselves. The behavior of these structural steel connections has had a major
influence on the material presented in the remaining sections of this chapter. Of the thou-
sands of failures evaluated by the authors, the overwhelming majority have occurred at
connections. Seldom do failures occur remote from connections.

Toughness and Fracture Toughness
Fracture toughness, when evaluated for structural steels, is usually performed using the
CVN impact test as defined by ASTM Specification E23. The locations of test pieces and
testing specimen sizes are also designated in ASTM Specification A673 for structural steels
(where toughness tests are frequently only a supplemental requirement) and ASTM