Page 451 - Failure Analysis Case Studies II

434

with big grains (up to -400 pm in diameter); intergranular but with small grains (-20 pm in
diameter); and mixed modes of fracture, transgranular and intergranular with large grains (see
Fig. 9).
In the center of the sleeve thickness, the fracture morphology was brittle cleavage (Fig. lo), and
at the inner diameter (ID) the fracture was ductile by nucleation and coalescence of microvoids,
indicating that the final fracture of the sleeve was produced in this zone.

4.2. Metallography
In the lower part of the sleeve a metallographic study was carried out on axial sections cut from
the fracture surface up to 70 mm below it. In all these sections no damage at the ID was found.
At the OD, the sections including up to 30 mm below the fracture, corresponding to the external
area of grooves, showed damaged material. This damage consisted of crack initiations, and loss
and dissolution of material (Fig. 12). EDS analysis were performed in some of these zones (see
Section 4.3). On the other hand, no chromium-plated areas were visible in these sections. After
etching the sections, different microstructural characteristics were observed (Fig. 13), indicating
that these areas have suffered heterogeneous heating at different temperatures and, even, at very
high temperatures as demonstrated by the large grain size existing in some zones. These changes
in the microstructure explain the different morphologies detected in the fracture surface.
In the metallographic samples containing the material from 30 to 70 mm under the fracture,
where externally no grooves were observed, the chromium plated layer was clearly detected
although with variable thickness (from 90 to 550 pm) and with some small cracks. In these sections
no changes in the microstructure were observed through the thickness of the sleeve.
Some metallographic sections were prepared from the upper part of the sleeve. The characteristics
observed were similar to the ones described above.

4.3. EDS analysis

EDS analyses were carried out in some of the damaged zones existing at the O.D. As an example
of the results obtained, Fig. 14 shows that generally the main element detected is chromium.
Taking into account the amount of this element in the base material, - 13%, it can be assumed
that these high concentrations are related to the original chromium-plated layer. There are also
significant amounts of copper in all the analyses and lead in almost all the analyses. These elements
may come from the initial bronze bearing. In addition, nickel is detected in some of the analyses,
an element that can be related to the graphite-nickel bearing placed after the bronze bearing.

4.4. Microhardness

Vickers microhardness with a load of 200 g were measured in zones close to the fracture surface
and in zones 30 mm from the fracture. In the last case the average value obtained was 325 HV, in
accordance with the hardness of this material. However in areas close to the fracture, the average
value at the center of the sleeve thickness was 420 HV while at the O.D., where different micro-
structural characteristics were observed, microhardness values from 250 to 375 HV were measured.