**4.3 Fractographic examination**

The aims of conducting fractographic analyses are to identify the origin and the propagation direction of a crack. This analysis is often performed using a combination of low and high-magnification microscopes. With the aid of a stereomicroscope, the origin/cause of a failure can be detected. At the same time, a scanning electron microscope can reveal detailed features and the identity of the failure cause. As shown in **Figure 7a**, the region identified as "Old fracture" is the fracture surface created during the in-service of the rotor, while the one marked as "New fracture" is the surface made during the splitting process. The "Old" fractured surface was critically inspected using a stereomicroscope to identify the failure's origin and root cause, as illustrated in **Figure 7a**. A thorough examination of the cracked surface showed some microstructural features variation from the primary matrix microstructure. The broken arrows in **Figure 7a** indicated some whitish gray phases, presumed to be corrosion

#### **Figure 7.**

*(a) Stereomicroscope analysis showing fracture surface features. The broken arrows indicate corrosion products; the circled areas indicate foreign structures; and (b) a micrograph showing the propagation of cracks along the graphite-pearlite interface.*

**Figure 8.** *SEM micrographs of the rotor's fracture surface.*

products. The region marked with a solid rectangle was further magnified to reveal its microstructural make-up. However, some distinct phases were noticed in the magnified micrograph (areas highlighted in broken circles), which were further examined through an SEM-EDS analysis for identification. Additionally, a metallographic sample was extracted from the region around the fracture and observed under the optical microscope, revealing that the crack, having initiated, propagated circumferentially along the flaky graphite-pearlite interface, as illustrated in **Figure 7b**.

**Figure 8** displays the SEM micrographs of the rotor's fracture surface. Different phases with varying textures can be identified from the micrographs. An inspection of the crack surface showed a relatively rough texture with a significant amount of oxidized products and foreign bodies inclusion, as displayed in **Figure 8**. The dark gray is the graphite phase, while the whitish-colored areas are the oxidized pearlite matrix due to the penetration of corrosive agents into the fractured surface leading to the corrosion of the cracked surface. Besides, some areas were recognized with a distinct appearance from the major constituent phases, as shown in **Figure 8b**. These critical features are labeled 1–6 (**Figure 8b**) and further investigated at high magnification to reveal their details. **Figure 9** presents the high magnification of some of the distinct microstructure features identified on the cracked surface of the rotor. A high number of foreign inclusions were also identified, some highlighted in **Figure 8b**. The foreign inclusions occasionally appear as a blocky (**Figure 9**, images 3–6) or lumpy (**Figure 9**, images 1, 2 and 7) structure. The location of the inclusions is approximately 1.620 ± 0.237 mm away from the disc in-board radial perimeter, as indicated in **Figure 8a**. **Figure 8** also displays that the cracked surface portrays layered structures suggesting brittleness, a characteristic fracture behavior exhibited by GCI materials [11, 29]. A network of fine microcracks can be seen on the fracture surface and even in the vicinity of the inclusions (**Figure 8**). Numerous microcracks are observed around these inclusions (**Figure 9**), suggesting potential nucleation sites for crack initiation.

However, an SEM-EDS spot and elemental mapping analysis were employed to identify the origin and composition of the foreign inclusions shown in **Figure 9**. The results of these analyses are presented in **Figures 10**–**13**. Furthermore, a close examination of the EDS analysis of Spots a and c in **Figure 10** reveals that the foreign materials in micrographs 3 and 5 (**Figure 9**) are silica sand (SiO2) inclusions. The shape of the sand inclusions is comparatively regular, as seen in the figure. However, some other oxides can be observed admixed with the silica as represented in the EDS microanalysis of **Figure 10d** and the SEM-EDS elemental mapping of micrograph 3 (**Figure 11**).

*Root Cause Failure Analysis of Castings: A Case Study of a Brake Rotor DOI: http://dx.doi.org/10.5772/intechopen.107950*

**Figure 9.** *SEM micrographs of the inclusions found on the fracture surface of the rotor.*

#### **Figure 10.**

*SEM-EDS spectra of spots (a-f) indicated in Figure 9.*

#### **Figure 11.**

*SEM-EDS elemental mapping of micrographs 1 and 3 of Figure 9***.**

Silica sand with binders is generally used for making mold and core in the iron casting process and can be differentiated from slag because it comprises a single phase (**Figure 10a** and **c**). The presence of silica inclusions can be attributed to the entrapment of mold material during metal pouring due to the erosion of or loose mold

*Root Cause Failure Analysis of Castings: A Case Study of a Brake Rotor DOI: http://dx.doi.org/10.5772/intechopen.107950*

**Figure 12.**

*SEM-EDS elemental mapping of micrographs 5 and 7 of Figure 9***.**

material. By analyzing the EDS microanalysis of spots b, e-f (**Figure 10**) and comparing it with the elemental mappings shown in **Figure 11** (micrograph 1) and 12 (micrograph 7), it can be seen that some of these lumpy inclusions are the oxides rich in calcium, silicon, zinc and aluminum with traces of admixture oxides of Mg, Ti, Mn, Na, K and Fe. The analysis also reveals minor phosphorous, sulphide and chloride concentrations (**Figure 10b, e, f, 11** and **13**).

