Page 358 - Biaxial Multiaxial Fatigue and Fracture

342 C. CALi R. CITARELLA AND M. PERRELLA

simplified computer programs. Load spectrum, threshold effects, environmental conditions,
microstructural effects, small crack effects, Multiple Site Damage (MSD) conditions, material
parameters scatter, mixed loading conditions (material flaws or pre-cracks, which may have
been introduced unintentionally during the manufacturing process, can have an arbitrary
orientation with respect to the loading applied to the component) and complex three
dimensional geometry, all complicate the process of predicting fatigue crack growth in real
word applications. This paper focuses on some of these complications (see also [I]): load
spectrum influence, complex three dimensional geometry, fatigue material parameters
assessment, mixed loading conditions and MSD conditions.
A series of laboratory tests have been designed and implemented to assess fatigue material
parameters and to evaluate three dimensional fatigue life prediction capabilities for a numerical
life prediction code (BEASY), based on Dual Boundary Element Method (DBEM). With such
code the geometry of the test specimen and the shapes of evolving crack fronts, are not
restricted to the simplified configurations found in the libraries of many commercial codes.
Many complications were purposely minimised: initial crack sizes and loading were such that
small crack and threshold effects had little consequence on the propagation phenomena;
environmental and microstructural effects were considered part of the experimental scatter.

EXPERIMENTS
Experimental fatigue tests were performed on complex geometry notched plates undergoing
cyclic axial load. The crack initiation process and the crack propagation were monitored and a
general traction load spectrum was applied on the specimen. Experimental crack growth rate
and crack path were compared with those obtained with a numerical procedure based on
DBEM and the correct simulation of the load introduction to the specimen was checked by
strain gauge measurements.

Simple geometry cracked plates and two-dimensional crack propagation

The fatigue data necessary for the numerical analysis were previously obtained by
experimental crack growth tests on simple geometry specimens [2]. All of the test aluminium
alloy specimens described in this paper were manufactured from a single lot. The first part of
the fatigue experimental tests was carried out on 3 simple notched (hole/slut) specimens
(Fig. l), in such a way to work out, with statistical significance (for R=O. I), the material fatigue
parameters for crack growth simulation. The initial through the thickness notch was cut by a
thin saw.
The plate geometry as well as the whole testing procedure was consistent with the ASTM E
647-91 specifications. A constant amplitude fatigue traction load (P-=14 kN
R=P,,JP-=O. I) was applied by a servo-hydraulic machine (Instron 8502), with a frequency
f=5 Hz, at ambient temperature.
During the experimental fatigue test on simple specimens, crack length data, measured by
optical systems, were recorded and used to work out, by analytical formula, the corresponding
SIF’s (Stress Intensity Factors) and crack growth rates. The overall specimen size was chosen
sufficiently large in order to get a mainly elastic behaviour, with plasticity effects confined to
the final part of crack propagation, whose monitoring can start only after a pre-cracking phase,
as prescribed by the ASTM E 647 standard.
The observed rectilinear crack propagation path turned out to be consistent with symmetric