**Figure 13.** *SEM-EDS elemental mapping of micrographs 8 and 9 of Figure 9***.**

#### *Root Cause Failure Analysis of Castings: A Case Study of a Brake Rotor DOI: http://dx.doi.org/10.5772/intechopen.107950*

The fractographic examination conducted on the failed rotor has identified indigenous and exogenous inclusions on the disc's cracked surface. The inclusions from external sources like slag, dross, ferroalloys, flux materials, ladle lining, and entrapments from eroded mold, core and refractory materials are categorized as exogenous. Those native or inherent in the molten metal treatment process are known as indigenous inclusions. Indigenous inclusions are the products of the liquid melt reactions with deoxidizers such as silicon, manganese, and aluminum particles or during desulphurization, creating some residual oxide inclusions in the casting [22]. However, previous studies have recognized ferroalloys as a significant source of inclusions in steel and cast-iron castings [11, 30]. Ferroalloys are pre-alloyed raw materials bonded with iron used in the treatment of molten iron and steel to produce castings with desired chemical composition. In general, the primary applications of ferroalloys in iron and steel makings include (i) alloying sources to enhance the mechanical properties and functional characteristics of iron and steel products (e.g., FeCr, FeMo, FeTi, FeMn, FeW), (ii) deoxidizers such as FeSi, FeMn, SiMn and FeAl, and (iii) reducing agents such as FeSi which can be used as a reducing agent to produce FeMo, FeV and other alloys [30]. Unfortunately, several impurities accompanied ferroalloys production, including H, N, O, S and P and other trace elemental impurities such as Mg, Al, Ti, V, Ca, etc. In the cast-iron melting process, FeSi ferroalloys are the common source of silicon addition to the molten iron to decrease its melting point, improve fluidity, and promote graphitization. These FeSi alloys are the source of Al and Ca impurities, which can significantly affect the quality of the castings by forming oxide inclusions. Unfortunately, these inclusions provide excellent nucleation sites for microcracks during cooling, which is why they were consistently found within microcracks [30]. Earlier studies [31] have shown that the predominant composition of slag formed from spheroidal and lamellar irons processing consists of several oxides, including FeO, MnO, SiO2, Al2O3 and MgO. In a similar investigation, Jonczy [32] concluded that the dominant component of the cast iron slag is silica (62.04%) in addition to 11.03% of Al2O3, 10.38% of MnO, 6.32% of MgO, 5.37% of CaO with the admixture of iron, sodium, potassium, barium and sulfur oxides. The chemical composition of slag can vary greatly depending on the level of impurities in the starting raw materials (e.g., pig iron and scraps) and other additives (e.g., flux agents and ferroalloys). For instance, the slags formed using steel scrap-based charges showed the highest zinc and aluminum contents and created a crystalline ZnAl2O4 (gahnite) phase in addition to other admixture oxides (e.g., SiO2, Al2O3, CaO, MgO and MnO). These high zinc and aluminum contents are due to the use of galvanized steel scrap as raw material (Zn and potentially Al) and of FeSi and SiC as additives (Al) [31]. This probably explains the high concentrations of Zn, Al and Ca found in some of the inclusions identified in **Figure 10(b, e** and **f)**.

The SEM-EDS elemental mappings shown in **Figures 11** and **13** depict the existence of chloride ions on the cracked surface of the rotor, which is also corroborated by the EDS microanalysis presented in **Figure 10f**. It can be observed that the chloride ions covered the entire disc's fractured surface, suggesting that the chloride ions possibly originate from an outside source, such as road salt. During winter, calcium and sodium chloride salts are generally used for roadway de-icing, thereby creating a high concentration of chloride ions on the motorways. Salt solution on the road can easily penetrate the crack surface of the rotor (**Figure 4**) and hasten the corrosion of the fracture surface (**Figure 8**). Chloride ions are a highly corrosive agent that serves as a catalyst and exacerbates the deterioration of steel and cast-iron components [11]. Analysis of the disc's fractured surface showed that the rusty products were only noticed in the pearlitic matrix areas, as displayed in **Figure 9**, micrographs 6–9, since graphite does not corrode.
